Tài liệu Báo cáo khoa học: Kinetic characterization of the first step of the ribozyme-catalyzed trans excision-splicing reaction docx

The rate constant for the substrate-cleavage reaction is 60-fold lower than that reported for the first step of the self-splicing reaction using a Tetrahymena thermo-philaribozyme, regar

ribozyme-catalyzed trans excision-splicing reaction

P Patrick Dotson II*, Joy Sinha* and Stephen M Testa

Department of Chemistry, University of Kentucky, Lexington, KY, USA

We previously reported that a group I intron-derived

ribozyme from Pneumocystis carinii can catalyze the

excision of a targeted sequence from within an RNA

transcript [1] This reaction, called trans

excision-splicing (TES), consists of two steps: substrate

cleav-age (an intramolecular transesterification reaction)

followed by exon ligation (Fig 1) In the

substrate-cleavage reaction, the phosphodiester backbone of an

intermolecular substrate is cleaved via nucleophilic

attack by the 3¢ terminal guanosine (G336),

generat-ing 5¢ and 3¢ exon intermediates [1a] In the

exon-liga-tion step, the nucleophilic 5¢ exon intermediate

attacks a phosphodiester backbone position within

the 3¢ exon intermediate, simultaneously ligating the

exons together and excising the internal segment The

substrate-cleavage reaction step is analogous to the 5¢ splice-site cleavage reaction in self-splicing [2], except that self-splicing utilizes an exogenous guanosine cofactor as the nucleophile The TES substrate-clea-vage reaction is also directly analogous to the natu-rally occurring self-cyclization reaction, which results

in the formation of full-length or truncated circular group I introns, in that they both utilize the 3¢ termi-nal guanosine of the intron (or ribozyme) as nucleo-philes [3–5]

Several studies have dissected the individual steps of RNA-catalyzed reactions through the establishment of kinetic frameworks [6–19] This approach has been mechanistically informative and has greatly advanced our understanding of the chemical basis of RNA

Keywords

group I intron; ribozyme; RNA; self-splicing;

trans excission-splicing

Correspondence

S M Testa, Department of Chemistry,

University of Kentucky, Lexington, KY

40506, USA

Fax: +1 859 323 1069

Tel: +1 859 257 7076

E-mail: testa@email.uky.edu

*These authors contributed equally to this

work

(Received 3 March 2008, revised 7 April

2008, accepted 14 April 2008)

doi:10.1111/j.1742-4658.2008.06464.x

Group I introns catalyze the self-splicing reaction, and their derived ribo-zymes are frequently used as model systems for the study of RNA folding and catalysis, as well as for the development of non-native catalytic reactions Utilizing a group I intron-derived ribozyme from Pneumocystis carinii, we previously reported a non-native reaction termed trans excision-splicing (TES) In this reaction, an internal segment of RNA is excised from an RNA substrate, resulting in the covalent reattachment of the flanking regions TES proceeds through two consecutive phosphotranseste-rification reactions, which are similar to the reaction steps of self-splicing One key difference is that TES utilizes the 3¢-terminal guanosine of the ribozyme as the first-step nucleophile, whereas self-splicing utilizes an exog-enous guanosine To further aid in our understanding of ribozyme reac-tions, a kinetic framework for the first reaction step (substrate cleavage) was established The results demonstrate that the substrate binds to the ribozyme at a rate expected for simple helix formation In addition, the rate constant for the first step of the TES reaction is more than one order

of magnitude lower than the analogous step in self-splicing Results also suggest that a conformational change, likely similar to that in self-splicing, exists between the two reaction steps of TES Finally, multiple turnover is curtailed because dissociation of the cleavage product is slower than the rate of chemistry

Abbreviations

GBS, guanosine-binding site; RE1, recognition element 1; RE2, recognition element 2; RE3, recognition element 3; TES, trans

excision-splicing.

catalysis The fine details regarding the mechanism by which the first step of the TES reaction occurs is lar-gely unknown In addition, little is known regarding the kinetics of 3¢ terminal guanosine-catalyzed reac-tions Therefore, a minimal kinetic framework for this substrate-cleavage reaction was established (Fig 2) There are multiple conclusions drawn from this kinetic framework as they relate to the TES reaction The rate constant for the substrate-cleavage reaction is

 60-fold lower than that reported for the first step of the self-splicing reaction using a Tetrahymena thermo-philaribozyme, regardless of whether an intermolecular

or intramolecular guanosine is being utilized as the first-step nucleophile [6,15] The rate constant for the first step of the TES reaction is only fourfold greater than that for substrate dissociation Furthermore, multiple turnover is curtailed because dissociation of the cleavage product is slower than the rate of cleavage Lastly, the results indicate that a conformational change exists between the two steps of the TES reac-tion Taken together, these results further demonstrate how group I intron-derived ribozymes exploit native recognition elements and catalytic sites to catalyze non-native, multi-step reactions

Results

A kinetic scheme for the substrate-cleavage reaction, which is reaction step 1 in Fig 1, is summarized in Fig 2 One complication in studying the substrate-cleavage reaction is that the second reaction step of TES (exon–ligation) proceeds immediately after the first step [20] To prevent the second reaction step while allowing the first, we previously utilized a sub-strate with a deoxyguanosine at the second step reac-tion site, [r(5¢-AUGACUdGCUC-3¢)], which prevents the second reaction step [1a] We found that the observed rate constant for the substrate-cleavage reaction using the deoxyguanosine substrate (kobs=

3 ± 0.5Æmin)1) is comparable, within error, to the normal substrate (kobs= 3.7 ± 0.2Æmin)1l; data not

G

(RE3)

U A

a

5′

U

U

A

a u g

a c u

U A G

G

A U

5′

G

c u c

a u g

a c u

U

A

G

(RE2)

5 ′ augacucuc 3′

Product dissociation

Ribozyme binding

Substrate cleavage

P1

(RE1)

P1

(RE1)

P10

(RE3)P10

(RE3) A

G

5′

G

C

5′

c u c c u c

3′

Ribozyme

P1

(RE1)

1

P1

P

(RE1)

(RE2)

u g a c u

A

G

U

G

5′

G

C

A

U

3′ (10-mer substrate)

Step 1

Ribozyme

P1

(RE1)

P1

(RE1)

P10

(RE3)

(RE2) A

G

U

G

5′

C

3′

Ribozyme

(9-mer product)

G

g1

g1

G

G

u c c

g1

u c c

3′

g1

g1

g1

Fig 1 Schematic of the two-step TES reaction The rPC ribozyme

is in uppercase lettering and, the 10-mer substrate is in lowercase lettering, and the ribozyme recognition elements recognition ele-ment RE1 and RE3 base pair with the substrate to form helices P1 and P10, respectively Note that recognition elements RE1, RE2 and RE3 are so named because they correspond to the regions in self-splicing introns that bind the exon substrates The sites of catalysis for the first step (substrate cleavage) and the second step (exon ligation) are shown with arrows, and the guanosine to be excised (G1) is circled The diagram shows only the recognition elements of the ribozyme.

shown) Note that the kobs value of the normal

reac-tion is in reasonable agreement with the previously

reported value of 4Æmin)1 [1] Therefore, the

deoxy-guanosine substrate reasonably mimics the normal

substrate as a first reaction step substrate

Impor-tantly, this substrate inhibits the exon-ligation step,

allowing us to isolate and analyze only the first

reac-tion step

Observed rate constants for substrate cleavage,

kobsand k2

Experiments under ribozyme excess conditions were

used to determine the pseudo-first-order rate constant

for the substrate-cleavage reaction Note that under

these reaction conditions the ribozyme–product

com-plex is denatured upon addition of stop buffer, and so

product dissociation is not observable Therefore, these

experiments measure the rate of substrate cleavage

from the ribozyme–substrate complex

The observed rate constants (kobs) were measured in

reactions containing various ribozyme concentrations

(5–300 nm) and 1.3 nm of 5¢-end radiolabeled substrate

(Fig 3A,B) As seen in Fig 3C, the observed rate

con-stants at the higher ribozyme concentrations (100–

350 nm) are independent of ribozyme concentration,

indicating that saturation of the ribozyme has

been reached Values of k2= 4.1 ± 0.5Æmin)1 and

KM= 102 ± 0.4 nm were obtained by fitting the

aver-age kobs values to the Michaelis–Menten equation

Herein, k2 represents the maximum first-order rate of

substrate cleavage under single turnover conditions

For lower ribozyme concentrations (5–40 nm) the kobs

values for the substrate-cleavage reactions increase

linearly with ribozyme concentration This linear

dependence reflects the apparent second-order rate

constant, k2⁄ KM, and the slope gives a value of (2.8 ±

0.5)· 107Æm)1Æmin)1 (Fig 3C, inset) Note that the

values obtained are similar to those reported

previ-ously for group I intron-derived ribozymes (Table 1)

[14,17,21]

Dependence of substrate cleavage on pH

It has been reported that the rate of the substrate-cleavage step in Tetrahymena [22–25], Anabaena [14] and Azoarcus [17] group I introns, as well as reaction steps for some small ribozymes [26–28], show a log-linear increase in the reaction rate constant with increasing pH in the acid range (up to pH 7) This is consistent with a single deprotonation step that takes place prior to the actual cleavage reaction [29] This is also consistent with the observed rate constant at a given pH being equivalent to the rate constant of the chemical step at that pH This was investigated for the Pneumocystis ribozyme by measuring the pH depen-dence of the observed rate constant of the substrate-cleavage reaction As seen in Fig 4, the logarithm of the observed rate constant increases linearly with pH

in the range 5–7 (slope = 0.5 ± 0.03), but not at higher pH values In the case of the Tetrahymena group I intron-derived ribozyme, such non-linear behavior was attributed to a pH-dependent conforma-tional change occurring within the ribozyme [24,25] This conformational change thus sets a limit on the observed rate constant of cleavage (k2), even though the rate constant of chemistry (kc) is expected to con-tinue to increase with increasing pH [24,25] Appar-ently, for our substrate-cleavage reaction, the rate of the chemical step is being masked by a conformational change, and so k2 is not equivalent to kc The rate of chemistry (kc), however, can be approximated by extrapolating the log-linear increase that occurs between pH 5 and 7 to higher pH values In our case,

kc is then approximately equal to 5.7 ± 1.1Æmin)1 at

pH 7.5 (Fig 4)

Control experiments were run to determine whether the observed rate constants shown in Fig 4 were being influenced by the specific buffers utilized in the respec-tive reactions We found that the values obtained using Mes and Hepes at pH 6.8 were within 1 SD of each other This was also true using Hepes and Epps at

pH 7.5 Apparently, there is not a buffer-specific effect

k1= 1 x 10 7

M–1· min –1

k–1= 0.9 min–1

k3= 3.5 x 10 3 M–1 ·min –1

E + P

E P

Fig 2 Kinetic scheme for the substrate-cleavage reaction E denotes the rPC ribozyme, S denotes the 10-mer substrate, and P denotes the 6-mer cleavage product All rate and equilibrium constants were measured or calculated (boxed values) in this report The scheme includes rate constants for substrate association (k1) and dissociation (k)1), cleavage (k2), and product association (k3) and dissociation (k)3) Note that the observed rate constant for the cleavage step (k 2 ) is distinguishable from the actual rate constant for chemistry (k c ) The scheme also includes equilibrium constants for substrate (Kd) and product (Kd) dissociation.

on the observed rate constants (kobs) Note that we

have not examined the rates of substrate cleavage

outside the pH range depicted because protonation or

deprotonation of nucleotides is expected to cause general chemical denaturation of the ribozyme [30]

Rate constant for substrate dissociation, k)1 The upper limit of the rate constant for substrate dissociation was measured in a pulse–chase experiment (Fig 5A) In this experiment, the time chosen for t1 (30 s) was such that a significant fraction of substrate would remain unreacted After the addition of the chase, which in this case is dilution with buffer, aliqu-ots were removed at designated times (defined as t2) up

to 15 min An otherwise identical reaction, but without the added chase, was carried out in parallel The ribo-zyme–substrate complex will decay through substrate cleavage (k2) and dissociation (k)1) Therefore, measur-ing the observed rate constant durmeasur-ing the chase phase will reflect both substrate cleavage and dissociation This is summarized by: kobs, chase= k2+ k)1 [6,9] Note that in this experiment k2= kobs, no-chase

The observed rate constants for the chase reaction (kobs, chase= 2.5 ± 0.04Æmin)1) and in the reaction without added chase (kobs, no-chase= 1.5 ± 0.01Æmin)1) were obtained from a single-exponential fit of product formation against t2 (Fig 5B) The substrate dissocia-tion rate constant (k)1= 0.9 ± 0.04Æmin)1) was then determined using Eqn (2) (see Experimental proce-dures) Note that k)1 is comparable in value to the cleavage step (k2), implying that the ribozyme– substrate complex does not reach equilibrium with free ribozyme prior to the cleavage step

Rate constant for substrate association, k1

The kinetic data indicate that substrate dissociation is comparable in value to the cleavage step This implies that the second-order rate constant for substrate cleavage, k2⁄ KM, will be a combination of substrate association (k1), dissociation (k)1) and cleavage (k2) steps Thus, the second-order rate constant can be represented as k2⁄ KM= k1k2⁄ (k)1+ k2) [31] As discussed earlier, a value of 2.8· 107Æm)1Æmin)1 was obtained for the second order rate constant k2⁄ KM Using this value of k2⁄ KM and the values of k2 and

k)1 (4.1Æmin)1 and 0.9Æmin)1 respectively), the calcu-lated value of k1is 3.4· 107Æm)1Æmin)1

For confirmation, k1 was directly measured in a pulse–chase experiment (Fig 5A) In this case, various concentrations of ribozyme and radiolabeled substrate were combined for varying times, tl(15–120 s) During the pulse phase, t1, the concentrations of the ribozyme, substrate and ribozyme–substrate complex are predicted to approach equilibrium, where the rate of

Time (min)

0.25 min 0.5 0.75 1 2 3 4 5 10 15 min Substrate

Product

Ribozyme (n M )

kobs

–1 )

0

1

2

3

0 0.5 1 1.5

kobs

-1 ))

A

B

C

Fig 3 Substrate-cleavage reactions All reactions were conducted

in H10Mg buffer (A) Representative polyacrylamide gel with the

5¢-end labeled substrate and 166 n M rPC ribozyme The positions of

the substrate and the substrate-cleavage product on the gel are

labeled The lane marked (+) buffer contains a 15-min reaction in

the absence of the ribozyme (B) Representative plot of the

sub-strate-cleavage reaction at ribozyme concentrations of 5 n M ( ),

10 n M (s), 20 n M (h), 40 n M (e), 166 n M (D) and 300 n M (d).

Observed rate constants (k obs ) were obtained from these plots and

are the average of two independent assays All data points between

the two independent assays have a standard deviation < 15%.

(C) Non-linear least squares fit to the Michaelis–Menten equation of

the average kobs values from (B) versus ribozyme concentration

(0–350 n M ) The plot resulted in a value of k2= 4.1 ± 0.5Æmin)1and

K M = 102 ± 0.4 n M respectively These values are the average of

the two independent assays The inset shows a representative plot

of the average kobsvalues from (B) versus ribozyme concentration

(5–40 n M ) The resulting k2⁄ K M value (2.8 ± 0.5 · 10 7 Æ M )1Æmin)1) is

the average of the two independent assays.

substrate association equals substrate dissociation [6,9].

For the chase phase, the mixtures were then incubated

for a time t2 = 15 min, which ensures that essentially

every substrate molecule that binds to the ribozyme

during tl is converted to product Therefore, the

amount of product formed during the chase period is

representative of the amount of ribozyme–substrate

complex formed during t1 Note, however, that if

k)1 k2, then both processes will be occurring during

t2 The amount of product formed was plot against time t1 (Fig 6A) The kobs values reflect the rate of approach to equilibrium of the ribozyme–substrate complex formation, which is represented by kobs= k1 [E] + k)1 [6,9] The slope of the plot of kobs versus ribozyme concentration gives the rate of substrate association, kl= (1 ± 0.01)· 107Æm)1Æmin-l (Fig 6B), which is in reasonable agreement (for ribozyme reac-tions) with the calculated value above

Reversibility of the substrate-cleavage reaction Under single-turnover conditions, the first-order rate constant (k2) of the substrate-cleavage reaction is 4.1Æmin)1(Fig 2), with a typical end point of 70–80% Over the period of 15–60 min, this end point does not change, indicating that either an internal equilibrium

Table 1 Kinetic parameters for group I intron-derived ribozyme

reactions The k cat values correspond to the k 2 values reported

throughout this text.

Ribozyme origin

kcat (min)1)

kcat⁄ K M

( M )1Æmin)1)

KMS

(l M )

kc (min)1) Pneumocystis carinii a 4.1 2.8 · 10 7 0.102 5.7

Tetrahymena thermophila b 0.1 9 · 10 7 0.001 350

Azoarcus sp BH72d 0.38 8.5 · 10 5

a

Substrate-cleavage reaction (endogenous guanosine-mediated) for

the Pneumocystis ribozyme (rPC) with 10 m M MgCl2, 25 m M Hepes

(pH 7.5) and substrate (5¢-AUGACUdGCUC-3¢) at 44 C b

Substrate-cleavage reaction (exogenous guanosine-mediated) of the

Tetrahy-mena ribozyme (L21-ScaI) with 0.5 m M guanosine, 10 m M MgCl2,

50 m M Mes (pH 7) and substrate (5¢-G 2 CCCUCUAAAAA-3¢) at 50 C

[6].cSubstrate-cleavage reaction (exogenous guanosine-mediated)

of the substrate (5¢-CUUAAAAA-3¢) using the Anabaena ribozyme

(L-8 HH) with 2 m M guanosine, 15 m M MgCl2, 25 m M Hepes

(pH 7.5) at 32 C [14] d

Substrate-cleavage reaction (exogenous guanosine-mediated) of Azoarcus ribozyme (L-10 HH) with 1 m M

guanosine, 15 m M MgCl2, 25 m M Hepes (pH 7.5) and substrate

(5¢-CAUAAA-3¢) at 30 C [17].

–1

–0.5

0

0.5

1

kobs

pH

Fig 4 pH dependence of the observed rate of substrate cleavage.

Values for k obs were obtained from single-turnover reactions at

44 C using 166 n M ribozyme, 1.3 n M 5¢-end labeled substrate in

buffer containing 135 m M KCl and 10 m M MgCl2 The buffers used

for this study were 50 m M Mes (pH 5.0–6.8), 50 m M Hepes

(pH 7.0, 7.5) or 50 m M Epps (pH 8.0, 8.5) Each rate is the average

of two independent measurements and has a standard deviation

< 15% The solid line represents the log-linear increase in the data

set from pH of 5–7 (slope = 0.5 ± 0.03) The extrapolation of the

line to pH 7.5 (depicted by dashed lines) gives a value of

0.75 ± 0.1 which corresponds to rate of chemistry (k c ) of

5.7 ± 1.1Æmin)1.

Time (min)

t2

0 10 20 30

A

B

Fig 5 Determination of the rate constant for substrate dissocia-tion, k)1 (A) Scheme of the pulse–chase experiment, which was conducted in H10Mg buffer at 44 C and 166 n M ribozyme The chase was initiated by diluting the reaction mixture with H10Mg buffer (B) Representative plot of cleaved substrate, after t1, versus time (t2) with chase (closed circles) and without added chase (open circles) The resultant first-order rate constants obtained with (k obs, chase = 2.5 ± 0.04Æmin)1) and without (k obs, no-chase = 1.5 ± 0.01Æmin)1) the chase are the average of two independent assays All data points between the two independent assays have

a standard deviation < 10% From this data, the rate of substrate dissociation, k)1, is 0.9 ± 0.04Æmin)1.

exists between ribozyme–substrate and ribozyme–

product complexes or only 70–80% of the substrate is

reactive Such an internal equilibrium has previously

been identified in a G-dependent substrate-cleavage

reaction with Anabaena and Tetrahymena ribozymes

[14,15] Therefore, a pulse–chase experiment was

con-ducted such that this equilibrium, if occurring, could

be disturbed and thus detected [15] In this assay, the

substrate-cleavage reaction was allowed to proceed to

completion and then an excess of unlabeled 5¢ exon

mimic was added (Fig 7A) Addition of a large excess

of unlabeled 5¢ exon mimic is expected to prevent

rebinding of any dissociated radiolabeled substrate or

radiolabeled 5¢ exon reaction product The result

(Fig 7B) shows that a substantial fraction of the

bound radiolabeled product can be converted back to

radiolabeled substrate, hence an internal equilibrium

exists Furthermore, the results imply that product

dissociation is slower than or similar to substrate dis-sociation [15]

Equilibrium dissociation constant of the substrate-cleavage product, Kd and substrate, Kd

A trace amount of 5¢-end radiolabeled substrate-cleavage product (the 6-mer) was incubated with vari-ous concentrations of ribozyme for 90 min at 44C in H10Mg buffer, and the ribozyme–product complex was then partitioned from the unbound product

on a native polyacrylamide gel [10] The equilibrium dissociation constant of the 5¢ exon product (Kd = 69 ± 6 nm) was then determined from a plot (Fig 8) of the fraction product bound versus ribozyme concentration [32,33] For the equilibrium dissociation constant of the substrate, Kd , an estimated value can

be obtained from the equation Kd = (k)1⁄ k1) =

90 nm [19]

Time (min)

0 20 40 60 80 100

A

B

Fig 7 Substrate-cleavage products undergo the reverse reaction (A) Scheme of the pulse–chase experiment, which was conducted using 166 n M ribozyme and a trace amount of 5¢-end labeled sub-strate in H10Mg buffer at 44 C The reaction was allowed to pro-ceed for 15 min (t 1 ), followed by the addition of excess unlabeled 5¢ exon product as the chase (B) A plot of the disappearance of substrate in a normal reaction (no chase, closed circles) and reap-pearance of the substrate in the presence of chase (open circles) Each point on the plot is the average of two independent experi-ments, and have a SD of < 15% Note that the error bars present

on the graph are too small to be statistically relevant.

Time (t 1 ) (min)

Ribozyme (n M )

kobs

–1 )

0

20

40

60

0

1

2

3

A

B

Fig 6 Determination of the rate constant for substrate

associa-tion, k 1 (A) Representative plot of pulse–chase experiments in

H10Mg buffer at 44 C with five different ribozyme concentrations:

30 n M (s), 50 n M ( ), 100 n M (e), 150 n M (r) and 200 n M (d).

All data points between the two independent assays have a

standard deviation < 10% (B) Representative plot of the k obs

values against ribozyme concentration The line is fit to the

equation kobs= k1[E] + k)1 and the substrate association rate

(k 1 = 1 ± 0.01 · 10 7

Æ M )1Æmin-l

) was calculated from the slope Note that the error bars present on the graph are too small to be

statisti-cally relevant.

Rate constant for dissociation of the 5¢ exon

product, k)3

The product dissociation rate constant (k)3) was

deter-mined using a pulse–chase assay (Fig 9A), combined

with native PAGE In this assay, an excess of

ribo-zyme was mixed with 1.3 nm 5¢-end labeled 5¢ exon

mimic, which was then incubated in H10Mg buffer

containing 3.4% glycerol at 44C for 30 min An

excess amount of unlabeled 5¢ product was then added

to initiate the chase, and aliquots were removed at

des-ignated times These aliquots were directly loaded onto

a running native polyacrylamide gel to isolate the

bound and unbound fractions For quantification, the

amount of product not bound after the chase was

sub-tracted from that at time t1, which yields the amount

of product dissociated due to the chase The rate of

product dissociation (k)3= 0.09 ± 0.05Æmin)1) was

then obtained from fitting Eqn (1) to a single

expo-nential function (Fig 9B) Apparently, product

diss-ociation is slower than substrate dissdiss-ociation, which

has previously been shown for a Tetrahymena

ribo-zyme [6]

Discussion

In this report, a kinetic framework for the first step of

the TES reaction was obtained Although the TES

reaction is not known to occur in nature, the

full-length circularization reaction, which does occur

natu-rally, has mechanistic similarities [3–5] Perhaps most

importantly, both reactions utilize a 3¢ terminal

guano-sine as a nucleophile to attack the 5¢ splice site (sub-strate-cleavage site) Furthermore, neither reaction requires an exogenous guanosine cofactor, which is standard for self-splicing reactions Finally, neither reaction is dependent on the formation of helix P10 for the 5¢ splice site cleavage reaction (see Fig 1) Note that in these studies, deoxyribose-containing substrates were used to isolate the first reaction step (substrate cleavage) by preventing the second reaction step (exon ligation) In addition, the product of the first reaction step is actually the intermediate in the full TES reaction

Substrate binding The rate constant for the substrate binding the Pneu-mocystisribozyme, k1, is far below the diffusional limit

of 1011Æm)1Æmin)1 for the collision of small molecules [34] Thus, unlike classical enzymes which react near diffusion-controlled limits [31,35–37], the Pneumocystis

Time (min)

0 5 10 15 20

0 10 20 30 40 50 60

A

B

Fig 9 Determination of the rate constant for dissociation of the 5¢ exon product, k)3 (A) Scheme of the pulse–chase experiment con-ducted with rPC ribozyme and 5¢-end labeled 5¢ exon mimic in H10Mg buffer containing 3.4% glycerol at 44 C In this reaction

t1= 30 min Excess unlabeled 5¢ exon mimic was added to initiate the chase, and product dissociation was followed by native band-shift gel electrophoresis (B) Representative plot of the fraction of unbound product versus chase time, t2 The rate of product dissoci-ation, k)3, is 0.09 ± 0.05Æmin)1, which is the average of two inde-pendent assays with each data point having a standard deviation typically < 20%.

Ribozyme (n M )

0

20

40

60

Fig 8 Determination of the equilibrium dissociation constant of

the substrate-cleavage product, Kd In the reaction, various

con-centrations of ribozyme were mixed with trace amounts of 5¢-end

radiolabeled substrate-cleavage product in H10Mg buffer containing

3.4% glycerol Shown is a representative plot of the percent

sub-strate-cleavage product bound to the ribozyme versus ribozyme

concentration The resultant value of Kd is 69 ± 6 n M is the

aver-age of two independent assays with each data point having a

stan-dard deviation < 15%.

ribozyme is not under diffusion control This value,

however, is within the range (107–109Æm)1Æmin)1)

expected for the formation of RNA duplexes [38–42],

as seen with other ribozymes [6,8,13,18,19,43] Thus,

the rate of assembly of the Pneumocystis ribozyme–

substrate complex appears to be limited by the process

of helix formation Nevertheless, because k2⁄ KM(k2=

4.1Æmin)1 and KM= 102 nm respectively) approaches

the rate of substrate association, catalysis can be

expected to occur about as fast as base-pairing

between the ribozyme and substrate This is typical of

ribozymes that bind their substrates through double

helices [6,13,16,19,44]

Substrate cleavage

The observed rate constant for the substrate-cleavage

reaction, k2, under single turnover conditions is

4.1Æmin)1 Although the true rate constant for the

actual chemical step is being masked, probably by a

local conformational change that occurs after substrate

binding and before the actual chemical step, this rate

is approximately four times faster than the rate

con-stant for substrate dissociation (k)1=0.9Æmin)1)

Therefore, although the substrate is more likely to

react than it is to dissociate, the similar order of

magnitude suggests that a non-trivial fraction of the

substrate will dissociate before the substrate-cleavage

reaction occurs

The ‘catalytic power’ of an RNA-cleaving ribozyme

can be estimated by comparing the observed rate

constant of a catalyzed reaction with that of an

equivalent uncatalyzed reaction Under simulated

physiological conditions, the uncatalyzed rate constant

of the phosphotransesterification reaction (knoncat) is

estimated to be 10)9Æmin)1 [6,45] Thus, a rate of

4.1Æmin)1 for the substrate-cleavage reaction

repre-sents a catalytic rate enhancement (k2⁄ knoncat) of

 109-fold This rate enhancement also corresponds

to  13 kcalÆmol)1 of transition-state stabilization

according to the following equation: DG = )RT

ln (k2⁄ knoncat), as discussed [6]

It was previously reported that a Tetrahymena

ribo-zyme can also catalyze a 3¢ terminal

guanosine-medi-ated substrate-cleavage reaction [3,4,15,46] In one

such study [15], the 3¢ terminal guanosine catalyzed

reaction was reported to behave similar to the

exoge-nous guanosine catalyzed reaction, for which

kc= 350Æmin)1 [6] In comparison, the P carinii

endogenous reaction is  60-fold slower (kc=

5.7 min)1; Table 1) This substantial difference might

be due to the Tetrahymena ribozyme being faster than

the Pneumocystis ribozyme, the difference in reaction

conditions, or that the intramolecular guanosine nucle-ophile in the Pneumocystis ribozyme, although bound

to the guanosine-binding site (GBS), is not bound in

an ideal orientation Indeed, this last idea may be sup-ported in that proper alignment of the intramolecular guanosine nucleophile with respect to the Pneumocystis ribozyme could be hindered by the absence of a P9.0 helix interaction, which is predicted to align the intra-molecular guanosine into the GBS [15]

The observed rate constant of substrate cleavage shows pH independence between pH 7 and 8.5, implying that in this range the rate of chemistry associated with substrate cleavage is masked by a conformational change The simplest interpretation

of this result is that the rate of substrate cleavage is not equivalent to the rate of chemistry, and that the rate of chemistry (extrapolated to be kc= 5.7Æmin)1)

is faster than the rate of substrate cleavage (mea-sured to be k2= 4.1Æmin)1) Note that the nature of the conformational change is unknown with respect

to the substrate-cleavage reaction, including any spe-cific rate constants associated with it, and so it is not included as a separate step in the reaction scheme (Fig 2)

Product dissociation For the fraction of substrates that do undergo the reaction, the resultant products dissociate from the ribozyme relatively slowly on the time scale of the reaction Furthermore, dissociation of the 5¢ exon product is slower than the cleavage step (by  75-fold), which significantly impedes the ribozyme from catalyzing multiple turnover reactions Of course, the 5¢ exon product of the cleavage reaction is an inter-mediate in the complete TES reaction, and so slow product dissociation is beneficial for the TES reac-tion as a whole In addireac-tion, the product off-rate,

k)3, is 20-fold slower than the substrate off-rate, k)1

It was also found in a Tetrahymena ribozyme [6,47] that the product off-rate is slower than the substrate off-rate, although in Tetrahymena there is only a twofold difference Apparently, there are additional

or more stable interactions that the ribozyme uses to bind the product relative to the ribozyme binding the substrate This is perhaps due to destabilization

of substrate binding via positioning of the 3¢-bridg-ing phosphoryl oxygen at the cleavage site next to a required Mg ion in the ground state [47] As the negative charge develops on the 3¢ oxygen upon entering the transition state, this interaction will become more favorable This transition state stabil-ization is thought to be an important

stabiliza-tion⁄ destabilization factor in ribozyme-substrate

bind-ing [47]

A conformation change exists between the two

steps of the TES reaction

The substrate guanosine to be excised (G1) and its

2¢-OH group are required for the second step of

TES [48], similar to (if not the same as) the role of

the xG in the second step of self-splicing [48–56]

This suggests that the guanosine to be excised is

likely binding to the GBS of the ribozyme for the

exon-ligation step of TES In the substrate-cleavage

step, however, the 3¢ terminal guanosine (G336) of

the ribozyme (Fig 1) is binding to that same GBS

Therefore, for the TES reaction, there is likely a

local conformational change between the two

reaction steps that sees G1 displace the ribozyme’s

3¢ terminal guanosine for binding into the GBS (see

Fig 1) The local conformational change that occurs

in TES is likely similar to the local conformational

change that occurs in self-splicing, with the

displace-ment of the intermolecular guanosine by the xG of

the intron [57–60] Nevertheless, because TES uses

an intramolecular nucleophile and self-splicing uses

an intermolecular nucleophile, the local

conforma-tional changes between the two steps of each

reac-tion can not be identical

Implications for TES applications

TES substrates, once bound, are four times more likely

to undergo the substrate-cleavage reaction than they

are to dissociate Therefore, to make more effective

TES ribozymes, one could decrease the rate of

sub-strate dissociation relative to that for the subsub-strate-

substrate-cleavage reaction Potential strategies for achieving this

are to increase the strength of helix P1, either through

target selection or elongation of helix P1 Note,

how-ever, that this strategy could result in a decrease in the

substrate cleavage rate

Results also suggest that the Pneumocystis ribozyme

catalyzes the substrate-cleavage reaction (catalyzed by

either an intermolecular or intramolecular guanosine)

 60-fold slower than the Tetrahymena ribozyme

Therefore, it appears that there is room for

improve-ment in terms of the rate of reaction This would be

beneficial not so much in terms of the rate of the

over-all reaction, as the cleavage reaction is not the limiting

step (binding is slower), but in terms of decreasing the

amount of substrate that dissociates from the ribozyme

before reactivity, effectively increasing the yield of the

reaction

Experimental procedures Oligonucleotide synthesis and purification RNA oligonucleotides were obtained from Dharmacon (Lafayette, CO, USA), deprotected following the manufac-turer’s protocol, and stored in sterile water Unlabeled RNAs were used without further purification The substrate RNAs were 5¢-end radiolabeled with T4 polynucleotide kinase (New England Biolabs, Beverly, MA, USA) and

NJ, USA) and gel purified on a 20% nondenaturing polyacrylamide gel [33]

Transcription The ribozyme precursor plasmid was generated as described previously [33] Prior to run-off transcription, the ribozyme plasmid was linearized with XbaI and purified using a QIA-quick PCR Purification kit (Qiagen, Valencia, CA, USA) The ribozyme, rPC, was then synthesized by run-off tran-scription and isolated as described previously [1] After-wards, the ribozyme was precipitated with 2-propanol, with ethanol, dissolved in sterile water, and quantified using a Beckman DU-650 UV-Vis spectrophotometer (Beckman Coulter Inc., Fullerton, CA, USA) at 260 nm

Measurement of observed substrate cleavage rate constants (kobs and k2)

was measured under single-turnover conditions, in which case the release of product would not affect the observed

H10Mg buffer, which consists of 50 mm Hepes (25 mm

reaction conditions appear to be optimal for the TES reac-tion [1] For the pH-dependence studies, Hepes (pH 7.5) was replaced with Mes (pH 5.0–6.8), Hepes (pH 6.8–7.5) or Epps (pH 7.5–8.5) Reactions were initiated by adding 5 lL

of an 8 nm solution of 5¢-end radiolabeled substrate [r(5¢-AUGACUdGCUC-3¢)] in the appropriate buffer (at

folding of the ribozyme prior to the addition of the radio-labeled substrate Aliquots (3 lL) were removed at specified

sepa-rated on a 12.5% denaturing polyacrylamide gel The bands were visualized on a Molecular Dynamics Storm 860 Phosphorimager and quantified using imagequant software (Molecular Dynamics, GE Healthcare, Piscataway, NJ,

USA) Data were fit using the kaleidagraph curve-fitting

program (Synergy Software, Reading, PA, USA) The final

concentration of the radiolabeled substrate in all reactions

is 1.3 nm A typical reaction utilized H10Mg buffer and a

final ribozyme concentration of 166 nm Pseudo-first-order

rate constants for the appearance of products were fit using

the following single exponential equation:

formed at time t and at the end point, respectively, and k is

the first-order rate constant

Measurement of the substrate dissociation rate

constant (k)1)

Pulse–chase experiments [6,61] were used to measure the

experi-ments, 10 lL of 200 nm ribozyme in H10Mg buffer was

combined with 2 lL of 8 nm 5¢-end radiolabeled substrate

5 lL of the reaction mixture and diluting the reaction

labeled substrate from the ribozyme is essentially

irrevers-ible Aliquots were removed at various times during the

chase phase and the reaction was quenched by adding an

reaction, but without adding the chase (which in this case is

buffer), was carried out in parallel The first-order observed

from a single-exponential fit of this data using Eqn (1) (as

differ-ence between the two measured observed rate constants:

k1¼ kobs;chase kobs;nochase ð2Þ

Measurement of the substrate association rate

constant (k1)

using a series of pulse–chase experiments In each reaction,

5 lL of a ribozyme stock (from 36 to 240 nm) in H10Mg

buffer was combined with 1 lL of 8 nm 5¢-end labeled

sub-strate and allowed to react in a total volume of 6 lL The

concen-tration, several chase reactions were initiated In each

chase, 1 lL of the original reaction mixture was removed

times ranging from 15 to 120 s The addition of chase ren-ders the dissociation of the substrate essentially irreversible

15 min, at which point the substrate-cleavage reaction was essentially complete The reaction was quenched with an

data to Eqn (1) This observed rate constant measures the rate of approach to equilibrium where substrate association

is equal to substrate dissociation Hence, the rate of

against ribozyme concentration and fitting to the equation:

Measurement of the dissociation constant, Kd

of the ribozyme–product complex

mimic binding to the ribozyme was determined using native PAGE [8,12,33,62] In this assay, several concentrations of ribozyme, ranging from 1.5 to 300 nm, were preannealed in

5 lL total volume containing 3.4% glycerol and H10Mg

integrity of the bound species during gel electrophoresis, the gel and the running buffer were made of H10Mg buffer

loaded The bound and unbound 5¢ exon mimics were sepa-rated from each other by running 6 lL of each reaction on

a 10% nondenaturing polyacrylamide gel The gel was placed on chromatography paper (Whatman 3MM CHR)

were visualized on a Molecular Dynamics Storm 860 Phos-phorimager and quantified using imagequant software (Molecular Dynamics) Data were fit using the

the 5¢ exon mimic, h is the fraction of 5¢ exon mimic bound

unbound ribozyme in the reaction

Measurement of rate constant of substrate-cleavage product dissociation (k)3) The dissociation rate constant of the 5¢ exon intermediate

of 300 nm ribozyme in 10 lL H10Mg buffer containing