ICANCERRESEARCH57, 4015-4022, September15, 1997]
`
`Cobalamin Analogues Modulate the Growth of Leukemia Cells in Vitro1
`
`Gary R. McLean, Pradip M. Pathare, D. Scott Wilbur, A. Charles Morgan, Clive S. Woodhouse,John W. Schrader,
`and Hermann
`J. Ziltener@
`
`and Departments of Medicine (J. W. S.J and Pathology and Laboratory Medicine (H. J. ZJ. University of British
`The Biomedical Research Centre (G. R. M., J. W. S., H. J. ii
`Columbia,
`Vancouver, British Columbia, Canada V6T 113; Department
`of Radiation Oncology. University
`of Washington.
`Seattle. WA 98103 fP. M. P.. D. S. WI:
`and
`Receptagen Corp.. Edmonds, WA 98020 (A. C. M., C. S. W.J
`
`(5, 6). Cbl deficiency may be
`enzymes
`tioning of the Cbl-dependent
`brought on by a lack of dietary Cbl, dysfunction of Cbl uptake via
`abnormalities
`in the binding proteins,
`including TCII (7), or errors of
`intracellular Cbl metabolism (8). Because Cbl deficiency can result
`in
`decreased cell proliferation, as evidenced in megaloblastic
`anemia, we
`have been investigating
`new methods to interfere with Cbl metabo
`lism as part of a program to develop
`antiproliferative
`agents.
`There are many naturally occurring analogues of Cbl (9), as well as a
`variety of Cbl analogues that have been synthesized by different
`labors
`tories (10, 11). Analysis ofCbl analogues in vitro and in vivo have shown
`that some interfere with Cbl metabolism, as evidenced by increased levels
`of homocysteine
`and methylmalonic
`acid,
`the substrates of the two
`mammalian enzymes dependent on Cbl (10, 12). More recenfly,
`it has
`been shown that relatively high doses of the c-lactam of CN-Cbl can
`inhibit
`the in vitro growth of HL6O cells (13), further promoting Cbl
`metabolism as a potential antiproliferative target.
`In vitro cultures in which growth is dependent on Cbl have been
`reported (14, 15). Recently, we have described in vitro growth con
`ditions in which the proliferation of human and munne leukemia cells
`were dependent on Cbl and recombinant human TCII (16). We have
`used this CbI1TCII-dependent
`proliferation
`assay to evaluate
`the
`changes in growth characteristics
`of leukemic cells brought about by
`modifications
`in the chemical structure of Cbl. Here, we show that the
`modification of Cbl generally resulted in reduced ability to support
`cell growth.
`In particular, modifications
`of the propionamide
`side
`chains of the Cbl corrin ring resulted in reduced or complete loss of
`activity.
`In many cases,
`the loss of bioactivity
`of Cbl analogues
`correlated with a capacity to inhibit Cbl-dependent
`cell proliferation
`in a dose-dependent
`manner.
`
`MATERIALS AND METHODS
`
`ABSTRACT
`
`Analogues of cyanocobalamin (CN-Cbl), with functional groups at
`tached to either the various propionamide groups of the comn ring or to
`the ribose-nucleotide linker arm, have been evaluated in a cobalamin
`(Cbl)-dependent
`in vitro cell growth assay.
`In this bioassay, CN-Cbl
`supported,
`in a dose-dependent manner,
`the growth of the murine lym
`phoma BW5147 and the Cbl carrier protein, human apo-transcobalamin
`II, reduced the required concentration of Cbl by 190-1000-fold. Any
`chemical modification
`of Cbl decreased
`its ability
`to support
`cellular
`of
`viability
`and
`proliferation,
`with
`several
`the modifications
`abrogating
`activity completely. All of the Cbl analogues that promoted growth re
`quired the presence of apo-transcobalamln II for the optimal support of
`cell growth. Generally, Cbl analogues modified at the d-posltion of the
`corrln ring and, to a lesser degree, analogues modified at the b- position
`supported cell growth, whereas analogues with modifications at the e-po
`sitlon
`did not
`support
`cell growth. Mixing
`experiments
`demonstrated
`an
`Inverse order of potency of Cbl analogues to inhibit cell growth. Thus, CbI
`analogues with modifications at
`the e-positlon were potent
`inhibitors,
`whereas b-analogues exhibited only partial
`inhibitory activity at high
`molar excess, and d-analogues had no Inhibitory activity at all. These
`results
`indicate
`that modifications
`at
`the c-position
`of Cbl abolish
`the
`ability
`of Cbl
`to support
`cell growth
`and generate
`potent
`Inhibitors
`of
`Cbl-dependent cell growth.
`
`INTRODUCTION
`
`is essential
`vitamin (vitamin B 12) that
`CN-Cbl3 is a water-soluble
`for cell growth. Naturally occurring Cbl analogues
`are required as
`coenzymes by two mammalian enzymes
`that catalyze metabolically
`critical monocarbon transfer
`reactions
`(I). One reaction involves the
`methylation of homocysteine
`in the de novo synthesis of methionine
`and is catalyzed by methionine
`synthase. The other
`reaction rear
`ranges L-methylmalonyl-C0A to succinyl-CoA and is catalyzed by
`L-methylmalonyl-C0A mutase. Cbl-binding
`proteins
`(R-binders
`and
`intrinsic factor) aid in its absorption from food and in its transport
`(2).
`The cellular uptake of Cbl is facilitated by the plasma protein TCII
`(3), which, when complexed to Cbl, binds to specific high affinity
`receptors on the surface of cells (4). The Cbl-TCII complex is inter
`nalized by receptor-mediated
`endocytosis,
`and Cbl
`is thought
`to be
`released from TCH via lysosomal
`action,
`followed by enzymatic
`modification to the forms that are active as coenzymes
`(1).
`In humans, deficiencies
`of
`the vitamin or perturbations of
`intracellular
`metabolism
`can result
`in a variety
`of cell growth-related
`disorders,
`including megaloblastic
`anemia, methylmalonic
`aciduria,
`and central nervous system abnormalities due to the improper func
`
`its
`
`Received 3/26/97; accepted 7/17/97.
`The costs of publication of this article were defrayed in part by the payment of page
`charges. This article must therefore be hereby marked advertisement in accordance with
`18U.S.C.Section1734solelyto indicatethisfact.
`
`from American
`cells were obtained
`lymphoma
`BWSI47 mouse
`Materials.
`Type Culture Collection
`(Rockville, MD). RPM! 1640 culture medium and
`RPM! 1640 culturemediumdeficient in Cbl and folate were obtainedfrom
`Stem Cell Technologies (Vancouver, Canada). FCS was from Life Technolo
`gies, Inc. (Grand Island, NY). QUSO was a gift from Degussa Corp.
`(Ridge
`field, NJ). CN-Cbl, 5-methyl
`tetrahydrofolate, MTF, and D,L-homocysteine
`were obtained from Sigma Chemical Co. (St. Louis, MO).
`Recombinant Human TCII. Recombinant protein (apo form), kindly pro
`vided by E. V. Quadros
`(Veteran Affairs Medical Center and State University
`of New York Health Science Center, Brooklyn, NY), had been produced by
`infection of SF9 cells with baculovirus
`containing
`human TCII cDNA and
`purified as described (17).
`Cell CultUre. BW5147 cells were maintainedin complete RPMI 1640
`medium supplemented with 10% FCS. Cells were grown in 60 X 15 mm
`culture dishes (Fisher Scientific Co., Nepean, Canada) in a humidified atmo
`sphere
`(5% CO2. 95% air) at 37°C. Cells used in the CbIITCII
`bioassay were
`grown to late logarithmic phase, then washed three times in PBS before
`resuspension in Cbl-deficient bioassay medium.
`CbVI'CHBioassay. To measureCbIITCII-dependentcell growth,we used
`RPM! 1640deficient in Cbl and in which the folic acid was replaced with 1 @M
`5-methyltetrahydrofolate
`and
`I p.si homocysteine.
`FCS was pretreated
`with
`QUSO to reduce interference of endogenous bovine TC!I/Cbl in the bioassay
`(16). In brief, 30 mg of QUSO were added per ml of FCS, mixed well, and
`removed by centrifugation as described previously (18). Washed cells were
`4015
`
`1 This
`
`work
`
`was
`
`supported
`
`by
`
`Receptagen
`
`Corp.
`
`and
`
`the Medical
`
`Research
`
`Council
`
`of
`
`Canada.
`2 To whom
`requests
`for
`reprints
`should
`be addressed,
`at The
`Biomedical
`Research
`Centre,2222HealthSciencesMall,Universityof BritishColumbia,Vancouver,British
`Columbia, Canada V6T 173. Phone: (604) 822-7834; Fax: (604) 822-7815; E-mail:
`hermann@brc.ubc.ca.
`transco
`TCII,
`cobalamin;
`Cbl,
`cyanocobalamin;
`used are: CN-CbI,
`3 The
`abbreviations
`balamin II; QUSO, microfine precipitated silica; MT1', 3-[4,5-dimethylthiazol-2-ylj-2,5-
`diphenyltetrazolium bromide;
`Ihipp,
`iodohippurate.
`
`Lilly Ex. 2058
`Sandoz v. Lilly IPR2016-00318
`
`

`
`COBALAMIN ANALOGUES INHIBIT CELL PROLIFERA11ON
`
`plated out in 96-well microtiter plates at 2000 cells per well in 100 @lof the
`above medium, supplemented with 10%QUSO-treated FCS. Following 5 days
`in culture with various
`concentrations
`of CN-Cbl or Cbl analogues
`in the
`presence and absence of 25 ng/ml apo-TCII, cell viability was assessed with
`MU reductionby measuringtheabsorbanceat 550 nm (A550nm)'as described
`previously
`(19, 20).
`Inhibition Bioassay. To test the inhibitory effects of the Cbl analogues on
`Cbl-dependent
`cell growth, BW5147 cells were cultured in the above bioassay
`medium. The tested Cbl analogues were titrated in the presence of a constant
`
`of 25
`or presence
`(0.5, 5, or 50 riM) in the absence
`of CN-Cbl
`concentration
`ng/ml apo-TCII. Following 5 days in culture, cell viability was assessed by
`MTT reduction.
`Cbl Derivatives. The syntheses of Cbl analogues 2—24have been de
`scribed previously (21—23),and their structures are shown in Figs. 1 (mono
`meric analogues) and 2 (dimeric analogues). The Cbl analogues were purified
`to >99% by high-performance liquid chromatography separation (21). Impor
`tantly, no free Cbl could be detected in any of the purified Cbl analogue
`preparations.
`Cbl analogues
`(2 mg) were dissolved
`in 100—200 @lof DMSO
`
`p@a-Iodohippurste-
`
`Biotin
`
`Compd #
`
`Sidechain Structures (R1-R6)
`
`1 R1 = NH2; R2 = NH2; R3 = NH2; 1@.4NH@ R@= H
`
`2 Ri = OH; R2 = NH2;
`
`R3 = NH2;
`
`Li = NH2;
`
`R@ = H
`
`3 Ri = N1-!2; R2 = OH*; R3
`
`NH2; Li = NH2; R@ = H
`
`4 R1 = NH@; R2 = NH2; R3
`
`OH; LI = NH2; R5
`
`H
`
`S Ri =NH2; R2 =N}12; R3 =NH2; R4 =OH; R@=H
`
`6 R1 = NH(CH2)12NH2;
`
`R2 = NH2; R3 = NH2;
`
`Li
`
`NH2; R@ H
`
`7 R1 = NH2;
`
`R2 = NH(CH2)i2N}12;
`
`R3 = NH2;
`
`Li = NH2;
`
`R@ = H
`
`8 R1 = NH@;R2 NH2; R3 = NH(CH@)l2NH2;R4 = NH2; R@= H
`
`9 R1= NH@;R2= NH2; R3 NH2; Li = NH(CH2)12NU2;R5= H
`
`10 R1 = NH(CH2)12NH-Biotin; R2 = NH2; R3 = Nl-12;R@= H; R5 = CN
`
`11 Ri = NH2; R2 = NH(CH2)12NH-Biotin; R3 NH2; Li = NH2; R@= H
`
`12 R1 = NH2; R2 = NH2; R3 = NH(CH2)12NH-Biotin; R@= NH2; R5 = H
`13 Ri = NH2; R2= NH2; R3 NH2; R4 NH(CH2)12NH-Biotin;R@= H
`
`Designation
`
`(CN-Cbl)
`
`(b-COOH)
`
`(*c4actone)
`
`(d-COOH)
`
`(e-COOH)
`
`(b-NH@)
`
`(c-NH@)
`
`(d-NH@)
`
`(e-NH@)
`
`(b-biotin)
`
`(c-biotin)
`
`(d-biotin)
`(e-biotin)
`
`14 Ri = NH(CH@)@@NH-p-Iodohippurate;
`
`R2 = NH2;
`
`R3 = NH2;
`
`Li = NH2;
`
`R@ = H
`
`(b-Ihipp)
`
`
`15 R1 =NH2; R2 =NH@;R3 =N1-l(CH@)@@NH-p-1odohippurate;R@=NH@;R5 =H
`16 R1= NH2; R2= NH2; R3 NH2; R4= NH(CH@)1@NH-p-Iodohippurate;R5= H
`
`17 R1 = NH2; R2 = NH2; R3 = NH2; R4 = NH2; R5 = COCH2CH2COOH
`
`(e-Ihipp)
`
`(5-COOH)
`
`18 Rj = NH2; R2 NH2; R3 = NH2; R4 NH2; R5 = CO(CH2)@CONH(CH2)i2NH2
`
`(5@NH2)
`
`Fig. 1. Monomeric CN-CbI derivatives
`
`evaluated in cell growth assays. Circled letters, corrin ring side chain nomenclature
`
`system.
`
`4016
`
`Lilly Ex. 2058
`Sandoz v. Lilly IPR2016-00318
`
`

`
`COBALAMIN ANALOGUES INHIBIT CELL PROLIFERATION
`
`b-propionamide dimer
`
`@
`
`@J
`
`x
`
`
`
`1:@x:?:@;H
`
`HOOt3
`
`19: XH
`20: X =p-iodobenzoate (Ibz)
`
`d-propionamide dimer
`
`21: XH
`22: X =p-iodobenzoate
`
`e-propionamide dimer
`
`•0
`
`23: XH
`24: X = p-iodobenzoate
`
`‘H
`
`Fig. 2. Dimeric CN-Cbl derivatives
`
`evaluated in cell growth assays.
`
`and then diluted to 1 ml with double-distilled water, and aliquots were stored
`at4°Cin thedarkuntiltested.The Cbl concentrationof thesepreparationswas
`determined by measuring A3@nmusing the experimentally derived m@iextinc
`tion coefficient, as described previously (21). The highest final concentration
`of DMSO in the bioassay medium was approximately 0.2%. This concentra
`tion and lower concentrations of DMSO do not interfere with CbIIFCII
`dependent cell growth.
`
`RESULTS
`
`by CN-Cbl. The in vitro cell growth assay
`Cell Growth Supported
`we previously established (16) has allowed us to monitor the effects of
`4017
`
`Cbl and its plasma transport protein, TCII, on cell growth. Figs. 3 and 4
`show that CN-Cbl supports the growth of the murine leukemic cell line
`BW5147 in a dose-dependent manner and that this growth is enhanced by
`the addition of 25 ng/ml of recombinant human apo-TCH. Maximal cell
`growth was achieved with 200-1000 n?4@CN-Cbl, with detectable effects
`on cell growth at concentrations as low as 1—10ni@iCN-Cbl. When
`optimal amounts of apo-TCII were added, the concentrations of CN-Cbl
`required were approximately 1000-fold lower. The levels of CbIITCII
`that were present were comparable to those found in vivo, indicating that,
`as expected, the natural (11 transport protein TCII aids in delivering 0,1
`to proliferating cells.
`
`Lilly Ex. 2058
`Sandoz v. Lilly IPR2016-00318
`
`

`
`COBALAMINANALOGUESINHIBITCELLPROLIFERATION
`
`0
`Background
`
`A
`
`1.2
`
`1
`
`0.8
`
`0.6
`
`0.4
`
`0.2
`
`EC0L
`
`O
`L()
`
`O0
`
`0.001
`
`0.01
`
`100
`10
`1
`0.1
`Cblanalogue(nM)
`
`1000
`
`internal
`detectable activity (Table 1), indicating that TCH-mediated
`ization occurred. These
`include
`the c-lactone
`(3) and d- and b
`carboxylate conjugated analogues with various substitutions
`(6, 8, 14,
`and 15) and the ribose-conjugated
`analogues
`(17 and 18). Dimeric
`analogues
`of Cbl-b-carboxylates
`and Cbl-d-carboxylates,
`linked
`through an isophthaloyl moiety (19—22),were able to support cell
`growth to a low level. However,
`the Cbl-b-carboxylate
`and Cbl-d
`carboxylate
`analogues
`conjugated
`with
`biotin
`(10 and
`12) were
`unable
`to support cell growth, either in the absence or presence of apo-TCII.
`Analogues with monocarboxylic
`acid substitutions at the d-, b-, and e
`positions of the corrin ring (2, 4, and 5) were also ineffective
`sup
`porters ofcell proliferation. Generally, any change of the Cbl structure
`at the e-position of the corrin ring (5, 9, 13, 16, 23, and 24) resulted
`in a complete
`loss of ability to support cell growth,
`even in the
`presence of TCH, suggesting this position of the molecule may be
`important
`in the intracellular
`function of Cbl. Also,
`ring opening
`conjugations with the c-lactone resulted in the loss of cell growth (7
`and 11).
`The growth response of BW5147 cells in the presence of Cbl
`isophthaloyl-linked
`dimers
`(19, 20, and 23) compared to that with
`CN-Cbl
`is shown in Fig. 4. Again, apo-TCII was absolutely required
`for dimeric Cbl analogue support of cell growth at the concentrations
`tested (Fig. 4B). Interestingly,
`the d- and b-linked Cbl dimers (21, 19)
`
`B
`
`1.2
`
`0.8
`
`0.6
`
`0.4
`
`0.2 -
`
`EC0L
`
`O
`LO
`
`0
`
`0
`Background
`
`A
`
`0.8
`
`0.6
`
`0.4•
`
`0.2
`
`EC0L
`
`C)
`LO
`
`00
`
`@
`
`0-@.-,-,.--@
`
`‘i
`
`—I
`
`I
`
`0.001
`
`0.01
`
`0.1
`
`1
`
`10
`
`100
`
`1000
`
`Cbl analogue (nM)
`
`in the
`cell proliferation
`b-Ihipp and d-Ihipp support
`analogues
`Fig. 3. Cbl-Ihipp
`absence and presence of TCH. BW5147 cells were seeded at 2000 cells/well
`in bioassay
`medium containing the indicated concentrations of CN-Cbl (1; 0), d-Ihipp (15; I),
`b-Lhipp (14; U). and e-Ihipp (16; A) in the absence (A) and presence (B) of 25 ng/ml
`apo-TCII. 0, background cell viability in the absence of CN-Cbl and Cbl analogue. After
`5 days in culture, cell viability was assessed by MT'l' reduction. Data points, means of
`three replicate results; bars, SE.
`
`0
`0.001 0.01
`
`0.1
`
`1
`
`10
`
`100
`
`0
`Background
`
`Cbl analogue (nM)
`
`1
`
`E 0.8
`
`C0L
`
`O
`tO 0.6
`
`0.4
`
`0.2
`
`00
`
`0
`0.001 0.01 0.1
`1
`CbI analogue
`
`10
`100
`(nM)
`
`0
`Background
`
`Cell Growth Supported by Cbl Analogues. Various analogues of
`Cbl were tested in this assay for their ability to replace CN-Cbl,
`in
`both the absence and presence of apo-TCH, and their dose-response
`curves correlated with those of CN-Cbl
`(1). The results are shown in
`summary form in Table 1. None of the analogues were as potent
`in
`supporting growth as was CN-Cbl, either in the absence or presence
`of apo-TCII. Moreover, even at the highest concentration tested and in
`the presence of apo-TCH, none were able to support growth to the
`level achieved by CN-Cbl.
`the
`at
`analogues with substituents
`In the absence of apo-TCH,
`ribose-5'-OH (17 and 18) were the most active in supporting cell
`growth (Table 1). Fig. 3 shows bioassay data for CN-Cbl and
`@l
`analogues with Ihipp substitutions at the d-, b-, and e- positions of the
`corrin ring (15, 14, and 16, respectively). Two of this group of Cbl
`analogues were capable of supporting cell growth in the absence of
`TCII but only to 20% of the level observed with CN-Cbl and only at
`
`the concentration
`tested. Thus, at
`concentration
`the highest
`tested, only four of the Cbl analogues
`retained bioactivity
`absence of TCII.
`
`range
`in the
`
`In contrast,
`
`in the presence of apo-TCII, many analogues exhibited
`
`Fig.4. 6- andd-Cbl-isophthaloyldichloridedimeranaloguessupportcellproliferation
`in the presence ofTCII. BWS147 cells were seeded at 2000 cells/well
`in bioassay medium
`containingthe indicatedconcentrationsof CN-CbI(1;0), b-dimer(19;•),d-dimer(21;
`L ande-dimer(23;A) in the absence(A)andpresence(B)of 25 ng/mlapo-TCII.0,
`backgroundcellviabilityin theabsenceofCN-Cbl(1)andCblanalogue.After5 daysin
`culture,
`cell viability was assessed by MiT reduction. Data points, means of
`three
`replicateresults;bars, SE.
`4018
`
`Lilly Ex. 2058
`Sandoz v. Lilly IPR2016-00318
`
`

`
`@
`
`COBALAMINANALOGUESINHIBITCELLPROLIFERATION
`
`Table I Cell growth with CN-Cbl derivatives in the presence or absence of
`recombinant human apo-TCIf'
`Cobalminderivative
`
`CellGrowth
`
`Compoundno.
`1
`2
`3
`4
`S
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`Designation
`(CN-Cbl)
`b-COOH
`c-Lactone
`d-COOH
`e-COOH
`‘@@2
`c-NH2
`d-NH2
`c-NH2
`b-Biotin
`c-Biotin
`d-Biotin
`e-Biotin
`b-Ihipp
`d-Ihipp
`e-Ihipp
`5'-COOH
`5-NH2
`b-Dimer
`b-Dimer-Ibz@'
`d-Dimer
`d-Dimer-Ibz
`e-Dimer
`e-Dimer-Ibz
`
`—apo-rhTCll
`++++
`-
`—
`-
`-
`-
`—
`—
`—
`—
`—
`—
`—
`+
`+
`—
`+
`++
`—
`
`—
`—
`—
`—
`
`+apo-rhTCH
`++++
`-
`+ +
`-
`-
`+
`—
`++
`—
`—
`—
`—
`—
`++
`
`—
`+++
`
`+
`+
`++
`++
`—
`—
`
`of Cbl analogue
`in presence
`The cell growth
`to CN-Cbl.
`relative
`on cell growth
`a Effect
`is represented
`as fraction of maximal growth:
`—,no cell growth; and + + , 50% cell
`growth,
`relative to CN-Cbl.
`b itz,
`iodobenzoate.
`
`exhibited a biphasic dose response, with reduced growth-promoting
`activity at high concentrations.
`by Cbl Analogues. Three groups of
`Inhibition
`of Cell Growth
`Cbl analogues, each with a specific side chain coupled to the d-, b-, or
`e-propionamide
`group, were tested for
`their ability to inhibit
`the
`growth-promoting
`effects of CN-Cbl. Serial dilutions of each of the
`G@l analogues were tested at three concentrations
`of CN-Cbl
`in the
`presence or absence of apo-TCH. Data showing the inhibitory effec
`tiveness of analogues modified at each of these three positions are
`shown for the Cbl-Ihipp analogues (14—16;Fig. 5), for the Cbl-biotin
`analogues (10, 12, and 13; Fig. 6), and for the Cbl-isophthaloyl-linked
`dimers (19, 21, and 23; Fig. 7). The Cbl analogues with either type of
`substitution at the e-position (13, 16, and 23) proved to be effective
`inhibitors ofCbliTCII-dependent
`cell growth,
`inhibiting cell growth in
`the presence of 5 nr@tCbl and apo-TCII, with an IC50 of 5—15n@i.
`Increased concentrations of the Cbl analogues were required to inhibit
`cell growth supported with higher concentrations
`of CN-Cbl.
`Inter
`estingly,
`the e-analogues were only effective
`as inhibitors of Cbl
`dependent cell growth in the presence of TCH. This requirement for
`TCII
`is shown in data with the e-Ihipp analogue 16, demonstrating
`that, at all of the concentrations tested, this analogue had no inhibitory
`activity in the absence of apo-TCII
`(Fig. 5D). Cbl b-analogues
`(10, 14,
`and 19) exhibited inhibitory effects on cells maintained in low con
`centrations of CN-Cbl
`(0.5 aM) and TCII, but this was reversible with
`the addition of higher CN-Cbl concentrations. The d-analogues of Cbl
`were ineffective as inhibitors; only a single analogue (21) partially
`inhibited Cbl-dependent
`cell growth and only at low Cbl concentra
`tions.
`
`DISCUSSION
`
`Here, we have evaluated the changes in Cbl bioactivity following
`conjugation of chemical groups to different sites on the Cbl molecule.
`Five sites on the Cbl molecule,
`the b-, c-, d-, and e- amide side groups
`of the corrmnring, and the ribose-5'-OH group were modified by
`different substitutions
`(Fig. 1). These Cbl analogues were designed in
`4019
`
`a manner such that each site of modification allowed for conjugation
`of moieties with different properties
`(21—23).These included mono
`carboxylic acid functionalities
`(2, 4, 5), a lactone (3), ribose succinate
`adducts (17 and 18), carboxylate
`adducts of diaminododecane
`(6—9),
`biotin conjugates
`(10—13), and p-Ihipp
`groups
`(14—16). Dimeric
`analogues of 0,1,
`linked through the b- (19 and 20), d- (21 and 22),
`and e- (23, 24) carboxylic acids of the corrin ring, were also evaluated.
`All Cbl analogues
`that supported cell growth were more potent
`in
`the presence of apo-TCII,
`indicating that
`they retain some ability to
`bind to TCII and are transported into cells via the TCII receptor.
`In
`earlier work (21—23),we have shown that modifications
`at either the
`Co metal, the 5'-OH of the ribose, or the e-propionamide side group
`of the corrin ring had little effect on binding to TCII, whereas Cbl
`analogues with substitutions
`at the c- and d-side chains of the corrin
`B-pyrroline
`ring bound only poorly with TCII, and b-analogues dis
`played intermediate TCII binding. The fact
`that
`the presence
`of
`apo-TCH nevertheless increased the cellular growth supported by the
`b- and d-analogues
`suggests that, although they do not bind TCII as
`well as e-analogues,
`they do bind sufficiently well and are internalized
`via the receptor-mediated
`process. We did not detect free Cbl
`in any
`of the Cbl analogue preparations,
`excluding the possibility that small
`amounts of Cbl contributed to this support of cell growth.
`The fact that the e-analogues, which bind well to TCII, exhibited no
`growth-promoting
`activity, either in the absence or presence of TCII
`(Table 1, Fig. 3, and Fig. 4) may indicate that
`the structure at the
`e-propionamide
`group is crucial
`for the ability of Cbl to support cell
`growth.
`In contrast,
`substitutions
`at the d- and b-positions,
`although
`negatively affecting interaction with TCII, appear
`to interfere
`less
`severely with Cbl function. Furthermore, our data indicate that alter
`ations of any type at the d-position have less effect on Cbl function
`than do alterations
`at the b-position of the corrin ring. Derivatives
`linked via the 5'-OH of the ribose (17 and 18) behaved comparably to
`CN-Cbl. However,
`this may reflect
`the fact
`that
`the ester
`linker
`is
`susceptible
`to hydrolysis,
`resulting in the release of Cbl
`into the
`medium or inside of the cell.
`The inhibitory activity of Cbl analogues correlated positively with
`their ability to bind TCH but negatively with their ability to promote
`growth. Thus, e-analogues
`such as 13, 16, and 23 were potent
`inhib
`itors of Cbl-dependent
`cell growth, whereas d- and b-analogues were
`relatively ineffective inhibitors. The fact that e-analogues
`require TCII
`to show inhibitory effects indicates that they may act as competitive
`inhibitors ofTCII-mediated Cbl uptake. These data are consistent with
`previous reports that
`in vivo application of Cbl analogues with mod
`ifications at the e-position resulted in the inhibition of methylmalonyl
`CoA mutase and methionine synthase activities, whereas
`in vivo
`treatment with d- and b-analogues did not affect
`the activity of these
`enzymes
`(10). Also concurring are data showing that
`in vitro treat
`ment of human fibroblasts with an e-analogue
`of Cbl
`increased
`homocysteine
`levels 5-fold and methylmalonic
`acid levels up to
`200-fold in the culture medium (12). These increases
`in metabolites
`occurred after 3—4weeks in culture and in the presence of high (10
`@tg/ml)concentrations
`of the Cbl analogue. The c-lactam of Cbl has
`also been shown to antagonize Cbl in vivo (10) and, more recently,
`in
`vitro (13). However,
`the in vitro effects were only seen in the absence
`of methionine
`and at analogue concentrations
`approximately
`1000-
`fold greater
`than we show for the e-analogues
`in the present study.
`Taken together,
`it appears that the inhibitory Cbl analogues may act,
`both in vitro and in vivo, by depleting intracellular Cbl. In the case of
`e-analogues, which retain high affinity for TCII but do not support cell
`proliferation,
`it
`is likely that
`they act as simple competitors with
`CN-Cbl
`for binding to TCII and thus reduce the intracellular
`level of
`functional Cbl. It has been shown previously that many analogues of
`Cbl (including b-, C-, d-, and e-analogues) bind to and activate human
`
`Lilly Ex. 2058
`Sandoz v. Lilly IPR2016-00318
`
`

`
`4'
`
`0
`
`Background
`
`A
`
`0.7-
`
`A
`
`COBALAMIN ANALOGUES INHIBIT CELL PROLIFERATION
`
`C
`
`EC L
`
`C)
`
`00
`
`EC0t
`
`o
`to
`
`00
`
`,,,,,
`
`,,,,,,,
`
`,
`
`,,,,@
`
`0.001
`
`0.01
`
`0 1
`.
`
`1
`
`10
`
`100
`
`‘‘ Background
`
`0
`
`
`
`
`
`@@@@
`
`analogue #15 (nM)
`
`0.001 0.01
`
`0.1
`
`1.2-
`
`B
`
`,,.,,,i,,.,.,.
`
`‘I
`
`‘I
`
`I
`
`10
`1
`analogue #16
`
`100
`
`
`
` :@@
`
`A
`
`E
`
`:1
`
`
`0.001
`
`0.01
`
`0.1
`
`1
`
`10
`
`100
`
`@-__‘__‘_‘_“_9,,,,,‘9,,,,,,,,,,,,,,,1'''â€(cid:157)“I
`
`S.
`
`0
`Background
`
`
`
`
`
`Background
`
`
`
`: ,@,,,,,,,,,,,,,,,,@,,,i,,, !
`
`EC0t
`
`oto
`
`a0
`
`
`@@
`@
`
`@
`
`
`
`0.001
`
`0.01
`
`‘1 ‘“
`
`‘‘
`
`0.1
`1
`10
`100
`analogue #1 4 (nM)
`
`analogue #1 6 (nM)
`
`@
`
`in bioassay medium
`in the presence of TCII. BW5147 cells were seeded at 2000 cells/well
`inhibits Cbl supported cell proliferation
`(e-lhipp)
`Fig. 5. The Cbl-Ihipp c-analogue
`(D). CN-Cbl at 0.5
`containing 25 ng/ml apo-TCII and the indicated concentrations
`ofd-Ihipp
`(15; A), b-Ihipp (14; B), and e-Ihipp (16; C) and e-Ihipp (16) in the absence ofapo-TCII
`(1; 0), 5 (1; 0), or 50 nM(1;
`was also added to the cultures. Cell viability in the absence ofCbl analogue is shown with 0.5 (1; •),5 (1;
`), and 50 nst CN-Cbl (1; A) and without
`CN-Cbl(1;0). After5 daysin culture,cell viabilitywasassessedby MiT reduction.Datapoints,meansof threereplicateresults;bars. SE.
`
`then be hypothesized that
`synthase in vitro (24). It must
`methionine
`the e-analogues of Cbl remain inside the cell but cannot be converted
`to active coenzyme forms. There are a number of intracellular steps
`that are required to generate functional Cbl coenzymes (8). Cbl must
`first be released from the internalized TCH receptor complex within
`the lysosome,
`transported to the cytoplasm, where it is reduced from
`the +3 oxidation state to the +2 state, and then further modified to the
`coenzyme
`forms. Cbl must
`then bind to and activate methionine
`synthase
`in the cytoplasm or methylmalonyl CoA mutase
`in the
`mitochondria
`to act in true coenzyme fashion.
`Intracellular metabo
`lism of 0,1 could be altered at any of the above steps by e-analogues.
`Importantly, we have shown that
`the inhibitory
`activity of the
`e-analogues may be abrogated by adding higher concentrations
`of
`CN-Cbl
`in the bioassay. This shows
`that the observed inhibitory
`effects of the e-analogues
`are due to specifically targeting Cbl trans.
`port and not nonspecific
`toxic effects of the substituted group. This
`finding is further strengthened by the fact that the inhibitory activity
`of Cbl analogues correlates with the position of modification,
`i.e., all
`e-analogues are inhibitory, whereas the nature of the side chain added
`does not correlate with inhibition [e.g., Ihipp analogues b- (14) and d
`(15) do not inhibit, whereas the e-analogue (16) is a potent inhibitor].
`Furthermore,
`the e-analogues
`required TCII
`to be present
`in our
`bioassay to show inhibitory activity,
`indicating that cell uptake of
`these inhibitors
`is mediated by TCII. This suggests
`that
`these ana
`
`logues may be suited particularly well for in vivo application because
`their cellular uptake will be enhanced by the presence of endogenous
`apo-TCII
`in the plasma.
`and inhibitory
`for growth-promoting
`Analysis of the 0,1 dimers
`activity showed similar dependence on structure as for the Cbl mo
`nomeric analogues and required TCII for activity. An interesting
`feature of Cbl dimers 19 and 21 is that, at high concentrations
`(250
`aM),
`they
`exhibited
`a reduced
`capacity
`to support
`cell
`growth
`(Fig.
`4).
`These Cbl analogues thus displayed biphasic effects whereby, at lower
`concentrations
`(<100 riM), they exhibited growth-promoting
`activity,
`and at concentrations above 100 n@i,this activity was lost.
`In conclusion, we have demonstrated in an in vitro bioassay system
`that alterations of the e-propionamide side chain, which did not affect
`its ability to bind TCH, resulted in a complete loss of its ability to
`promote growth,
`indicating that this site is critical for proper mete
`bolic function of Cbl. Modification
`of Cbl at other sites resulted in
`reduced capacity for support of cell growth. Furthermore, we have
`shown that
`the e-analogues
`of Cbl can act as inhibitors
`of Cbl
`supported cell growth, supporting previous reports that e-analogues of
`0,1 result
`in the reduction of methylmalonyl CoA mutase and methi
`onine synthase activities and the increase in methylmalomc
`acid and
`homocysteine
`levels (10, 12). This suggests that Cbl metabolism may
`be a useful target for antiproliferative drugs. In support of this hy
`pothesis, we have demonstrated earlier that blocking cellular uptake of
`
`4020
`
`Lilly Ex. 2058
`Sandoz v. Lilly IPR2016-00318
`
`

`
`COBALAMIN ANALOGUES INHIBIT CELL PROLIFERATION
`
`to be a useful strategy in cancer therapy, and the potential problem of
`Cbl antagonism inducing neurotoxicity and other side effects should
`be rapidly reversed through the administration of pharmacological
`doses of 0@l. It has been shown that
`inactivation of Cbl by nitrous
`oxide had a therapeutic effect on patients with leukemia (28, 29).
`Furthermore, nitrous oxide treatment of human and munne leukemic
`cells in vitro has been shown to induce apoptosis (30). Cbl analogues,
`such as the e-analogues described here, which inhibit growth through
`interference of Cbl processing rather
`than simply acting as antago
`nists, should prove to be effective
`antiproliferative
`agents
`in vivo.
`
`0
`Background
`
`0.6.
`
`A
`
`0.5 -
`
`10
`1
`0.1
`01 0.01
`analogue #21 (nM)
`
`100
`
`0
`Background
`
`0.6- B
`
`0.5 -
`
`0.2-
`
`0.1 -
`0--s.
`0.001
`10
`1
`0.1
`0.01
`analogue #19 (nM)
`
`-@‘ @-‘@“‘@.@.@,,, “I
`
`0
`Background
`
`100
`
`0.5.
`
`C
`
`0.4 -
`
`0.3 -
`
`0.2 -
`
`0.1 -
`
`EC0t
`
`O
`tO
`
`a0
`
`0.2
`
`A
`
`0.15@
`
`0.1
`
`0.05.
`
`,,,,.. ,,I,,,,,, “1,''â€(cid:157)
`0@ .,,,.,. .@, ...,...,
`1
`10
`0.001
`0.01
`0.1
`analogue #12 (nM)
`
`100
`
`0.8
`
`B
`
`0.6
`
`0.4
`
`0.2
`
`EC0t
`
`O
`tO
`
`a0
`
`EC0t
`
`O
`tO
`
`a0
`
`0
`
`.,,.,, -I ‘‘“-I ‘“@‘‘
`,.,._.,
`.,,
`..,.,
`0.001 0.01
`0.1
`1
`10
`100
`analogue #10 (nM)
`
`Background
`
`0.7. C
`0.6.
`0.5.
`0@g 0.4-
`a
`0.3.
`0
`0.2-
`
`0.1.
`
`0
`0.001
`
`£
`
`I
`
`0
`Background
`
`10
`1
`0.1
`0.01
`analogue #13 (nM)
`
`100
`
`inhibits Cbl supported cell proliferation in
`Fig. 6. The Cbl-biotin c-analogue (e-biotin)
`the presenceof TCU.BW5147cellswereseededat 2000cells/wellin bioassaymedium
`containing the indicated concentrations
`of d-biotin (12; A), b-biotin (10; B), and e-biotin
`(13; C) in the presence of 0.5 (1; 0), 5 (1; 0). or 50 nMCN-Cbl (1; Li). Cell viability in
`the absence ofCbl analogue is shown with 0.5 (1; O), 5 (1; 5), and 50 nssCN-Cbl (1; A)
`and withoutCN-Cbl(1; 0). All wellscontained25 ng/mlapo-TCII.After5 days in
`culture, cell viability was assessed by MU reduction. Data points, means of three
`replicate results; bars, SE.
`
`@
`
`@
`
`@
`
`@
`@
`
`Cbl with anti-TCH antibodies is antiproliferative for leukemic cells
`(16). In addition, it has been recently shown that, in vivo, antibodies
`against the cellular TCII receptor resulted in the failure of rabbits to
`thrive through development
`of Cbl deficiency (25). Furthermore,
`a
`mixture of OH-Cbl and ascorbic acid, when injected into mice bearing
`inhibits Cbl supported cell
`dichloride dimer c-analogue
`Fig. 7. The Cbl-isophthaloyl
`proliferation in the presence of TCII. BW5147 cells were seeded at 2000 cells/well in
`ascites
`tumors,
`inhibited the growth of
`the tumor cells
`(26) and
`bioassay medium containing the indicated concentrations
`of d-dimer
`(21; A), b-dimer
`(19;
`increased the survival
`time of the mice (27). These Cbl analogues
`B), and e-dimer (23; C) in the presence of0.5 (1; 0), 5 (1; [J), or 50 nsi CN-Cbl
`(1, is).
`Cell viability in the absence of Cbl analogue is shown with 0.5 (1; •),5 (1; U), and 50
`were produced by the reaction of OH-Cbl and ascorbic acid (11), and
`nM CN-Cbl
`(1; A) and without CN-Cbl
`(1; 0). All wells contained 25 ng/ml apo-TCII.
`their exact structure is unclear.
`After5 daysinculture,cellviabilitywasassessedbyMU reduction.Datapoints,means
`Disruption of Cbl uptake and metabolism by inhibitors may prove
`of three replicate results; bars, SE.
`4021
`
`,,,,,,.@,,,,,,.@
`0-
`1
`10
`0.001
`0.01
`0.1
`analogue #23 (nM)
`
`“1
`
`‘‘‘““I ‘‘‘“I'
`
`0
`Background
`
`100
`
`Lilly Ex. 2058
`Sandoz v. Lilly IPR2016-00318
`
`

`
`COBALAMIN ANALOGUES INHIBIT CELL PROLIFERATION
`
`to be
`an anilide of Cbl has been shown previously
`Interestingly,
`successful
`in treating acute myelogenous
`leukemia in one patient (31).
`Further evidence,
`albeit anecdotal,
`suggesting that Cbl antagonism
`may be beneficial
`in leukemia treatment addresses an “experimentof
`natureâ€(cid:157)in a patient with both pernicious anemia (Cbl deficiency) and
`chronic myeloid leukemia, whose condition worsened with 0@l treat
`ment and improved when