(Received for publication, May 30, 1996, and in revised form, October 9, 1996)
From the Departments of Chemistry and Biochemistry,
University of Washington, Seattle, Washington 98195-1700 and
§ Warner-Lambert, Ann Arbor, Michigan 48106-1047
Protein geranylgeranyltransferase-I (PGGT-I) and
protein farnesyltransferase (PFT) attach geranylgeranyl and farnesyl
groups, respectively, to the C termini of eukaryotic cell proteins.
In vitro, PGGT-I and PFT can transfer both geranylgeranyl
and farnesyl groups from geranylgeranyl pyrophosphate (GGPP) and
farnesyl pyrophosphate (FPP) to their protein or peptide prenyl
acceptor substrates. In the present study it is shown that PGGT-I binds
GGPP 330-fold tighter than FPP and that PFT binds FPP 15-fold tighter
than GGPP. Therefore, in vivo, where both GGPP and FPP
compete for the binding to prenyltransferases, PGGT-I and PFT will
likely be bound predominantly to GGPP and FPP, respectively. Previous
studies have shown that K-Ras4B and the Ras-related GTPase TC21 are
substrates for both PGGT-I and PFT in vitro. It is shown
that TC21 can compete with the C-terminal peptide of the subunit of
heterotrimeric G proteins and with the C-terminal peptide of lamin B
for geranylgeranylation by PGGT-I and for farnesylation by PFT,
respectively. K-Ras4B competes in both cases but is almost exclusively
farnesylated by PFT in the presence of the lamin B peptide competitor.
Rapid and single turnover kinetic studies indicate that the rate
constant for the PGGT-I-catalyzed geranylgeranyl transfer step of the
reaction cycle is 14-fold larger than the steady-state turnover number, which indicates that the rate of the overall reaction is limited by a
step subsequent to prenyl transfer such as release of products from the
enzyme. PGGT-I-catalyzed farnesylation is 37-fold slower than
geranylgeranylation and is limited by the farnesyl transfer step. These
results together with earlier studies provide a paradigm for the
substrate specificity of PGGT-I and PFT and provide information that is
critical for the design of prenyltransferase inhibitors as anti-cancer
agents.
Modification of the C termini of specific eukaryotic proteins by
attachment of either 15-carbon farnesyl or 20-carbon geranylgeranyl groups is required for their proper membrane targeting and functional activation (1-9). Two closely related enzymes protein
farnesyltransferase (PFT)1 and PGGT-I
transfer prenyl groups from prenyl pyrophosphates to proteins that
contain a C-terminal CaaX motif (C is cysteine, a is usually
an aliphatic amino acid, and X is a variety of amino acids)
(10-13). The X residue of this motif plays a major role in
recognition by these two enzymes (14). PFT preferentially transfers a
farnesyl group to the cysteine residue of the CaaX motif
when X is serine, methionine, glutamine, or cysteine, and possibly other residues. PGGT-I preferentially geranylgeranylates proteins having a C-terminal leucine or phenylalanine. PFT and PGGT-I
consist of a common subunit and distinct
subunits (15-18). A
third enzyme, protein geranylgeranyltransferase-II, also known as Rab
geranylgeranyltransferase, transfers the 20-carbon prenyl group to both
cysteines of Rab proteins that have C-terminal sequences CXC, CC, or possibly CCXX (19, 20). Although the
C-terminal CaaX tetrapeptides are sufficient to be
recognized and prenylated by PFT and PGGT-I, there appears to be
additional determinants present in certain protein substrates that are
important recognition features. The polylysine domain near the C
terminus of K-Ras4B-CVIM and regions of
subunits of heterotrimeric
G proteins upstream of the CaaX sequence influence
prenylation patterns in vitro and in cell lines
over-expressing these proteins (21, 22). Rab geranylgeranyltransferase
does not detectably utilize short peptides as substrates but has a
unique subunit called Rab escort protein that binds to Rab proteins via
contacts with regions that lie far away from the C-terminal site
of prenylation (23, 24).
Steady-state kinetic analyses of PFT and PGGT-I show that both enzymes
can operate by a random sequential mechanism in which either prenyl
donor or acceptor binds first to the enzyme followed by binding of the
other substrate to form a ternary complex that goes on to products.
However, operationally, both enzymes show a strong tendency to bind
prenyl pyrophosphate first (25-28). Both PFT and PGGT-I have been
shown to tightly bind both FPP and GGPP (12, 29, 30). This result plus
the observation that H-Ras-CVLS protein can be covalently cross-linked
to the subunit of PFT suggest that the common
subunit binds
prenyl pyrophosphate and the distinct
subunits bind prenyl
acceptors (29). However, recent studies show that photoaffinity analogs
of FPP and GGPP exclusively label the
subunits of PGGT-I and PFT
(26, 31, 32), and cross-linking studies with photoreactive peptide
prenyl acceptors suggest that prenyl acceptors bind to an
subunit/
subunit interface (33).
PFT and PGGT-I show mixed specificity under certain conditions.
Purified PGGT-I can transfer both geranylgeranyl and farnesyl groups to
protein or peptide substrates ending with leucine and can
geranylgeranylate substrates ending with serine (26, 34). PFT can
transfer a geranylgeranyl as well as a farnesyl group to substrates
having a C-terminal methionine (25). The small GTP-binding protein Rho
B exists as a mixture of farnesylated and geranylgeranylated forms when
produced in transfected COS cells or in a reticulocyte lysate
translation mixture (35). Rho B is a poor substrate for PFT in the
presence of FPP or GGPP but can be efficiently geranylgeranylated or
farnesylated by PGGT-I in the presence of GGPP or FPP, respectively
(30). Some prenyl acceptors seem to be substrates for both
prenyltransferases. The oncogenic proteins K-Ras4B-CVIM and the
Ras-related GTP-binding protein TC21-CVIF (also known as R-Ras2) can be
farnesylated by PFT and geranylgeranylated by PGGT-I in
vitro (21, 36). Trueblood and co-workers (37) showed that
over-expression of the subunit of PFT or PGGT-I in yeast can
partially overcome the inactivating effect of mutation in the
subunit of the other enzyme (PGGT-I or PFT, respectively). This
suggests that under forcing conditions, both PFT and PGGT-I can
prenylate proteins that are normally prenylated by a single
prenyltransferase.
In this paper we report studies aimed at understanding the specificities of PFT and PGGT-I in detail with particular focus on determining the precise affinities of PFT and PGGT-I for both FPP and GGPP. These results, when combined with our earlier studies (26), provide a comprehensive description of the features of these enzymes that dictate their specificities. The results are important for predicting the consequences of inhibiting PFT or PGGT-I on protein prenylation in cells. This is of considerable interest because inhibitors of these enzymes are being developed as potential anti-cancer drugs due to their ability to block prenylation and thus transforming activity of oncogenic RAS proteins (38, 39). We also report kinetic studies of PGGT-I which probe the nature of the rate-limiting step for the prenyl transfer reactions, and our results are compared with the analogous study carried out with PFT (40).
Biotin-6-CAIL and biotin-lamin
B-CAIS were synthesized and analyzed by mass spectroscopy as described
previously (34). [3H]FPP and [3H]GGPP (15 Ci/mmol, labeled on carbon-1) and unlabeled FPP and GGPP were purchased
from American Radiolabeled Chemicals. The chemical and radiometric
purities of these prenyl pyrophosphates were routinely monitored by
thin layer chromatographic analysis as described (34), and compounds
were not used if either purity was less than 90%. Recombinant
baculoviruses that express the
and
subunits of rat PFT
(Professor Y. Reiss, Tel-Aviv University) and that express the
subunit of rat PGGT-I (Professor J. L. Goldstein, Southwestern Medical
Center) were obtained as gifts from the indicated individuals.
RhoB-CKVL was obtained as a generous gift from Professor J. L. Goldstein. H-Ras-CVLS produced in Escherichia coli was
prepared as described (41). K-Ras4B-CVIM was obtained from the cytosol
of baculovirus-infected Sf9 cells as described (42). TC21-CVIF was
obtained as a generous gift from Dr. V. Manne (Bristol-Myers Squibb,
Princeton, NJ).
Sf9 cells were cultured in IPL-41 medium supplemented with
10% fetal bovine serum, 4 mg/ml yeastolate, 0.1% Pluronic F-68, and
50 µg/ml gentamycin (all from Life Technologies, Inc.). A 6-liter
culture of Sf9 cells (log phase, about 1.5 × 106
cells/ml) was coinfected with recombinant viruses expressing the subunit of PFT/PGGT-I and either the
subunit of PGGT-I or the
subunit of PFT at multiplicities of infection of 1. All subsequent
steps were carried out at 4 °C unless otherwise noted. For PGGT-I
production, the cells were harvested 72 h after infection by
centrifugation and washed once with 20 mM Tris-HCl, 1 mM DTT, 0.13 M NaCl, pH 8.0. The cell pellet
was frozen at
80 °C. The frozen cells were thawed on ice in 300 ml
of 20 mM Tris-HCl, 1 mM DTT, and 0.05 M NaCl, pH 8.0 (buffer A) containing freshly added protease
inhibitors (1 mM phenylmethylsulfonyl fluoride, 30 µM tosyl-lysine-chloromethyl ketone, 30 µM
tosyl-phenylalanine-chloromethyl ketone, and 10 µg/ml each of
aprotinin, leupeptin, and pepstatin A, all from
Sigma). The cells were disrupted by six strokes in a
Dounce homogenizer. The lysate was centrifuged for 10 min at 10,000 × g, and the supernatant was centrifuged for
1 h at 100,000 × g. The resulting supernatant was
fractionated on a column (2.6 × 20 cm) of Q-Sepharose (Pharmacia
Biotech Inc.) equilibrated in buffer A as described previously (12). A
linear gradient (4-liter total volume) of NaCl concentrations from 0.05 to 0.6 M was applied, and PGGT-I was eluted at about 0.3 M NaCl. The enzyme peak fractions (assayed with 5 µM biotin-
6-CAIL and 1 µM
[3H]GGPP as described below) were combined, dialyzed
against buffer A, and then chromatographed on a Mono-Q HR10/10 column
(Pharmacia) as described previously (12). The yield of PGGT-I is 90 mg, and the purity was more than 90% as judged by SDS-polyacrylamide gel
electrophoresis. The specific activity is 4.7 microunits/µg, where 1 microunit is the amount of enzyme that produces 1 pmol of product per
min at 30 °C using 2 µM [3H]GGPP and 5 µM biotin-
6-CAIL peptide as substrates.
PGGT-I was stored at
80 °C.
For purification of PFT, infected Sf9 cells (3.6-liter culture) were
harvested 2-3 days after infection by centrifugation, and all
subsequent steps were carried out at 4 °C. The cells were washed
twice with phosphate-buffered saline, and the cell pellet was taken up
in 300 ml of 20 mM Tris-HCl, 50 mM
MgCl2, 20 µM ZnCl2, 1 mM DTT, pH 7.4. The cell suspension was transferred to a
pre-chilled cavitation bomb (in three 100-ml portions); the bomb was
charged with N2 to 600 p.s.i., and after 30 min, the
sample was drained from the bomb through a high pressure outlet valve.
The cell lysate was centrifuged at 100,000 × g for 45 min. To the supernatant was added solid ammonium sulfate to give 55%
saturation, and the suspension was centrifuged at 18,000 × g for 20 min. The protein pellet was taken up in 31 ml of
buffer B (20 mM Tris-HCl, 50 mM NaCl, 20 µM ZnCl2, 1 mM DTT, pH 7.5), and
the sample was dialyzed twice against 2 liters of buffer B. The sample
was centrifuged at 18,000 × g for 20 min, filtered
through a 0.22-µm filter cartridge, and loaded onto a Mono-Q HR10/10
column. The column was washed with 15 ml of buffer B and then developed
with a linear gradient of 0.05-1 M NaCl in buffer B (100 ml total). PFT (assayed as described below) elutes at about 0.5 M NaCl and was stored at 80 °C. The specific activity
of the final PFT preparation is 138 microunits/µg based on the assay
with 10 µM H-Ras-CVLS and 2 µM
[3H]FPP, and the purity was greater than 90% as judged
by SDS-polyacrylamide gel electrophoresis.
The absolute mole amount of [3H]GGPP-binding capacity of
PGGT-I was determined as the moles of PGGT-I·[3H]GGPP
that eluted in the void volume of a spin gel filtration column (12)
when a stock solution of PGGT-I was incubated with an excess of
[3H]GGPP of known specific radioactivity. The specific
radioactivity of [3H]GGPP, diluted with unlabeled GGPP,
was determined from the mass of GGPP measured by phosphate analysis
(43) and from the cpm measured by scintillation counting. In one case,
the mass of [3H]GGPP was determined by integration of
1H NMR resonances obtained from a sample containing
[3H]GGPP and a known amount of dimethyl sulfoxide as an
internal standard; the mass determined in this way is in agreement with the phosphate analysis. The absolute mole amount of
[3H]FPP-binding capacity of PFT was determined similarly
using [3H]FPP of known specific radioactivity. The
absolute mole amount of PGGT-I was determined from the absorbance at
280 nm using the extinction coefficient of 134 mM1 cm
1, which was calculated
from the amino acid sequence using the published formula (44).
To assay the geranylgeranylation
activity of PGGT-I, 1 µM [3H]GGPP (15 Ci/mmol) and 5 µM biotin-6-CAIL were
incubated with 0.1 µg of rat PGGT-I in a total volume of 20 µl
containing 30 mM potassium phosphate, 5 mM DTT,
0.5 mM MgCl2, 20 µM
ZnCl2, pH 7.7. After 5 min at 30 °C, the reaction was
terminated by boiling for 3 min, and 40 µl of avidin-agarose
suspension (50% aqueous suspension, Pierce) was added to the mixture
to measure radioactivity transferred to the biotinylated peptide as
described previously (12). For assaying the farnesylation activity of
PGGT-I, 1 µM [3H]FPP (15 Ci/mmol) and 20 µM biotin-
6-CAIL were used as substrates. The standard assay for PFT was carried out with 10 µM
H-Ras-CVLS and 3.3 µM [3H]FPP in a total
volume of 20 µl containing 30 mM potassium phosphate, 5 mM DTT, 10 mM MgCl2, 20 µM ZnCl2, 25 mM NaCl, pH 7.7. Mixtures were incubated at 30 °C for 10 min and terminated by the
addition of 200 µl of 10% HCl in ethanol. The amount of radiolabeled
farnesylated H-Ras-CVLS was quantified using the glass fiber filter
binding method (25). The enzymatic activities of the
PGGT-I·[3H]GGPP and PGGT-I·[3H]FPP
binary complexes isolated by spin column gel filtration (see below)
were assayed using the same conditions as above except that the
radioactive prenyl pyrophosphate was omitted, and the incubation time
was 15 or 60 s.
The standard assay mixture contains 2-20 nM PGGT-I or PFT, 1 nM to 40 µM [3H]GGPP or [3H]FPP, 30 mM potassium phosphate, 1 mM DTT, 0.1% n-octylglucoside, pH 7.7, and 2 µl of 70% ethanol, 30% 0.25 M NH4HCO3 (prenyl pyrophosphate stock solution solvent) in a total volume of 50 µl. The mixture was incubated at 30 °C for 15 min and applied to a 0.5 ml-column of Sephadex G-50 pre-equilibrated with 30 mM potassium phosphate, 1 mM DTT, 0.1% n-octylglucoside, pH 7.7. The column was spun for 2 min at 2,500 rpm in an HB-4 swinging-bucket rotor and then washed once with 50 µl of the same buffer by spinning in the same way. These two eluants that contain protein but negligible amounts of unbound prenyl pyrophosphate were combined, and the mixture was submitted to scintillation counting. Control experiments without enzyme were also carried out for each assay, and the amount of cpm eluted in the first two eluants was typically 10-50 cpm, and this value was subtracted from that measured in the presence of enzyme.
Rapid Quench Kinetic AnalysisPre-steady-state kinetic
measurements were made using a KinTek Quench Flow apparatus Model
RQF-3. Three syringes were driven by a step motor. PGGT-I or PFT alone
or together with [3H]GGPP or [3H]FPP
(Syringe 1) was rapidly mixed with biotin-6-CAIL alone or together with nonradioactive FPP or GGPP, or [3H]GGPP
(Syringe 2). Each set of samples (45-µl aliquot each) was incubated
at 30 °C for a series of time periods (0.25 to 50 s), and the
transfer reaction was terminated by rapid mixing with stop solution
(230 µl of 1.5 M MgCl2, 0.2 M
H3PO4, 0.5% bovine serum albumin, Syringe 3).
Aliquots (150 µl) of the quenched mixtures were mixed with 100 µl
of scintillant proximity assay beads suspension (20 mg/ml in
phosphate-buffered saline, Amersham Corp.) and subjected to
scintillation counting in a
counter (Wallac microbeta 1450 counter). Binding of the radioactive, prenylated and biotinylated peptide to streptavidin on the the beads places the radioactive prenyl
group in close proximity to the scintillant bead. Each set of data
obtained was well fit to the first-order equation, and the reaction
rate constants were calculated based on the first-order equation.
Additional details are given in Table II.
|
Reaction mixtures contained 0.05 µg of PGGT-I or PFT, 2 µM [3H]GPPP or [3H]FPP, 2 µM protein substrate (see Table III), 2 µM peptide substrate (see Table III) in 30 µl of 30 mM potassium phosphate, 5 mM DTT, 0.5 mM MgCl2, 20 µM ZnCl2, pH 7.7. After 15 min at 30 °C, a 10-µl aliquot was quenched with 2 × Laemmli loading buffer. After boiling, the sample was analyzed by SDS-polyacrylamide gel electrophoresis (12.5% gel). The gel was prepared for fluorography, and the radiolabeled bands were visualized using x-ray film (26). Gel slices containing radiolabeled proteins were solubilized in 0.5 ml of 30% H2O2 at 60 °C overnight, and the samples were submitted to scintillation counting. The cpm data were converted to pmol of prenylated protein by correcting for quenching. This was carried out by mixing a known amount of [3H]GGPP with a blank gel slice and determining the cpm as described above. To quantify the amount of prenylated peptide, a 10-µl aliquot of the same prenylation reaction mixture was mixed with 200 µl of methanol; the sample was applied to a disposable anion exchange column (SpeN+, 1 ml bed, Baker) to remove [3H]prenyl pyrophosphate, and the column was washed with five 200-µl portions of methanol. The eluant was collected in a single fraction, and solvent was removed in a Speed-vac concentrator (Savant Instruments). The residue was taken up in methanol and spotted onto a 20 × 20-cm plate of silica gel 60 (EM Science). The plate was developed with n-propanol:concentrated NH4OH:water (6:3:1 by volume) and sprayed with EN3HANCE (DuPont NEN) prior to exposure to x-ray film. The region of the plate that contained the radioactive prenylated peptide was scraped, and the silica was submitted to scintillation counting. Counting efficiency was determined by counting a known amount of [3H]GGPP together with blank silica gel.
|
We
have previously shown with in vitro reactions that purified
bovine PGGT-I is able to transfer both geranylgeranyl and farnesyl groups to substrates containing C-terminal CaaL motifs such as biotin-6-CAIL (26, 34). Steady-state kinetic studies of
PGGT-I showed that the KM values for the interaction
of GGPP and FPP with PGGT-I in the presence of a saturating amount of biotin-
6-CAIL are similar (0.6 µM) (26).
In the presence of saturating amounts of biotin-
6-CAIL
and prenyl donor, kcat for geranylgeranylation
and farnesylation are similar (26). Nevertheless, addition of 1 µM nonradiolabeled GGPP to PGGT-I reaction mixtures containing 1 µM [3H]FPP and 2 µM biotin-
6-CAIL nearly abolished the
transfer of the radiolabeled farnesyl group to peptide (Table
I and Ref. 34). In addition, in PGGT-I reaction mixtures
containing equimolar amounts of FPP and [3H]GGPP, the
amount of radioactive geranylgeranylated peptide formed was the same as
in reaction mixtures that did not contain FPP (Table I and Ref. 34).
This suggests that PGGT-I has different affinities for GGPP and
FPP.
|
To determine equilibrium dissociation constants (Kd)
for the PGGT-I·GGPP and PGGT-I·FPP complexes, a very low
concentration of PGGT-I was incubated with various concentrations of
either [3H]GGPP or [3H]FPP, and the binary
complexes were isolated by spin column gel filtration. As shown in Fig.
1A, a biphasic profile of
[3H]GGPP binding to PGGT-I was observed. This measurement
was carried out with 2 nM PGGT-I, the minimum that could be
used to obtain a reliable amount of cpm in the column eluant. From the
hyperbolic response in the range 0-50 nM
[3H]GGPP, it can be seen that the Kd
is approximately 3 nM. When the
PGGT-I·[3H]GGPP complex obtained with 20 nM
PGGT-I and 20 nM [3H]GGPP was isolated by
spin column gel filtration and then mixed with 5 µM
biotin-6-CAIL, 49 ± 2% (1, 100 cpm) of the bound
cpm was transferred after incubation periods of 10, 30, and 180 s. Thus the transfer reaction is completed in less than 10 s. When this transfer experiment was repeated in the presence of 4 µM unlabeled GGPP added together with
biotin-
6-CAIL, 48 ± 2% of the enzyme-bound cpm
was transferred. The similarity in the percent geranylgeranyl group
transferred in these two experiments indicates that both peptide
binding to PGGT-I·[3H]GGPP and subsequent prenyl
transfer occur before [3H]GGPP dissociates from both the
PGGT-I·[3H]GGPP binary complex and the
PGGT-I·[3H]GGPP·biotin-
6-CAIL ternary
complex. These experiments were repeated with the same amount of PGGT-I
but with a higher concentration of [3H] GGPP (1.3 µM) so that the second binding process shown in Fig. 1A is saturated. After spin column gel filtration, 37 ± 3% (2,900 cpm) of the bound cpm was transferred to
biotin-
6-CAIL after incubation periods of 10, 30, and
180 s. Again, the cpm transferred to peptide was not reduced if 4 µM unlabeled GGPP was added along with peptide.
It was initially thought that the lower affinity binding of
[3H]GGPP to PGGT-I (Fig. 1A) is due to
nonspecific interaction of this isoprenoid with the enzyme. However,
this is not the case. If it were, 1,100 cpm would be transferred to
peptide in the presence of excess unlabeled GGPP, i.e. the
same amount of radioactivity that is transferred from the high affinity
PGGT-I·[3H]GGPP to peptide. The fact that a total of
2,900 cpm is transferred from the complex prepared from 1.3 µM [3H]GGPP indicates that the lower
affinity PGGT-I·[3H]GGPP complex is capable of prenyl
transfer to peptide, and furthermore this transfer occurs without
exchange of [3H]GGPP with the pool of non-bound GGPP. For
both the high and low affinity PGGT-I·[3H]GGPP
complexes, only a portion of the bound [3H]GGPP is
transferred to peptide, 49 and 32%, respectively. The reason for this
is not known with certainty; however, it is probably due to partial
dissociation of [3H]GGPP from the
PGGT-I·[3H] GGPP binary complex or from the
PGGT-I·[3H]GGPP·biotin-6-CAIL ternary
complex that occurs before prenyl transfer.
As shown in Fig. 1A, PGGT-I that had been incubated with 5 mM DTT for 30 min at 30 °C at neutral pH leads to a form
of the enzyme that only shows that high [3H]GGPP binding
component and up to 56 fmol of [3H]GGPP can be bound.
This suggests that two forms of PGGT-I exist that differ in their
patterns of disulfides; the -chain of PPGT-I contains 15 cysteines
throughout its length (16). As will be shown by single turnover kinetic
studies presented below, the rate of transfer of the geranylgeranyl
group from GGPP to peptide acceptor catalyzed by these two forms of
PGGT-I are the same within experimental error.
As shown in Fig. 1A, only a small amount of PGGT-I·[3H]FPP is formed in the presence of 0-200 nM [3H]FPP. Significant amounts of PGGT-I·[3H]FPP were formed with 20 nM PGGT-I and concentrations of [3H]FPP higher than 0.2 µM (Fig. 1B). From this curve a Kd of 1 µM is obtained for the PGGT-I·[3H]FPP complex, and thus it is concluded that GGPP binds about 330-fold tighter to PGGT-I than does FPP.
Binding of [3H]FPP and [3H]GGPP to PFT was
also studied by spin column gel filtration. From the binding curves
shown in Fig. 2, A and B, values
of Kd of 2 nM and 30 nM
for [3H]FPP and [3H]GGPP, respectively,
were obtained. Thus, [3H]FPP binds about 15-fold tighter
than [3H]GGPP to PFT.
In contrast to the results with PGGT-I·[3H]GGPP, only
4% of the labeled farnesyl group was transferred to
biotin-6-CAIL in a reaction mixture containing 16 nM PGGT-I·[3H]FPP and 20 µM
biotin-
6-CAIL after 15 s, and no transfer of radioactivity to the peptide occurred if the reaction mixture also
contained excess unlabeled FPP or GGPP (10 µM).
Presumably the small amount of labeled farnesyl group transferred to
the peptide in the absence of unlabeled FPP or GGPP is due to the fact
that the concentration of PGGT-I·[3H]FPP used, 16 nM, is far below the Kd for this binary complex (1 µM), and thus most of the
[3H]FPP dissociated from the enzyme after the spin column
but prior to the addition of peptide. Indeed, when the transfer
reaction was carried out with 500 nM
PGGT-I·[3H]FPP, 35% of the radioactivity was
transferred to biotin-
6-CAIL after 15 s. Only 1.5%
of the radioactivity was transferred if this reaction mixture also
contained 50 µM unlabeled FPP, and no detectable
radioactivity was transferred in the presence of 50 µM
unlabeled GGPP. These observations indicate that within a 15-s period,
most of the [3H]FPP bound to PGGT-I has exchanged with
FPP or GGPP present in the non-enzyme-bound pool.
Values of Kd determined as described above were confirmed by competitive binding studies. As shown in Fig. 3A, addition of increasing amounts of unlabeled FPP to a mixture of 20 nM PGGT-I and a fixed amount of [3H]GGPP (20 nM) led to a decrease in the amount of enzyme-bound radioactivity as detected by spin column gel filtration. Since [3H]GGPP dissociates slowly from PGGT-I, it is important to verify that binding exchange has reached equilibrium. PGGT-I was incubated with a mixture of [3H]GGPP and unlabeled FPP at 30 °C for 30 or 60 min, and the amount of enzyme-bound [3H]GGPP was the same for both incubation periods (data not shown). Thus, the 60-min incubation period used in the experiments shown in Fig. 3A is sufficient to attain equilibrium. At equilibrium, the ratio of binary complexes is given by Equation 1.
![]() |
(Eq. 1) |
Single Turnover Rates of PGGT-I-catalyzed Geranylgeranylation and Farnesylation
Rapid quench experiments were carried out to
measure the rate of transfer of the geranylgeranyl or farnesyl group to
biotin-6-CAIL catalyzed by PGGT-I, and results are
summarized in Table II. A solution of 200 nM
PGGT-I·[3H]GGPP was rapidly mixed with a solution of 5 µM biotin-
6-CAIL, and the reaction mixture
was quenched with acid at various times after the mixing. The results
in Fig. 4A show that under these single
turnover conditions, the geranylgeranyl group is transferred to the
peptide with a first-order rate constant of 0.49 ± 0.06 s
1. Essentially the same rate constant was obtained when
200 nM PGGT-I·[3H]GGPP was mixed with 20 µM biotin-
6-CAIL instead of 5 µM, which indicates that prenyl transfer is slower than
binding of peptide to PGGT-I·[3H]GGPP, and thus the
rate of prenyl transfer from the ternary complex being measured is not
limited by peptide binding. Essentially the same observed rate constant
was seen when a solution of 200 nM PGGT-I was mixed with a
solution containing 5 µM peptide and 200 nM
[3H]GGPP, which indicates that binding of
[3H]GGPP to PGGT-I occurs faster than prenyl transfer.
When PGGT-I·[3H]GGPP was mixed with a solution
containing both 5 µM biotin-
6 and a
20-fold excess (4 µM) of unlabeled GGPP, the amount of
radioactive geranylgeranyl transfer at the end of the reaction was only
10% lower than that transferred in the absence of unlabeled GGPP, which indicates that the radiolabeled geranylgeranyl group was transferred to peptide without significant release from the enzyme, and
thus the observed prenyl transfer rate reflects single turnover of the
ternary complex.
Rapid quench experiments with the complex formed from 1.5 µM PGGT-I and 1 µM [3H]FPP
mixed with a solution of 40 µM
biotin-6-CAIL and 10 µM GGPP yielded an
observed first-order rate constant of 0.015 ± 0.004 s
1 (Fig. 4B, Table II), which is 37-fold
smaller than the rate constant for geranylgeranyl transfer. In these
experiments, higher concentrations of reactants were used since
[3H]FPP binds about 330-fold weaker to PGGT-I than does
[3H]GGPP. In addition, 10 µM unlabeled GGPP
was included so that single turnover from the
PGGT-I·[3H]FPP·biotin-
6-CAIL complex
is being measured owing to the fact that once [3H]FPP
dissociates from the enzyme, it will be irreversibly replaced by
GGPP.
Rapid quench experiments were also performed with PFT acting on
biotin-TKCVIM. A solution of PFT (150 nM) and
[3H]FPP (110 nM) was mixed with biotin-TKCVIM
(4 µM). The farnesyl group is transferred to the peptide
with a first-order rate constant of 1.9 ± 0.7 s1
(not shown). The rate was essentially the same when 2 µM
unlabeled FPP or GGPP was present in the peptide solution (1.8 ± 0.4 s
1 and 3.3 ± 0.3 s
1,
respectively). The rate constant for farnesyl transfer is similar to
the value of 0.44 s
1 previously reported for
PFT-catalyzed farnesylation of biotin-GLPCVVM (40).
In vitro prenylation studies with the prenyl acceptors K-Ras4B-CVIM, RhoB-CKVL, and TC21-CVIF have shown that these proteins can be farnesylated or geranylgeranylated (21, 30, 35, 36), and these reactions were studied in more detail in the present study. The relative affinities of PGGT-I and PFT for GGPP and FPP described above are preserved in the presence of these protein substrates. This is shown in Table I. For example, the geranylgeranylation of K-Ras4B-CVIM by PGGT-I in the presence of 1 µM [3H]GGPP is not inhibited if 1 µM FPP is also present, but farnesylation of K-Ras4B-CVIM by PGGT-I with 1 µM [3H]FPP is abolished if 1 µM GGPP is also present. Likewise, the farnesylation of the same protein, K-Ras4B-CVIM by PFT in the presence of 1 µM [3H]FPP is not inhibited if 1 µM GGPP is present, but the geranylgeranylation reaction in the presence of 1 µM [3H]GGPP is inhibited by 77% if 1 µM FPP is also present. Similar results are obtained with RhoB-CKVL and TC21-CVIF (Table I).
Competition between prenyl acceptors for geranylgeranylation by PGGT-I·[3H]GGPP and for farnesylation by PFT·[3H]FPP was also studied, and the results are summarized in Table III.
The results in this study clearly show that both PGGT-I and PFT display binding selectivity for their respective prenyl pyrophosphate substrates. The 330-fold higher affinity of GGPP versus FPP for PGGT-I does not contradict the published data showing that micromolar amounts of PGGT-I are capable of binding both GGPP and FPP (30) (also confirmed in our lab). Since the Kd values for the complexes of PGGT-I with GGPP and FPP are 3 nM and 1 µM, respectively, micromolar amounts of enzyme can be titrated with micromolar amounts of GGPP or FPP. It is only when the concentration of PGGT-I is near the Kd = 3 nM for GGPP, as is the case in the present study, that the very high affinity of PGGT-I for GGPP can be detected. The same applies to PFT (29), and in the present study using low concentrations of PFT it is found that FPP binds about 15-fold tighter to PFT than does GGPP.
As shown previously, values of kcat are similar
for PGGT-I-catalyzed geranylgeranylation and farnesylation of
biotin-6-CAIL under conditions where both prenyl donor
and acceptor are saturating (26). However, this in vitro
situation is not relevant to the situation that exists inside the cell
where both GGPP and FPP are present and compete for binding to PGGT-I.
Under competitive conditions, geranylgeranylation is greatly preferred
over farnesylation unless the concentration of FPP in the cell is more
than 330-fold higher than the concentration of GGPP, which seems
unlikely. Even if there is a small amount of PGGT-I·FPP in the cell,
the KM for the interaction of this binary complex
with biotin-
6-CAIL is 29-fold higher than the
KM for the interaction of PGGT-I·GGPP with the
same peptide (26), and this would disfavor the farnesylation reaction.
In the presence of saturating biotin-
6-CAIL, values of
KM for FPP and GGPP are similar (26). However, this
situation is also not relevant to the reaction in the cell or in
vitro since although prenyl acceptor can bind before prenyl donor,
the PGGT-I-catalyzed reaction cycle involves the preferential binding
of GGPP prior to prenyl acceptor (26). The same is true for PFT, which
preferentially binds FPP before prenyl acceptor (25), and binds FPP
15-fold tighter than GGPP. Since the selectivity of PFT for binding FPP
versus GGPP is not so rigid, it is difficult at this point
to say with confidence that PFT will be loaded only with FPP in
vivo, especially since the relative sizes of the FPP and GGPP
in vivo pools are not known.
For the reasons given above, it seems likely that in the cell PFT and PGGT-I are fully loaded with their respective prenyl pyrophosphate, and one may now ask which prenyl acceptors do these binary complexes choose. Such a selection is dictated by steady-state Equation 2.
![]() |
(Eq. 2) |
Kinetic data for PFT-catalyzed farnesylation of Ras-CVIM (farnesyl
acceptor) and Ras-CAIL (geranylgeranyl acceptor) are available (25).
According to Equation 2 and the published data, the PFT·FPP complex
prefers Ras-CVIM over Ras-CAIL by a factor of 40 (relative kcat/KM values).
Interestingly, the kcat for farnesylation of
Ras-CVIM is 2.5-fold smaller than the kcat for
farnesylation of Ras-CAIL. However, the 80-fold lower
KM for Ras-CVIM compared with Ras-CAIL
more than offsets the difference in kcat values,
and thus Ras-CVIM is the preferred substrate based on KM selectivity. These specificity features of PGGT-I and PFT are summarized in Fig. 5.
As described in the Introduction, the GTPases RhoB-CKVL, TC21-CVIF, and
K-Ras4B-CVIM, which are members of the Ras superfamily of proteins,
have been reported to undergo both farnesylation and
geranylgeranylation. The presence of these proteins in reaction mixtures did not influence the preference of PFT and PGGT-I in binding
FPP and GGPP, respectively (Table I). Thus, the prenylation selectivity
of these proteins will be dictated by their
kcat/KM values in comparison
to those of other prenyltransferase substrates present in the cell
according to Equation 2. When K-Ras4B-CVIM competes against an equal
concentration of biotin-6-CAIL (the
6
protein is known to be geranylgeranylated in vivo (45)) for geranylgeranylation by PGGT-I·GGPP, the peptide is preferred by 8.7-fold. Competition between K-Ras4B-CVIM and biotin-lamin B-CAIS (the
lamin B protein is known to be farnesylated in vivo (46)), for farnesylation by PFT-I·FPP leads to almost exclusive
farnesylation of the Ras protein. These results are consistent with
those of James et al. (21) who showed that the
KM for the interaction of K-Ras4B-CVIM with
PFT·FPP is about 10-fold smaller than the KM for
the interaction of the same protein with PGGT-I·GGPP; the
kcat for these two reactions are similar. These
results give only the relative velocities for the pair of prenyl
acceptors studied. In the cell K-Ras4B-CVIM is competing with many
other substrates, and the relative
kcat/KM values for these other substrates are not known. Nevertheless, the results suggest that
K-Ras4B-CVIM will normally become farnesylated in cells, but if a
PFT-selective inhibitor is used, K-Ras4B-CVIM can effectively compete
with geranylgeranyl acceptors for geranylgeranylation by PGGT-I. This
point has obviously important consequences for the design of
prenyltransferase inhibitors as anti-cancer drugs. Indeed, James
et al. (47) showed that a PFT-selective inhibitor does block
Ras-dependent activation of the mitogen-activated protein kinase cascade when the cell is activated by oncogenic H-Ras. In
contrast, when this cascade is regulated by endogenous Ras, the PFT
inhibitor has no effect. K-Ras4B-CVIM may be the endogenous activator
of mitogen-activated protein kinase in the cells studied, and
inhibition of PFT may cause K-Ras4B-CVIM to become geranylgeranylated by PGGT-I with restoration of its signaling function (47).
TC21-CVIF, which is a Ras-related protein with oncogenic potential, has
been shown in vitro to be farnesylated by PFT and geranylgeranylated by PGGT-I (36). In the present study, it is shown
that TC21-CVIF can be a substrate for PGGT-I and PFT because it
competes as well with biotin-6-CAIL for
geranylgeranylation by PGGT-I as it does with lamin B for farnesylation
by PFT (Table III), although the peptide substrates are 10-fold
preferred in both cases. The CaaX sequence of TC21 is
unusual in that it has a C-terminal phenylalanine. However, this is
probably not the reason that TC21 is a substrate for both PGGT-I and
PFT because another GTPase CDC42, which also has a C-terminal
phenylalanine, is a much better substrate for PGGT-I than for PFT (34).
Interestingly, the transformation of cells by oncogenic TC21-CVIF is
not inhibited by a selective PFT inhibitor, and the present results
support the suggestion of Carboni and co-workers (36) that TC21-CVIF may be geranylgeranylated in cells treated with PFT inhibitors. The
type of prenyl group attached to TC21-CVIF in cells treated or not
treated with prenyltransferase inhibitors remains to be determined.
Furfine and co-workers (40) measured the single turnover rate of
conversion of the PFT·FPP·peptide-CVVM complex to products and
obtained a value of 0.06 s1. Interestingly, this value is
7.3-fold smaller than the kcat for the
steady-state reaction when both prenyl donor and acceptor are
saturating. This implies that a step after the conversion of the
ternary complex to the enzyme product complex is rate-determining for
steady-state turnover. The slow step may be release of prenylated peptide from the enzyme (40). The same appears to be true for the
PGGT-I·GGPP·biotin-
6-CAIL complex, which undergoes
conversion to product with a single turnover rate constant of 0.56 s
1 under conditions where binding of GGPP or peptide to
the enzyme is not rate-determining (Table II). With the same
preparation of PGGT-I, the kcat for steady-state
turnover in the presence of saturating amounts of both
biotin-
6-CAIL and GGPP is 0.032 s
1. Thus
the single turnover geranylgeranyl transfer rate constant is 14-fold
faster than the steady-state turnover number, which establishes that a
step after prenyl transfer, such as product release, is
rate-determining. The conversion of
PGGT-I·FPP·biotin-
6-CAIL to products occurs with a
37-fold smaller rate constant (0.015 s
1) than the
analogous geranylgeranylation reaction, and this value is similar to
the steady-state kcat of 0.012 s
1
measured with saturating amounts of FPP and
biotin-
6-CAIS. Thus, the rate of PGGT-I-catalyzed
farnesylation is probably limited by the chemical step of prenyl
transfer. Thus, the similar observed values of
kcat for PGGT-catalyzed geranylgeranylation and
farnesylation (26) is not due to similar rate constants for the prenyl
transfer step.
The biphasic binding of [3H]GGPP to PGGT-I purified in
the presence of 1 mM DTT (Fig. 1A) was
unexpected. The weaker binding component is not due to nonspecific
adsorption of [3H]GGPP to PGGT-I since
[3H]GGPP bound in this weaker complex is transferred to
peptide without exchange with free GGPP in the buffer. Since the PGGT-I used in this study is recombinant and originates from plaque-purified baculoviruses expressing the and
subunits of this enzyme, the
low and high affinity binding of [3H]GGPP is an intrinsic
property of this enzyme and is not due to contamination by another
prenyltransferase. This enzyme can be converted into a form that shows
only monophasic, high affinity binding of GGPP by incubating it with
higher amounts of DTT (5 mM). PGGT-I does not seem to exist
as a mixture of multimers since enzyme handled with 1 mM
DTT or treated with 5 mM DTT elute in the same volume when
applied to a gel filtration column, and the elution volume is
reasonable for a heterodimer of
90 kDa. PGGT-I contains many
cysteines, and the two forms of PGGT-I may have a different pattern of
intramolecular disulfides. Prenyl transfer catalyzed by these two forms
occurs with similar kinetics since no evidence for a multi-exponential
process was seen in the rapid quench experiments (Fig. 4, A
and B).