- Email: [email protected]
Effects of DNA and synthetic oligodeoxyribonucleotides on the binding properties of a cGMP-binding protein from Dictyostelium discoideum
294
Biochimica et Biophysica Acta, 1040 (1990) 294-300 Elsevier
BBAPRO 33721
Effects of DNA and synthetic oligodeoxyribonucleotides on the binding properties of a cGMP-binding protein from Dictyostelium discoideum A m a d e o M. Parissenti * and M. Barrie Coukell Department of Biology, York University, North York (Canada)
(Received 29 January 1990)
Key words: cGMP-bindingprotein; DNA; Syntheticoligodeoxyribonucleotide;(D. discoideum)
Previously, we identified a cGMP-binding protein (cGBP) in Dictyostelium discoideum that can exist in two forms: a fast-dissociating (F-type) activity and a slow-dissociating (S-type) activity. Moreover, the F-type activity was converted effectively to S-type by the addition of nucleic acids, especially DNA (Parissenti, A.M. and Coukell, M B. (1989) J. Cell Sci. 92, 291-301). In this study, we examined the effects of heterologous DNA and various synthetic homooligodeoxyribonucleotides on the cGMP-binding properties of partially purified F-type activity. Equilibrium and kinetic binding experiments revealed that DNA increased the affinity of the protein for cGMP without altering the number of binding sites. However, the presence of DNA decreased only slightly the apparent K d of the protein for cGMP because the nucleic acid also reduced the rate of cGMP association. Addition of oligo(dGMP) s or oligo(dCMP) s to the protein increased both total cGMP binding and the conversion of F-type activity to S-type; in contrast, oligo(dAMP) s or oligo(dTMP) s, at the same concentration, had no effect. Oligodeoxycytidylic acids with chain lengths less than about eight nucleotides were also ineffective or inhibitory. Analysis of cGMP binding to intact, filipin-permeabilized cells revealed a binding activity with association and dissociation rates comparable to isolated S-type activity. This observation suggests that in vivo the cGBP might exist in its S-form.
Introduction Early development of amoebae of Dictyostelium discoideum is regulated by the autonomous, periodic production and secretion of cAMP [1-3]. The secreted cAMP binds to specific receptors on the surface of neighboring cells, and induces these cells to (1) move by chemotaxis to the source of the cAMP signals [3,4] and (2) regulate the expression of specific classes of genes [5-7]. Recently, a number of cellular and molecular processes thought to be associated with the cAMPinduced chemotactic response have been identified, and models for directed cell movement have been proposed [8,9]. In contrast, little is known about the biochemical pathways responsible for the regulation of gene expression by extracellular cAMP (see Ref. 10). Binding of cAMP to the cell-surface receptors results in the production or mobilization of several putative * Present address: Joslin Diabetes Center, Boston, MA, U.S.A.
Correspondence: M.B. Coukell, Department of Biology,York University, 4700 Keele Street, North York, Ontario, Canada M3J 1P3.
intracellular regulators of gene expression including cAMP [111, c G M P [12] and Ca 2+ [13,14]. To date, results of most studies on the role of intracellular c G M P i n early Dictyostelium development indicate that this molecule is involved primarily in chemotaxis [15-17]. However, a few observations suggest that cGMP might also function in the nucleus. For example, using antibodies to cGMP, Mato and Steiner [18] found that bound c G M P was present predominately in the nuclear area of the cells. Also, we recently identified a cGMPbinding protein (cGBP) in extracts of Dictyostelium cells that can exist in two forms: a fast-dissociating, F-type activity (/1/2 < 1 min) and slow-dissociating, Stype activity (tl/2 = approx. 68 min). Interestingly, the F-type activity bound to DNA-cellulose and the isolated F-form could be converted effectively to S-type activity by the addition of low concentrations of heterologous DNA [19]. To examine further the possibility that the cGBP might interact with D N A in vivo (perhaps to regulate gene expression), we have extended our earlier work by analyzing the effects of DNA and synthetic oligodeoxyribonucleotides on the binding properties of partially purified F-type activity. In ad-
016%4838/90/$03.50 © 1990 Elsevier SciencePublishers B.V. (BiomedicalDivision)
295 dition, we have examined the binding of cGMP to permeabilized Dictyostelium cells. Materials and Methods
Materials Chemicals used in this study were obtained from the following sources: [8,5'-3H]cGMP (30-50 Ci/mmol); 1 Ci = 37 GBq) (New England Nuclear); glycerol (BDH); dithiothreitol, Tris, cAMP, cGMP, dCMP, oligo (dCMP)4, oligo(dCMP)s, oligo(dCMP)10_ls, oligo (dCMP)24_36 , oligo(dTMP)8, oligo(dAMP)8, EGTA, DNA (Herring sperm), single-stranded DNA (Calf thymus) cellulose (Sigma); and DEAE-cellulose (DE52) (Whatman). Media components were from Difco or Scott laboratories except for the bacteriological peptone, which was from Oxoid. Oligo(dGMP)8 was synthesized by Dr. Safwat Awadallah (York University) using a Pharmacia gene assembler and purified by FPLC chromatography on a Mono-Q column (Pharmacia) using a 0.3-0.9 M NaC1 gradient (pH 12.0). Fractions corresponding to the oligo(dGMP)8 peak were desalted, lyophilized and stored at - 15 o C.
Growth and developmental conditions All experiments were conducted with D. discoideum strain AX2 [20]. The cells were grown axenically in HL-5 medium [21] at 22°C and harvested at a density of approx. 1.107 cells/ml. To induce cell differentiation, amoebae were washed by centrifugation and plated on PBS agar [22]. All experiments employed slug stage cells (approx. 16 h of development), except for the cell permeabilization studies in which vegetative amoebae were used.
Isolation and analysis of F-type cGMP-binding activity F-type binding activity was isolated from filterbroken cells by DEAE-ceUulose chromatography as described [23], except that the lysates were prepared from 3.0.109 cells suspended in 15 ml of stabilization buffer (20 mM Tris-HCl, 10% glycerol (v/v), 1 mM EGTA, 1 mM dithiothreitol (pH 7.5)). Gradient fractions possessing cGMP-binding activity were pooled and dialyzed overnight against stabilization buffer. In some experiments, the dialyzed F-type activity was fractionated further on a DNA-cellulose column (2.5 ml of packed resin) at 4°C. To do this, 5 ml of eluate was applied slowly to the column, the resin was washed with stabilization buffer and then the cGMP-binding activity was eluted by the addition of 10 ml vols. of 50-500 mM NaCI in stabilization buffer. As shown in Fig. 1, most of the activity eluted in the 100 mM salt wash, well resolved from the bulk of the protein. These fractions were desalted by dialysis (0°C) against stabilization buffer. The association/dissociation kinetics of cGMP binding to F-type activity (and to permeabilized cells) were measured as described previously [19]. Nonspecific [3H]cGMP binding was determined by preincubating the F-type activity (or cells) with 0.5 mM nonradioactive GMP. DNA, when present, was sheared, boiled, cooled on ice and added to the binding reactions to give a concentration of 100 gg/ml.
0
~o.
' .IO~. .;oo. ' J. 50 ." '.150."~-'.300: ",500. o 0.8 o3
0.3
'o
Measurement of F-type and S-type cGMP-binding activity The relative amounts of F-type and S-type activity present in fractionated cell extracts were determined in the following manner. First, total (F-type and S-type) binding activity in each sample was measured with 20 nM [3H]cGMP as described [23]. Then, a second aliquot of each sample was incubated with the same binding cocktail until equilibrium was reached (60 min), 0.5 mM nonradioactive cGMP was added and bound [3H]cGMP was measured 2 min later. Under these conditions, [3H]cGMP bound to fast-dissociating (F-type) sites in the preparation is displaced by unlabeled ligand, and is not detected in the assay. However, [3H]cGMP bound to slow-dissociating (S-type) sites remains bound after the addition of unlabeled ligand. Therefore, the percentage of S-type binding activity (%S) in a sample was calculated by dividing the cpm bound after addition of nonradioactive cGMP by the total cpm bound ( × 100), while the percentage of F-type activity (%F) was 100 %S. All cGMP binding assays were performed in duplicate.
x
o) 0.6
§
ft. 0.40 ~ t~
01 [ ~
•
¢q 0.2 le
0 0
5
,"'.~--~,~.,~ 15 20 Fraction number
10
25
Fig. 1. DNA-cellulos¢ chromatography of F-type cGMP-binding activity. Binding activity from a DEAE-cellulosc column was dialyzed, bound to DNA-cellulose and duted as described in Materials and Methods. Fractions were assayed for cGMP-binding activity ( o ) as described in Ref. 23, and protein (e) by the method of Bradford [35]. The numbers at the top of the figure indicate the salt concentrations (mM) of the washes.
296
Cell permeabilization
calculated from the linear portion of the plot, was 2.1 nM. Thus, DNA appears to increase the affinity of the F-form of the cGBP for cGMP without altering the number of binding sites. The relatively small decrease (2-3-fold) in the K d of the F-type activity upon addition of DNA was unexpected, since nucleic acids had been observed to reduce the tl/2 for cGMP dissociation by more than 100-fold (Ref. 19, A.M. Parissenti, unpublished results). Since the K 0 is a function of both the rates of cGMP association and dissociation, a small change in the Kd is possible if the interaction of DNA with the binding protein reduces both rates. Therefore, we examined the effects of DNA on the initial rates of association (Ra) and dissociation (Ro) of cGMP to partially purified F-type binding activity. Since it was difficult to measure the rates accurately at 20 nM cGMP, these experiments were performed at 1 nM. From the experiment shown in Fig. 3, apparent R a values in the presence and absence of DNA were calculated to be 19.4 and 81.6 fmol cGMP bound/min, respectively. The corresponding values for R d ( + D N A ) and R o ( - D N A ) were 3.0 and 70.0 fmol cGMP released/min, respectively. Thus, the addition of DNA reduced the R a by 4.2-fold and the R a by 23.3-fold. These values are probably underestimated due to the difficulty of measuring accurately initial rates of association and dissociation in the absence of DNA. It should also be noted that under these conditions (1 nM cGMP), the level of cGMP binding at equilibrium was 5-6-fold greater in the presence of DNA than in the absence of DNA (Fig. 3B; 0 min).
Vegetative amoebae (grown to approx. 1.107 cells/ml) were permeabilized with filipin as described by Milne and Coukell [24], except that KCI was omitted from the HMK buffer and the permeabilized cells were washed and resuspended in stabilization buffer to a density of 1.5 • 108 cells/ml. Results
Effects of DNA on the cGMP-binding properties of F-type activity Preliminary experiments revealed that cGMP binding to F-type activity is very complex. Fig. 2A shows the equilibrium binding of 1-17 nM cGMP to partially purified F-type activity in the presence and absence of DNA. A narrow cGMP concentration range was chosen for this analysis, since the level of cGMP binding under both conditions decreased substantially at ligand concentrations above 20 nM. In the presence of DNA, cGMP binding saturated at a lower cGMP concentration than in the absence of DNA. Fig. 2B shows the same data analyzed according to Scatchard [25]. In the absence of DNA, the data produced a linear plot with a slope corresponding to a dissociation constant (K a) of 6.1 nM. However, in the presence of DNA, the plot was only linear for cGMP concentrations up to approx. 7 nM; at higher concentrations, the plot curved sharply downwards and intercepted the x-axis at approximately the same point as the plot in the absence of DNA. The apparent K d for the cGBP in the presence of DNA,
!
ii
i
i
i
!
t
B
A 2.5 o x
2.0
o
oo ~ ° ' " ° ~
o ~3o
1.5
~-rn
ft.
o o = 1.0
o
o o
20
0.5
10
t ;
1'0
15
b.°,,° a.P]. n.
20
,'o
,'o
8'o
Bound
Fig. 2. (A) Equilibrium binding of different concentrations of cGMP to F-type activity in the presence (o) and absence (O) of DNA. Samples of F-type activity, purified through DEAE-cellulose and DNA-cellulose, were assayed for cGMP binding as described [23], except the concentration of [3H]eGMP used in the total and nonspeeific binding assays was varied from 1-17 nM. (B) A Seatchard plot of the data from (A). Bound is the fmol of eGMP bound while free is the free cGMP concentration (in nM). The results are representative of three experiments.
297 6
A
oB
!
J
!
I
~
|
|
I
I
!
O X
E 4 (.,) 'ID e-I O (1.
:[
/
.u 2
/
o
I~,..[
0
i
i
10
|
i
I
20
I
OL i
30 0 Time (rain)
10
20
--~
30
Fig. 3. Effects of DNA on (A) the association of cGMP with and (B) dissociation of cGMP from F-type binding activity. F-type activity was purified through DEAE-ceUulose and DNA-cellulose as described in Materials and Methods. The binding protein was then incubated with 1 nM [3H]cGMP in the presence (o) or absence (O) of DNA. After 60 min, 0.5 mM unlabeled cGMP was added to initiate dissociation. At the times indicated, samples were removed to determine the amount of bound ligand. The results are representative of three experiments.
Effects of oligodeoxyribonucleotides of different chain lengths on cGMP binding by F-type activity It was observed previously [19] that, like DNA, oligodeoxycytidylic acid molecules with chain lengths of 24-36 nucleotides could convert F-type binding activity to S-type and increase slightly total cGMP-binding activity. In contrast, mononucleotides had no effect on the cGMP-binding properties of this protein. To determine the minimum oligonucleotide chain length capable of inducing these changes, preparations of F-type activity were preincubated with oligodeoxycytidylic acids of various lengths and then assayed for total cGMP binding and %S-type activity (see Materials and Methods). Since the effects of nucleic acids on cGMP binding were concentration-dependent (see later), nucleotides and oligonucleotides were added to give a final concentration of 50/~g/ml. This was the lowest concentration of oligo(dCMP) 24- 36 capable of converting 100% of F-type activity to S-type. It can be seen from Fig. 4 that a chain length of eight nucleotides or more was necessary to induce an appreciable conversion of F-type activity to S-type and to increase total cGMP binding. Shorter oligonucleotides were unable to convert F-type activity to S-type, and actually inhibited total cGMP binding (Fig. 4, panel C). With different preparations of F-type activity, even oligonucleotides 8 units in length were sometimes inhibitory (data not shown); in contrast,
A
B
C
D
E
F
Fig. 4. Effects of dCMP and oligodeoxycytidylic acids of different chain lengths on the binding of cGMP to F-type activity. Samples of F-type binding activity, purified through DEAE-cellulose, were preincubated on ice with 50 # g / m l of (B) dCMP, (C) oligo(dCMP)4, (D) oligo(dCMP)s, (E) oligo(dCMP)]0_l~ or (F) oligo(dCMP)24_36. After 1 h, total and S-type binding activities were determined as described in Materials and Methods. In each case, total (left bar) and S-type (right bar) binding activities are expressed relative to an unsupplemented control (A). The numbers above the right bars indicate the percentage of total binding activity that is S-type. Results are an average of values obtained in two experiments.
298 longer oligonucleotides were never inhibitory. These results suggest that for oligodeoxycytidylic acids, a minimum chain length of about eight nucleotides is necessary to enhance cGMP binding by F-type activity.
200
Base specificity of the oligodeoxyribonucleotide-induced alteration of cGMP binding
150
To determine if cGMP binding to F-type activity is modulated preferentially by oligodeoxyribonucleotides containing certain bases, binding activity was preinc u b a t e d with o l i g o ( d G M P ) 8 , o l i g o ( d C M P ) 8, oligo(dTMP)8 or oligo(dAMP)8 (each at a concentration of 25/~g/ml or approx. 10/~M), and then assayed for total binding and %S-type activity. As shown in Fig. 5, only oligo(dCMP)8 and oligo(dGMP)8 were able to convert F-type activity to S-type, and only oligo(dGMP)8 increased appreciably total cGMP-binding activity. Since the experiment presented in Fig. 5 was performed with relatively high concentrations of oligonucleotides, the apparent differences in the effects of oligo(dCMP)8 and oligo(dGMP)8 on cGMP binding might have been due, in part, to oligonucleotide concentration rather than base composition. Therefore, the effects of these oligonucleotides were examined over a lower concentration range. Since the response of F-type activity to oligo(dCMP) 8 was somewhat variable (see above), in these experiments oligo(dGMP)8 was compared to oligo(dCMP)]0_ts. Fig. 6 shows that at all
~5 lO0 t~
50
C
A
B
C
D
E
Fig. 5. Effects of different homooligodeoxyribonucleotides with the same chain length on the binding of cGMP to F-type activity. Samples of F-type activity, purified through DEAE-cellulose, were preincubated with 25/~g/ml (approx. 10/~M) of (B) oligo(dGMP)s, (C) oligo(dCMP) s, (D) oligo(dAMP)8 or (E) oligo(dTMP)s. (A) was an unsupplemented control. After 1 h, total and S-type binding activities were determined as described in Materials and Methods. Results are presented as described in the legend to Fig. 4, and are an average of values obtained in two experiments.
200[
•
O0
A
.
°
.
°
B
75
150L
.~ 50
lOft
g,
0 rr
25
5(2
0
I
0
I
2
I
I
,
0
i
i
4 r O l i g o n u c l e o t i d eJ,.u M
Fig. 6. Effects of increasing concentrations of oligo(dGMP)s and oligo(dCMP)10_ls on cGMP binding to F-type activity. Samples of F-type activity, purified through DEAE-cellulose, were preincubatod for 1 h with various concentrations of oligo(dGMP) 8 (o) or oligo(dCMP)t0_15 (O). Total (A) and S-type (B) binding activities were then determined as described in Materials and Methods. In (A), total binding is expressed relative to an unsupplemented control. This experiment was repeated once with similar results.
299 concentrations examined, oligo(dGMP)8 was more effective than oligo(dCMP)10_ls at converting F-type activity to S-type and stimulating total binding activity. Moreover, oligo(dGMP)8 was able to alter both properties of the cGBP at concentrations as low as 90 nM.
tl/2 for cGMP dissociation in these cells (approx. 70 rain) was very similar to the tl/2 for the dissociation of cGMP from isolated S-type activity (approx. 68 rain) [19] and F-type activity in the presence of DNA (approx. 65 rain, data not shown).
cGMP binding by permeabilized amoebae
Discussion
To determine if a cGMP-binding activity with S-type or F-type properties exists in intact amoebae, we examined the association/dissociation kinetics of cGMP binding by filipin-permeabilized cells. Since cGMPbinding activity in crude cell extracts often exhibits a fast-dissociating component which can be converted to an S-form by nucleic acids [19], these experiments were performed in the presence and absence of exogenous DNA. Also, vegetative (rather than slug stage) cells were used in these experiments because filipin permeabilization at low antibiotic concentrations is more efficient with growth-phase cells [24] and extracts of these cells possess substantial levels of cGMP-binding activity [23]. To reduce the possibility of cell lysis, cGMP association and dissociation were only followed for 15 min. Under these conditions, there was no detectable cell lysis during the permeabilization procedure or cGMP binding. As shown in Fig. 7, cGMP binding by permeabilized cells after incubation for 15 min was 6-7-fold higher than by nonpermeabilized cells. In the permeabilized cells, the rates of both cGMP association and dissociation were relatively slow, and they were unaffected by the presence of DNA. Interestingly, the
5f X
Assoc.
4
Dissoc. !
"
o
o 0
g
i 0
10
20 Time (rain)
30
Fig. 7. Association/dissociation kinetics of cGMP binding by permeabilized cells. Vegetative amoebae were permeabilized with filipin, washed in stabilization buffer and the kinetics of cGMP association and dissociation were analyzed in the presence ( o ) or absence (e) or exogenous D N A as described [19]. Dissociation was initiated after 16 min of incubation by the addition of 0.5 m M unlabeled cGMP (dashed line). (11) indicates the amount of cGMP bound by nonpermeabilized cells after 15 rain. Similar results were obtained in two additional experiments.
It was observed previously [19] that nucleic acids, and, in particular DNA, greatly reduced the rate of dissociation of cGMP from the F-form of the cGBP. Therefore, we examined in more detail the effects of DNA and synthetic homooligodeoxyribonucleotides on the cGMP-binding properties of this protein. Although binding of cGMP to the cGBP is complex, results of equilibrium binding studies indicated that, at low ligand concentrations (1-20 nM), DNA increases appreciably the affinity of the protein for cGMP without changing the number of binding sites (Fig. 2). Interestingly, the dramatic reduction in dissociation rate and the increase in affinity of the cGBP for cGMP in the presence of DNA are accompanied by only a small decrease (2-3fold) in the apparent K d of the protein for cGMP. This can probably be explained by the finding that DNA also reduces considerably the rate of cGMP binding to the protein (Fig. 3). Thus, interaction of the cGBP with nucleic acid appears to induce a conformational change in the cGMP-binding site(s) of the protein that restricts the rate of both cGMP association and dissociation. Although the binding properties of the cGBP are altered by RNA and heterologous DNA, and the protein binds to heterologous DNA-cellulose (Fig. 1), there is some evidence that the cGBP might recognize specific DNA sequences. For example, (1) DNA converted Ftype activity to S-type much more efficiently than RNA when added at the same concentration [19]. (2) Only oligodeoxycytidylic acids with chain lengths of about eight nucleotides or more were able to transform F-type activity to S-type and increase total cGMP binding (Fig. 4). In fact, shorter oligonucleotides (but not mononucleotides) consistently inhibited cGMP binding. This latter observation suggests that the short oligonucleotides transform the protein into a conformation with fewer binding sites a n d / o r reduced affinity for cGMP. (3) At low homoohgodeoxyribonucleotide concentrations, only oligo(dGMP) and oligo(dCMP) were able to alter the properties of the binding protein, and oligo(dGMP) was most effective (Figs. 5 and 6). Together, these observations suggest that the cGBP might possess a binding site which recognizes G-C rich regions of DNA, eight nucleotides or more in length. The fact that the cGMP-binding properties of this protein are altered by RNA and heterologous DNA does not preclude the existence of a specific DNA-binding site; many sequence-specific DNA-binding proteins bind nucleic acids and other polyanions nonspecifically [26].
300 It should be noted that although the F-type activity appears to be modulated preferentially by G-C rich DNA sequences, there is no direct evidence at present that the protein preferentially binds to these sequences. Our studies on cGMP binding to filipin-permeabilized cells suggest that the cells possess a binding activity with slow association and dissociation kinetics (Fig. 7). This activity might correspond to the S-form of the cGBP because the tl/2s for cGMP dissociation from the two binding activities are very similar (i.e., approx. 70 rain). What is the role of this cGMP-binding activity in Dictyostelium? It probably does not function to mediate rapid cellular response to the extracellular cAMP signals, such as chemotaxis, because the rates of binding and dissociation are too slow [3,4,27,28]. These responses might be regulated by other cGMP-binding proteins [19]. However, the slow-associating/dissociating activity could function to mediate long-term responses to a series of cAMP signals, such as gene expression. A nuclear location for this activity is suggested by immunohistochemical studies which detected a slow-dissociating cGMP-binding activity in nuclei of these cells [18], and cells of another cellular slime mold, Polysphondylium violaceum [29]. Interestingly, the genome of Dictyostelium is very A-T rich except for coding sequences and putative regulatory regions [30]. Recently, it has been found that normal regulation of expression of two cysteine proteinase genes by extracellular cAMP is dependent on the presence of certain G-rich, enhancer-like elements [31-33]. Similar G-rich sequences are also present upstream of other Dictyostelium genes (see Ref. 34). Due to the dramatic effects of D N A on the cGMP-binding properties of the cGBP and the likelihood that cGMP plays some regulatory role in these cells, it is tempting the speculate that this protein might function to modulate gene expression by interacting with certain of these G-rich sequences, and perhaps other transcription factors. We are presently attempting to purify the cGBP and to examine its interactions with endogenous putative regulatory sequences. Acknowledgements A.M.P. was a recipient of an Ontario Graduate Scholarship. This work was supported by a grant from the NSERC of Canada.
References 1 Shaffer, B.M. (1975) Nature 255, 549-552. 2 Darmon, M., Brachet, P. and Pereira Da Silva, L.H. (1975) Proc. Natl. Acad. Sci. USA 72, 3163-3166. 3 Gerisch, G., Hi~lser, D., Malchow, D. and Wick, U. (1975) Phil. Trans. Roy, Soc. London B 272, 181-192. 4 Devreotes, P.N. (1982) in The Development of Dictyostelium discoideum (Loomis, W.F., ed.), pp. 117-168, Academic Press, New York. 5 Kimmel, A.R. and Carlisle, B. (1986) Proc. Natl. Acad. Sci. USA 83, 2506-2510. 6 Mann, S.K.O. and Firtel, R.A. (1987) Mol. Cell. Biol. 7, 458-469. 7 Franke, J., Podgorski, G.J. and Kessin, R.H. (1987) Dev. Biol. 124, 504-511. 8 Devreotes, P.N. and Zigmond, S.H. (1988) Annu, Rev. Cell Biol. 4, 649-686. 9 Spudich, J.A. (1989) Cell Regul. 1, 1-11. 10 Mann, S.K.O., Pinko, C. and Firtel, R.A. (1988) Dev. Biol. 130, 294-303. 11 Gerisch, G. and Wick, U. (1975) Biochem. Biophys. Res. Commun. 65, 364-370. 12 Mato, J.M., Krens, F.A., Van Haastert, P.J.M. and Konijn, T.M. (1977) Proc. Natl. Acad. Sci. USA 74, 2348-2351. 13 Bumann, J., Wurster, B. and Malchow, D. (1984) J. Cell Biol. 98, 173-178. 14 Abe, T., Maeda, Y. and Iijima, T. (1988) Differentiation 39, 90-96. 15 Mato, J.M., Krens, F.A., Van Haastert, P.J.M. and Konijn, T.M. (1977) Biochem. Biophys. Res. Commun. 77, 399-402. 16 Ross, F.M. and New~,ll, P.C. (1981) J. Gen. Microbiol. 127, 339-350. 17 Liu, G. and Newell, P.C. (1988) J. Cell Sci. 90, 123-129. 18 Mato, J.M. and Steiner, A.L. (1980) Cell Biol. Int. Rep. 4, 641-648. 19 Parissenti, A.M. and Coukell, M.B. (1989) J. Cell Sci. 92, 291-301. 20 Watts, D.J. and Ashworth, J.M. (1970) Biochem. J. 119, 171-174. 21 Cocucci, S. and Sussman, M. (1970) J. Cell Biol. 45, 399-407. 22 Coukell, M.B., Cameron, A.M., Petrie, C.M. and Mee, J.D. (1984) Dev. Biol. 103, 246-257. 23 Parissenti, A.M. and Coukell, M.B. (1986) Biochem. Cell Biol. 64, 528-534. 24 Milne, J.L. and Coukell, M.B. (1989) Exp. Cell Res. 185, 21-32. 25 Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 51,660-672. 26 Kumar, S.A. and Dickerman, H.W. (1983) in Biochemical Actions of Hormones (Litwack, G., ed.), Vol. 10, pp. 259-301, Academic Press, New York. 27 Wurster, B., Schubiger, K., Wick, U. and Gerisch, G. (1977) FEBS Lett. 76, 141-143. 28 Van Haastert, P.J.M., Van Walsum, H. and Pasveer, F.J. (1982) J. Cell Biol. 94, 271-278. 29 Pan, P. and Wedner, H.J. (1979) Differentiation 14, 113-118. 30 Kimmel, A.R. and Firtel, R.A. (1982) in The Development of Dictyostelium discoideum (Loomis, W.F., ed.), pp. 233-324, Academic Press, New York. 31 Datta, S. and Firtel, R.A. (1987) Mol. Cell. Biol. 7, 149-159. 32 Pears, C.J. and Williams, J.G. (1987) EMBO J. 6, 195-200. 33 Driscoll, D.M., Pears, C.J. and Williams, J.G. (1988) Dev. Genet. 9, 455-468. 34 Podgorski, G.J., Franke, J., Faure, M. and Kessin, R.H. (1989) Mol. Cell. Biol. 9, 3938-3950. 35 Bradford, M.M. (1976) Anal. Biochem. 72, 248-254.
Biochimica et Biophysica Acta, 1040 (1990) 294-300 Elsevier
BBAPRO 33721
Effects of DNA and synthetic oligodeoxyribonucleotides on the binding properties of a cGMP-binding protein from Dictyostelium discoideum A m a d e o M. Parissenti * and M. Barrie Coukell Department of Biology, York University, North York (Canada)
(Received 29 January 1990)
Key words: cGMP-bindingprotein; DNA; Syntheticoligodeoxyribonucleotide;(D. discoideum)
Previously, we identified a cGMP-binding protein (cGBP) in Dictyostelium discoideum that can exist in two forms: a fast-dissociating (F-type) activity and a slow-dissociating (S-type) activity. Moreover, the F-type activity was converted effectively to S-type by the addition of nucleic acids, especially DNA (Parissenti, A.M. and Coukell, M B. (1989) J. Cell Sci. 92, 291-301). In this study, we examined the effects of heterologous DNA and various synthetic homooligodeoxyribonucleotides on the cGMP-binding properties of partially purified F-type activity. Equilibrium and kinetic binding experiments revealed that DNA increased the affinity of the protein for cGMP without altering the number of binding sites. However, the presence of DNA decreased only slightly the apparent K d of the protein for cGMP because the nucleic acid also reduced the rate of cGMP association. Addition of oligo(dGMP) s or oligo(dCMP) s to the protein increased both total cGMP binding and the conversion of F-type activity to S-type; in contrast, oligo(dAMP) s or oligo(dTMP) s, at the same concentration, had no effect. Oligodeoxycytidylic acids with chain lengths less than about eight nucleotides were also ineffective or inhibitory. Analysis of cGMP binding to intact, filipin-permeabilized cells revealed a binding activity with association and dissociation rates comparable to isolated S-type activity. This observation suggests that in vivo the cGBP might exist in its S-form.
Introduction Early development of amoebae of Dictyostelium discoideum is regulated by the autonomous, periodic production and secretion of cAMP [1-3]. The secreted cAMP binds to specific receptors on the surface of neighboring cells, and induces these cells to (1) move by chemotaxis to the source of the cAMP signals [3,4] and (2) regulate the expression of specific classes of genes [5-7]. Recently, a number of cellular and molecular processes thought to be associated with the cAMPinduced chemotactic response have been identified, and models for directed cell movement have been proposed [8,9]. In contrast, little is known about the biochemical pathways responsible for the regulation of gene expression by extracellular cAMP (see Ref. 10). Binding of cAMP to the cell-surface receptors results in the production or mobilization of several putative * Present address: Joslin Diabetes Center, Boston, MA, U.S.A.
Correspondence: M.B. Coukell, Department of Biology,York University, 4700 Keele Street, North York, Ontario, Canada M3J 1P3.
intracellular regulators of gene expression including cAMP [111, c G M P [12] and Ca 2+ [13,14]. To date, results of most studies on the role of intracellular c G M P i n early Dictyostelium development indicate that this molecule is involved primarily in chemotaxis [15-17]. However, a few observations suggest that cGMP might also function in the nucleus. For example, using antibodies to cGMP, Mato and Steiner [18] found that bound c G M P was present predominately in the nuclear area of the cells. Also, we recently identified a cGMPbinding protein (cGBP) in extracts of Dictyostelium cells that can exist in two forms: a fast-dissociating, F-type activity (/1/2 < 1 min) and slow-dissociating, Stype activity (tl/2 = approx. 68 min). Interestingly, the F-type activity bound to DNA-cellulose and the isolated F-form could be converted effectively to S-type activity by the addition of low concentrations of heterologous DNA [19]. To examine further the possibility that the cGBP might interact with D N A in vivo (perhaps to regulate gene expression), we have extended our earlier work by analyzing the effects of DNA and synthetic oligodeoxyribonucleotides on the binding properties of partially purified F-type activity. In ad-
016%4838/90/$03.50 © 1990 Elsevier SciencePublishers B.V. (BiomedicalDivision)
295 dition, we have examined the binding of cGMP to permeabilized Dictyostelium cells. Materials and Methods
Materials Chemicals used in this study were obtained from the following sources: [8,5'-3H]cGMP (30-50 Ci/mmol); 1 Ci = 37 GBq) (New England Nuclear); glycerol (BDH); dithiothreitol, Tris, cAMP, cGMP, dCMP, oligo (dCMP)4, oligo(dCMP)s, oligo(dCMP)10_ls, oligo (dCMP)24_36 , oligo(dTMP)8, oligo(dAMP)8, EGTA, DNA (Herring sperm), single-stranded DNA (Calf thymus) cellulose (Sigma); and DEAE-cellulose (DE52) (Whatman). Media components were from Difco or Scott laboratories except for the bacteriological peptone, which was from Oxoid. Oligo(dGMP)8 was synthesized by Dr. Safwat Awadallah (York University) using a Pharmacia gene assembler and purified by FPLC chromatography on a Mono-Q column (Pharmacia) using a 0.3-0.9 M NaC1 gradient (pH 12.0). Fractions corresponding to the oligo(dGMP)8 peak were desalted, lyophilized and stored at - 15 o C.
Growth and developmental conditions All experiments were conducted with D. discoideum strain AX2 [20]. The cells were grown axenically in HL-5 medium [21] at 22°C and harvested at a density of approx. 1.107 cells/ml. To induce cell differentiation, amoebae were washed by centrifugation and plated on PBS agar [22]. All experiments employed slug stage cells (approx. 16 h of development), except for the cell permeabilization studies in which vegetative amoebae were used.
Isolation and analysis of F-type cGMP-binding activity F-type binding activity was isolated from filterbroken cells by DEAE-ceUulose chromatography as described [23], except that the lysates were prepared from 3.0.109 cells suspended in 15 ml of stabilization buffer (20 mM Tris-HCl, 10% glycerol (v/v), 1 mM EGTA, 1 mM dithiothreitol (pH 7.5)). Gradient fractions possessing cGMP-binding activity were pooled and dialyzed overnight against stabilization buffer. In some experiments, the dialyzed F-type activity was fractionated further on a DNA-cellulose column (2.5 ml of packed resin) at 4°C. To do this, 5 ml of eluate was applied slowly to the column, the resin was washed with stabilization buffer and then the cGMP-binding activity was eluted by the addition of 10 ml vols. of 50-500 mM NaCI in stabilization buffer. As shown in Fig. 1, most of the activity eluted in the 100 mM salt wash, well resolved from the bulk of the protein. These fractions were desalted by dialysis (0°C) against stabilization buffer. The association/dissociation kinetics of cGMP binding to F-type activity (and to permeabilized cells) were measured as described previously [19]. Nonspecific [3H]cGMP binding was determined by preincubating the F-type activity (or cells) with 0.5 mM nonradioactive GMP. DNA, when present, was sheared, boiled, cooled on ice and added to the binding reactions to give a concentration of 100 gg/ml.
0
~o.
' .IO~. .;oo. ' J. 50 ." '.150."~-'.300: ",500. o 0.8 o3
0.3
'o
Measurement of F-type and S-type cGMP-binding activity The relative amounts of F-type and S-type activity present in fractionated cell extracts were determined in the following manner. First, total (F-type and S-type) binding activity in each sample was measured with 20 nM [3H]cGMP as described [23]. Then, a second aliquot of each sample was incubated with the same binding cocktail until equilibrium was reached (60 min), 0.5 mM nonradioactive cGMP was added and bound [3H]cGMP was measured 2 min later. Under these conditions, [3H]cGMP bound to fast-dissociating (F-type) sites in the preparation is displaced by unlabeled ligand, and is not detected in the assay. However, [3H]cGMP bound to slow-dissociating (S-type) sites remains bound after the addition of unlabeled ligand. Therefore, the percentage of S-type binding activity (%S) in a sample was calculated by dividing the cpm bound after addition of nonradioactive cGMP by the total cpm bound ( × 100), while the percentage of F-type activity (%F) was 100 %S. All cGMP binding assays were performed in duplicate.
x
o) 0.6
§
ft. 0.40 ~ t~
01 [ ~
•
¢q 0.2 le
0 0
5
,"'.~--~,~.,~ 15 20 Fraction number
10
25
Fig. 1. DNA-cellulos¢ chromatography of F-type cGMP-binding activity. Binding activity from a DEAE-cellulosc column was dialyzed, bound to DNA-cellulose and duted as described in Materials and Methods. Fractions were assayed for cGMP-binding activity ( o ) as described in Ref. 23, and protein (e) by the method of Bradford [35]. The numbers at the top of the figure indicate the salt concentrations (mM) of the washes.
296
Cell permeabilization
calculated from the linear portion of the plot, was 2.1 nM. Thus, DNA appears to increase the affinity of the F-form of the cGBP for cGMP without altering the number of binding sites. The relatively small decrease (2-3-fold) in the K d of the F-type activity upon addition of DNA was unexpected, since nucleic acids had been observed to reduce the tl/2 for cGMP dissociation by more than 100-fold (Ref. 19, A.M. Parissenti, unpublished results). Since the K 0 is a function of both the rates of cGMP association and dissociation, a small change in the Kd is possible if the interaction of DNA with the binding protein reduces both rates. Therefore, we examined the effects of DNA on the initial rates of association (Ra) and dissociation (Ro) of cGMP to partially purified F-type binding activity. Since it was difficult to measure the rates accurately at 20 nM cGMP, these experiments were performed at 1 nM. From the experiment shown in Fig. 3, apparent R a values in the presence and absence of DNA were calculated to be 19.4 and 81.6 fmol cGMP bound/min, respectively. The corresponding values for R d ( + D N A ) and R o ( - D N A ) were 3.0 and 70.0 fmol cGMP released/min, respectively. Thus, the addition of DNA reduced the R a by 4.2-fold and the R a by 23.3-fold. These values are probably underestimated due to the difficulty of measuring accurately initial rates of association and dissociation in the absence of DNA. It should also be noted that under these conditions (1 nM cGMP), the level of cGMP binding at equilibrium was 5-6-fold greater in the presence of DNA than in the absence of DNA (Fig. 3B; 0 min).
Vegetative amoebae (grown to approx. 1.107 cells/ml) were permeabilized with filipin as described by Milne and Coukell [24], except that KCI was omitted from the HMK buffer and the permeabilized cells were washed and resuspended in stabilization buffer to a density of 1.5 • 108 cells/ml. Results
Effects of DNA on the cGMP-binding properties of F-type activity Preliminary experiments revealed that cGMP binding to F-type activity is very complex. Fig. 2A shows the equilibrium binding of 1-17 nM cGMP to partially purified F-type activity in the presence and absence of DNA. A narrow cGMP concentration range was chosen for this analysis, since the level of cGMP binding under both conditions decreased substantially at ligand concentrations above 20 nM. In the presence of DNA, cGMP binding saturated at a lower cGMP concentration than in the absence of DNA. Fig. 2B shows the same data analyzed according to Scatchard [25]. In the absence of DNA, the data produced a linear plot with a slope corresponding to a dissociation constant (K a) of 6.1 nM. However, in the presence of DNA, the plot was only linear for cGMP concentrations up to approx. 7 nM; at higher concentrations, the plot curved sharply downwards and intercepted the x-axis at approximately the same point as the plot in the absence of DNA. The apparent K d for the cGBP in the presence of DNA,
!
ii
i
i
i
!
t
B
A 2.5 o x
2.0
o
oo ~ ° ' " ° ~
o ~3o
1.5
~-rn
ft.
o o = 1.0
o
o o
20
0.5
10
t ;
1'0
15
b.°,,° a.P]. n.
20
,'o
,'o
8'o
Bound
Fig. 2. (A) Equilibrium binding of different concentrations of cGMP to F-type activity in the presence (o) and absence (O) of DNA. Samples of F-type activity, purified through DEAE-cellulose and DNA-cellulose, were assayed for cGMP binding as described [23], except the concentration of [3H]eGMP used in the total and nonspeeific binding assays was varied from 1-17 nM. (B) A Seatchard plot of the data from (A). Bound is the fmol of eGMP bound while free is the free cGMP concentration (in nM). The results are representative of three experiments.
297 6
A
oB
!
J
!
I
~
|
|
I
I
!
O X
E 4 (.,) 'ID e-I O (1.
:[
/
.u 2
/
o
I~,..[
0
i
i
10
|
i
I
20
I
OL i
30 0 Time (rain)
10
20
--~
30
Fig. 3. Effects of DNA on (A) the association of cGMP with and (B) dissociation of cGMP from F-type binding activity. F-type activity was purified through DEAE-ceUulose and DNA-cellulose as described in Materials and Methods. The binding protein was then incubated with 1 nM [3H]cGMP in the presence (o) or absence (O) of DNA. After 60 min, 0.5 mM unlabeled cGMP was added to initiate dissociation. At the times indicated, samples were removed to determine the amount of bound ligand. The results are representative of three experiments.
Effects of oligodeoxyribonucleotides of different chain lengths on cGMP binding by F-type activity It was observed previously [19] that, like DNA, oligodeoxycytidylic acid molecules with chain lengths of 24-36 nucleotides could convert F-type binding activity to S-type and increase slightly total cGMP-binding activity. In contrast, mononucleotides had no effect on the cGMP-binding properties of this protein. To determine the minimum oligonucleotide chain length capable of inducing these changes, preparations of F-type activity were preincubated with oligodeoxycytidylic acids of various lengths and then assayed for total cGMP binding and %S-type activity (see Materials and Methods). Since the effects of nucleic acids on cGMP binding were concentration-dependent (see later), nucleotides and oligonucleotides were added to give a final concentration of 50/~g/ml. This was the lowest concentration of oligo(dCMP) 24- 36 capable of converting 100% of F-type activity to S-type. It can be seen from Fig. 4 that a chain length of eight nucleotides or more was necessary to induce an appreciable conversion of F-type activity to S-type and to increase total cGMP binding. Shorter oligonucleotides were unable to convert F-type activity to S-type, and actually inhibited total cGMP binding (Fig. 4, panel C). With different preparations of F-type activity, even oligonucleotides 8 units in length were sometimes inhibitory (data not shown); in contrast,
A
B
C
D
E
F
Fig. 4. Effects of dCMP and oligodeoxycytidylic acids of different chain lengths on the binding of cGMP to F-type activity. Samples of F-type binding activity, purified through DEAE-cellulose, were preincubated on ice with 50 # g / m l of (B) dCMP, (C) oligo(dCMP)4, (D) oligo(dCMP)s, (E) oligo(dCMP)]0_l~ or (F) oligo(dCMP)24_36. After 1 h, total and S-type binding activities were determined as described in Materials and Methods. In each case, total (left bar) and S-type (right bar) binding activities are expressed relative to an unsupplemented control (A). The numbers above the right bars indicate the percentage of total binding activity that is S-type. Results are an average of values obtained in two experiments.
298 longer oligonucleotides were never inhibitory. These results suggest that for oligodeoxycytidylic acids, a minimum chain length of about eight nucleotides is necessary to enhance cGMP binding by F-type activity.
200
Base specificity of the oligodeoxyribonucleotide-induced alteration of cGMP binding
150
To determine if cGMP binding to F-type activity is modulated preferentially by oligodeoxyribonucleotides containing certain bases, binding activity was preinc u b a t e d with o l i g o ( d G M P ) 8 , o l i g o ( d C M P ) 8, oligo(dTMP)8 or oligo(dAMP)8 (each at a concentration of 25/~g/ml or approx. 10/~M), and then assayed for total binding and %S-type activity. As shown in Fig. 5, only oligo(dCMP)8 and oligo(dGMP)8 were able to convert F-type activity to S-type, and only oligo(dGMP)8 increased appreciably total cGMP-binding activity. Since the experiment presented in Fig. 5 was performed with relatively high concentrations of oligonucleotides, the apparent differences in the effects of oligo(dCMP)8 and oligo(dGMP)8 on cGMP binding might have been due, in part, to oligonucleotide concentration rather than base composition. Therefore, the effects of these oligonucleotides were examined over a lower concentration range. Since the response of F-type activity to oligo(dCMP) 8 was somewhat variable (see above), in these experiments oligo(dGMP)8 was compared to oligo(dCMP)]0_ts. Fig. 6 shows that at all
~5 lO0 t~
50
C
A
B
C
D
E
Fig. 5. Effects of different homooligodeoxyribonucleotides with the same chain length on the binding of cGMP to F-type activity. Samples of F-type activity, purified through DEAE-cellulose, were preincubated with 25/~g/ml (approx. 10/~M) of (B) oligo(dGMP)s, (C) oligo(dCMP) s, (D) oligo(dAMP)8 or (E) oligo(dTMP)s. (A) was an unsupplemented control. After 1 h, total and S-type binding activities were determined as described in Materials and Methods. Results are presented as described in the legend to Fig. 4, and are an average of values obtained in two experiments.
200[
•
O0
A
.
°
.
°
B
75
150L
.~ 50
lOft
g,
0 rr
25
5(2
0
I
0
I
2
I
I
,
0
i
i
4 r O l i g o n u c l e o t i d eJ,.u M
Fig. 6. Effects of increasing concentrations of oligo(dGMP)s and oligo(dCMP)10_ls on cGMP binding to F-type activity. Samples of F-type activity, purified through DEAE-cellulose, were preincubatod for 1 h with various concentrations of oligo(dGMP) 8 (o) or oligo(dCMP)t0_15 (O). Total (A) and S-type (B) binding activities were then determined as described in Materials and Methods. In (A), total binding is expressed relative to an unsupplemented control. This experiment was repeated once with similar results.
299 concentrations examined, oligo(dGMP)8 was more effective than oligo(dCMP)10_ls at converting F-type activity to S-type and stimulating total binding activity. Moreover, oligo(dGMP)8 was able to alter both properties of the cGBP at concentrations as low as 90 nM.
tl/2 for cGMP dissociation in these cells (approx. 70 rain) was very similar to the tl/2 for the dissociation of cGMP from isolated S-type activity (approx. 68 rain) [19] and F-type activity in the presence of DNA (approx. 65 rain, data not shown).
cGMP binding by permeabilized amoebae
Discussion
To determine if a cGMP-binding activity with S-type or F-type properties exists in intact amoebae, we examined the association/dissociation kinetics of cGMP binding by filipin-permeabilized cells. Since cGMPbinding activity in crude cell extracts often exhibits a fast-dissociating component which can be converted to an S-form by nucleic acids [19], these experiments were performed in the presence and absence of exogenous DNA. Also, vegetative (rather than slug stage) cells were used in these experiments because filipin permeabilization at low antibiotic concentrations is more efficient with growth-phase cells [24] and extracts of these cells possess substantial levels of cGMP-binding activity [23]. To reduce the possibility of cell lysis, cGMP association and dissociation were only followed for 15 min. Under these conditions, there was no detectable cell lysis during the permeabilization procedure or cGMP binding. As shown in Fig. 7, cGMP binding by permeabilized cells after incubation for 15 min was 6-7-fold higher than by nonpermeabilized cells. In the permeabilized cells, the rates of both cGMP association and dissociation were relatively slow, and they were unaffected by the presence of DNA. Interestingly, the
5f X
Assoc.
4
Dissoc. !
"
o
o 0
g
i 0
10
20 Time (rain)
30
Fig. 7. Association/dissociation kinetics of cGMP binding by permeabilized cells. Vegetative amoebae were permeabilized with filipin, washed in stabilization buffer and the kinetics of cGMP association and dissociation were analyzed in the presence ( o ) or absence (e) or exogenous D N A as described [19]. Dissociation was initiated after 16 min of incubation by the addition of 0.5 m M unlabeled cGMP (dashed line). (11) indicates the amount of cGMP bound by nonpermeabilized cells after 15 rain. Similar results were obtained in two additional experiments.
It was observed previously [19] that nucleic acids, and, in particular DNA, greatly reduced the rate of dissociation of cGMP from the F-form of the cGBP. Therefore, we examined in more detail the effects of DNA and synthetic homooligodeoxyribonucleotides on the cGMP-binding properties of this protein. Although binding of cGMP to the cGBP is complex, results of equilibrium binding studies indicated that, at low ligand concentrations (1-20 nM), DNA increases appreciably the affinity of the protein for cGMP without changing the number of binding sites (Fig. 2). Interestingly, the dramatic reduction in dissociation rate and the increase in affinity of the cGBP for cGMP in the presence of DNA are accompanied by only a small decrease (2-3fold) in the apparent K d of the protein for cGMP. This can probably be explained by the finding that DNA also reduces considerably the rate of cGMP binding to the protein (Fig. 3). Thus, interaction of the cGBP with nucleic acid appears to induce a conformational change in the cGMP-binding site(s) of the protein that restricts the rate of both cGMP association and dissociation. Although the binding properties of the cGBP are altered by RNA and heterologous DNA, and the protein binds to heterologous DNA-cellulose (Fig. 1), there is some evidence that the cGBP might recognize specific DNA sequences. For example, (1) DNA converted Ftype activity to S-type much more efficiently than RNA when added at the same concentration [19]. (2) Only oligodeoxycytidylic acids with chain lengths of about eight nucleotides or more were able to transform F-type activity to S-type and increase total cGMP binding (Fig. 4). In fact, shorter oligonucleotides (but not mononucleotides) consistently inhibited cGMP binding. This latter observation suggests that the short oligonucleotides transform the protein into a conformation with fewer binding sites a n d / o r reduced affinity for cGMP. (3) At low homoohgodeoxyribonucleotide concentrations, only oligo(dGMP) and oligo(dCMP) were able to alter the properties of the binding protein, and oligo(dGMP) was most effective (Figs. 5 and 6). Together, these observations suggest that the cGBP might possess a binding site which recognizes G-C rich regions of DNA, eight nucleotides or more in length. The fact that the cGMP-binding properties of this protein are altered by RNA and heterologous DNA does not preclude the existence of a specific DNA-binding site; many sequence-specific DNA-binding proteins bind nucleic acids and other polyanions nonspecifically [26].
300 It should be noted that although the F-type activity appears to be modulated preferentially by G-C rich DNA sequences, there is no direct evidence at present that the protein preferentially binds to these sequences. Our studies on cGMP binding to filipin-permeabilized cells suggest that the cells possess a binding activity with slow association and dissociation kinetics (Fig. 7). This activity might correspond to the S-form of the cGBP because the tl/2s for cGMP dissociation from the two binding activities are very similar (i.e., approx. 70 rain). What is the role of this cGMP-binding activity in Dictyostelium? It probably does not function to mediate rapid cellular response to the extracellular cAMP signals, such as chemotaxis, because the rates of binding and dissociation are too slow [3,4,27,28]. These responses might be regulated by other cGMP-binding proteins [19]. However, the slow-associating/dissociating activity could function to mediate long-term responses to a series of cAMP signals, such as gene expression. A nuclear location for this activity is suggested by immunohistochemical studies which detected a slow-dissociating cGMP-binding activity in nuclei of these cells [18], and cells of another cellular slime mold, Polysphondylium violaceum [29]. Interestingly, the genome of Dictyostelium is very A-T rich except for coding sequences and putative regulatory regions [30]. Recently, it has been found that normal regulation of expression of two cysteine proteinase genes by extracellular cAMP is dependent on the presence of certain G-rich, enhancer-like elements [31-33]. Similar G-rich sequences are also present upstream of other Dictyostelium genes (see Ref. 34). Due to the dramatic effects of D N A on the cGMP-binding properties of the cGBP and the likelihood that cGMP plays some regulatory role in these cells, it is tempting the speculate that this protein might function to modulate gene expression by interacting with certain of these G-rich sequences, and perhaps other transcription factors. We are presently attempting to purify the cGBP and to examine its interactions with endogenous putative regulatory sequences. Acknowledgements A.M.P. was a recipient of an Ontario Graduate Scholarship. This work was supported by a grant from the NSERC of Canada.
References 1 Shaffer, B.M. (1975) Nature 255, 549-552. 2 Darmon, M., Brachet, P. and Pereira Da Silva, L.H. (1975) Proc. Natl. Acad. Sci. USA 72, 3163-3166. 3 Gerisch, G., Hi~lser, D., Malchow, D. and Wick, U. (1975) Phil. Trans. Roy, Soc. London B 272, 181-192. 4 Devreotes, P.N. (1982) in The Development of Dictyostelium discoideum (Loomis, W.F., ed.), pp. 117-168, Academic Press, New York. 5 Kimmel, A.R. and Carlisle, B. (1986) Proc. Natl. Acad. Sci. USA 83, 2506-2510. 6 Mann, S.K.O. and Firtel, R.A. (1987) Mol. Cell. Biol. 7, 458-469. 7 Franke, J., Podgorski, G.J. and Kessin, R.H. (1987) Dev. Biol. 124, 504-511. 8 Devreotes, P.N. and Zigmond, S.H. (1988) Annu, Rev. Cell Biol. 4, 649-686. 9 Spudich, J.A. (1989) Cell Regul. 1, 1-11. 10 Mann, S.K.O., Pinko, C. and Firtel, R.A. (1988) Dev. Biol. 130, 294-303. 11 Gerisch, G. and Wick, U. (1975) Biochem. Biophys. Res. Commun. 65, 364-370. 12 Mato, J.M., Krens, F.A., Van Haastert, P.J.M. and Konijn, T.M. (1977) Proc. Natl. Acad. Sci. USA 74, 2348-2351. 13 Bumann, J., Wurster, B. and Malchow, D. (1984) J. Cell Biol. 98, 173-178. 14 Abe, T., Maeda, Y. and Iijima, T. (1988) Differentiation 39, 90-96. 15 Mato, J.M., Krens, F.A., Van Haastert, P.J.M. and Konijn, T.M. (1977) Biochem. Biophys. Res. Commun. 77, 399-402. 16 Ross, F.M. and New~,ll, P.C. (1981) J. Gen. Microbiol. 127, 339-350. 17 Liu, G. and Newell, P.C. (1988) J. Cell Sci. 90, 123-129. 18 Mato, J.M. and Steiner, A.L. (1980) Cell Biol. Int. Rep. 4, 641-648. 19 Parissenti, A.M. and Coukell, M.B. (1989) J. Cell Sci. 92, 291-301. 20 Watts, D.J. and Ashworth, J.M. (1970) Biochem. J. 119, 171-174. 21 Cocucci, S. and Sussman, M. (1970) J. Cell Biol. 45, 399-407. 22 Coukell, M.B., Cameron, A.M., Petrie, C.M. and Mee, J.D. (1984) Dev. Biol. 103, 246-257. 23 Parissenti, A.M. and Coukell, M.B. (1986) Biochem. Cell Biol. 64, 528-534. 24 Milne, J.L. and Coukell, M.B. (1989) Exp. Cell Res. 185, 21-32. 25 Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 51,660-672. 26 Kumar, S.A. and Dickerman, H.W. (1983) in Biochemical Actions of Hormones (Litwack, G., ed.), Vol. 10, pp. 259-301, Academic Press, New York. 27 Wurster, B., Schubiger, K., Wick, U. and Gerisch, G. (1977) FEBS Lett. 76, 141-143. 28 Van Haastert, P.J.M., Van Walsum, H. and Pasveer, F.J. (1982) J. Cell Biol. 94, 271-278. 29 Pan, P. and Wedner, H.J. (1979) Differentiation 14, 113-118. 30 Kimmel, A.R. and Firtel, R.A. (1982) in The Development of Dictyostelium discoideum (Loomis, W.F., ed.), pp. 233-324, Academic Press, New York. 31 Datta, S. and Firtel, R.A. (1987) Mol. Cell. Biol. 7, 149-159. 32 Pears, C.J. and Williams, J.G. (1987) EMBO J. 6, 195-200. 33 Driscoll, D.M., Pears, C.J. and Williams, J.G. (1988) Dev. Genet. 9, 455-468. 34 Podgorski, G.J., Franke, J., Faure, M. and Kessin, R.H. (1989) Mol. Cell. Biol. 9, 3938-3950. 35 Bradford, M.M. (1976) Anal. Biochem. 72, 248-254.