LITERATURE CITED i. 2. 3. 4. 5.

6. 7. 8. 9. I0. ii. 12. 13. 14.

V. A. Dobrynina, D. V. Ioffe, S. N. Kuznetsov, et al., Khim.-Farm. Zh., No. 6, 14-17 (1974). V. G. Gandel', "Study of ways of improving the production of tableted preparations containing toxic and highly active substances," Dissertation, Moscow (1967). T. Bano, T. Szarvos, and L. Aradi, Pharm. Zentral., I00, 221-225 (1961). Yu. P. Adler, E. V. Markova, and Yu. V. Granovskii, Experiment Planning during the Search for Optimal Conditions [in Russian], Moscow (1971). V. G. Belikov, V. D. Ponomarev, and N. I. Kokovkin-Shcherbak, Application of Mathematical Planning and Treatment of Experimental Results in Pharmacy [in Russian], Moscow (1973), p. 232. S. A. Minina, N. K. Zubkova, L. S. Efimova, et al., Med. Prom. SSSR, No. 4, 60-63 (1966). V. I. Egorova, Yu. N. Slavyanov, and 0. A. Bartashevich, Med. Prom. SSSR, No. I, 48-52 (1961). N. K. Zubkova and S. A. Minina, in: Trudy, Leningradsk. Khim.-Farm. In-ta, No. 24, 159-164 (1969). E. E. Kol'man-lvanov, V.A. Belousov, E. E. Borzunov, et al., Tableting Machines in Medicinal Industry [in Russian], Moscow (1975). L. Kzowczynci, Farm. Pol., Nos. 15-16, 351-354 (1962). V. G. Belikov and S. Kh. Mutsueva, in: Synthesis and Analysis of Drugs [in Russian], L'vov (1966), p. 16. G. S. Tereshin, Zh. Anal. Khim., No. 4, 389-395 (1959). G. S, Tereshin, Zh. Anal. Khim., No. 5, 516-522 (1959). V.I. Egorova, Med. Prom. SSSR, No. i, 36-38 (1964).

STABILITY OF WATER-SOLUBLE VITAMINS AND COENZYMES. HYDROLYSIS OF PYRIDOXAL-5-PHOSPHATE

IN ACIDIC, NEUTRAL, AND

WEAKLY ALKALINE SOLUTIONS I

E. I. Kozlov and M. Sh. L'vova

UDC 615.356:577.164.13].011.4

Pyridoxal-5'-phosphate (PLPh) is the coenzyme form of group B6 vitamins, and is included in the composition of various enzymic systems which catalyze different biochemical processes. Therefore, PLPh preparations can be applied in medicine during the treatment of various diseases which progress with disturbances of the protein, carbohydrate, and fatty acid metabolism, as with nervous and cardiovascular diseases. For the preparation of ready-for-use medicinal forms with PLPh, information must be obtained on the decomposition mechanism and factors and conditions influencing its stability. This information is necessary when technique is worked out for the production of the preparations with the coenzyme, selection of storage conditions, and the therapeutic use of the drug. In the present work we studied the hydrolysis of PLPh in acidic, neutral, and weakly alkaline aqueous and aqueous--alcoholic solutions, and its kinetic features. EXPERIMENTAL For the investigation we used chromatographically pure pyridoxal hydrochloride (PL) and PLPh hydrate, mp 139-142QC (dec). The buffer solutions were prepared according to [i] from reagents of "kh.ch." ["chemically pure"] and'hh.d.a." ["analytic"] grades. The pH values were adjusted according to the temperature. To study the kinetics of the hydrolysis of PLPh, we used the following buffer mixtures:

All-Union Scientific-Research Vitamins Institute, Moscow. Translated from Khimiko-Farmatsevticheskii Zhurnal, Vol. ii, No. ii, pp. 109-115, November, 1977. Original article submitted February 22, 1977. 0091-150X/77/IIII-1543507.50

9 1978 Plenum Publishing Corporation

1543

Ig ~-lx40

kef f - 104 sec -1 ~I

14 T3121110

9

-

b y i~

!I

4,0

/:

8,O/ ~ 2,0 .I0

O

50

100

150

200

I I

250 3OOmin

I 2.

I 3

Fig. i

4

5

6

7

8 pH

Fig. 2

Fig.l. Semilogarithmic anamorphoses of kinetic curves of the hydrolysis reaction of PLPh in aqueous buffer solutions at 98~ at various pH values. I) 7.73; 2) 7.37; 3) 6.64; 4) 6.24; 5) 5.91; 6) 5.23; 7) 4.27; 8) 3.72; 9) 3.4; i0) 3.0; ii) 2.6; 12) 1.0; 13) 1.6; 14) 2.05. Fig. 2. Dependence of the rate constant of the hydrolysis of PLPh in aqueous buffer solutions on the pH value at 98~

pH 0.65--1,0

1.0--2.2 2.2--3.6 3,7--5.5 5.9--8.0

Componenm CHsCOONa, HC1 KCI, HCt Potassium biphthalate, HCI

CH~COOH, CHsCOONa KH~PO4, Na~HPO4

The phosphate salts were crystallized from distilled water and dried in a dessicator over calcium chloride. The concentration of the buffer components was not less than 50 times that of PLPh. It was found that at the given value of the ionic strength, the rate constant is practically independent of the type of buffer. The water-ethanol and water--isopropanol mixtures (pH method given in [2]; the pH of these mixtures, containing ride and 0.002 mole/liter of sodium chloride, was measured aqueous buffer solutions and corrections were introduced pH [3].

-2.8-2.9) were prepared by the 0.002 mole/liter of hydrogen chloby a pH-meter standardized for into the numerical values of the

The ionic strength of the solution (~) was maintained constant (~ = 0.2), or was varied by adding potassium chloride. The hydrolysis of PLPh was studied over the temperature range of 37-98~ (•176 The coenzyme was determined spectrophotometrically on a Specord UV-Vis instrument using rectangular l-cm cuvettes. Method. An aliquot portion is withdrawn from a solution of 0.015 g (0.056 mmole) of PLPh hydrate in 25 ml of aqueous--alcoholic solution previously saturated with admixture-free nitrogen, and placed in a i00 ml volumetric flask. A calculated amount of potassium chloride is added, and the solution is brought to the mark with the corresponding buffer solution. The solution obtained with an initial concentration of PLPh of (i.0-4.5).i0 -4 mole/liter is placed in a thermostated, light-protected vessel equipped with a magnetic stirrer , reflux condenser, inlet tube for the introduction of nitrogen, and a thermometer. Probes of the reaction solution cooled to 20~ (if necessary, diluted with the corresponding buffer solution) are withdrawn periodically, and the optical density (D) is measured at 388-392 nm for solu-

1544

tions at pH 3.7-8.0, at 333-340 nm for solutions at pH 0.65-3.4,* and at 370-380 nm for aqueous--alcoholic solutions (pH 2.8). The corresponding buffer solution is placed in the reference cuvette. By special experiments it was shown that in the wavelength regions found, PL (a hydrolysis product of PLPh) has no absorption maximum and does not contribute to the optical density. The values of the effective rate constant (kef f) were calculated from the following equation (in the calculation, the data used were obtained at a degree of conversion of from I0 to 70%): 2.303 .

1

keff=~lg

(1)

1--x '

in which x

=

Do--Dr Do

,

where D o i s the optical density at the initial moment of time; D t is the optical density at the moment of time t in the absorption maximum of PLPh. RESULTS AND DISCUSSION Reaction Order. Figure i shows the dependence over the pH range studied (0.65-8.0). During the computation of the kef f value in the experiments with different concentrations of the coenzyme in the aqueous buffer solution at 98~ (pH 5.91, ~ 0.2), the following data were obtained: Concentration of I PI2h 9 104 mole/] liter I kerr't05 ~ c ' 1 ' I

0.79 4.86

i

1,13 5.56

1.58 5.37

II

2,25 1 5.05

6.70 5.55

I

] ke,ff. me = ] =5.28.10 - i sec-1

From the above results it is seen that the hydrolytic cleavage of PLPh is a first-order reaction with respect to the starting compound. This result agrees with the reaction order obtained during the hydrolysis of other phosphate esters of organic compounds [5]. Dependence of the Kate Constant of the Hydrolysis of PLPh on the pH of the Medium. We studied the influence of the hydrogen-ion concentration on the rate Of decomposition of PLPh at the same initial concentration of the coenzyme and at a constant value of ~ equal to 0.2. It follows from the dependence given in Fig. 2 that with decrease in pH from 8.0 to 4.0, there is a gradual increase in the rate of hydrolysis of PLPh, and then in the pH range of 4.0-2.0, the kef f sharply increases, and then decreases up to the boundary of the region studied (pH 0.65). The character of the observed dependence is analogous to that obtained by other authors [6], who studied the hydrolysis of PLPh in an acidic medium, using the polarographic method to determine the initial compound. In contrast to most other monophosphate esters, which give a maximum of the rate of the hydrolysis reaction at pH 4.0 [7-9], the highest rate of decomposition of the coenzyme is here reached at pH 2.05. Influence of Temperature. The dependence of the rate constant of the hydrolysis of PLPh on temperature in the 37-98~ range is governed by the Arrhenius equation (Fig. 3). The effective activation energy calculated by the method of least squares is 22.98 • 0.7 kcal/ mole at pH 2.05 and 22.24 • 0.5 kcal/mole at pH 6.98; the value of the preexponential factor is equal to 10 I~ and 108.2 liter.mole-sec-*, respectively. Influence of Ionic Strength. The dependence of the effective rate constant of the hydrolysis of PLPh on ~ was studied at constant pH (6.47) (at a PLPh concentration of 1.58-10 -~ mole/liter). A certain decrease in pH, observed during the addition of potassium chloride, was compensated for hy the addition of a Na2HPO~ solution. It follows from the data obtained that when the pH is maintained constant, the primary salt effect is observed, which indicates thr~arallelism of the abscissa axis and the line representing the dependence of log kef f on y~ (the parameters of the Debye--H~ckel equation for dilute solutions) (Fig. 4, curve i). *The pH-dependent shift of the absorption maximum of PLPh is explained by the contribution of different forms of PLPh to the total absorption. This process is discussed in detail in [4]. 1545

log kef f + q 4~

log keff + 5 0,8 0,7

3

2

0,6

2

0,5 !

0,4

2,6

+403

1

[

i

2,8

8~0

3a2

-

-

-

0,3

1

0,4

1

-

I

0.6

0,8

I 1,0

I

1~2

Fig. 4

Fig. 3

Fig. 3. Dependence of the rate constant of the hydrolysis of PLPh on temperature. i) 6.98; 2) 2.05. Fig. 4. Dependence of the rate constant of the hydrolysis of PLPh on the ionic strength of the solution, i) Change in pH was compensated for by the addition of NaaHPO,, 2) without the addition of Na2HP04. TABLE i. Dependence of the Effective Rate Constant of the Hydrolysis of PLPh on the Dielectric Constant

Medium

~, '-'' ~ ~I ' "2s ~ C' ",..';' -

!O_~ ) I

'

l~

,,, .

Isopropanol-- water 38,0 Ethanol-water i32,5

49,9 53,6

1,26 1,64

Isopropanol- water 27,8 Ethanol-water !21,1

61,6 67,0

3,51 6,79

Isopropanol--water 13,5 Water ] 0,0

70,2 78,3

7,19 ll,l

When there was no pH compensation, the addition of potassium chloride to the reaction solution led to a decrease in pH, which caused an increase in keff, i.e., a secondary salt effect was noted (Fig. 4, curve 2). Dependence on the Dielectric Constant. The influence of the dielectric constant of the medium (E) on the rate constant of the hydrolysis of PLPh (1.58.10 -4 mole/liter) was studied in mixtures of ethanol-water and isopropanol-water solvents under the following conditions: temperature 820C, pH 2.8, B 0.2. The dielectric constants of the alcohol-water mixtures were calculated from the Mohr formula [i0] taking into account the e of each component [Ii]. The table lists the solvents used, the c values*, and the experimentally obtained values of kef fThe results obtained show that the rate constant of the hydrolysis of PLPh decreases sharply with increase in e of the medium. The dependence of log kef f on i/e, plotted according to the Kirkwood equation, is linear (Fig. 5). Reaction Mechanism. It is known that PLPh exists in aqueous solution in the form of a complex mixture of ionic and molecular forms [4, 12]. Recently it was shown that PLPh is present in the form of a free aldehyde at high pH values, a completely hydrated aldehyde at pH 1.7, and a mixture of the two forms at pH 4.0 [13]. Each of the molecular forms exists up to pH 5.0 in the form of a cation (K), dipolar ion (D), and anion (A). At higher values *The change in ~ with a change in temperature is fairly small for alcohols into account in the calculations.

1546

and was not taken

keff.lO 4 log kef f + 5

sec -t

4

3 0,8

.

~

NN,

~ 1,0

N,

|

l

I

I

I

I

1,2

1,4

1,6"

1~8.

2=0

I

2,2

I I0 2

~"

f

I

!

~

!

0,2

0j4

0,6

0,8

%0'

Fig. 5

C

Fig. 6

Fig. 5. Dependence of the rate constant of the hydrolysis of PLPh on the dielectric constant of the medium at pH 2.8. Fig. 6. Dependence of the rate constant of the hydrolysis of PLPh on the mole fraction of the dipolar io~ in the pH range 0.65-2.5. of pH, two other ionic forms appear -- di- and trianions (DA and TA) [4, 12]. Thus, the effective rate constant of the hydrolysis of the coenzyme can he represented in the form of a sum of specific constants according to the n,mher of molecular forms of PLPh in the solution. The general picture of the processes taking place in aqueous solution of PLPh can be represented by the following scheme* OliO

O

II

O

CHO

,I

-

_-

OHO

O

II

_

HO..,~.~C~OP-0H ~, ~O~CHzOP,--O K~ o . . L CH~0P,-O CH$ ~

CH 3 N

CH5

CHO

-

OHIO _-O-

lie

TA

OH?.

OHo

O

KD -

l-[

t-'l

PL

DA

The ionization constants of the first three forms of PLPh in water are given in ~6, 14, 15]. Since the pK value can increase with increase in temperature, while the kef f value is independent of the ionic strength (see Fig. 4, i), for the calculations we used the pK~ and pK2 values obtained in [6] at a temperature of 87.8~ At low and medium values of pH (0.65 = 5.0) in the solution, three ionic forms of PLPh~ K, D, A, are present in equilibrium. Thus, the effective rate constant of the hydrolysis can he represented as a sum of the three component hydrolysis constants;

[K]

[R]_.

[A]

keff=kK-~-q-k D--~-t~A-~-.

(2)

Depending on the pK~ and pK2 values, we can assume that at pH values up to 2.5, a mixture of K + D forms exists preferentially in the solution, at pH from 2.5 to 3.5, K + D + A *For the sake of simplicity, PLPh and PL are represented as free aldehyde forms.

1547

are present, and the pH range 3.5-5.0 is characterized by the presence of the D + A forms. Thus, in the pH range from 0.6 to 2.5, Eq. (2) is reduced to the following expression:

where C = [K] + (D) pK,

<2.5 1.4 1.64 0.9

pK,

pK,

4.14 3.44 3.58 3.15

6.20 5.75 5.75

Ref.

t~C

P,

25

0.16 0.I

[14]

2,0

1151 [6]

25 25 87.8

1.0

1151

If we express the concentration of the dipolar ion D using the equilibrium constants of PLPh, K, and K2, we obtain the following expression: C I/i; =

1 + [ H + I KV ~+ K 2 [ H + I - - T "

Equation (3) shows that the specific rate constants of the hydrolysis of the ionic forms of PLPh can he obtained graphically from the linear dependence of kef f on the mole fraction Qf the dipolar ion (Fig. 6), and the slope of the llne gives the numerical value k D - - k K = 3.91-10-~*sec-*, while the section cut on the ordinate, k K = 0.03,10 -4 sec -2, whence k D = 3.96-10 -~ sec-*.* For the pH 3.5-5.0 range, where PLPh exists mainly in the form of a dipolar ion and anion, Eq. (2) acquires the form ~ f f = k D ~ + kA ~=kAJr(kD--kA)~, where C = (D) + (A). From a calculation s~m~lar to that given above, the specific rate constant of the hydrolysis reaction of PLPh is kA = 0.23.10-a,sec -*. A comparison of the values of the specific rate constants of the hydrolysis reaction shows that the principal reactive species in pH 0.65-5.0 range is the dipolar ion D, whose predominating contribution also reflects the maximum on the curve for the kef f vs pH dependence. The Kef f value calculated from Eq. (2) at pH 2.05, when the ion D is preferentially present in the solution, agrees well with the experimental value found within the experimental error. The hydrolysis of the dipolar ion of PLPh proceeds about 20 times faster than that of the anion A, and almost i00 times faster than the decomposition of cation K. The assumption put forward in [6] states that the stage determining the rate of reaction is the monomolecular heterolysis of the P-O bond in PLPh, facilitating the transition of the hydrogen atom from the hydroxyl group of the phosphoric acid residue to the oxygen atom of the ester linkage. We believe that the disadvantage of the proposed scheme of the reaction mecha-~sm is that the active influence of the aldehyde group of PLPh in the composition of the intermediate complex is not taken into account. Meanwhile, it has been noted in the literature that the reactivity of the aldehyde group [18], as well as the pH-dependent change in the spectrum of PLPh [19], are related to the inductive influence of the 5-phosphate residue of the coenzyme. This influence can be exerted as the result of sterlc factors also, since the molecular models of PLPh allow for an approach of the phosphate group and the carbonyl group within atomic distances. If we consider these data, the mechanism of the hydrolysis of PLPh in acidic media can be represented hy the following scheme. *The results of the kinetic study of the hydrolysis of ascorbic acid-3-phosphate were treated s~milarly I17]. 1548

HO,/OH CH

.......

0I-

0 ---P - - 0

slow

c% ~ H

p

-

o.~ +

PL + PO5

H~O fast =- H~PO~

Additional data that we obtained on the hydrolytic decomposition of the coenzyme agree well with the suggested reaction mechanism. The zero slope of the line in the log kef f vs ~ d e p e n d e n c e indicates either a reaction of positive and negative ions with the neutral molecule, or intramolecular transformations of ions or neutral molecules. The high polarity of the activated complex explains the increase in kef f during increase in the dielectric constant of the medium. It is possible that the anomalous pH-dependent shift of the maximum of the rate constant of the hydrolysis of PLPh into the more acidic region, compared with simple alkyl phosphates, is also related to the influence of the carbonyl group of the coenzyme. The pH region, where the coenzyme exhibits the highest stability to hydrolysis, is interesting from the practical point of view. As in the case of simple alkyl phosphates, the rate of decomposition of PLPh decreases markedly under neutral and weakly alkaline conditions (see Fig. 2). With increase in pH from 5.0 to 8.0, the concentration of the dipolar ion in the solution falls practically to zero, and other ionic forms of PLPh, mainly the DA and TA anions, then play the role of the principal species and are characterized by a low reactivity. In the continuation of this work, we propose to study the factors determining the stability of PLPh, in particular, under neutral and alkaline conditions. LITERATURE CITED i. 2. 3. 4.

E. N. Vinogradova, Methods for the Determination of the Hydrogen Ion Concentration [in Russian], Moscow (1956), pp. 88-89. R. G. Bates and G. Schwarzenbach, Helv. Chim. Acta, 38, 699 (1955). R. G. Bates, Determination of pH. Theory and Practice, Wiley--lnterscience, New York (1964) [Russian translation: Leningrad (1968)]. Yu. V. Morozov, N. P. Bazhulina, L. P. Cherkashina, et al., Biofizika, No. 3, 397-406

(1967). 5. 6. 7. 8. 9. I0. ii. 12. 13. 14. 15. 16. 17. 18. 19.

J. R. Cox and O. B. Ramsay, Chem. Rev., 64, 317 (1964). P. Zuman and O. Manu~ek, Coll. Czech. Chem. Commun., 26, 2134-2143 (1961). J. D. Chanley, E. M. Gindler, and H. Sobotka, J. Am. Chem. Sot., 74, 4347 (1952). L. Kugel and M. Halmann, J. Am. Chem. Soc., 88, 3566 (1966). L. Kugel and M. Halmann, J. Org. Chem., 342, 642 (1967). W. Moore, J. Am. Pharm. Assoc., 47, 856 (1958). Chemists Handbook [in Russian], Vol. i, Moscow-Leningrad (1963), pp. 948-956. N. P. Bazhulina, A. Ya. Lomakin, Yu. V. Morozov, et al., Mol. Biol., No. 6, 899-905 (1970). R. D. Lapper, H. H. Mantsch, and I. C. P. Smith, Can. J. Chem., 53, 2406 (1975). V. R. Williams and J. B. Neilands, Arch. Biochem., 53, 56-70 (1954). F. J. Anderson and A. E. Martell, J. Am. Chem. Soc., 86, 715-719 (1964). Yu. V. Morozov, N. P. Bazhulina, V. I. Ivanov, et al., Biofizika, i0, 595-601 (1965). H. Nomura, M. Kuwayama, T. Ischiguro, et al., Chem. Pharm. Bull., 19, 341-354 (1971). M. V. Buell and R. E. Hansen, J. Am. Chem. Soc., 82, 6042-6049 (1960). N. P. Bazhulina, V. I. Ivanov, Yu. V. Morozov, et al., Biofizika, 16, 991 (1971).

1549