5364
`
`J. Am. Chem. Soc. 1999, 121, 5364-5372
`
`Kinetics of RNA Degradation by Specific Base Catalysis of
`Transesterification Involving the 2′-Hydroxyl Group
`
`Yingfu Li and Ronald R. Breaker*
`Contribution from the Department of Molecular, Cellular and DeVelopmental Biology, Yale UniVersity,
`PO Box 208103, New HaVen, Connecticut 06520-8103
`ReceiVed February 24, 1999
`
`Abstract: A detailed understanding of the susceptibility of RNA phosphodiesters to specific base-catalyzed
`cleavage is necessary to approximate the stability of RNA under various conditions. In addition, quantifying
`the rate enhancements that can be produced exclusively by this common cleavage mechanism is needed to
`fully interpret the mechanisms employed by ribonucleases and by RNA-cleaving ribozymes. Chimeric DNA/
`RNA oligonucleotides were used to examine the rates of hydroxide-dependent degradation of RNA
`phosphodiesters under reaction conditions that simulate those of biological systems. Under neutral or alkaline
`pH conditions, the dominant pathway for RNA degradation is an internal phosphoester transfer reaction that
`is promoted by specific base catalysis. As expected, increasing the concentration of hydroxide ion, increasing
`the concentration of divalent magnesium, or raising the temperature accelerates strand scission. In most instances,
`the identities of the nucleotide bases that flank the target RNA linkage have a negligible effect on the pKa of
`the nucleophilic 2′-hydroxyl group, and only have a minor effect on the maximum rate constant for the
`transesterification reaction. Under representative physiological conditions, specific base catalysis of RNA
`cleavage generates a maximum rate enhancement of ∼100 000-fold over the background rate of RNA
`transesterification. The kinetic parameters reported herein provide theoretical limits for the stability of RNA
`polymers and for the proficiency of RNA-cleaving enzymes and enzyme mimics that exclusively employ a
`mechanism of general base catalysis.
`
`Introduction
`Ribonucleic acids perform critical functions in modern
`biological systems that range from information storage and
`transfer to structure formation and catalysis. Polymers of RNA
`are most commonly composed of the four standard ribonucleo-
`tides (G, A, U, and C) which are joined via 3′,5′-phosphodiester
`linkages (Scheme 1, 1). The stability of RNA toward chemical
`and enzymatic degradation plays a central role in the biological
`function of this polymer. One of the greatest risks to the
`molecular integrity of RNA under conditions that approximate
`those found inside cells (near neutral pH and the presence of
`alkali metals and alkali-earth metals) is the inherent chemical
`instability of its RNA phosphodiester bonds. The close proximity
`of the adjacent 2′-hydroxyl group to the phosphorus center of
`each internucleotide linkage permits facile transesterification to
`occur, particularly under strongly acidic or strongly basic
`conditions.1 Both acid-catalyzed and base-catalyzed reactions
`proceed via an SN2 mechanism wherein the 2′ oxygen attacks
`the adjacent phosphorus center. It is largely the protonation state
`of the 2′ oxygen that dictates the rate at which each internucleo-
`tide linkage cleaves, although higher-ordered RNA structure also
`contributes significantly to the rate of
`transesterification.
`Alkaline conditions favor specific base catalysis, in which the
`2′-hydroxyl group is deprotonated by hydroxide to generate the
`more nucleophilic 2′-oxyanion group (Scheme 1, 1). In this
`process, the P-5′O bond of the phosphodiester linkage is
`cleaved upon formation of a new P-2′O bond to yield 2′,3′-
`
`* Address correspondence to this author. E-mail: ronald.breaker@yale.edu.
`(1) Oivanen, M.; Kuusela, S.; Lo¨nnberg, H. Chem. ReV. 1998, 98, 961-
`990.
`
`Scheme 1. RNA Cleavage by Alkali-Promoted
`Transesterificationa
`
`a Structures 1 and 2 represent the 3′,5′-phosphodiester linkage in
`the ground-state configuration and the pentacoordinate intermediate,
`respectively. Cleavage products carrying a 2′,3′-cyclic phosphodiester
`terminus (3) and a 5′-hydroxyl terminus (4) are generated following
`nucleophilic attack by the 2′ oxygen. B represents any of the four natural
`nucleotide base moieties. Dashed lines depict the continuation of the
`RNA chain via additional 3′,5′-phosphodiester linkages or alternatively
`chain termination with hydrogen.
`cyclic phosphate (3) and 5′-hydroxyl (4) termini. This cyclizing
`mechanism for RNA cleavage, which was first proposed and
`subsequently established by investigators during the early
`1950s,2-6 is the primary pathway for the uncatalyzed degradation
`of RNA polymers under typical cellular conditions.7
`(2) Brown, D. M.; Todd, A. R. J. Chem. Soc. 1952, 52-58.
`(3) Bacher, J. E.; Kauzmann, W. J. Am. Chem. Soc. 1952, 74, 3779-
`3786.
`(4) Brown, D. M.; Todd. A. R. J. Chem. Soc. 1953, 2040-2049.
`(5) Lipkin, D.; Talbert, P. T.; Cohn, M. J. J. Am. Chem. Soc. 1954, 76,
`2871-2872.
`(6) Brown, D. M.; Magrath, D. I.; Neilson, A. H.; Todd, A. R. Nature
`1956, 177, 1124-1125.
`
`10.1021/ja990592p CCC: $18.00 © 1999 American Chemical Society
`Published on Web 05/25/1999
`
`CUREVAC EX2028
`Page 1
`
`

`

`Kinetics of RNA Degradation
`
`J. Am. Chem. Soc., Vol. 121, No. 23, 1999 5365
`
`Although a tremendous quantity of data concerning this
`reaction has been generated over the years,1 no systematic
`examination of the reaction kinetics has been reported for
`conditions that are of physiological relevance. Most importantly,
`we wanted to determine the kinetic parameters needed to more
`accurately estimate the rate of RNA transesterifcation under
`neutral pH conditions and ambient temperatures. A detailed
`understanding of the kinetics of RNA transesterification under
`high pH conditions makes possible a quantitative assessment
`of the chemical stability of RNA in various aqueous environ-
`ments. In addition, these data provide kinetic parameters that
`would be useful for investigating the catalytic strategies used
`by RNA-cleaving enzymes and for designing RNA-cleaving
`enzymes, ribozymes, and nuclease mimics.8,9
`
`Results and Discussion
`RNA Substrates and the RNA Cleavage Assay System.
`All
`internucleotide linkages within an RNA polymer can
`undergo cleavage by transesterification. Therefore, the kinetic
`analysis of RNA phosphoester cleavage in the context of large
`RNA polymers can be problematic. To examine the cleavage
`of an individual RNA phosphodiester, we synthesized five
`different 22-nucleotide DNAs that contain a single embedded
`RNA linkage located between nucleotides 12 and 13 (e.g.,
`substrate S(ApG); Figure 1A). Degradative mechanisms that
`spontaneously act upon DNA, such as phosphoester hydrolysis,10
`oxidative cleavage,11 and cleavage as a result of depurination,12
`are all at least several orders of magnitude slower than RNA
`transesterification under alkaline conditions. In this context, the
`cleavage of the lone RNA phosphodiester linkage can be
`assessed without complications resulting from unwanted cleav-
`age of the flanking DNA phosphodiesters. As a result, these
`chimeric “embRNA” substrates13 provide an excellent alternative
`to the use of dinucleotides (e.g., see refs 7 and 14) for model
`compounds to study RNA degradation.
`Higher-ordered nucleic acid structure could significantly affect
`the rate at which an internucleotide linkage undergoes cleavage
`by transesterification (see below). However, none of the five
`chimeric substrates used in this study are predicted15,16
`to form unimolecular or bimolecular secondary structures that
`could bias cleavage rates and therefore complicate data collec-
`tion and analysis (the DNA mfold server can be accessed on
`the Internet at www.ibc.wustl.edu/∼zuker/dna/form1.cgi). The
`reaction conditions most frequently used for data collection
`involved high pH or high temperature, both of which are
`expected to disfavor the formation of higher-ordered structures.
`Moreover, absorbance/temperature profiles conducted with the
`all-DNA version of S(ApG) at pH 7 and 10.6 indicate that no
`significant higher-ordered structures are formed that otherwise
`could bias the kinetics of RNA cleavage (data not shown).
`Cleavage of a 5′ 32P-labeled version of the substrate S(ApG)
`at the RNA phosphodiester produces a single new 32P-labeled
`that corresponds to the 12-nucleotide 5′-cleavage
`product
`(7) Ja¨rvinen, P.; Oivanen, M.; Lo¨nnberg, H. J. Org. Chem. 1991, 56,
`5396-5401.
`(8) Sigman, D. S.; Mazumder, A.; Perrin, D. M. Chem. ReV. 1993, 93,
`2295-2316.
`(9) Trawick, B. N.; Daniher, A. T.; Bashkin, J. K. Chem. ReV. 1998, 98,
`939-960.
`(10) Guthrie, J. P. J. Am. Chem. Soc. 1977, 99, 3991-4001.
`(11) Carmi, N.; Balkhi, S. A.; Breaker, R. R. Proc. Natl. Acad. Sci. U.S.A.
`1998, 95, 2233-2237.
`(12) Lindahl, T. Nature 1993, 362, 709-715.
`(13) Jenkins, L. A.; Bashkin, J. K.; Autry, M. E. J. Am. Chem. Soc.
`1996, 118, 6822-6825.
`(14) Koike, T.; Inoue, Y. Chem. Lett. 1972, 569-572.
`
`Figure 1.
`(A) Substrate oligonucleotides used for cleavage rate
`determinations. Each 22-nucleotide DNA contains a single ribonucleo-
`tide (underlined) that forms an embedded RNA phosphodiester linkage
`(triangle). Oligonucleotides are radiolabeled at the 5′ terminus with
`32P (*p). Nucleotide changes made to the parent sequence “S(ApG)”
`are identified by the arrows, where the sequence variations for each
`substrate are identified within the parentheses. (B) Representative
`autoradiogram of a typical separation of 32P-labeled substrate (S) and
`32P-labeled product (P; 12-nucleotide 5′-cleavage fragment) made by
`denaturing 10% PAGE. Transesterification reactions with S(ApG) were
`conducted at pH 13.4 for the times indicated (see Experimental Section).
`The 5′-cleavage fragments carry heterogeneous termini consisting of
`2′,3′-cyclic phosphate, or the subsequent hydrolysis products 2′
`phosphate or 3′ phosphate. Fragments with these different chemical
`configurations are not resolved with this analysis. (C) Representative
`plot depicting the decay of the RNA linkage as derived from the data
`in part B. For this plot, the negative slope of the line represents the
`rate constant for the reaction in 3.16 M KCl, pH 13.4 (k ) 0.014 min-1).
`fragment as expected. Denaturing polyacrylamide gel electro-
`phoresis is used to separate substrate and product molecules,
`and the resulting 32P-labeled reaction products are imaged by
`autoradiography (Figure 1B). The relative amounts of each
`radiolabeled oligonucleotide were quantitated using a phosphor-
`based imaging system, which provides an extremely sensitive
`and accurate determination of product yields. A plot of the
`natural log of the fraction of substrate remaining at various
`incubation times forms a straight line over 3 half-lives when
`incubated at pH 13.4 (Figure 1C). This result indicates that the
`RNA transesterification reaction proceeds to near completion
`without variation in reaction kinetics. The negative slope of the
`resulting line provides the rate constant (k) for RNA trans-
`esterification under the specific reaction conditions surveyed.
`This same approach was used to establish rate constants under
`all reaction conditions.
`Dependence of RNA Transesterification on pH. The rate
`constant for cleavage of the substrate S(ApG) (Figure 1A) was
`determined at different pH values ranging between 8.75 and
`14.5. At pH values below 12, 2-(cyclohexylamino)ethanesulfonic
`CUREVAC EX2028
`Page 2
`
`

`

`5366 J. Am. Chem. Soc., Vol. 121, No. 23, 1999
`
`Li and Breaker
`
`is predicted to have a slope of 0, as is seen with the data
`generated with S(ApG) near pH 14 (Figure 2, inset). The pKa
`for the 2′-hydroxyl group can be inferred from the pH-dependent
`rate profile by determining the pH at which k is half its
`maximum value. Under these reaction conditions (3.16 M K+,
`23 °C), the maximum value for k or kmax for S(ApG) equals
`∼0.022 min-1. The half-maximal rate constant or 1/2kmax of
`∼0.011 min-1 is obtained at pH of ∼13.1, which establishes
`the apparent pKa for the 2′-hydroxyl group of the labile
`internucleotide linkage of S(ApG) under these assay conditions.
`A pKa value of 13.1 is consistent with the assumption that the
`pKa of a lone 2′-hydroxyl group will be greater than ∼12.4,
`which is the pKa value observed21-24 for a terminal hydroxyl
`of an RNA monomer when the vicinal 2′ and 3′ oxygen atoms
`remain unesterified.
`Another critical parameter that can be established from these
`data is the rate constant for RNA transesterification at pH values
`that approximate physiological conditions. Unfortunately, ex-
`perimental determination of rate constants below pH 9 becomes
`difficult due to the extremely slow rate of the reaction. Although
`long incubation times can be used to generate measurable
`amounts of cleavage product, the background rate of substrate
`cleavage due to radiolysis begins to contribute to the total
`observed rate for substrate cleavage, and interferes with data
`interpretation. Due to the linear nature of the pH-dependent rate
`profile, extrapolating the line toward acidic pH values should
`reveal the rates for RNA transesterification under biologically
`relevant conditions. This can be done with confidence through
`pH 6, assuming that pH only affects the protonation state of
`the 2′-hydroxyl group. With reaction conditions below pH 6,
`specific base catalysis (measured at elevated temperatures)
`becomes a minor mechanism relative to the competing mech-
`anism of specific acid catalysis for RNA transesterification.1
`As a result, data extrapolation for specific base-catalyzed RNA
`cleavage under these acidic conditions also is problematic.
`Equation a1 was derived from a simple slope-intercept form
`(y ) mx + b; see Experimental Section for details on the
`derivation of equations).
`
`log k ) 0.983(pH) - 14.8
`
`(a1)
`
`This equation can be used to estimate the contribution that
`specific base catalysis makes toward the rate constant for RNA
`transesterification at any pH value below the pKa of the 2′-
`hydroxyl group. A slope (m) of 0.983 for the cleavage of
`S(ApG) was derived by plotting log k versus pH (Figure 2).
`The y-intercept (b) is determined as the y-intercept of the line
`formed by plotting pH-dependent rate constants for pH 9 to 13
`using linear regression analysis. For ApG, a y-intercept of -14.8
`min-1 is obtained at pH 0. The rate constant (y) for RNA
`transesterification under different buffer conditions can be
`predicted by substituting different values for pH (x).
`By employing a more simplified equation (eq a2) with the
`above-described parameters, specific base-catalyzed cleavage
`of an ApG RNA linkage at 23 °C, pH 6.0, and in 3.16 M K+
`is projected to proceed with an otherwise uncatalyzed rate
`constant (kbackground) of 1.30 × 10-9 min-1.
`(21) Levene, P. A.; Simms, H. S.; Bass, L. W. J. Biol. Chem. 1926, 70,
`243-251.
`(22) Izatt, R. M.; Hansen, L. D.; Rytting, J. H.; Christensen, J. J. J. Am.
`Chem. Soc. 1965, 87, 2760-2761.
`(23) Izatt, R. M.; Rytting, J. H.; Hansen, L. D.; Christensen, J. J. J. Am.
`Chem. Soc. 1966, 88, 2641-2645.
`(24) Birnbaum, G. I.; Giziewicz, J.; Huber, C. P.; Shugar, D. J. Am.
`Chem. Soc. 1976, 98, 4640-4644.
`
`CUREVAC EX2028
`Page 3
`
`Figure 2. The pH dependence of the rate of S(ApG) RNA cleavage.
`Data were collected in reaction mixtures containing CHES (shaded
`circles) and CAPS (open circles) or in unbuffered (filled circles)
`reactions. Inset: High-resolution profile depicting the pH range that
`encompasses the transition from first-order to zero-order kinetics that
`occurs near the pKa for the 2′-hydroxyl group. The apparent pKa for
`the 2′-hydroxyl group of the ApG internucleotide linkage under these
`reaction conditions (3.16 M KCl, 23 °C) is ∼13.1, as determined by
`the pH value that provides a value for k that is half-maximal (kmax ∼
`0.02 min-1).
`
`acid (CHES; pH 8.75 to 10) and 3-(cyclohexylamino)-1-
`propanesulfonic acid (CAPS; pH 10.25 to 11.75) were used to
`stabilize the hydroxide ion concentration. At pH values above
`12, appropriate amounts of KOH were added to unbuffered
`reaction mixtures to establish the desired hydroxide ion con-
`centration. No buffering agent was deemed necessary at these
`more alkaline pH conditions due to the high concentrations of
`hydroxide ions and the short incubation times that were needed
`to accurately establish rate constants for transesterification. In
`most instances, ionic strength was maintained constant in all
`reactions by supplementing each reaction mixture with KCl such
`that the total concentration of ionic potassium (K+) was 3.16
`M. This concentration was chosen to normalize the amount of
`K+ delivered to each reaction with that delivered to the reaction
`at pH 14.5 in the form of KOH.
`A linear relationship was revealed by plotting the logarithm
`of the rate constants for RNA transesterification between pH 9
`and 13 (Figure 2). This correlation, which has been well
`documented by other investigators,7,14,17-20 is due to the
`increasing fraction of 2′-oxyanion groups relative to 2′-hydroxyl
`groups that is produced under progressively higher hydroxide
`concentrations. Specific base catalysis generates a 10-fold
`increase in the number of reactive 2′-oxyanion groups for each
`unit increase in pH. This effect is reflected in Figure 2, where
`the line depicting the relationship between rate constant and
`pH maintains a slope of ∼1 below pH 13. This linear
`relationship holds until the pH of the reaction mixture ap-
`proaches the pKa of the 2′-hydroxyl group. When the pH
`matches the pKa, half of the substrate oligonucleotide already
`carries the more powerful 2′-oxyanion, and further deprotonation
`at most can have a 2-fold effect on the rate constant for the
`reaction.
`The line formed when plotting the rate constants obtained at
`pH values significantly greater than the pKa for the 2′-hydroxyl
`(15) Zuker, M. Science 1989, 244, 48-52.
`(16) SantaLucia, J., Jr.; Allawi, H. T.; Seneviratne, P. A. Biochemistry
`1996, 35, 3555-3562.
`(17) Bock, R. M. Methods Enzymol. 1967, XIIA, 218-221.
`(18) Liu, X.; Reese, C. B. Tetrahedron Lett. 1995, 36, 3413-3416.
`(19) Matsumoto, Y.; Komiyama, M. J. Chem. Soc., Chem. Commun.
`1990, 15, 1050-1051.
`(20) Weinstein, L. B.; Earnshaw, D. J.; Cosstick, R.; Cech, T. R. J. Am.
`Chem. Soc. 1996, 118, 10341-10350.
`
`

`

`Kinetics of RNA Degradation
`
`J. Am. Chem. Soc., Vol. 121, No. 23, 1999 5367
`
`k ) 10{-14.8+0.983(pH)}
`
`(a2)
`
`At pH 6, both specific base catalysis and specific acid
`catalysis are near a minimum,1 although cleavage by a depu-
`rination/(cid:2)-elimination mechanism becomes increasingly sig-
`nificant.12 However, the kbackground value at pH 6 is of particular
`interest as it reflects the rate at which transesterification will
`occur under aqueous conditions that are presumed to be the
`safest for the long-term storage of RNA.
`the rate
`Equation a2 can be used to accurately predict
`constants for RNA transesterification for pH values that range
`from 6 to values that approach the pKa for the 2′-hydroxyl group.
`This equation, however, does not take into consideration that
`rate constants begin to plateau at pH values near and above the
`pKa. Equation a3 is a modified form of a2 that can be used to
`
`k ) 10{-14.80-0.983log(Ka+[H+])}
`
`(a3)
`
`approximate the rate constants for pH values beginning at pH
`6 and ranging beyond pH 13.1, which equals pKa determined
`for the 2′-hydroxyl group of S(ApG). Values for the equilibrium
`constant Ka and [H+] are the negative antilogarithms of the pKa
`for the 2′-hydroxyl group and the pH of the reaction mixture,
`respectively.
`Power of 2′-Hydroxyl and 2′-Oxyanion Groups as Nu-
`cleophiles. Nucleophilic attack by the 2′ oxygen is the rate-
`limiting step for RNA transesterification under alkaline reaction
`conditions. Therefore, deprotonation of the 2′-hydroxyl group
`to yield the more nucleophilic 2′ oxyanion directly influences
`the rate constant for transesterification. Beginning above pH 5,
`increasing the pH of the reaction mixture results in a corre-
`sponding linear increase in the logarithm of the rate constants
`for RNA transesterification.1 This finding indicates that, even
`at pH values lower than 6, the deprotonated 2′-oxyanion form
`of the nucleophile dictates the rate constant observed for RNA
`cleavage.
`Conversely, the data indicate that the protonated form of the
`nucleophile, the 2′-hydroxyl group, makes a negligible contribu-
`tion to cleavage kinetics, even at pH values near 5. This is true
`despite the fact that at pH 5 the 2′-hydroxyl group is present in
`more than 108-fold excess over the deprotonated form. There-
`fore, to play the dominant role in nucleophilic attack, the 2′-
`oxyanion group must be at least 108-fold more powerful than
`its protonated counterpart. For comparison, the oxyanion group
`derived from methanol is ∼106-fold more nucleophilic than its
`corresponding hydroxyl form when attacking a methyl iodide
`substrate in methanol solvent.25 The ∼100-fold discrepancy
`between these two values most likely reflects the influences
`that reaction parameters such as solvent and substrate types
`(among many other factors) have on the relative nucleophilic
`power of related attacking groups.26
`Dependence of pKa on the Concentration of Potassium
`Ions. Several reaction parameters other than pH also may affect
`the rate constants for RNA transesterification. We find that K+
`concentration has a significant influence on the rate constants
`at a given pH. Specifically, increasing K+ concentrations yields
`progressively higher rate constants at pH values below 13 (data
`not shown). Unfortunately, we could not test whether the original
`kmax of 0.022 min-1 attained at 3.16 M [K+] (Figure 2) changes
`with different concentrations of ionic potassium, because
`(25) Pearson, R. G.; Sobel, H.; Songstad, J. J. Am. Chem. Soc. 1968,
`90, 319-326.
`(26) Pross, A. In Theoretical and Physical Principles of Organic
`ReactiVity; John Wiley & Sons: New York, 1995; pp 232-234.
`
`Influence of [K+] on the apparent pKa of the 2′-hydroxyl
`Figure 3.
`group at 23 °C.
`establishing pH values above the pKa for the 2′-hydroxyl group
`using KOH would introduce excessive amounts of K+ into the
`reaction mixture. However, ionic potassium is expected to
`stabilize the deprotonated form of the 2′-hydroxyl group through
`ionic interactions, thereby lowering its pKa and producing the
`increases in the observed rate constants. As a result, we speculate
`that the pKa, and not the maximum rate constant, is most likely
`to be influenced by K+.
`If this assumption is correct, then observed increases in rate
`constants with increasing [K+] exclusively reflect decreasing
`pKa values for the 2′-hydroxyl group. Apparent pKa values for
`the 2′-hydroxyl group at different concentrations of K+ were
`determined by establishing the pH values needed to achieve
`1/2kmax. A comparison of apparent pKa values versus [K+] forms
`a linear plot with a slope of -0.24 (Figure 3), which reflects
`the ∼0.6 unit shift in apparent pKa over the concentration range
`of potassium used in this study. Although an apparent pKa of
`13.1 was established at high potassium concentration, a value
`of 13.7 is expected with potassium concentrations that approach
`those observed inside cells (∼0.25 M). This relationship holds
`even when taking into consideration the activity coefficients
`for KCl. At 25 °C the activity coefficients that more accurately
`reflect the free concentration of K+ drop significantly between
`0 and 0.3 M.27 However, the coefficients decrease only slightly
`over the concentration range used in this study and, as a result,
`corrections for K+ activities were not made.
`Rate constants can be calculated for RNA transesterification
`at different concentrations of K+ using eq b. The background
`rate constant (kbackground) used for this equation (1.30 × 10-9
`min-1) is the value projected for RNA cleavage at pH 6.0 and
`23 °C in the presence of 3.16 M K+ using eq a2 (see above).
`The value for [K+] should reflect the total molar concentration
`present under the reaction conditions of interest. Equation b
`begins with a kbackground under the maximum [K+] tested (3.16
`M), and the additional factor adjusts this value for lower K+
`concentrations.
`kprojected ) kbackground × 10{-0.24(3.16-[K+])}
`Transesterification Rates and the Concentration of Di-
`valent Magnesium. Cationic metals have profound effects on
`the folded structures and chemical stability of RNA molecules.28
`It has been known for many years that various divalent metal
`ions strongly promote RNA degradation by transesterification.29-32
`(27) Weast, R. C. In CRC Handbook of Chemistry and Physics; CRC
`Press: Cleveland, OH, 1976; p D-153.
`(28) Pan, T.; Long, D. M.; Uhlenbeck, O. C. In The RNA World;
`Gesteland, R. F., Atkins, J. F., Eds.; Cold Spring Harbor Laboratory Press:
`New York, 1993; pp 271-302.
`(29) Butzow, J. J.; Eichorn, G. L. Biopolymers 1965, 3, 95-107.
`(30) Farkas, W. R. Biochim. Biophys. Acta 1968, 155, 401-409.
`CUREVAC EX2028
`Page 4
`
`(b)
`
`

`

`5368 J. Am. Chem. Soc., Vol. 121, No. 23, 1999
`
`Li and Breaker
`
`magnesium hydroxide complex,34 and this would complicate
`the interpretation of any data that were derived under more
`alkaline conditions. As expected, increasing concentrations of
`MgCl2 between 0.005 and 0.05 M bring about progressively
`faster rates of RNA cleavage. A plot of log k versus log [Mg2+]
`at pH 9.5 reveals a linear relationship with a slope of 0.80 over
`the range of metal
`ion concentrations used (Figure 4B),
`reflecting the substantial impact that this particular divalent metal
`has on the rate of RNA transesterification.
`In addition, we found that the concentration of K+ also affects
`the rate enhancement caused by the addition of Mg2+. At
`concentrations of K+ in excess of 0.5 M, the addition of 0.005
`M Mg2+ results in an ∼3-fold increase in the rate constant.
`the influence of Mg2+ increases sharply as the
`However,
`concentration of K+ approaches zero (Figure 4C). The loss of
`Mg2+-promoted RNA cleavage at higher ionic strengths may
`be due to K+-mediated shielding of the negatively charged
`phosphates which may preclude the formation of RNA-Mg2+
`complexes. Mg2+ may bind more strongly to the RNA phos-
`phodiester backbone at low [K+], thereby accelerating trans-
`esterification by any of several mechanisms.
`The effects of Mg2+ that are made apparent in Figure 4 could
`be due to general base catalysis, general acid catalysis, catalysis
`by direct Lewis acid function, or in part may be due to the
`increase in ionic strength resulting from increasing [Mg2+].
`Regardless of the mechanism,
`the results described above
`indicate that rate constants for physiological pH conditions and
`concentrations of Mg2+ can be predicted with reasonable
`accuracy. Equation c can be used to project the rate constants
`for RNA cleavage with Mg2+ concentrations between 0.005 and
`0.05 M and with concentrations of K+ between 0.03 and 3.16
`M. However, for reaction conditions with ion concentrations
`that are outside these ranges, this empirical equation may not
`provide accurate projections.
`kprojected ) kbackground × 69.3([Mg2+]0.8) × 3.57([K+])-0.419
`(c)
`
`The values for the first factor (kbackground) can be established
`using eq a between pH 6 and 13. The second factor represents
`the enhancement that Mg2+ contributes to the overall rate
`constant as determined by the data in Figure 4B. The third factor
`reflects the influence that K+ has on the RNA cleavage activity
`of Mg2+.
`Transesterification Rates and Temperature. Temperature
`of the reaction mixture was found to have a substantial effect
`on the rate of RNA transesterification. In a reaction buffered
`with 50 mM CAPS (pH 10.7 at 23 °C), a linear trend between
`the rate constant and temperature is observed with progressively
`higher temperatures producing modest increases in the rate
`constant (Figure 5). However, the value for each rate constant
`must be corrected by considering the variation in pH that occurs
`when altering the temperature of the reaction mixture.35 Specif-
`ically, the relevant pKa of CAPS changes with increasing
`temperature (d{pKa}/dT) by a value of -0.032 unit per degree
`Celsius (see Experimental Section).
`After correction for the changes in pH, the temperature-
`dependent variation in rate constants becomes much more
`pronounced. An increase in the rate constant of greater than 3
`orders of magnitude is apparent between 4 and 50 °C after
`correcting the data for the inherent variation in pH (Figure 5A).
`A linear trend with a slope of 0.07 is observed for the influence
`(34) Dean, J. A., Ed. In Lange’s Handbook of Chemistry, 13th ed.;
`McGraw-Hill: New York, 1985; pp 4-73.
`(35) Ellis, K. J.; Morrison, J. F. Methods Enzymol. 1982, 87, 405-426.
`CUREVAC EX2028
`Page 5
`
`Figure 4.
`Influence of [Mg2+] on the rate constant for RNA
`transesterification. Assay reactions for data depicted in part A were
`conducted at 23 °C in 3.16 M K+ in the absence or presence of Mg2+
`as indicated, where the buffer (50 mM CHES) was adjusted to various
`pH values as shown. Similarly, reactions for data depicted in part B
`were conducted in 50 mM CHES (pH 9.5 at 23 °C) with various
`concentrations of MgCl2. The effect of [K+] on Mg2+-promoted RNA
`cleavage (C) is expressed as the ratio (Δk) of the rate constants in the
`presence of Mg2+ versus the absence of Mg2+, at various concentrations
`of K+.
`
`The catalytic roles of metal ions are diverse. For example,
`magnesium cations presumably could promote the cleavage of
`RNA by facilitating proton transfer or by serving as a Lewis
`acid catalyst.28,33
`We examined the effects of divalent magnesium cations on
`the rate constant for RNA transesterification using concentra-
`tions of Mg2+ that are of physiological relevance. At pH 9.5
`and 3.16 M K+, the addition of divalent magnesium to 0.005
`M induces an ∼3-fold increase in rate constant over that
`observed in the absence of divalent metal ions. Moreover, this
`relationship holds for pH values ranging between 8.8 and 9.8
`(Figure 4A). Assays under higher pH values were not attempted
`because higher hydroxide ion concentrations are expected to
`promote the formation of substantial amounts of insoluble
`(31) Eichorn, G. L.; Tarien, E.; Butzow, J. J. Biochemistry 1971, 10,
`2014-2019.
`(32) Zago´rowska, I.; Kuusela, S.; Lo¨nnberg, H. Nucleic Acids Res. 1998,
`26, 3392-3396.
`(33) Zhou, D.-M.; Taira, K. Chem. ReV. 1998, 98, 991-1026.
`
`

`

`Kinetics of RNA Degradation
`
`J. Am. Chem. Soc., Vol. 121, No. 23, 1999 5369
`
`Figure 5. The effects of temperature on the rate constant for RNA
`transesterification. (A) Temperature dependence of rate constants plotted
`before (open circles) and after (filled circles) correction for pH variation.
`(B) Arrhenius plot of the corrected data depicted in part A.
`
`of temperature on RNA transesterification in the presence of
`3.16 M K+. The rate constant for a particular temperature can
`be calculated using eq d, where T1 is the temperature of interest,
`T0 is a temperature where the rate constant is known, and kT0 is
`the value for the known rate constant.
`× 10[0.07(T1-T0)]
`
`(d)
`
`kT1
`
`) kT0
`
`Cleavage of S(ApG) also produces a linear relationship
`between the reciprocal of the temperature and the natural
`logarithm of the rate constant (Figure 5B). This standard
`Arrhenius relationship indicates that the reaction kinetics are
`elementary and are not complicated by structure formation or
`other possible factors. The activation energy (Ea) of the reaction
`is 29 kcal mol-1, compared with an Ea of 13.1 kcal mol-1
`determined for cleavage of RNA by a hammerhead ribozyme.36
`Transesterification Rate and Nucleotide Base Composi-
`tion. The precise structural orientation adopted by RNA linkages
`is known to be an important factor that influences the rate of
`cleavage by transesterification. It has been proposed37 that the
`alignment of the attacking 2′ oxygen must make its approach
`to the phosphate from the side opposite that of the 5′-oxygen
`leaving group for productive nucleophilic attack to occur.
`Indeed, this principle is consistent with the observed hydrolytic
`stability of duplex RNA, whose base stacking precludes the 3′,5′-
`internucleotide linkages within its helical structure from adopting
`this permissive “in-line” orientation.38,39 In contrast, the analo-
`gous 2′,5′-internucleotide linkage becomes hypersensitive to
`strand scission when positioned within an otherwise normal
`RNA helix.38 This latter observation is consistent with the
`hypothesis that a 2′,5′-phosphodiester linkage favors an in-line
`conformation when embedded within a typical A-form RNA
`helix.
`Likewise, it follows that base stacking within a single-stranded
`oligonucleotide that simulates the formation of an A-form helix40
`would also contribute to chemical stability. Therefore one would
`expect that nucleotide base composition, with its influence on
`the magnitude of single-stranded base stacking, may affect the
`rate at which an otherwise unconstrained RNA linkage under-
`goes cleavage by transesterification. Several early reports
`indicate that certai