Biochemical

Vol. 44. No. 2. pp. M-204.

Pharmacoology,

1992

ooo6-2952/92

Printed in Great Britain.

s5.00 + 0.00

Pergamon Press Ltd

RAPID COMMUNICATION SUBSTRATE SPECIFICI.IY OF ESCHENCHIA COLI THYMIDINE PHOSPHORYLASE NUCLEOSIDES

WITH ANTI-HUMAN

IMMUNODEFICIENCY

F. Schinazi’t,

Peck* and Jean-Pierre

Raymond

Annette

Sommadossi§

*Veterans Affairs Medical Center, Decatur, GA 30033, and Laboratory Department BDepartment

of Pediatrics,

Emory University

of Pharmacology, University

of Alabama

at Birmingham,

(Accepted ABSTRACT.

Various nucleoside

moiety

Nucleosides

remain

than saturated

one of the most

syndrome

to their 5’-triphosphate

triphosphate

against

human

chain terminators

of human

of these

compounds

by various enzymes

present

phosphorylase.

Nucleosides

nucleosides

complex

The increasing

virus (HIV) reverse

a variety

thymidine

phosphorylase

of pyrimidine

The compounds dideoxycytidine

inhibitors

of natural

transcriptase

(RT). Their selectivity

[I]. These nucleotides

studied

(AzddU,

didehydrothymidine deoxythymidine

were

have

demonstrated

(D2T),

from

human

their

potential

phosphorylases

[2-51. In this study, we

for their

resistance

(AZT; Zidovudine;

2’,3’-dideoxyuridine

since this compound treated

In addition, is a recently

with AZT [17,18].

the involvement

the

to bacterial

bone

marrow

discovered

Studies

using

(D2U) (D4LJ)

compared

to AZT.

[7],

[6], 2’,3’-

3’-azido-2’,3’-

[7],

3’-fluoro-3’-

(FTU), and

catabollte

of AZT that was found liver microsomes

increases

and erythroid

in

have

in the presence

enzyme system [3]. Of significance

Similarly,

2’,3’-

3’-amino-3’-deoxythymidine

rat and human

of AMT and AMT-glucuronide

of an NADPH-dependent when

Retrovir)

(D4C) 17,111, 3’-deoxy-2’,3’-

nucleoside

finding that AMT was 5- to 7-fold more toxic to granulocyte-macrophage derived

is

to act as

has been placed

addressed

[15,16], 2’,3’-dideoxy5-fluoro-3’-thiauridine

(FTC) [16].(

that the rate of formation

NADPH, suggesting

activity

[7,12,13], 2’,3’-dideoxy-2’,3’-didehydrouridine

(FLT) [14], dioxolane-T

and humans

anti-HIV

3’-azido-3’-deoxythymidine

3’-deoxythymidine

dideoxy-5-fluoro-3’-thiacytidine

monkeys

with

CS-87) [8-101, 2’,3’-dideoxy-2’,3’-didehydrocytidine

(D4T)

(AMT) was evaluated,

are thought

much emphasis

few studies

are

nucleoside-5’-

in certain human cells. Indeed, as recently reported,

nucleosides

of acquired

(dThdPase).

(D2C),

dideoxyuridine

to this enzyme.

for the treatment

may have an impact on the efficacy and toxicity of certain 2’,3’-dideoxynucleosides evaluated

was a substrate,

were resistant

by the viral RT. Although

in cell culture,

order

in the sugar

(ARC) [I]. Most of these compounds

competitive

DNA polymerases

for their ability

that were unsaturated

class of compounds

form and are potent

of the viral DNA catalyzed

on the metabolism catabolism

promising

immunodeficiency

based on their weak inhibition

were examined

ones. 3’-Deoxy-2’,3’-didehydrothymidlne

(AIDS) and AIDS-related

metabolized

AL 35294, U.S.A.

1992)

3’-azido-, 3’-fluoro-, 3’-0x6 and 3’-thiapyrlmidlne

immunodeflciency

Atlanta, GA 30322; and

agents and their metabolites

to cleavage was U > T >> C derivatives.

were more susceptible

whereas

Birmingham,

5 June

Pharmacology,

Division of Clinical Pharmacology,

bond by Escherichia coli thymidine

to be cleaved across the glycosidic of susceptibility

antiviral

of Biochemical

School of Medicine,

Center for AIDS Research,

FOR PYRIMIDINE

VIRUS ACTIVITY

of

was the

bone precursor

cells

3’-amino-2’,3’-dideoxyuridine

+ Corresponding author. Tel. (404) 728-7711;Fax (404) 728-7726. 1 Schinazi RF, Liotta DC, Choi WB, Peck A, McClure H, Boudinot FD, Sommadossi J-P, Davis M, Furman PA and Painter G. Selective inhibition of human immunodeflciency virus and hepatitis B virus by 2’,3’-dideoxyb-fluoro-3’thiacytidine (FI’C). National Collaborative Drug Discovery Group, Frontiers in HlV Therapy. San Diego, CA, Nov. 3-7, 1991. 199

200

Rapid communication

(AMDU)

was evaluated,

rhesus monkeys

since small amounts

that were treated with AzddU

Racemic FTC is a potent anti-HIV-l have potent

anti-hepatitis

converted

to (+)-FTU

thymidine

phosphorylase,

FUra is a highly phosphorylase

B virus

agent in human

activity

by cytidine

of this compound

cytotoxic

compound.

lymphocytes,

in liver microsomes

A concern

which was recently

reported

to also

of FTC was

is that if FTC and FTU are substrates

(5-FCyt) and 5-fluorouracil

Therefore,

and

results].

in 2.2.15 cells 1201.1 Only the (+I-enantiomer

deaminase.8

then 5-fluorocytosine

was examined

are formed

[19; Schinazi R, unpublished

for

(5-FUra) would be produced.

the susceptibility

of FTC and

5-

FTU to thymidine

in this paper.

MATERIALS AND METHODS Materials. prepared

The 2’,3’-dideoxy analogues

according

1 H-NMR

to published

spectroscopy

reported

procedures

and elemental

herein and racemic dioxolane-T,

[7-161. All the compounds dThdPase

analysis.

Chemical Co. (St. Louis, MO). HPLC grade solvents

reaction

by melting point,

was obtained

from

and all other chemicals,

the Sigma

analytical

grade,

from J.T. Baker (Phillipsburg, NJ).

were obtained Enzyme

E. di

from

and KH2P04,

FTC, and FTU were

were characterized

Stock solutions

assays.

was performed

reaction was initiated was 200 FL. Since concentration

of the compounds

in a buffer by adding

of uracii present

enzyme

already

by filtering

contains

water.

The enzyme

pH 7.4, at room temperature.

to 100 PM compound

in the enzyme mixture

of the enzyme

in deionized

of 100 mM KH2PO4,

1.1 units of enzyme

the commercial

and 60 mln after addition

consisting

were prepared

2 mM uracil

was 10 FM. Reactions

the sample through

The

in the buffer. The final volume as a stabilizer,

the final

were terminated

at 30 min

a 0.45 pm Acrodlsc@ LCl3 PVDF

filter (Gelman Sciences, Ann Arbor, MI); then the eluent was stored frozen at -70” until analysis by HPLC. HPLC analysis. instrument. separation detected

A 40 pL portion of each sample was analyzed

A Hypersil

ODS-18 5 pm column

of the various nucleosides at a UV wavelength

listed below as Methods

Lakewood,

and bases. The flow rate was 1 mlfmin,

of 262 nm. Different

A through

by HPLC using a Hewlett

(Jones Chromatography,

and the compounds

mobile phases and conditions

10 mM KH2P04.

cytoslne

the approximate containing

the concentrations

C: isocratic

by a 10 min linear gradient

elution

concentration

nucleosides,

are

from 7 to 30% MeOH.

all HPLC peaks were compared of compound

thymine

at different

by a 1 min

for 10 min with 10 mM KH2PO4 and Method

Method E: lsocratic efution for 5 min with 10 mM KH2fO4,

from 0 to 30% MeDH. For consistency, obtain

for

E. Method A: isocratic elution with 10 mM KH2PO4 (pH 6.61 and MeOH

from 7 to 20% MeOH. Method

MeOH (9371, followed

1050 were

used in these analyses

(93:7). Method B: isocratic elution for 6 min with 10 mM KH2PO4 and MeOH (93:7), followed linear gradient

Packard

CO) was used

D isocratic elution with

then a 10 mln linear gradient

relative to the uracil peak in order to

time points.

(10 FM) was used as an internal

In the case of uracil

standard

and

in order to determine

of the base and nucleoside.

RESULTS AND DISCUSSION

dThdPase

is an enzyme

that catalyzes

the reversible phosphorolysis

of several natural and synthetic

pyrimidme nucleosides. Niedzwicki et at. (21f have studied the st~c~~ac~~t~ of nucleosides as substrates or inhibitors of dThdPase derived from mouse liver. compounds with electron-withdrawing

relationship of a number They determined

or hydrophobic groups at the 5-position of the pyrimidine

that

moiety

201

Rapid communication

enhance binding to dThdPase. The authors indicafe that dThdPase is highly specific for the 2’-deoxyribosyl

moiety of nucleoside ligands, and that ribo- or arabino-nucleosides

are not cleaved by dThdPase. Similarly,

Kono et al. [22] determined the substrate specificity of dThdPase derived from human liver tumor using a variety of uridine and fhymidine modified compounds.

if was found that an intact 3’-hydroxyl function

was essential for cleavage. 5-Promo, tlrforo, and -fh.rorouridme were substrates for dThdPase, whereas 2’-, 3’- or ~-deoxyuridine deoxyuridine

were not. ~-Deoxy-5-fluoro~dine

was also a better substrate

(FdUrdf for this enzyme. However, no ~,3’~deoxypy~midine

than 5-fluoro-Z-

nucieosides were studied by

these groups, Since this class of compounds is currently clinically important for the treatment of HIV infertions, if was essential to determine if these compounds are cleaved by dThdpase that is present in various cells We have successfully applied a relatively simple HPLC system to determine the degree of glycosidic bond cleavage of pyrimidine nucleosides Rabie 1). The results of the enzyme assays and the HPLC method are shown in Table 1. The data demonstrated

that among the ~,~-dideoxypyrimidine

nucleosides, the U

derivatives were better substrates than either the T or C derivatives. The increasing order of susceptibility to cleavage was U > T >> C derivatives. nucleosides

Consistent

with previous

findings with cytosine containing

f23], D4C was markedly less susceptible than D4T to dThdPase (Table 11. The 2Jdideoxy

derivatives ID2U, D2C or D2T) were more resistant than the D4 compounds (D4U and D4T& D2C, AZT, and AzddU were completely

resistant

to dThdPase.

susceptible to cleavage by E. c& dThdPase.

The 3’-amino derivatives,

AMT and AMDU, were

Small amounts (~3%) of thymine were detected wifh PLT. The

3’-0x0 derivative related to thymidine was also a poor substrate for this enzyme and substitution hydroxy

function

phosphorylases.

by a 3’-Ruoro or 3’-azido group in thymidine

markedly

increased

of a 3’-

resistance

to

FTC and FLU were not substrates for this enzyme, suggesting that the potential generation

of 5-PUra from FTC is unlikely in cell culture or in z&o, Studies with purified human liver dThdPase and in rhesus monkeys, as well as metabolic studies using radiolabeled mononuclear

FTC in human

cells and HepG2 cells, suggest that cleavage to 5-PUra is unlikely

[Schinazi R, unpub~~h~

peripheral

in mammalian

blood cells

results].~ As expected, 99% of the FdUrd (used as a positive eontroll was cleaved

by dThdPase affer 1 hr, whereas 2’deoxycyfidine {used as a negative control) was not Gable If. We have demonstraf~

that certain R4-pyrimidine

derivatives

are substrates

for dThdPase.

It is

significant that D4T, a compound undergoing Phase l/IL clinical trials in HIV-infecfed individuals, is a good substrate for this enzyme. D4T has a good oral bioavaibbility

in animals $141;however, only about 50% of

the dose is recovered in urine [241. it is conceivable that a portion of this drug could be converted to thymine and its metabolifes in humans. Of significance is the finding that intracellular metabolic studies in bone marrow cells indicated thymidine

that 70% of the radioactivity

derived from [3H]D4T is associated with

in nucleic acids (41. This conversion could be mediated by dThdPase and may have important

consequences

since the intracellular

levels of this enzyme vary in different

cells. Furthermore,

fhymidine produced by the degradafive enzyme may compete for the phosphorylation D4T. Lower levels of dThdPase are found in T-~yrnphoc~~

the

and degradation of

than in bone marrow cells, which may explain

its selective antiviral activity in human lymphocytes and lack of bone marrow toxicity in humans [4]. It is significant

that both D4T and ~,~-dideoxy~osine

are cleaved by dThdPase

and purine

nucleoside

phosphorylase, respectively, and dso do not cause bone marrow toxicity in nihm and in viaa, it should be stressed that the bacterial enzyme used in these studies may not have the same substrate

specificity requirement

for cleavage as the human-derived

enzyme.

AMT is a good substrate for E. coli dThdPase (Table 11, but a poor substrate for mouse liver enzyme [21]. Studies are in progress to determine the substrate specificity of human liver dThdPase towards AMT. Similar studies on the effect of human dThdPase on D4T are consistent with our findings, suggesting a good correlation between bacterial and mammalian

enzymes.

Rapid communication

202 Table 1. Effect of E. coli thymidine Nucleoside

Base

HPLC method

E

Urd Uracil

A

FdUrd 5-F&a

C

D2U Uracil

C

D4U Uracil

D

AMDU Uracil

C

AzddU Uracil

A

FTU 5-FUra

C

dThd Thymine

C

D2T Thymine

A

D4T Thymine

A

AMT Thymine

C

FLT Thymine

B

AZT Thymine

C

Dioxolane-T Thymine

E

dCyd Cytosine

E

D2C Cytosine

E

D4C Cytosine

E

FTC 5-FCyt * None of the compounds + Relative to thyrnidine

phosphorylase

Retention

on various pyrimidine

% Nucleoside

nucleosides

and base at:

time (min)

0 hr”

0.5 hr

1 hr

11.1 5.7

100.0 0.0

0.0 100.0

100.0

22.5 4.14

100.0 0.0

1.0 99.0

99.0

100.4

15.4 3.8

100.0 0.0

67.8 32.2

44.7 55.3

56.1

12.0 3.8

97.3 2.7

43.0 57.0

14.6 85.4

86.7

10.1 5.5

97.3 2.7

29.5 70.5

9.0 91.0

923

21.0 3.75

100.0 0.0

100.0 0.0

100.0 0.0

0.0

22.5 4.14

100.0 0.0

100.0 0.0

100.0 0.0

0.0

13.6 6.6

99.3 0.7

1.4 98.6

1.4 98.6

100.0

20.2 6.9

100.0 0.0

87.1 12.9

68.1 31.9

324

23.5 6.9

100.0 0.0

43.1 57.9

27.1 72.9

74.0

7.9 6.6

100.0 0.0

62.6 37.4

43.2 56.8

57.6

21.8 6.9

100.0 0.0

98.7 1.3

97.4 26

26

25.0 6.6

100.0 0.0

100.0 0.0

100.0 0.0

0.0

16.1 6.8

98.4 1.6

96.4 1.6

98.4 1.6

0.0

11.2 4.2

100.0 0.0

100.0 0.0

loo.0 0.0

0.0

13.8 4.3

loo.0 0.0

loo.0 0.0

100.0 0.0

0.0

13.2 4.3

99.5 0.5

99.5 0.5

99.5 0.5

0.0

15.8 4.4

100.0 0.0

100.0 0.0

100.0 0.0

0.0

were cleaved after incubation at 1 hr (minus background).

without

enzyme

Relative

%+

0.0 101.5

1.0

for 60 mm at room temperature.

203

Rapid communication Thus, it appears isozyme

that the mammalian

and that data obtained

nucleosides

that are cleaved

by the bacterial

Studies with other related nucleosides could provide rationale

an understanding

for designing

All the anti-HIV

nucleosides

bacterial

nucleosides.

This possibility

clinically relevant

should

appears

are ongoing.

based upon the finding nucleosides

isozyme.

Such studies

of these compounds

with minimal

inactive

Consequently,

with the human

preparation

of HIV replication

than the bacterial

cautiously.

and lack of toxicity

are given orally to adult

pyrimidine

specificity

also be studied

in the gut to produce unlikely

orally bioavailable

activity

inhibitors herein

found

patients

and could

compounds

and a

mammalian

toxicity.

interact

or competing

that most of the currently

such as AZT, D2C, AzddU,

with

natural available

and FLT are very

for bacterial dThdPase.

ACKNOWLEDGEMENTS. Liotta (Emory University) supported

enzyme

substrate

should be interpreted

using a human liver enzyme

described

enzymes

has a greater enzyme

of the selective

other cell-specific

degradative

poor substrates

enzyme

with the bacterial

We are grateful

for providing

by the Department

to Dr. C.K. Chu (University

some of the compounds

of Veterans

described

Affairs, and the United

25899 (R.F.S) and HL 42125 (J-P.S)J. J-P.S. is the recipient

of Georgia, Athens) and Dr. D. in this paper.

This work was

States Public Health Service Grants [AI

of a Faculty Research

Award

from the American

Cancer Society.

REFERENCES 1. Schinazi

RF, Mead JR and Feorino

PM, Insights

into HIV chemotherapy.

AIDS Res Hum Retroviruses

8: 553-579,1992. 2. Zhu Z, Ho H-T, Hitchcock dideoxythymidine

MJ and Sommadossi

(D4T) in human

peripheral

J-P, Cellular

pharmacology

of 2’,3’-didehydro-2’,3’-

cells. Biochem Pharmacol

blood mononuclear

39: R15-

R19,1990. 3. Cretton

EM, Xie M-Y, Bevan RJ, Goudgaon

azido-3’-deoxythymidine

in hepatocytes

amino-3’-deoxythymidine,

a highly

NM, Schinazi and liver

toxic catabolite

RF and Sommadossi

microsomes for human

with

J-P, Catabolism

of 3’-

of formation

of 3’-

evidence

bone marrow

cells. Mel Phnrmacol 39:

258-266,199l. 4. Zhu Z, Hitchcock dideoxythymidine 5. Sommadossi 6. Mitsuya

MJM and Sommadossi

JP, Metabolism

in human bone marrow

J-P, Nucleoside

H, Weinhold

effect of human

and differences.

PA, St. Clair MH, Lehrman

and Broder S, 3’-Azido-3’-deoxythymidine and cytopathic

cells. Mol Pharmacol40:

analogs: Similarities

KJ, Furman

and DNA interaction

838-845, 1991. Clin Infect Dis (in press).

SN, Gallo RC, Bolognesi

(BW A509U): an antiviral

T-lymphotropic

of 2’,3’-didehydro-2’,3’-

agent that inhibits

virus type III/lymphadenopathy

D, Barry DW the infectivity

associated

virus in

vitro. Proc Nut1 Acad Sci USA 82: 7096-7100, 1985. 7. Chu CK, Schinazi activity

RF, Arnold

of 2’,3’-saturated

immunodeficiency

BH, Cannon

and

unsaturated

virus type 1 in peripheral

DL, Doboszewski pyrimidine

and

blood mononuclear

B, Bhadti VB and Gu Z, Comparative purine

nucleosides

cells. Biochem

against

Pharmacol37:

human 3543-

3548,1988. 8. Chu CK, Schinazi

RF, Ahn MK, Ullas GV and Gu ZP, Structure-activity

nucleosides

as antiviral

mononuclear

cells. J Med Chem 32: 612-617, 1989.

agents

for human

immunodeficiency

virus

relationships type

of pyrimidine

1 in peripheral

blood

Rapid communication

204

9. Eriksson BFH, Chu CK and Schinazi

RF, Phosphorylaiio~

of 3’-azido-2’,3’-dideoxyuridine

virus reverse transcriptases by its 51..

preferential inhibition of human and simian immunodeficiency Anfimicrob Agents Chemofber 33: 1729-1734, 1989.

triphosphate. 10. Lin T-S,

GUO

and

J-Y#Schinazi RF, Chu CK, Xiang J-N and prusoff WI-L Synthesis and antiviral activity of

various 3’-azido analogues

of pyrimidine

deoxyribonucleosfdes

against human immunodeflciency

virus (HIV-I, HTLV-III/LAV). J Ned Chem 31: 336-340, 1988, 11. Lin T-S, Schinazi RF, Chen MS, Kinney-Thomas

E and Prusoff WH, Antiviral activity of 2’,3’-clideoxy-

cytidln-2’-ene ~2’~-dideoxy-~,~-dld~ydr~y~dine~ ~iachem P~~cot

against human ~~odeficiency

virus in r&o.

36: 311-316, 1987.

12 Lin T-S, Schinazi RF and Prvsoff WH, Potent and selective in vitro activity of 3’-deoxythymidin-2’-ene (3’-deoxy-~,3’-didehydrothymidine)

against human immunodeficiency

virus. &o&em P~~~cot

36:

2713-2718,1987. 13. Cosford NDP and Schinazi RF, Selenium nucleophiles

for the preparation

of antiviral nucleosides. 1

Org Chem 56: 2161-2165,1991. 14, Schinazi

RF, Boudinot

deoxythymidine

FD, Doshi

KJ and

McClure

and 3’-deaxy-2’-3’-didehydrothymidine

HM, Pharmacoklnetics in rhesus

of 3’-fluoro-3’Anfimicrob

monkeys.

Agents

Chemofher 34: 1214-1219, 1990, 15. Chu CK, Ahn SK, Kim I-IO, Beach JW, Alves AJ, Jeong LS, Islam Q, Van Roey P and Schinazi RF,

Asymmetric

synthesis

of en~ntiomerically

pure ~-~-~~~R,~~~-dioxolan~-th~ine

and its anti-HIV

activity. Tetrahedron left 32: 3791-3794, 1991. 16, Choi W-B, Wilson

stereochemistry

LJ, Yeola S, Liotta DC and Schinazi

of N-glycosylation

in the synthesis

RF, In situ complexation

of oxathioianyl

and dioxolanyl

directs

the

nucfeoside

analogues. ] Am Chem Sot 113: 9377-9379, 1991. 17. Cretton EM, Schinazi RF, McClure HM, Anderson DC and Sommadossi J-P, Pha~aco~netics azido-~-de~xythymidine

and its catabolites and interactions

with probenecid

of 3’-

in rhesus monkeys.

Anfimicrob Agents Chemafher 35: 801-807, 1991. 18. Stagg MP, Cretton EM, Kidd L, Diasio RB and Sammadossi J-P, Clinical pharmacokinetics

deoxythymidine deoxythymid~ne

fAZT) and catabolites with formation (AMT). Ciin ~~~~1 Ther (in press).

of a toxic catabolite,

of 3’-azido-3’3’-amino-3’-

19. Cretton EM, Xie M-Y, Schinazi RF, Chu CK and Sommadossi J-P? Cataboiic disposition of 3’-azido-2’,3’dideoxyuridine

in hepatocytes with evidence of azido reduction being a general catabolic pathway of 3’-

azido-2’,3’-dideoxynucleosides.

~io~em

Ph~~acul

(in press).

20. Doong S-L, Tsai C-H, Schinazi RF, Liotta DC and Cheng Y-C, Inhibition of the repfication of hepatitis B virus in vitro by ~,~-dideoxy-3’-thiacytld~e

and refated analogues. PYOCNat1 Acad Sci USA 88: 849s

8499,199l. 21. Niedzwicki JG, el Kouni MH, Chu SH and Cha S, Structure-activity

pyrimidine

22. Kono A, Hara Y, Sugata

phospho~lase

S, Matsushima

of ligands of the

Y and Ueda T, Substrate

specificity

of a thymidine

in human Iiver tumor, Gem Phurm Buff ~~0~0~ 32: 1919-1921, 1984.

23. Cooney DA, DalaI M, Mitsuya H, Mc~on studies on the cellular pharmacology gioch~

reIationship

Biochem Pharmacot 32: 399-415, 1983.

nucleoside phosphorylases.

P~rmacol35:

JB, Nadkami M, Balzarini J, &coder S and Johns DG, Initial

of 2’,3’-dldeoxycytidine,

an inhibitor

of HTLV-III infectivity,

2065-2068, 1986.

24. Russell JW, Whiterock VJ, Marrero D and KIunk LJ, Disposition in animals of a new anti-HIV agent:

2’,3’-didehydro-3’-deoxythymidine.

Drug Mefab Dispos IS: 153-157, 1990.