Indian Journal of ChemistryVol. 29A. October 1990, pp. 1002-1007

Interaction of metal ions with 5 '-guanosine monophosphate-Evidencefor binding through N( 1) in solution

P Rabindra Reddy*t & T K Adharani

Department of Chemistry. Osmania University. Hyderabad 500 007. India

Received 24 November 1989; revised 5 March 1990; accepted 16 April 1990

The stability constants of Cu(ll), Ni(II}, Zn(II} and Co(II} binary complexes (1:1) with 5-guanosinemonophosphate (5-GMP) as the primary ligand and of the corresponding ternary complexes (1:1:1)with glycine, histidine and histamine as the secondary ligands have been determined potentiometriallyat 35°C and 0.10 M (KN03) ionic strength. In the acidic medium, N (7) and phosphate oxygen of 5'-GMP are found to be more favoured coordination sites; the preference for N( 1} increases with increasein pH. This multi-site bonding behaviour of 5'-GMP towards cations in solution has been investigatedand the influence of pH on the coordinating abilities of 5 '-GMP assessed. The influence of secondaryligands on the stabilities of M(II}-5 '-GMP sysems is also discussed. Species distribution curves havebeen obtained using the computer programme BEST.

Metal ion interactions with ligands having differentdonor atoms provide an opportunity to study theircontributions to the stabilities of metal complexes.This type of investigations become much more rele-vant when the metal ion interactions with ligands ofbiological importance are considered. We have beeninterested for some time in these interactions, inparticular with nucleosides and nucleotides, becauseof their multisite binding behaviour towards cationsin solution. These have been several attempts 1-3bothin solution and solid state to probe this behaviour ofnucleic acid components. The present study is yet an-other attempt in this direction.

Earlier work in solution" on 5' -guanosine mon-ophosphate (5'-GMP) relates to the interaction ofCu(II) at pH 7. 13C and 31p NMR studies' of 5'-GMP with Mn2 + indicated the presence of multi-ple binding sites. Earlier studiesv" have shownthat the metals are bonded to 5' -guanosine mono-phosphate (5'-GMP, structure I) through N (7) andthe oxygen atom of H20 molecules (which are hy-drogen bonded to phosphate group). Aoki et al.9have shown that Cu(II) is directly bonded to N (7)of the base and to the phosphate moiety in[Cu(5'-GMPh(H20)s] compound. Pt(II) and Mg(II)were also found to be N (7 )-bonded with5-GMplO. However, only Mg(H) was found to bedirectly coordinated to phosphate at lower pH.

tDBT !India Overseas Fellow at Harvard University

(1986-88).

1002

Recently Perno et al.II have studied the interactionof Ni(II) and cis-[(NH3hPt]2+ with deoxy-guano-sine monophosphate.

However, there are discrepancies between theresults obtained in solution and solid states. Themajor reason for this seems to be the high de-pendence of metal binding sites on pH. Obviously,the experimental pH has to be monitored verycarefully in order to get reliable information.

In addition to the ambiguity due to base versusphosphate binding, purine nucleotides display adichotomy between binding through N (1) and N(7)12. In 5'-GMP, N (7) is deprotonated in slightlyacidic solutions and therefore metal binds prefer-entially through N (7). In neutral and basic solu-tions metal ions bind through N (1).

Though the NMR spectroscopy has proved tobe one of -the most powerful techniques for deter-mining the binding sites of the metal ions to nucle-otides, the rapid exchange of acidic protons is aproblem and many of these investigations havebeen made in solvents such as DMSO. In generalthese solutions are very different from the aqueousones. Similarly, crystal structure determinationshave been confined mainly to samples grown inacidic solutions where N (1) is protonated. In viewof this it was thought important to investigate themetal ion interactions with 5' -GMP in aqueous so-lution covering the pH range 3-10.

In the present paper interactions of Cu(II),Ni(II), Zn(II) and Co(II) with 5'-guanosine mono-phosphate (binary and ternary systems) have been

REDDY et al: INTERACTION OF METAL IONS WITH 5'-GUANOSINE MONOPHOSPHATE

studied to get further information on the metalbinding sites in absence and presence of biologi-cally important secondary ligands. The secondaryligands used in the ternary systems are glycine,histidine and histamine. The structures of primaryand secondary ligands showing the dissociationsteps are depicted in Charts I and II.

Materials and Methods5' -Guanosine monophosphate (5 '-GMP) was

obtained from Fluka (Switzerland), and glycine(gly), histidine (histi) and histamine (hista) wereobtained from Sigma chemical Company (USA).

The experimental method consisted of a poten-tiometric titration of solutions containing metalion, nucleotide and secondary ligands in (1:1) and(1:1:1) ratios at 35±0.I°C with standard NaOHsolution. The ionic strength was maintained con-stant by using O.IOM (KN03) as the supportingelectrolyte and relatively low concentration of theligand and metal ion (1 x 10- 3 M). Other experi-mental details can be found elsewhere'.

Calculations

Binary systemsIn M(II)-5'-GMP (1:1) binary systems two types

of interactions have been observed. Cu(lI) andZn(lI) form one type of complexes and the rest ofthe metal ions another type.

In the case of Cu(lI)- and Zn(II)-5'-GMP sys-tems, the following equation was used in the buf-fer region m = 1-3.

· .. (1)

For Ni(lI) and Co(lI), the following equations wereemployed. In the buffer region m = 1-2, Eq. 2 wasused,

· .. (2)

and in the buffer region m = 2-3, Eq. 3 was used.

MHL.:ML+H+ · .. (3)

Ternary systems

M(11)-5'GMP-gly(I:I:I) systemEqs 4 and 5 were used to calculate stability

constants for Cu(II).

... (4)

Il"ae

Chart 1- Dissociation steps involved in the ionisation of5' -guanosine monophosphate

CH -CH-COO-~ 2, +1-\ NH)

(b) HN NHt~

CH2-CH-COO- CH2

-CH-COO-~ I + ~ I

Ka r=»; NH) KZO r=; NH2~HN N ~HN N

'-..:P' ~

CHART 2

Chart II-(a) Dissociation steps involved in the ionization ofglycine, (b) dissociation steps involved in the ionization of his-tidine and (e) dissociation steps involved in the ionization of

histamine.

For the rest of the metal ions, equations 6 and 7were used. In the buffer region m = 0-2, Eq. 6 wasused,

... (6)

and in the buffer region m = 2-5, Eq. 7 was used,

M(H2L)HA.:MlA+3H+ ... (7)

M(11)-5!GMP- histilhista (1:1:1) systemsEquations 4 and 5 were employed to calculate

the stability constants of Cu(lI) complexes of 5'-GMP with histidine and histamine as secondary li-gands.

For Ni(II), Zn(lI) and Co(lI), Eqs 8 and 9 wereused; in the buffer region m = 1-4, Eq. 8 was used,

M(HJ..(A ~=3~ MAL+2H+

M(H2L)H2A'=M(HL)A+ 3H+

... (5) while Eq. 9 was used between m = 4 and 5.

... (8)

1003

INDIAN J CHEM, SEC A, ocrOBER 1990

M(HL}A~MLA+ H+ ... (9)

It should be noted here that in the case of histi-dine and histamine systems [except for Cu(I1}),thebuffer region m = 0-1 was not taken into accountfor the calculation of constants as the titrationcurves overlapped with the free ligand curve.

All calculations were performed using the ex-perimental data based on the trends in the titrationcurves on a CASIO-PB-lOO personal computerwith the aid of suitable material balanced equ-ations. The species distribution curves for varioussystems were obtained with the help of the com-puter programme BESTI3.

Results and Discussion

Binary M(Il)-5!GMP systemThe titration curve of Cu(I1}-5'-GMP system

(Fig. l e) showed an inflection at m = 1 followed bya buffer region. Accordingly, it was assumed that anormal (1: I) complex was formed in the buffer re-gion m= 1-3. The constant (K~L) was calculatedusing Eq. (I). The behaviour of Zn(II} was similarto that of CU(I1}.

The titration curve of Co(I1} (Fig. Id) showedinflections at m = 1 and m = 2; the correspondingprotonated (K~d and normal (K~) (1:1) con-

10

9

8

0>o 6-..J

I

4

stants were calculated in the above buffer regionusing Eqs 2 and 3. Similar trends were obtained inthe case of Ni(II).

For all the above metal ions no attempt wasmade to calculate the constants in the buffer re-gion m = 0- 1 as the titration curves in this regionoverlapped with the free ligand curve. All the.above binary constants are presented in Table 1.

Ternary (1:1:1) systems

Glycine systemThe mixed ligand titration curve of Cu(II)-5'-

GMP-gly (Fig. lc) in a 1:1:1 ratio showed an in-flection at m = 3 followed by a buffer region. For-mation of protonated and normal (1:1:1) com-plexes was assumed in the buffer regions beforeand after m = 3 respectively. The constantsKM d KM(H,L)A al led' EqM(H,L)A an MAL were c cu at usmg s4 and 5. For Ni(I1}, Zn(II) and Co(ll) systems, aninflection was observed at m = 2 followed by abuffer region. The corresponding constants werecalculated using equations 6 and 7.

The various constants pertaining to the aboveinteractions are listed in Table 1.

Histidine/histamine systemThe trends in the CU(ll)-5' -GMP-histi/hista titra-

b

a or mFig. I-Potentiometric titration curves for the interaction of S'-GMP with Cu(D) and Co(D) in both binary (1:1) and ternary(1:1:1) systems at 3S"C, /=0.01 mol dm? (KN03) [(a) free S'-GMP, (b) Co(D)-S!QMP-histi, (c) Cu(D)-S'-GMP-gly, (d) Co(II)-S'-

GMP and (e) Cu(D)-5.:0MPj

1004

REDDY et al: INTERACTION OF METAL IONS WITH S'-GUANOSINE MONOPHOSPHATE

Table I-S~bility constants" of the (1:1) binary and (1:1:1) ternary complexes of S'-GMP with glycine [Temp = 3SOC;/= 0.01 mol dm -) (KNO))]

Metal M(II)-s'-GMP (1:1) M(II)-s'-GMP-Gly (1:1:1)

K~L K~HL K"MHL K~(H2LIA KMM(H1LIHA K i:ti'LLIA K i:ti'i.2LIHAML

Cu(n( 6.64 l3.90 5.25

Ni(II) 2.77 4.04 9.25 16.85

Zn(II) 5.29 9.21 12.33

Co(II) 1.94 3.85 8.88 11.40

*The constants are accurate to ±0.0610gKunits. Acid dissociation constants of s'-GMP: p K, = 2.68; pK2.; pK2a = 6.21; pK)a = 9.21.

Table 2-Stability constants" of the (1:1: 1) ternary complexes of 5' -GMP with histidine and histamine [Temp = 35°C;/=0.10 mol dm -) (KNO))]

Metal M(II(-s '-GMP-Histi (1:1:1) M(II)-s'-GMP-Hista (1:1:1)ion

K~(H2LIAKM(H2L)H2A KM(H2LIA KM(HLIA

K~(H2L)AKM(H1L)H1A KM(HLIA K~LIAM(HL)A MLA MLA M(HLIA MLA

Cu(II) 14.95 4.92 13.93 4.34

Ni(II) 12.97 3.22 11.88 3.92

Zn(II) 11.35 4.20 11.21 3.67

Co(II) 11.l3 2.65 10.70 3.30

·The constants are accurate to ± 0.06 log K units.

tion curves were found to be similar to those inCu(II)-S~GMP-gly system.

The mixed ligand titration curve of Co(II)-S'-GMP-histidine (Fig, Ib) or histamine showed in-flections at m = 1 and m = 4. The correspondingconstants were calculated using Eqs 8 and 9. Simi-lar results were observed in the case of Ni(II) andZn(ll) also (Table 2).

The titration curve of free S'-GMP showed(Fig. 1a) inflections at m = 1 and 2 followed by abuffer region indicating the stepwise dissociation ofits protons. These are assigned to dissociations ofN(7)-H, secondary phosphate hydrogen andN)l )-H respectively. The values of dissociationconstants agreed well with the literature values forS' _GMP14,15as well as those for its nucleoside'S"and base? which further confirmed our assignmentsabout proton dissociation.

The stability constants pertaining to the Cu(II),Zn(II), Ni(II) and Co(II)-5'-GMP interaction arepresented in Table 1. No interaction was observedwith Mg(II) and Ca(II). The stability constants ofS'-GMP complexes were higher than the corre-sponding constants of guanosine" complexes.The higher stabilities in case of former can be at-tributed to the involvement of phosphate moiety

in metal bonding in addition to the base coordina-tion. It can be seen from Table 1 that Cu(IIl andZn(II) differ significantly in their interactions with5'-GMP as compared to Ni(II) and Co(II). Thistype of selective interactions are an indication ofthe differences in the mode of bonding in thesecomplexes. However, no interaction was observedfor these metal ions upto pH 4.5. These resultsare not surprising when the pH-dependent coordi-nating ability of 5' -GMP is considered.

Among the three potential donor sites of 5'-GMP [N(7), phosphate oxygen and N(1)], N(7) andphosphate oxygen are found to be more favouredsites in acidic medium and the preference for N( 1)increases as pH increases. Since no interaction wasobserved in acidic medium, the possibility of in-volvement of N(7) in coordination to metal can beruled out. This leaves only secondary phosphateoxygen and N(1) as potential donor sites. The ex-perimental pH range of the present study confirmsthis assumption which is further corroborated bythe species distribution curves.

The species distribution curves for Ni(1I)-5'-GMP indicate that the formation of protonated(1:1) complex starts at pH - 5.5 and reaches amaximum of 26% around pH 7.S. The predomi-

100S

INDIAN J CHEM, SEC A, OCTOBER 1990

nant species in this system also seems to be thenormal (1:1) complex which attains a maximumvalue of 68% at pH around 9. Based on these da-ta it is proposed that in the protonated complexonly phosphate oxygen is involved in metal bind-ing and in the normal (1:1) complex, Ni l ) coordi-nates along with hydrogen bonded phosphate oxy-gen.

The ternary constants of various systems arelisted in Tables 1 and 2. It may be seen from thetables that in these systems also different types ofcomplexes are formed.

The stability constants for the protonated ter-nary Cu(II system decrease in the order: histi >hista :::::gly. This suggests that the action of histi-dine is neither glycine-like nor histamine-like; itacts as a terdentate ligand coordinating through allits available donor groups (i.e. carboxylate, imida-zole and amino groups). This shows that in thesesystems only N(7) of 5'-GMP is involved in bond-ing in addition to donor sites of histamine and his-tidine. However, in the case of normal complexesthe differences in the stabilities are narroweddown indicating that histidine acts as a bidentateligand. This is not surprising because of the factthat in this buffer region (m = 3 to 5) two donorsites [phosphate oxygen and N(l)] from the 5'-GMP are involved in bonding. A species distribu-tion curve for Cu(II)-5'-GMP-gly (1:1:1) system(Fig. 2) further confirms this assumption.

When the protonated ternary complexes ofNi(II) and Co(II) with histidine, histamine and 5'-GMP are compared, it is observed that histidineforms more stable complexes than histamine. Thismay be attributed again to the differences in theirmodes of bonding. In the case of Zn(II) the stabil-ity constants for both the systems are almost same.The reduced stability of Zn(II)-5'-GMP-histidineternary complex may be due to the lower coordi-nation number of Zn(II) where histidine may actas a bidentate ligand. However, in the case of nor-mal complexes of Ni(II) and Co(II) the trend is re-versed. In the buffer region m = 4-5, histidine mayact as a bidentate ligand coordinating only throughimidazole and amino nitrogens. It is likely that ahydrogen bonding type of interaction betweenCOO- of histidine and - NH2 of 5'-GMP takesplace as a result of which the COO - group is un-able to coordinate to the metal. Such an interac-tion disfavours the formation of ternary complexesin solution due to the entropy factor as the ligandsbecome more rigid as a consequence of bridgeformation between two non coordinating groupsand also due to the different degrees of solvent in-

1006

teraction with the metal ion. This may be the rea-son for the lower stabilities of histidine complexescompared to those of histamine complexes. Basedon these conclusions, a tentative structure forNi(II)-5'-GMP-histidine is proposed (Fig. 3).

100

90

-o~

M

60::'uX-V) 50

40

30

20

10

3 4

Fig. 2-Species distribution curve for Cu(II)-5.!GMP-gly (1:1:1)ternary system

Fig. 3- Tentative structure for Ni(II)-5'-GMP-histidine com-plex (1:1:1) in solution

REDDY et al.: INTERACTION OF METAL IONS WITH 5'-GUANOSINE MONOPHOSPHATE

Finally, it is clear from the above investigationsthat the metal ion interactions with nucleic acidcomponents are highly selective, in particular withmononucleotides where the presence of phosphatemoiety adds a new dimension.

AcknowledgementThe financial assistance from UGC, New Delhi

through Scheme No. (F-12-87/84 SR-III) is grate-fully acknowledged,References

1 Rabindra Reddy P, Harilatha Reddy M & Venugopal Red-dy K, Inorg Chem; 23 (1984) 974.

2 Martin R B & Mariam Y H, Metal ions in biological sys-tem, Chap 2 (Marcel Decker, New York) 1979.

3 Marzilli L G, Kistenmacher J J & Eicchorn G L, Nucleicacid-metal ion interaction, Chap 5 (Wiley-Interscience,New York) 1980.

4 TuA T & Friedrich C G, Biochemistry, 7 (1968) 4367.5 Kotowycz G & Suzuki 0, Biochemistry, 12 (1973) 3434.6 Mansy S & Tobias R S, JAm chem Soc, 96 (1974) 6874.7 De Meester P, Goodgame D M L, Skapski A C & Smith

B T, Biochem Biophys Acta, 340 (1974) 113.8 De Meester P, Goodgame D M L, Jone T J & Skapski A

C,BiochemJ,139(1974)791.9 Aoki K, Biochem Biophys Acta, 425 (1976) 369.

10 Tajrnir A J & Theophanides T, Can J Chem, 61 (1983)1813.

11 Perno J T, Cwikel D & Spiro G T, Inorg Chem, 26 (1987)400.

12 Martin R B, Acc Chem Res, 18 (1985) 32.13 Motekaitis R J & Martell A E, Can J Chem, 60 (1982)

2403.14 Bock R M, Ling N S, MareD S A & Lipton S H, Arch Bi-

ochem Biophy, 62 (1956) 253.15 AlbenA,BiochemJ,54(1953)646.16 Fiskin AM & Beer M, Biochemistry, 4 (1965) 1289.

1007