Sovember 1969 COENZYME
111 1019
to lactate and in this process ail a-orieiitcd H atom is
transferred.
k R
I
R
I I1 I11
Sormally t.he S A D H produced in step 1 can be uti-
lized in step 2 . If) however, the H atom attached to
C-4 of the pyridine ring of S A D is replaced by some
&her group R, st'ep 1 would produce a dihydropyridine
derivat,ive (111)which cannot be used in step 2 since it
cont,airis no a-hydrogen at,om. The presence of a 4-
substit,uted NAD could therefore affect the glycolytic
process which involves recycling of KAD. Sumerous
subst,it,utedriicot'inanlides are known to be incorporated
int,o XAD in vivo and it should be possible to achieve
the desired effect by administering a subst,ituted nico-
tinamide to the t)umor-bearinghost..
This approach will be effect.ive if (1) the substitut'ed
nicotinamide (N*) is incorporat'ed int.0 the NAD mole-
cule to give the analog (N*AD) ; ( 2 ) t.he K*AD is a co-
enzyme for the glyceraldehyde t'o glyceric acid oxida-
tion, ot,herwise respiration in normal cells may be hin-
dered; ( 3 ) H addition to the K*AD is 1,4 and also if the
st,ereospecificity of N*AD-mediat'ed H-transfer reac-
tions is the same as for those of NL4D;(4) the rate of H
transfer is comparable with that of the NAD-NADH
system; and (5) N*AD competes favorably wit'h YAD
for association with the apoenzyme.
These aspects are being systematically exarnined'~~
and it has been shown that reactionof dithionite mit'h
4-methyl-substitut,ed nicotinamide derivatives leads to
1,4 addition' just as in the reduction of X24Dto KADH
by the same reagent,5 which mimics the enzymic pro-
cess. Most of t'hese points can only be tested when 4-
substituted S A D derivatives are available ; the syn-
thesis of 4-11e-NAD has been achieved on a small
scale.6 I n the meantmimeit is possible to study some of
t'he reactions involved using model substances. The
quaternary compounds used in this study have been the
X-benzyl, -propoxymethyl, and -tetraacetyl-,B-D-gluco-
pyranosyl salts derived from nicotinamide and 4-
methylnicotinamide, t'here being indications in the
literature' that the rates of react'ion of the propoxy-
methyl and tetraacetyl-P-D-glucopyranosyl salts more
closely approach those of t'he natural coenzyme (XAD).
Point 4 (above) is also important because the glycero-
phosphate "shuttle" mechanism which is available in
normal cells and involves a p-specific H t'ransfer from
N.4DH8 is lacking in cancer which have a lower
S.4D content. T o establish whet.her H transfer from
4-hle-substituted dihydronicotiriamide derivatives
takes place at a rate comparable with that of dihydro-
(3) S . 0. Kaplan, "The Enzymes." Vol. 3, Academic Press, New York,
N. Y.. 1960, Chapter 12.
(4) W.C. J. Ross, J . Chem. Soc.. C . 1816 (1966).
( 5 ) >I. B. Tarmolinsky and S. P. Colowick, Biochim. Biophys. Acta,
ao, 117 (1956).
(6) F . Searle (Institute of Cancer Research), private communication,
1969.
(7) K. Wallenfels. "Steric Course of Microbiological Reactions," J. 6-
.I, Churchill Ltd.. London, 1959, p 10.
( 8 ) H. R . Levy a n d B. Tennesland, J . B i d . Chem., 288, 85 (195i).
( Y ) G. E. Boxer and T. AI. Devlin, Science, 184, 1495 (1961).
IEHIBITOHY.
nicotinamides, the rates of H transfer from some sub-
stituted dihydronicotinamide derivatives to 2,B-di-
chlorophenolindophenol (IV) have now been studied.
The results are discussed in the latter section of this
work.
0 0
Iv
V VI
Cyanide Addition Reactions.-The affinity of nicotin-
amide salts for anions (CN- in particular) has been
used to obtain an indication of the chemical reactivity
of the pyridine ring.' Many 3-substituted pyridinium
salts react with C S - to form adducts which are in equi-
librium with the reactants. 1,4-Dihydropyridine struc-
tures (V, R1 = H) have been proposed by San Pietro'O
on the basis of D-exchange reactions. The 4-H only
undergoes exchange with the medium, and it was rea-
soned that this H is acidic because the cyanide group is
in the (Y position to the H atom. Furthermore, the ab-
sorption spectra of the cyanide adducts from various
pyridinium salts were similar to those of 1,4-dihydro-
However, as Kosower has pointed
the D exchange may not uniquely indicate a 4 location
for the cyanide group in the adduct since there is no proof
that it is the adduct itself which undergoes the ex-
change, and no evidence bearing directly on the point of
attachment of the cyanide group to the pyridine nu-
cleus has so far been available.
Walter and Kaplan14 have found that the CN- reac-
tion with 4-Me-SAD was very slow (250 times smaller
than for SAD) and they have suggested that the 4-
RIe substituent may interfere sterically with the addi-
tion to the 4 position, and that the reaction may not
proceed in the usual way, but that addition may occur
to other positions on the ring. Little C S - addition
takes place with 3-carbamoyl-l-(2,B-dichlorobenzyl)-4,-
6-dimethylpyridinium bromide (VI). This result has
also been attributed to steric hindrance by the 4-
Rle'2*'jz16 and it is therefore important to ascertain the
position of the cyanide group in the products from the
reactions between CY- and quaternary nicotinamide
salts, and also the cyanide position in the 4-Rle-sub-
stituted adducts. The spectroscopic data for the CN-
addition products derived from the nicotinamide and
4-methylnicotinamide salts studied in the present work
are given in Table I. The uv spectra presented in
Table I shorn that 2-cyano-1,2-dihydronicotinamide
structures are inconsistent with the recorded physical
(10) A. S a n Pietro, J . B i d . Chem., 217, 579 (1959).
(11) M. Marti, &I. Viscontini, and P. Karrer, Helo. Chim. Acta, 89, 1451
(1956).
(12) K. Wallenfels and H. Schuly, Ann., 621, 215 (1959).
(13) E. hl. Kosower, ref 3, Chapter 13.
(14) P. Walter a n d S . 0. Kaplan, J . Bid. Chem., 288, 2823 (1963).
(15) E. 11. Kosover, "hIolecular Blochemistry," JIcGraiv-Hill Book Go.,
Inc., Ken. York, N. I-,, 1962, Section 2.13.
(16) J. Biellmann and H. J. Callot, Bull. S o t . Chzm. Frunce, 1159 (1968).
Sovcmber 1909 ISHIBITORS.
COE~VZYME I11 1021

TABLEI1 dronicotinamide derivatives studied in the present work

EQUILIBRIUM CONSTANTS AT 25.0" are presented in Table 111.
xrnax. Rate constants for the benzyl compound (VII, R =
IbU R2 s mp ks mole benzyl, R1 = H) and the glucopyranosyl compound (VII,
1'm H CN 328 513 zt 51 R = tetraacetyl-P-D-glucopyranosyl, R1 = H) a t pH 7
Pm 31e CS 331 0.36 f 0.04 a t 25' have also been determined by Wallenfels and
Tg H CN 319 4030 f 500
Gellrich,13 and are approximately of the same order of
Tg JIe CS 323 1.98 + 0 . 2 3
BL H CX 344 6.46 i 0.29
magnitude as the corresponding values in Table 111.
Bz 3Ie cs 344 0 0056 + 0 . 0 0 0 3 The reaction mechanism of the oxidation process has
Pm, propoxgmethyl; Tg, tetraacetyl-P-n-glucopyranosgl; been discussed in detail by these authors.23 As the
BL, benzyl. ease of H- transfer is related to the electron density in
the heterocyclic ring, the rate constants should be mark-
the corresponding compounds lacking 4-Xe. These edly affected by the groups attached to S-1. I n agree-
results indicate that 4-3le-SAD derivatives which ment x i t h this, the values in Table I11 increase in the
might be formed in vivo from 4-methylnicotinamide same order as for the electron-donating properties of the
compounds may provide only weak competition with 1 substituents, i.e., benzyl > propoxymethyl- > tetra-
the natural coenzyme in the oxidative stage of glycol- acetyl-/3-D-glucopyranosyl.
ysis, and the resulting concentration of 1,4-dihydro-4- I n compounds with the same N substituent, thr
methylnicotiriamide derivatives produced could be highest rate constants are displayed by the l-alkyl-l,6-
much lower than the S A D H concentration in the tumor dihydro-4-methylnicotinamidederivatives (PIII, R1 =
cells. However, by incorporating substituents, which Ale). The high rates of H transfer are predictable be-
could form covalent bonds with groups adjacent to the cause these structures resemble allylic systems where
active site of the enzyme, into 4-Me-substituted nico- the methyl group facilitiates double-bond rearrange-
tinamides, the derived S*ADs should be preferentially ment (e.(/., as in the acid-catalyzed interconversion of
bound to the enzyme (compare the work of Baker21on cis- and trans-crotyl alcohol via but-3-en-2-01) . 2 4 The
the design of irreversible antagonists). This approach significant effect of the 4-Me substituent in accelerating
to increasing the involvement of substituted S*ADs and the 6-H release can be seen from a comparison with the
+

the use of ring substituents which confer greater reac- rate constants for the 1,B-dihydronicotinamide deriva-
tivity toward 1,4 addition i5 now being investigated. tives (VIII, R, = H) without the 4-1\Ie, where the oxi-
dation rate is much lower.
There is little difference between the rate constant
values for the 1,4-dihydronicotinamide derivatives
when Rl is Me or H. I n these cases, H- elimination
may involve electron displacement from either of the
two double bonds conjugated to the S-1 lone pair.
R R Such electromeric effects should be more important than
VI1 VI11 the weaker electron-repelling effect of the 4-3Ie26 in
determining the rate constant value. The results show
H-Transfer Reactions of Dihydronicotinamide Deriv- that the 4-Me does not exert an effective inhibiting in-
atives.-The dihydronicotinamide derivatives investi- fluence upon oxidation as the H-transfer rates are little
gated were the 1,4- and 1,B-dihydro compounds (VII, affected by the 4-Me substituent. It is therefore pos-
R1 = Ale or H) and (VIII, R1 = Me or H) where the iible that the rates of H transfer from NADH and its
S substituents mere benzyl, propoxymethyl, and 4-Me analog could be of the same order. Such a result
tetraacetyl-/3-i)-glucopyr~Iio~yl.I n view of the poten- is desirable for the chemotherapeutic approach dis-
tial biological involvement of the enzymatically reactive cusqed earlier, since otherwise respiration in normal
1,6-SADH,22the 1,B-dihydro compounds were studied. cells may be hindered after administration of a 4-
I n order t o investigate the effect of an electron-with- methylnicotinamide derivative.
drawing group at the 4 position, upon the rate constant, Some a ~ t h o r s have ~ ~ ? ~investigated
~ 2,B-dichloro-
the 4-cyano-1,4-dihydronicotinamidederivatives (V, phenolindophenol oxidations over various ranges of H +
R1 = H or Me, R = propoxymethyl) were also studied. concentrations and the influence of the pH of the
The spectroscopic method used mas that described by medium upon the value of the rate constant has been
Wallenfels and G e l l r i ~ h .observing
?~ the change of oxi- discussed by Wallenfels and G e l l r i ~ h . ~ The
~ reactions
dizing agent concentration with time. 2,B-Dichloro- are pH dependent and considerable variations in rate
phenolindophenol (redox potential +0.217 Y) was may be achieved by slight alterations in the H + concen-
chosen a$ the oxidizing agent because the rate of oxida- tration of the reaction medium. The rate constants
tion for tlihytlroriicoti ri:tmitlc- is gcncr:illy \low cnoiigh for thc H-trunhfer re:wtions at various pH values wew
to provide accurate absoiptiuu ineasuienients, a i d the tlieiefoie recorded in the present work. When R1 was
relatively fast side reactions which are present in other J I e or H , the reaction rate increased as the pH de-
reagents are absent in 2,6-dichlorophenolindophenol.7~23creased, indicating that the oxidation process was facil-
Rate constants for the H-transfer reactions of the dihy- itated in more acidic media. With an electron-with-
(24) W. G. Young a n d J. S. Franklin, J . Amer. Chern. Soc.. 88, 785
(21) B. R. Baker, "Design of Active-Site-Directed Irreversible Enayrno (1966).
I i d ~ i t i i t w ~ ,.John
" Wilez. and Sons,I n c . , N e a York, K. Y . . Cliauter 8. ( 2 5 ) C. IC. Ingold, "Structure a n d Mechanism in Organic Chemistry,"
122) K. Chakraverts and 9. Chaykin, Biochim. B i o p h y r . Res. Commun., Cornell University Press, N e n York, K.Y.,1953, p 70.
15, 262 (1964). (26) S. J. Leach, J. H. Baxendale, and 11. G. Evans. 4 u d . J . Chem., 6,
( 2 3 ) K. TVallenfels and 11.Gellrich. .4nn.,621, 149 (1959). 409 (1953).
Sovrmher 1969 COENZYME
t!ie corresponding quaternary salts. I t was found that
the oxidation rate constants decreased in the same order
of substituents in which the affinity constants increased.
If the oxidation rate constants ( K ) for the 1,4-dihydro-
nicotinamide derivatives at pH 7 (Table 111) are com-
pared with the C S - addition constants ( k ) for the cor-
responding quaternary salts given in Table 11,it is seen
that the 4-Me derivatives, and the three compounds
lacking 4-3Ie, form two separate series, where the rate
conqtant K decrease.; in the substituent order R =
benzyl. propoxyniet hyl, tetraacetyl-6-u-glucopyranoqyl,
while the C S - :iffinity con;.tailit k increases. These re-
w l t y :ire therefore hirnilnr to thoqe of W.~;nllenfels.~
Experimental Section
The H-transfer reaction is represented by the equation, P y H
Ind + H + +. Py' + IndHz. P y H represents the dihydro-
nicotinamide derivative and Ind represents the 2,6-dichloro-
phenolindophenol. When the dihydronicotiriamide compound is
i n excess, the reaction becomes kinetically of the first order, and
the rate eqiiation is t,hen2'
[Ind10
K = - 1 In -~
t[I'yH]o [Ind],
where [IndIo is the initial concentration of 2,6-dichlorophenol-
indophenol, [Ind] is t.he roncentration after t#imet, and [PyHIo
i y the initial concent ration of the dihgdloriic.otiriamide derivative.
Hy rrarrangement
, - [PyH'oKt
log [ I ~ l d ]=
2.303
+ log [Indlo
The corresponding equation given by Wallenfels and Gellrichzs
is in error, as the firht, term on the right-hand side has no negat,ive
sign attached. The reaction was followed by observing the de-
vrease in 2,6-dichlorophenolindophenol visible absorption a t 640
m p with time, and the rate constant, K was calculated from the
dope of the graph of log [IndI1 against, t,ime. K was also ob-
tained by recording the time to.5 at, which [ I n d ] was
~ reduced by
one-half, and substituting [IndIo,:2for [IndIt in eq 1 to give eq
:3. The spectra were recorded at constant H + concentrations by
the u\e of phozphate buffer solutionz of known pH.28 Absorption
spectra were recorded on a Cnicam SP 800A spectrophotometer
linked to an S P 21 .lave recorder. The temperature was main-
tained at 25.0" by a Shandon K 2 Ultra-Thermostat. Dihydro-
nicotinamide derivatives were prepared and used on the same day
for the oxidation experiment.. Reactant concentrations were
0.37 X mol of 2,6-dichlorophenolindophenoland 2.5 X
t o 4.0 X 10-3 mol of dihydronicotinamide derivative in 0.007 mol
(27) S. Glasstone, "Texthook of Physical Chemistry," Macmillan a n d
Co., L t d . , London, 1964. p 1057.
(28) G. Kortum and J. Bockris, "Texthook of Electrochemistry," El-
sevier Puhlishing Co., Amsterdam, 1951, p 744.
ISHIRITORS. I11 1028
of 1: 1 aqueous-methanolic phosphate buffer solution.28 Melt.ing
points were determined in open capillary tubes and are corrected.
Compounds whose elemental analyses are indicat,ed only by
symbols showed values within 0.4% of the theoretical values.
The nmr spectra were recorded with a Perkin-Elmer R 10 spec-
trometer (60 lfcps), TMS = 0. Cyanide equilibrium constants
were obtained by the met,hod of Wallenfels,iz*Oquat,ernary salt
concent.rations being J 4 in H20, CN- concentrations being
6X t o 2 X lo-' Ai' in H2O. Evaporat.ions were carried out
under reduced pressure.
-
1 (Tetraacetyl-~-~-glucopyranosy~)-3-carbamoy~-4-cyano-l,4-
dihydropyridine.-A solution of 3 g (0.0058 mole) of 1-(tetra-
acetyl-p-~-gliicopyranosyl)-3-carbamoylpyridinium bromide29 in
20 ml of H i 0 x-as added to a solution of 10 g of K C S in 20 ml of
HgO at, 0". Anhydrocis Na&04 (100 g ) was added and the mix-
tlire was stirred. The slurry was extracted with hIeCS, the ex-
t,racts were filtered, and the filtrate was evaporated. The residue
was recrystallized from Et20-31eOH to give 1.8 g (90%) of yel-
low cubes, mp 39-41'. Anal. (C2iH?&8010) C, H. X.
+ I n a similar way, l-(tetraacetyl-p-n-glucopyranosyl)-3-car-
bamoyl-4-cyano-l,4-dihydro-4-methylpyridine was prepared from
1 g (0.0019 mole) of l-(tetraacet~yl-~-n-glucopyranonyl)-3-car-
bamoyl-4-methylpyridinium bromide' and gave 0.65 g (94%) of
creamy needles, mp 72-74". Anal. (C22H&3010) C, H, N.
1-Benzyl-3-carbamoyl-4-cyano-l,4-dihydropyridine.-Asolu-
tion of 2 g (0.0069 mole) of 1-benzyl-3-carbamoylpyridiniiim
bromide30 in 100 ml of H,O m-as added to a solution of 12 g of
K C S in 100 ml of H I O at 0". The mixture was shaken for 10
min and the precipitate was collect,ed and washed with ice-H20
followed by Et,O. Rec tallizat,ion from EtOH yielded 1.4 g
(887)of colorless needles, mp 141-142'. $nul. (C14H18S~O)
C, H, X',
3-Carbamoyl-4-cyano-l,4-dihydro-l-propoxymethylpyridine.
-A solution containing 0.4 g (0.0017 mole) of 3-carbamoyl-l-
propoxymethylpyridinium chloride' in 100 ml of H20 at 0' was
(2) added to a solution of 10 g of KCN in 100 ml of H20 at 0". The
mixture was extracted u-ith CHCb and the organic ext,ract,i were
washed with H20, dried (NagSO,), and evaporated. Ilecrystal-
lization of the residue from Et20-P\IeOH gave 0.2 g (51%) of
colorless needles, mp 86-87". Anal. (C11HljN80~) C, H, N.
I n a similar way, 3-carbamoyl-4-cyano-l,4-dihydro-4-methyl-
1-propoxymethylpyridine was prepared from 0.3 g (0.002 mole)
of 3-carbamoyl-4-methyl-1-propoxymethylpyridinium chloride'
and gave 0.1 g (207,) of colorless needles, mp 185-186". Anal.
(CnH17?;302) C, H, K .
Acknowledgment -The author is indebted t o Pro-
fessor W. C. J. Ross for many helpful discussions and is
(3) grat'eful to the Nedical Research Council for financing
the purchase of the up- spectrometer and accessory ap-
paratus used in this work. The investigation was sup-
ported by grants to the Chester Beatty Research In-
stitute (Instit'ute of Cancer Research, The Royal
Cancer Hospital) from the Medical Research Council
and the British Empire Cancer Campaign and by the
Public Health Service Grant No. CA-03188 from the
Xationnl Cancer Institute, U. S. Public Health Service.
(29) L. J. Haynes and A. Todd, J . Chem. Soc., 303 (1950).
(30) C . Ukita, D. Mizuno, and S. Kosaka, J . Pharm. SOC.J a p a n , I S ,
111 (1953).