An Overview of Foiate etabolism: Features Relevant to the
`
`Hilary Calvert
`
`S INCE the observati~m of reduced folate levels
`
`in children with leukemia made by Farber et
`al~ in the 1940s, the study of folic acid metabolism
`and the action of antifolate drugs has been inti-
`mately linked to the development of cancer ther-
`apeutics, Folic acid plays a role in a wide range of
`metabolic pathways in various species. In humans
`it is an essential vitamin and functions primarily in
`the processes involved in cellular proliferation and
`amino acid metabolism. This review will focus
`mainly on those aspects of mammalian folate me-
`tabolism relevant to cell proliferation since these
`are the most germane to the use of antifolates in
`cancer therapy. The textbook by R.L. Blakleyz is a
`comprehensive work covering all aspects of folate
`metabolism.
`
`ASPECTS OF FOLATE METABOLISM
`
`Fotate Pathways Associated With Cell Proliferation
`
`Folic acid fi~nctions mainly in its fully reduced
`form, 5,6,7,8-tetrahydrofolate (FH4; Fig 1), FH4
`serves as a carrier for one-carbon moieties within
`the cell. These are obtained from a variety of
`sources that include serine, In this reaction, serine
`hydroxymethyl tmnsferase forms 5,10-methylene
`tetrahydrofolate (CHxFtt4) while converting
`serme to glycine (Fig 2). CH2FH4 may be con-
`vetted within the cell to one-carbon carrying
`late derivatives of various oxidation states. One of
`these, 10-formyl tetrahydrofolate, is the substrate
`for two enzymes involved in the de novo synthesis
`of purines. These are glycinamide ri~onucleotide
`formyl transferase (GARFT) and aminoimidazole
`carboxmnide ribonucleotide formyl transferase
`(AICARFT). Thus, two of the carbon atoms in
`the purine skeleton are derived from folate. The
`folate-dependent reactions of purine synthesis use
`the carbon atom from the 10~formyl group and
`release unsubstituted tetrahydrofolate as tbe folate
`product. Thus, the folate molecule can then ac-
`quire another carbon atom from serine and con-
`tinue to cycle through GARFT and AICARFT,
`allowing continued purine synthesis without any
`overall consumption of folate. CH~FIt4 is also the
`substrate for the enzyme thymidylate synthase
`(TS). Thymidylate synthase converts deoxyuri-
`
`Seminars in Oncotogy, Vol 26, No 2, Suppl 6 (April), 1999: pp 3-10
`
`dine monophosphate into thymidine m~mophos-
`phate and is a key enzyme involved in cell prolif-
`eration because it is the rate-limiting step in the de
`novo synthesis of thymidylate, which is required
`exclusively for DNA synthesis. The folate product
`of TS is not tetrahydrofolate, but the oxidized
`form, dihydrofolate (FH2). This product cannot
`continue to function in folam ~netabolism until it
`is converted back to FH4 by the enzyme dihydro-
`folate reductase (DHFR),
`
`The Rt~le of Folate and Antifolate Polygtutamates
`
`Folic acid possesses a glutamate residue shown
`at the right-hand side of the folate structures in Fig
`1. Naturally occurring folates within the cell are
`converted to polyglutamate forms by the addition
`of glutamate residues via a T-peptide linkage. An-
`tifolates that possess a glutamate residue (known
`as classical antifolates) are also frequently con-
`vetted into their corresponding polyglutamate
`forms. The process of polyglutarnation is accom-
`plished by the enzyme folylpoly-y-glutamate syn-
`thetase, This reaction is illustrated in Fig 3 using
`the antifolate LY231514 (MTA) as an example.
`The process is analogous for natural folates and
`many other classical antifolates, tn Fig 3, the car-
`boxylate groups of the glutamic acid residue are
`shown in their ionized form, carrying a negative
`charge, showing that polyglummation increases
`the overall negative charge on the folate molecule
`by one unit for each additional glutamate, The
`negatively charged polyglutamates cannot cross
`the cell membrane and are therefore retained and
`concentrated within the cell. This is probably the
`major physiologic role of polyglutamation. Cells
`that are deficient in folylpoly-T-glutamate syn-
`thetase are auxotrophic for the end products of
`
`From the Cancer Research Unit, DOartment of Oncalogy,
`University of Newcastle upon Tyne.
`Sponsored by Eli Lilly and Company.
`Dr Calvert is a consultant for and has received research funding
`from Eli Lilly and Company and Zeneca.
`Address reprint requests to Hilary Calvert, MD, Cancer Re-
`search Unit, Department of Oncatogy, Fremting~on Place, Univer.
`sity of Newcastle upon Tyne, NE2 4HH.
`Copyright © 1999 by W.B. Saunders Company
`0093.7754/99/2602-0602510.00/0
`
`Sandoz Inc.
`Exhibit 1007-0001
`
`JOINT 1007-0001
`
`

`
`4
`
`HILARY CALVERT
`
`Folic Acid
`
`Dihydrofolic
`Acid (FH2)
`
`COOH
`
`COOH
`
`~ COOH
`
`Tetrahydrofolic
`Acid (FH4)
`
`~
`
`COOH
`
`NH=
`
`COOH
`
`5,1 0-Methylene
`tetrahydrofolic
`Acid (CH2FH4)
`
`1 0-formyl
`tetrahydrofolic
`Acid (CHOFH4)
`
`5-methyl-
`tetrahydrofolic
`acid
`
`Fig 1. Forms of folio acid.
`
`folate metabolism (thymidine, hypoxanthine, and
`glycir~e). In addition to being retained within the
`cells, the polyglutamate forms of natural folates
`also may be better substrates for the various folate
`metabolizing enzymes.
`The formation of polyglutamates of those anti-
`folates that are substrates for folylpoly-y-glutamate
`synthetase also has profound effects on their ac-
`tivity. The polyglutamates may be retained within
`the cell for very long periods? thus increasing the
`potency of the cytotoxic action of these com-
`pounds. In addition, the addition of glutamate
`residues frequently renders the compounds much
`more potent inhibitors of their target enzyme. For
`example, raltitrexed pentaglutamate is roughly
`
`100-fold more potent ~s a TS inhibitor than the
`parent moleculeA The effects of polyglutamation
`on the potency of molecules such as these is so
`profound that they may be considered as prodrugs
`for their polyglutamam forms. Indeed, cellular
`sistance to antifolates can be caused by a reduction
`in the ability of the cell to forra the polyglutamate
`derivatives.5 A more complete and in-depth re-
`view of polyglutamation and its relevance to can-
`cer therapy is given by Richard G. Moran else-
`where in this supplement.
`
`Cell Membrane Transport of Fotates a~d Antifolates
`
`Folates do ~ot cross the cell membrane to an
`appreciable extent by passive diffusion but require
`
`Sandoz Inc.
`Exhibit 100%0002
`
`JOINT 1007-0002
`
`

`
`OVERVIEW OF EOLATE METABOLISM
`
`5
`
`PYRIMiDINE SYNTHESIS
`
`PURINE SYNTHES~S
`
`dUMP CHFH,~ ~
`
`C H~ F~H4 C H OFH4.~,.~
`
`GARFT
`
`TMP
`
`/
`
`DNA
`
`~ FH,~
`
`DHFR
`
`/
`
`DNA
`
`\
`
`RNA
`
`Fig 2. Netabolic pathways of folate metabolism.
`
`specific transport mechanisms. Them are several
`mechat~isms that have been characterized; these
`are reviewed in depth by Sierra and Goldman in
`this supplement. Of these, the most extensively
`characterized mechanism involves the ]~duced fo-
`late carrier (RFC1).’ This is an anion exchange
`concentrative process and is known to be capable
`of transporting methotrexate and a number of
`other antifoIates as well as tetrahydrofolam itself.
`
`Changes in this carrier that alter its relative affin-
`ity for antifolates have been shown to be a cause of
`drug resistanceA The reduced folate carrier has a
`relatively low afIinity for natural folates (1 to 5
`/xmol/L) compared with their physiologic extracel-
`lular concentrations (typically in the nanomolar
`region). A second mechanism of folate transport,
`the folate receptor, has a much higher affinity for
`folates which, after binding, am internalized
`
`~~,,foo- p%~tAa.qlutamate
`
`MTA
`
`~c~-o-
`o
`
`Fig 3, Formation of polyglutamates.
`
`Sandoz Inc.
`Exhibit 1007-0003
`
`JOINT 1007-0003
`
`

`
`6
`
`HILARY CALVERT
`
`within a membrane vesicle and subsequently re-
`leased into the cytoplasm. Three genes for folate
`receptors have been cloned (see Sierra and Gold-
`man). Folate receptors may be responsible for the
`transport of some antifolates, for example, lome-
`trexol. In addition to these two mechanisms, the
`level of folates and antifolates within the cell may
`be affected by an energy-dependent efflux pump
`and by a low pH tr~sporter.
`
`Actions of Various Antifolates
`Having been introduced nearly 50 years ago,
`methotrexate (Fig 4) is the antifolate with the
`
`longest history. It acts mainly by inhibition of
`DHFR. The result of this inhibition is that in-
`tracellular folate accumulates in the form of
`dihydrofolate. There is a consequent inhibition
`of the de novo synthesis both of purines and
`thymidine. This may be due in part to a dimi-
`nution in the intracellular pools of tetrahydro-
`folates, but additionally, methotrexate polygl.u-
`tamates and the accumulated dihydrofolate
`polyglutamates are capable of inhibiting both
`TS and AICARFT directly,v’9 Characteristi-
`cally, the intracellular pools of dihydrofolate
`and deoxyuridine will increase following expo-
`
`NH
`
`O
`
`COOH
`
`COOH
`
`~oo~
`
`COON
`
`Methotrexate
`
`Raltitrexed
`(TomudexTM,
`ZD 1694)
`
`NH % COOH
`
`CB 3717,
`PDDF
`
`Lometrexol
`
`DDATHF
`
`LY 309887
`
`LY 231514
`
`MTA
`
`COOH
`
`Fig 4. Structures of various
`antifolates.
`
`Sandoz Inc.
`Exhibit 1007-0004
`
`JOINT 1007-0004
`
`

`
`OVERVIEW OF FOLATE METABOLISM
`
`7
`
`oxyuridine and, thus, to a disproportionate in-
`crease in the concentration of deoxyuridine. It has
`been shown that this increase in the intracellular
`pool of deoxyuridine monophosphate is mirrored
`by a corresponding increase in the extracellular
`pool of deoxyuridin@~ presumably due to intra-
`cellular phosphatases allowing the release of de-
`oxyuridine from the cells (Fig 7). This provides a
`useful surrogate for in vivo TS inhibition. The
`plasma deoxyuridine levels can be monitored and
`an elevation compared with baseline indicates the
`inhibition, in viw~, of TS..4
`In addition, selective inhibitors of GARFT, the
`first folate-dependent enzyme involved in the
`pathway of de novo purine synthesis, have been
`developed, Examples of these are lometrexol and
`LY309887 (Fig 4). These cornpounds have good
`antitumor activity in preclinical systems with the
`suggestion that their activity may be preserved
`tumor cells that have a nonfunctional p53 path-
`way. The clinical toxicity of many antifolates is,
`not surprisingly, affected by the pretreatment fo-
`late status of the patient. In the case of the
`GARFT inhibitors, the effect of the folate status is
`particularly marked, with the maximum tolerated
`dose being at least 10-fold higher in patients who
`have received folate supplementation compared
`with those who have not. Is
`
`PURINE SYNTHESIS
`
`GARFT
`
`AICARFT
`
`sure to methotrexate~°; these effects are illus-
`trated in Fig 5.
`It has been argued that the effects of methotrex-
`ate on reduced folate pools, and consequently, the
`indirect inhibition of purine de novo synthesis and
`amino acid interconversions, may be detrimental
`to its main antiproliferative action, namely, the
`inhibition of the synthesis of thyrnidylate that is
`required exclusively for DNA synthesis.~ For this
`reason, many researchers have developed antifo-
`lares designed to inhibit TS directly while not
`affecting other folate enzymes. The first of these to
`be used clinically was CB 3717,~2 but this has been
`superseded by raltitrexed (Tomudex, ZD 1694;
`Zeneca Pharmaceuticals, Cheshire, England), which
`is lice,~ed for the treatment of colon cancer in some
`countries. These specific TS inhibitors produce the
`elevation of the deoxyuridine pool in a rnam~er sim-
`ilar to that observed following methotrexate but,
`importantly, dihydmfolate pools do not increase and
`purine synthesis is mmffected (Fig 6).
`Both direct TS inhibitors (such as raltitrexed
`and CB 3717) and drugs that inhibit TS indirectly
`(such as methotrexate) lead to a marked increase
`in the intraceltular pool of deoxyuridine mono-
`phosphate. The reduction in thymidine nucleo-
`tides caused by these drugs leads to activation of
`the pyrimidine synthetic pathways producing de-
`
`PYRIMIDINE SYNTH,~IS
`
`~
`
`Increased ~ool size
`
`Reduced flux
`
` ’dUMP
`
`!
`
`TMP ,
`/
`’
`
`’~/ DHFR
`FH2
`
`~ FH4
`
`DNA
`
`Methotrexate
`
`DNA
`
`RNA
`
`Fig S. Effects of DHFR inhibition.
`
`Sandoz Inc.
`Exhibit 1007-0005
`
`JOINT 1007-0005
`
`

`
`PYRIMIDINE SYNTHESIS
`
`PURINE SYNTHESIS
`
`HILARY CALVERT
`
`RaltitmxCd !
`
`CB 3717
`
`~_
`
`1
`
`. ’dUMP CHFH4
`CHOFH4 "--.., ’f
`~
`11
`--, --’,
`, z~ ~2P~4
`
`/
`
`~. ~ I
`X,.[
`
`FH~
`
`Kev:
`
`DNA
`
`FH~
`
`/ Increased Dool size
`
`Reduced flux
`
`Fig 6. Effects of TS inhibition.
`
`I
`/ \
`
`DNA
`
`RNA
`
`Clinical Measurement of Functional Folate Status
`
`Although the effect of folic acid supplementa-
`tion on reducing the toxicity of antifolate drugs
`(particularly the GARI~T inhibitors) is clear, it
`always has been difficult to correlate antifolate-
`induced toxicity with pretreatment folate levels.
`One possible explanation for this is that the folate
`levels do not adequately reflect the functioning of
`folic acid within proliferating cells at the time of
`measurement. In addition to the pathways dis-
`cussed so far, folic acid is also involved in cellular
`methyIation reactions by virtue of its role in me-
`
`thionine synthesis. CHzFH4 can be reduced to
`5-methyltetrahydrofolate (Fig 1). This is a sub-
`strate for the enzyme methionine synthase, which
`uses the methyl group to convert homocysteine to
`methionine. Methionine in turn takes part in cel-
`lular rnethylation reactions regenerating homocys~
`teine. Methionine synthase is B~2-dependent but
`also uses 5-methyltetrahydrofolate as the co-sub-
`strate. Thus, any functional deficiency either in
`B~z or folate will result in reduction in the flux
`through methionine synthase and a consequent
`increase in the plasma level of homocysteinO6
`
`Plasma
`
`UdR
`
`Intracellular fluid
`
`UdR~ dUMP
`
`:ell Membrane
`
`TS
`
`TMP
`
`Fig 7. Plasma deoxyuridine." a
`surrogate for TS inhibition,
`
`Sandoz Inc.
`Exhibit 1007-0006
`
`JOINT 1007-0006
`
`

`
`OVERVIEW OF FOLATE METABOLISM
`
`Folate or B12
`Deficiency
`
`Homocvsteine
`
`CHaFH4 ~
`
`CH2FH4
`
`Cellular Methvlation
`
`roactiogs
`
`S-adenosvl-
`methionine
`
`~""methionine
`
`Fig 8, Role of 5-methyl tetra-
`hydrofolic acid: a reduction in
`functional folate increases
`plasma homocystelne levels,
`
`lethionine Svnthetase ~
`
`(Fig 8). The measurement of pretreatment plasma
`homocysteine has proved m be a sensitive way of
`predicting the toxicity of MTA.t7
`
`LY23t514 (MTA)
`MTA was developed by Eli Lilly and Company
`(Indianapolis, IN), initially as a TS inhibitor.
`However, it rapidly became clear that, unlike any
`of the other antifolates discussed, MTA is capable
`of inhibiting two other enzymes involved in folate
`metabolism, GARFT and DHFR (see Mendelsohn
`et al, this supplement). MTA also has a broad
`spectrum of preclinical activity, displays different
`patterns of cross-resistance m other antifolates,
`and has an encouraging level of activity docu-
`mented in early phase Ii clinical trialsAs It is
`possible that its capability of inhibiting more than
`one locus contributes to these resul.ts by increasing
`the spectrum of biochetnical profiles of tumors
`potentially sensitive to the drug and discouraging
`the development of drug resistance. Reports that
`folIow in this supplement address these issues in
`detail.
`
`CONCLUSIONS
`
`Naturally occurring folates have complex met-
`abolic pathways and are involved in a number of
`biochemical processes essential to life, including
`cell proliferation. In addition to their direct role in
`various metabolic pathways, a number of other
`phenomena will significandy affect the actions
`both of natural folates and their analogues acting
`as antifolates. These include cell membrane trans-
`port, the formation of polyglummates, and the
`pretreatment folate status of the patient con~
`cerned. The very complexity of the processes in-
`volved suggests ~vays in which the action of anti-
`
`fo[ates could be tuned to have a selective
`advantage agaiz~st tumors compared with normal
`tissues. Several clinically active drugs have already
`been developed. LY231514 (MTA) may establish
`itself as an important addition and advance those
`currently available.
`
`REFERENCES
`
`1. Farber S, Diamond LK, Mercer RD, et al: Temporary
`refiaissions in acute leukaemia in children produced by folic
`acid antagonist 4~amino pteroyl-glutamic acid (aminopterin).
`N Engl J Med 238:787-793, 1948
`2. Blakeley RL: "IT~e Biochemistry of Folic Acid and Related
`Pteridine~s. Amsterdam, The Netherlands, North Holland Pub-
`lishing Co, 1969
`3. Sikora E, ~ackman AL, Newel~ DR, et ab Formation and
`retention and biological activity of Nm-propargy[-5,8-dide-
`azafolic acid (CB3717) polyglutamates in Li210 cetls in vitro.
`Biochem Pharmacol 37:4047-4054, 1988
`4. Bisset GMF, Pawekzak K, Jackman AL, et al: The
`synthesis and ~ymidylate synthase inhibitory activity of the
`poly-g-glummyl conjugates of N-[5.[N-(3,4-dihydro-2-methyl-
`4-oxoquinazlolin-6-yhnethyl)-N-methylamino]-2-thenoyl]-L-
`glutamic acid (ICI D1694) and other quinazoline antifoiates.
`J Med Chem 35:859-866, 1992
`5. Takemura Y, Kobayashi H, Miyachi H, et al: Biological
`acti~,ity and inttacellular metabolism of ZDI694 in human
`leukemia cell lines with different resistmme mechanisms to
`antifolate drugs, Jpn J Cancer Res 87:773-780, 1996
`6. Schuetz JD, Matherly LH, Westin EH, et aI: Evidence for
`a functional defect in the tramlocation of the methotrexate
`transport carrier in a methotrexate-resistant umrine LI210
`leukemia cell line. J Biol Chem 263:9840~9847, 1988
`7. Allegra CJ, Chabner BA, Drake JC, et al: Enhanced
`inhibition of thymidylate synthase by methotrexate polygluta-
`mates. J Biol Chem 260:9720-9726, 1985
`8. Allegro CJ, Fine ~, Drake JC, et al: The effect of
`methotrexate olx intracellular folate pools in human MCF-7
`breast cancer cells. Evidence for direct inhibition of purine
`synthesis, J Biol Chem 261:6478-6485, 1986
`9. A[legra cJ, Drake JC, Jolivet J, et al: Inhibition of phos-
`phoribosylaminoimidazolecarboxamide transformylase by
`
`Sandoz Inc.
`Exhibit 1007-0007
`
`JOINT 1007-0007
`
`

`
`10
`
`HILARY CALVERT
`
`methotrexate and dihydrofolic acid po]yglutamates. Proc Natl
`Acad Sci U S A 82:4881-4885, 1985
`10. Jackson RC, Jackman AL, Calvert" AH: Biochemical
`effects of the quinazoline inhibitor of thymidylate synthetase,
`CB3717, on human lymphoblastoid cells. Biochem Pharmacol
`32:3783-3790, 1983
`11. Jones TR, Calvert AH, Jackman AL, et al: A potent
`antitumour quinazoline inhibitor of thymidylate synthetase:
`Synthesis, biological properties and therapeutic results in mice.
`Eur J Cancer 17:11-19, 1981
`12. Calvert AH, Alison DL, Harland SJ, et al: A phase l
`evaluation of the quinazoline antifolate thymidylate synthetase
`inhibitor Nl°-propargyl-5,8-dideazafolic acid. J Clin Oncol
`4:1245-1252, 1986
`13. Taylor GA, Jackman AL, Calvert AH, et al: Plasma
`nucleoside and base levels following treatment with the new
`thymidylate synthetase inhibitor, CB3717, in De Bruyn C,
`Simmons HA, Muller M (eds): Purine Metabolism in Man IV,
`Part B: Biochemical, Immunological and Cancer Research.
`New York, NY, Plenum, 1983, pp 379-382
`
`14. Raft l, Taylor GA, Calvete JA, et al: Clinical pharma-
`cokinetic and pharmacodynamic studies with the non-classical
`antifolate thymidylate synthase inhibitor 3,4-dihydro-2-amino-
`6-methyl-4-oxo-5-(pyridylthio)-quinazoline dihydrochloride
`(AG337) given by 24 hour continuous intravenous infi_~sion.
`Clin Cancer Res 1:1275-1284, 1995
`15. Laohavinij S, Wedge SR, Lind MJ, et al: A phase I
`clinical study of the antipurine antifolate lometrexol
`(DDATHF) given with oral folic acid. Invest New Drugs 14:
`325-335, 1996
`16. Savage DG, Lindenbaum J, Stabler SP, et al: Sensitivity
`of serum methylmalonic acid and total homocysteine determi-
`nations for diagnosing cobalamin and folate deficiencies. Am J
`Med 96:239-246, 1994
`17. Niyikiza C, Walling J, Thornton D, et al: LY231514
`(MTA): Relationship of vitamin metabolite profile to toxicity.
`Proc Am Assoc Clin Oncol 34:2139, 1998 (abstr)
`18. Calvert AH, Walling JM: Clinical Studies with MTA.
`Br J Cancer 78:35-40, 1998 (suppl 3)
`
`Sandoz Inc.
`Exhibit 1007-0008
`
`JOINT 1007-0008