Download Polyamines and other charged amines bind to RNA by hydrogen

volume 5 Number 4 April 1978
N u c l e i c A c i d s Research
Hydrogen bonding interactions of polyamines with the 2' OH of RNA
Philip H. Bolton and David R. Kearns
Department of Chemistry, University.of California, San Diego, La Jolla, CA 92093, USA
Received 26 January 1978
ABSTRACT
Polyamines and other charged amines bind to RNA by hydrogen bonding to
the 3' phosphate and to the 2' OH. This mode of binding suggests a mechanism
by which DNA and RNA might be distinguished by enzymes.
INTRODUCTION
The biochemical importance of polyamines has recently become more widely
recognized1~1*. Polyamines bind strongly to polynucleotides and stabilize both
secondary and tertiary structure, they are associated with nucleic acids in_
vivo» and they may be important in regulating the biological function of some
nucleic acids1"1*. The studies of Gosule and Schellman 5 and of Flink and Pettijohn6 indicate that spermine, NH 2 (CH 2 ) 3 NH(CH 2 ) 1) NH(CH 2 ) NH 2 , may play an important role in packaging DNA in bacterial cells and in phage.
Spermine affects
the properties of tRNA in solution2 and the crystallization of tRNA for x-ray
diffraction studies was accomplished using solutions containing spermine3.
A number of methods have been used to study polyamine interactions with
nucleic acids, but there is little direct evidence concerning the precise manner in which polyamines bind to RNA and DNA.
Tsuboi 8 and Liquori, et al., 9
proposed models for spermine and spermidine, NH 2 (CH 2 ) NH(CH2) NH2,binding to
DNA in which phosphate groups interact with each positively charged amino
group, the tetramethylene portion of the polyamine bridges the narrow (minor)
groove of the helix between the two strands and the trimethylene portion
bridges adjacent phosphate groups. Gabbay and co-workers 1 0 > n proposed that
diamines bind in the minor groove of RNA and DNA by hydrogen bonding of the
amines to phosphate oxygens.
The model of Gabbay and co-workers is of interest since we have presented
pmr evidence that the 2' OH of RNA is involved in hydrogen bonding to a water
© Information Retrieval Limited 1 Falconberg Court London W1V5FG England
1315
Nucleic Acids Research
Base
Base
Figure
the 2'
and 3 1
bonded
1. Schematic drawing of the water molecule which interacts with both
OH and 3' phosphate. The interaction of a charged amine with the 2' OH
phosphate is also shown. Note that the 2' OH proton is not hydrogen
to the amine.
molecule which is simultaneously hydrogen bonded to the 3' phosphate12 (Fig.
1).
Gabbay's model implies that the binding of amines to RNA would displace
the bound water molecule and profoundly affect the 2' OH-phosphate interaction,
and hence the resonance of the 2' OH proton.
In the present investigation, high resolution pmr is used to study the
interaction of polyamines with RNA.
This method is very sensitive to changes
in nucleic acid conformation and can be used to monitor both hydrogen bonding
and stacking interactions.
Furthermore, we recently found conditions in which
the resonance from the 2' OH proton of RNA can be observed (at 6.8 ± 0.2 ppm)
and this provides an additional monitor of the RNA backbone conformation and
minor groove binding sites 12 .
As we shall show, the binding of certain poly-
amines has a pronounced and specific effect on the resonance of the 2' OH proton which can be used to characterize the nature of the polyamine-RNA interaction.
To evaluate the influence of binding strength and ionic charge on the
nature of the polyamine-RNA interactions, several classes of compounds were
investigated.
Mono- and di-amines were included because they are ionic amines
which bind less strongly to RNA than polyamines, whereas urea was included to
explore the effect of an uncharged amine.
The binding of magnesium was studied
to determine how a divalent cation, which is believed to specifically interact
1316
Nucleic Acids Research
with the phosphate groups, affects the behavior of the resonance from the 2'
OH proton.
RESULTS
Interaction of RNA with Amines and Folyamines.
The results in Fig. 2
show the effect of the polyamine spermine on the spectra of poly(A), poly(C)
poly (U)
T=I2°
3mM spermine
9
I mM spermine
8
poly(C)
Figure 2. 300 MHz pmr spectra of (a) poly(A), (b) poXy(U) and (c) poly(C) in
the presence and absence of spermine. The level of spermine and temperature
are indicated in the figure. The samples were prepared as described elsewhere 12 and the concentration of the samples, in monomer units, was 40-45 mM.
The chain lengths of the samples were about 70. Samples were extensively dlalyzed against 0.1 M NaCl and either 10 mM cacodylate or phosphate buffer at
pH 7.0. The pmr spectra were obtained by use of correlation spectroscopy17
using a Varian HR-300 spectrometer. Assignments of resonances 12 * 18 are also
indicated in the figure.
1317
Nucleic Acids Research
and on poly(U) at neutral pH and low temperature.
In this condition poly(A)
and poly(C) form single strand ordered helices 10 and poly(U) forms a double
helix with U'U base pairs 1 3 .
In each case the addition of about one spermine
per ten phosphates eliminates the resonance of the 2' OH proton at 6.8 ± 0.2
ppm without affecting any other RNA resonances.
The observation that polya-
mines eliminate the 2' OH resonance at levels less than one amine per phosphate
is consistent with the fact that, while the binding constant of polyamines is
rather large, the polyamines are quite mobile when bound to nucleic acids 19 .
The spectra in Fig. 2 show also that the resonance is not lost by gradual
shifting of the resonance position with increasing levels of spermine and that
the addition of spermine beyond the level needed to eliminate the resonance
from the 2' OH proton has no other effect on the spectra.
Significantly, re-
sonances from the exchangeable amino protons are not affected by spermine.
The effects of spermine on the spectra of double helical poly(I)'poly(C)
and on tRNA are shown in Fig. 3.
In both cases the addition of less than one
spermine per ten phosphates induces the loss of intensity near 6.8 ppm from 2'
OH proton resonances with some small shifts (less than about 0.1 ppm) and
sharpening of the spectrum of poly(I) •poly(C) . The spectrum of E_. coli
tRNA M i x e d is too diffuse to observe any small shifts.
The results presented above show that the addition of spermine or spermidine to ENA specifically eliminates the resonance from the 2' OH proton.
The
diamines putrescine, cadaverine and lysine (Fig. 4) also exhibit this specific
effect, but higher levels are required (about one amine per phosphate) to
eliminate the 2' OH resonance of the homo-ribopolynucleotides.
Monoamines
such as glycine will also eliminate the 2' OH resonance, but only at levels
of about three amines per phosphate.
The amine urea, which is uncharged in
the conditions used in these experiments, has no effect on the 2' OH resonance
up to at least 200 mM.
In summary, all of the positively charged amines inves-
tigated can eliminate the 2 1 OH resonance.
To examine the effect on the 2' OH resonance when other positively charged
ions bind to the phosphate group, the spectra of poly(U) and poly(A) were investigated in the presence of magnesium.
As shown in Fig. 4, the addition of
magnesium does not alter the intensity of the 2' OH resonance, but it does induce a small downfield shift (about 0.1 ppm) in the resonance position and some
broadening of all the resonances.
Interaction of Mononucleosides with Amines, Polyamines and Magnesium.
The
effect of various amines and magnesium on the pmr spectrum of uridine and other
mononucleosides was investigated.
1318
To reduce the rate of exchange of the 2' OH
Nucleic Acids Research
polyd)-poly(C)
°
I
4 m M spermine
H6 H8 H2
NHj
NH 2
2'0Hv
I
I
8
7
L
6
E.coli tRNA'mixed
T=l°
ppm
Figure 3. 300 MHz pmr spectra of (a) poly(I)*poly(C) and (b) mixed £. coli
tRNA in the presence and absence of spermine. The level of spermine and the
temperature are indicated in the figure. The samples and spectra are described
in the caption to Fig. 2. The tRNA sample, 1 mM, was dialyzed against a buffer
containing 10 mM magnesium in addition to the above constituents. Assignment
of the poly(I)*poly(C) spectrum will be discussed elsewhere.
1319
Nucleic Acids Research
poly(A)
T=O°
IOmM Mg
2'OH
8
7
ppm
poly (U)
l5mM Mg
2'OH
8
1320
7
ppm
6
Nucleic Acids Research
poly(C)
URIDINE
Me S O / H O
T= I 2°
15=1 (v/v)
T = 2O°
lOmM Mg
lOmM spermine
i
10 m M
8
7
ppm
' ppm
6
Figure 4. 300 MHz pmr spectra of (a) poly(A) and (b) poly(U) in the presence
and absence of magnesium. The temperature and magnesium level are indicated
in the figure. Also shown, (c), is the spectrum of poly(C) in the spectrum of
poly(C) in the presence and absence of lysine. The level of lysine and temperature are indicated in the figure. The samples and spectra are as described
in the caption to Fig. 2. The 300 MHz pmr spectrum of uridine in Me2SO-H2O
mixed solvent, 15:1 (v/v), in the presence and absence of the indicated levels
of magnesium and spermine, is shown in 4d. The concentration of uridine was
about 80 mM, and the spectra were obtained as in Fig. 2.
1321
Nucleic Acids Research
proton of the mononucleosides with water 12 , these experiments were carried out
in a mixed solvent system containing dimethyl sufloxide (Me2S0) and H2O 15:1
(v/v).
Under these conditions the line width of the 2' OH resonance is deter-
mined by exchange with water and any increase in the exchange rate, as monitored by the line width, is easily detected.
The spectrum of uridine in the
presence of either spermine or magnesium is shown in Fig. 4.
Neither sper-
mine nor magnesium exhibits any effect on the uridine resonances.
These re-
sults are important in demonstrating that at the concentrations used in the
above mentioned polynucleotide experiments the-amines and magnesium do not
appreciably affect the rate of exchange of the 2' OH proton of mononucleosides
with water.
This is to be contrasted with the effect of certain other ions.
For example, the addition of 10 mM KH 2 P0 4 (pH 7.0) eliminated the 2' OH resonance of mononucleosides, but is without effect on the corresponding resonance
in polynucleotides.
DISCUSSION
Independent of any specific model for the polyamine-RNA interaction, the
pmr results clearly indicate that the presence of polyamines eliminates the
resonance of the 2' OH proton without affecting the resonances of other protons.
This result strongly suggests that polyamines interact in some manner
with the 2' OH when bound to RNA.
This selective effect on the 2' OH proton
could be attributed to a variety of mechanisms such as a structural change in
the RNA double helix induced by the amines, a specific catalysis of the exchange of the 2' OH proton
or to some other mechanism.
A structural change
seems to be ruled out by circular dichoism observations on RNA in the presence
and absence of spermine which showed that spermine does not disrupt the A form
of RNA11*'15.
Furthermore, our pmr results indicate that spermine and other
amines do not appreciably affect the resonance positions of the other RNA protons.
It appears that, at the concentrations examined here, the amines do not
significantly affect the exchange rate of the 2' OH proton of mononucleosides.
From this we infer that, in the absence of phosphate groups, there is no special (strong) interaction between the amines and the 2' OH which catalyzes
the exchange of the 2' OH proton.
The possibility that binding of any posi-
tively charged ion to the phosphates of RNA catalyzes the exchange of the 2'
OH proton with water seems unlikely since high levels of magnesium or sodium
do not eliminate the 2' OH proton resonance.
Thus, it appears that some
other mechanism has to be invoked to account for the highly specific effect
of amine binding on the 2 1 OH proton resonance.
1322
In an earlier study 12 we
Nucleic Acids Research
presented pmr evidence that the resonance from the 2' OH proton can only be
observed in aqueous solutions in ordered polynucleotldes where it is both
protected from exchange with water and shifted downfield by a hydrogen bonding interaction with a water molecule that is simultaneously hydrogen bonded
to a 3' phosphate oxygen.
In this scheme, depicted in Fig. 1, a bridging
water molecule acts as a proton acceptor for the 2' OH and as a proton donor
to the phosphate. The results on the interaction of amines with RNA presented
here and elsewhere1 >3,1t,l0,11
are
consistent with the notion that the amines
substitute for the bridging water molecules in the manner shown in Fig. 1.
In this scheme, the amine acts as proton donor to both the 3 1 phosphate and
the 2' OH oxygens. This effect is not observed for uncharged amines presumably due to their relatively weak binding to RNA. As recently discussed by
Tse and Newton 16 , these conditions are especially favorable for strong hydrogen binding.
In our model for the amine-RNA complex the 2' OH proton is no
longer hydrogen bonded to a bridging water molecule, and can rapidly exchange
with bulk water molecules.
This accounts for the experimental observation
that the amines eliminate the 2' OH proton resonance. The proposed substitution of the amine for the bridging water is consistent with circular dichroism
14
» 1 5 and pmr observations which indicate that amines do not appear to affect
the manner of stacking of the bases.
If the amines bound directly to the nu-
cleic acid bases the stacking interactions could have been affected. The
substitution of an amine for the bound water would not require change in the
backbone to accommodate the new hydrogen bonding arrangement involving the
amine.
This model for the binding of amines to RNA is similar to that of Gabbay
10,11
alu j
co-workers. The main difference is that in the model proposed here
the amine hydrogen bonds to both the phosphate oxygen and 2' OH oxygen, and
displaces a bound water molecule. The conformation of the amines bound to
RNA is essentially the same in both proposals.
Since the above results suggest that the binding of amines to RNA involves
hydrogen bonding to the phosphates as well as to the 2' OH, the binding of
amines to RNA and DNA may be different because of the additional hydrogen
bonding offered by RNA.
Furthermore, the interstrand separations are differ-
ent in RNA than in B form DNA.
Presumably, electrostatic interactions con-
tribute much of the binding energy of the polyamines to RNA and DNA, but the
present study suggests that there is an additional hydrogen bonding interaction which contributes in the binding to RNA, but not DNA. This interaction
1323
Nucleic Acids Research
with the 2 1 OH of the RNA might be utilized by certain proteins to distinguish
between RNA and DNA.
ACKNOWLEDGMENT
The support of the U.S. Public Health Service (Grant GM 22969 and, in
part, RR 00757) is most gratefully acknowledged.
We would like to thank Dr.
J. Kraut for use of his Kendrew model of a ten base pair segment of double
helical RNA, and for helpful discussions.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
1324
Tabor, H. and Tabor, C.W. (1972) in Methods in Enzymology, Meister, A.,
Ed., Vol. 36, pp. 203-268. John Wiley and Sons, Inc., New York.
Sakai, T.T. and Cohen, S.S. (1976) Prog. Nucleic Acids Res. 17, 15-42.
Bachrach, U. (1973) Function of Naturally Occurring Polyamines, pp. 7481. Academic Press, New York.
Cohen, S.S. (1971) Introduction to the Polyamines, pp. 129-136. Prentice
Hall, Inc., New Jersey.
Gosule, L.C. and Schellman, J.A. (1976) Nature 259, 333-335.
Flink, I. and Pettijohn, D.E. (1975) Nature 25J3, 62-63.
Rich, A. and RajBhandary, U.L. (1976) Ann. Rev. Biochem. 45_, 805-860.
Liquori, A.M., Constantino, L., Crescenzi, V., Ella, V., Giglio, E.,
Puliti, R., De Santis Sarino, M., and Vitagliano, V. (1967) J. Mol. Biol
24, 113-122.
Tsuboi, M. (1964) Bull. Chem. Soc. Japan 37_, 1514-1522.
Gabbay, E.J., Adawadkor, P.D. and Wilson, W.D. (1976) Biochemistry 15,
•146-151.
—
Gabbay, E.J. (1969) J. Am. Chem. Soc. <n, 5136-5150.
Bolton, P.H. and Kearns, D.R. (1977) Blochim. Biophys. Acta (in press).
Guschlbauer, W. (1976) Nucleic Acid Structure, pp. 70-98. SpringerVerlag, New York.
Maestre, M.F. and Tinoco, I. (1967) J. Mol. Biol. 23_, 323-335.
Thrierr, J.L., Deubel, V. and Leng, M. (1972) Biochemie ,54, 1115-1119.
Tse, Y.C. and Newton, M.D. (1977) J. Am. Chem. Soc. £9, 611-613.
Dadock, J. and Sprecher, R.F. (1974) J. Mag. Res. 13_, 243-248.
Ts'o, P.O.P. (1974) in Basic Principles in Nucleic Acids Research, Ts'o,
P.O.P., Ed., Vol. II, pp. 305-469. Academic Press, New York.
Gabbay, E.J., Glasser, R. and Gaffney, B.L. (1970) Annals New York Acad.
Sci. 171, 810-826.