Specificity of Poly(ADP-Ribose) Synthetase Inhibitors
MICHAEL R. PURNELL!, WILLIAM R. KIDWELL2 , LINDSAY MINSHALL3 ,
and WILLIAM J.D. WHISH3
Introduction
Poly(ADP-ribose) synthetase inhibitors, and in particular 3-aminobenzamide, have been used extensively in recent years as probes to elucidate the function of poly(ADPribose) in the cell. Our initial report [1] on substituted benzamides as physiologically specific inhibitors was based on the observation that cells grew at an unchanged rate in the presence of 2 mM 3-aminobenzamide, a concentration 1,000 times the Ki value. In general, high concentrations are needed to elicit a cellular response compared to assays in vitro. Interpretation of results is complicated by the possibility of affecting a target other than poly(ADP-ribose) synthetase. Recently, processes other than ADPribosylation have been reported to be altered by benzarnides. A major criticism of most work is that the studies entail the use of only one inhibitor, usually 3-aminobenzamide.
We have shown that poly(ADP-ribose) synthetase is inhibited by a series of benz amides (Table 1 [1]) and analysis of the structure of the inhibitors revealed that potency of inhibition did not correlate solely with anyone property. On the assymption that another enzyme is unlikely to have the same spectrum of sensitivity as poly(ADPribose) synthetase to these compounds, then by using a series of inhibitors, it should be possible to answer definitively whether poly(ADP-ribose) synthesis regulates a cel-
Table 1. Inhibitors of poly(ADP-ribose) synthetase
Compound Abbreviation Symbol Ki(J,LM)
3-Acetamidobenzamide AAB • 0.4 3-Aminobenzamide AB 0 2.6 Benzamide B • 1.0 3-Hydroxybenzamide HB 0 1.0 3-Methoxybenzamide MB ... 0.6 3-N itrobenzamide NB 6 9.8
Cancer Research Unit, Royal Victoria Infirmary, Newcastle-upon-Tyne NEI 4LP, Great Britain 2 Laboratory of Pathophysiology, NCI, NIH, USA 3 Biochemistry Group, University of Bath, Bath, Great Britain
ADP-Ribosylation of Proteins (ed. by F.R. Althaus, H. Hilz, and S. Shall) © Springer-Verlag Berlin Heidelberg 1985
Specificity of Poly(ADP-Ribose) Synthetase Inhibitors
50
~ 20 g 'u
== :. 10 c 'il Ii.
.~ 5 .. a; 1:1:
Fig.l
2
Concentration (mM)
AB
B HB
100 ... 0 ..
MB
50
~ 20 c .. ~
NB 8' 10 'il
AAB
Ii.
.~ '.i: 5 a; 1:1:
2
Fig. 2
•
5
o .. 0"
10
~
.0
15 20 25
Concentration ki x 103
Fig. 1. Colony formation by CHO-Kl cells after 24 h exposure to benzamides
Fig. 2. Colony formation vs inhibitory potency towards poly(ADP-ribose )synthetase in vitro
99
lular process. For competitive inhibitors such as benzamides, the Km term in the denominator of the Michaelis-Menten equation is modified by (1 + I/Ki)' Different compounds should therefore give the same biological response at the same I/Ki value. Using this criterion, we have measured a number of parameters to determine whether ADP-ribosylation is involved in their regulation.
The Effect of Benzamides on Colony Formation
In order to test for the cytotoxicity of benzarnides, we chose an exposure time of 24 h. This was to minimise any cytostatic, but not cytotoxic lesions. Exponentially growing CHO-Kl cells, doubling time approx. 12-14 h, were added to six well dishes containing different concentrations of inhibitor for 24 h. Colonies were allowed to form during a subsequent 5 day incubation in normal growth medium (Ham's F12 + 10% FCS). Figure 1 shows that the inhibitors displayed a wide range of toxicity. When, however, data were expressed as relative plating efficiency vs poly( ADP-ribose) synthetase inhibition (Fig. 2), a remarkably good correlation was obtained for five of the six inhibitors; 3-nitrobenzamide is clearly different (see Discussion). It is apparent from Fig. 3 that these cells tolerate the presence of 10 mM 3AAB for 8 h with only minimal toxicity, suggesting either the lethal lesion can be repaired or it does not manifest itself except on prolonged exposure.
100
50
20 > u c .!! u
~ 10 8' j ... .. . ~ ... .. 'il =
2
1
Fig. 3
0 4 8 12 24
Hours
30
g 20 '! .. e-.~ .. c :2 f 10 :c ...
Fig. 4
30 Minutes
Fig. 3. Colony formation after treatment with 10 mM 3AAB for different times
Fig. 4. Inhibition of uridine incorporation by 0-5 mM 3AAB
The Effect of Benzamides on RNA Precursor Incorporation
M.R. Purnell et al.
o
2
60
In order to investigate the molecular basis for cytotoxicity, we examined the effect of benzamides on macromolecular syntheses. The most dramatic effect observed was on [3 H]-uridine incorporation (Fig. 4). The reaction time course was non-linear but extrapolation back to the origin gave a common intersection. For comparison of different benzamides, we measured incorporation for 30 min. Figure 5 shows the dose response curves for a number of compounds. 5-Methylnicotinamide also inhibited. It is interesting to note that all compounds gave biphasic curves. At present, we distinguish between two processes being affected or one process having a biphasic sensitivity. Uri dine is not the only nucleoside whose incorporation is inhibited by benzamides. Adenosine incorporation (in the presence of hydroxyurea to inhibit DNA synthesis) is also inhibited by 3-acetoamidobenzamide and the two dose response curves are superimposable (Fig. 6); this suggests that the target is probably common to both pathways of incorporation, since it is unlikely that two different kinases, for example, are inhibited to the same extent by 3-acetamidobenzamide.
Analysis of the inhibition produced by 10 mM benzamides showed very poor correlation with a dose response curve for 3-acetamidobenzamide when the data were normalized for poly(ADP-ribose) synthetase inhibition (Fig. 7 A). 5-Methylnicotinamide (T), 3-nitrobenzamide and 3-methoxybenzamide deviated markedly from the AAB dose response curve. A much better correlation was obtained between inhibition
Specificity of Poly(ADP-Ribose) Synthetase Inhibitors 101
AB
HB MN
c
" c 0.5 .~ " 0.2 '! " ... " l; e-" .. . 5 " 0 .. 0.1 AAB .: c .. ! ....
'w; = .. .. ..
0.2 > '" c:;
'.;::l 0.05 = .. c '11 .. cr:: '" NB >
'.;::l .. "ii cr::
0.02 0.1 0 2 4 6 8 10 0 2 4 6 8 10
Concentration (mM) Concentration (M) Fig. 5 Fig. 6
Fig. 5. Dose response curves for benzamides and 5-methylnicotinamide (MN)
Fig. 6. Inhibition of uridine (e) and adenosine (2 roM hydroxyurea) (0) incorporation by 3 AAB
and electron withdrawal by the substituent group (Harnrnet constant) as can be seen from Fig. 7B. Thus the effects observed are probably not poly(ADP-ribose)-mediated.
Figure 8 shows that one of the targets affected by benzamides is transport of nucleosides. At this preliminary stage, we cannot say whether transport is inhibited in a biphasic manner or another component of RNA metabolism is affected.
c .. '! .. e-.. .5 .. c :; ·c = ~
i cr::
A B
0.5 0
• • o o ,.
0.2 • v •
0.1 .. ..
0.05
0.02 L----'L----''-----'_---'_---I
o 5 10 15
Concentration
ki x 103
20 25 -0.02 0 02 0.4 0.6 0.8
Hammet constant
Fig. 7 A,B. Inhibition of uridine incorporation by 10 roM benzamides vs (A) inhibition of poly(ADP-ribose)synthetase or (B) Hammet constant
102 M.R. Purnell et al.
11 0
to c • " '+:l 0.5 I!
0.8 " e-" NB N ... ... .5
l! 0.6 .. ... c = ;g .. c E :i 0.4
>-
~ oS 0.2 .. : 10 mM > ... '+:l AAB ..
'ii a:
0.1 '--_.L..-._.L..-._'--_L...-----''----'_ 10 20 30 40 50 o 2 3 4 5 6
Fig. 8 Seconds Fig. 9 Hours exposure to inhibitor
Fig. 8. Inhibition of uridine transport by lO mM AAB
Fig. 9. Inhibition of thymidine incorporation by 10 mM AAB or 6 mM NB. Data expressed as incorporation in·30 min compared to controls
The Effect of Benzamides on Thymidine Incorporation
When [3 H]-thymidine incorporation was measured in the presence and absence of 10 mM 3AAB, no significant difference was observed. If, however, cells were preincubated in the presence of 10 mM 3AAB for increasing periods of time, a distinctly biphasic time course was observed (Fig. 9). At present we have no simple explanation for the initial slow decrease in incorporation. Blockage of cell cycle progress in G, at approx. 2 h before S-phase could account for the second portion of the curve. The observation that 3-nitrobenzamide at 6 mM (which gave an approx. equivalent inhibition of uridine incorporation) did not give this effect suggests such a block could be poly(ADP-ribose) synthetase mediated. To test whether this was indeed a G, block, we treated cells with 10 mM AAB for 8 h, a treatment which caused minimal toxicity (Fig. 3), and then released the block and monitored the recovery of [3 H)-thymidine
10
il 8 ... ... " :;; !3 6
1 1! 4 -" " 'il 2 a:
o 2 4
H ours after medium change
Fig. 10. Recovery of thymidine incorporation after treatment with lO mM AAB for 8 h. Medium on treated cultures was replaced with fresh
6 medium + AAB. Labelled thymidine was added for 30 min thereafter
Specificity of Poly(ADP-Ribose) Synthetase Inhibitors 103
Control ••••
A C
B D
DNA/cell
Fig. llA-D. FMF analysis of cell cycle blockage by 10 mM NB: A 4 h treatment; B 8 h; or 10 mM AAB: C4 h;D 8 h
incorporation. The data in Fig. 10, which showed an increase in incorporation 2 h after removal of inhibitor, are consistent with a G1 blockade. A third method, viz. FMF analysis, also showed that entry of cells into S-phase was blocked by 10 mM AAB. In contrast, 10 mM 3NB caused an accumulation of cells in the S-phase (Fig. 11).
The Effect of 3-Aminobenzamide on DNA Methylation
Measurement of methylation of C4 C]-deoxycytidine incorporated into DNA in the presence of 5 mM 3AB showed no significant change (Fig. 12, Table 2). This is in contrast to the report of Morgan and Cleaver [2] who reported inhibition but did not present any experimental data.
104
C U mC 10
8 0 .
C'? Q
6 ><
• '" 4
CI ... Ie
E 2 ~ &oJ
0 5 10 15 20 25 30
Minutes
Table 2. DNA methylation in CHO-Kl cells grown in the presence or absence of 5 mM 3AB for 48 h
Addition mSC/C
0 2.61 2.63 2.66
5 mM3AB 2.63 2.63
Discussion
M.R. Purnell et al.
GT Fig. l2. HPLC resolution of Cyd and m5Cyd following isolation of 14 C-dCyd labelled cellular DNA and enzymic diges-tion to deoxynucleosides. Elution posi-tions of markers are shown
35
It has clearly been shown that inhibition of poly(ADP-ribose) synthetase was responsible for the cytotoxicity exhibited by five of the six benzamides tested. 3-Nitrobenzamide was markedly more cytotoxic than would be expected on the basis of its Ki .
Nitro-compounds have been shown to be cytotoxic and genotoxic. This is thought to arise by reduction to reactive intermediates which can then modify DNA [3]. When uridine incorporation was correlated with the Hammet constant, 3-nitrobenzamide fitted very well. The shorter exposure time in this assay minimises any metabolite effects.
It is of interest to note that such high levels of poly(ADP-ribose) synthetase inhibitors are required to exhibit cytotoxicity. One possible explanation is that in normal cycling cells, the poly(ADP-ribose) mediated step can take place over an extended period of time. For enhancement of cytotoxicity of alkylating agents which generally take place at lower inhibitor concentrations, the timing presumably is more critical.
Although probably not contributing to the cytotoxicity of the benzamides, inhibition of nucleoside transport is nevertheless important. A number of workers have used
Specificity of Poly(ADP-Ribose) Synthetase Inhibitors 105
nucleoside incorporations in conjunction with inhibitors to define the role of ADPribosylation in the cell, e.g. UDS. Caution should, therefore, be exercised when evaluating such results, particularly as there may be synergy between agents which affect membrane function, e.g. bleomycin, and inhibitors of transport.
The most interesting new finding to emerge is the arrest of cells in G1 and G2 by high concentrations of 3-acetamidobenzamide, but not 3-nitrobenzamide. This opens the possibility of synchronizing cells at a poly(ADP-ribose)-mediated restriction point. Release of such a population would allow one to define which events are dependent on release. It should also make isolating cell-cycle specific acceptor proteins from cells considerably easier.
References
1. Purnell MR, Whish WJD (1980) Novel inhibitors of poly(ADP-ribose) synthetase. Biochem J 185:775-777
2. Morgan WF, Cleaver JE (1983) Effect of 3-aminobenzamide on the rate of ligation during repair of alkylated DNA. Cancer Res 43:3104-3107
3. Biaglow JE (1981) Cellular electron transfer and radical mechanisms for drug metabolism. Radiat Res 86:212-242