Modular total chemical synthesis of a human immunodeficiency virus type 1 protease

Article abstract of DOI:10.1021/ja072870n

As part of our ongoing studies of the human immunodeficiency virus type 1 (HIV-1) protease enzyme, we set out to develop a modular chemical synthesis of the protein from multiple peptide segments. Our initial attempts were frustrated by the insolubility of intermediate peptide products. To overcome this problem, we designed a synthetic strategy combining the solubility-enhancing properties of C-terminal (Arg)_n tags and the biological phenomenon of autoprocessing of the Gag-Pol polyprotein that occurs during maturation of the HIV-1 virus in vivo. Synthesis of a 119-residue peptide chain containing 10 residues of the reverse transcriptase (RT) open reading frame plus an (Arg) ₁₀ tag at the C-terminus was straightforward by native chemical ligation followed by conversion of the Cys residues to Ala by Raney nickel desulfurization. The product polypeptide itself completed the final synthetic step by removing the C-terminal modification under folding conditions, to give the mature 99-residue polypeptide. High-purity homodimeric HIV-1 protease protein was obtained in excellent yield and had full enzymatic activity; the structure of the synthetic enzyme was confirmed by X-ray crystallography to a resolution of 1.07 A. This efficient modular synthesis by a biomimetic autoprocessing strategy will enable the facile synthesis of unique chemical analogues of the HIV-1 protease to further elucidate the molecular basis of enzyme catalysis.

Full text of DOI:10.1021/ja072870n

Published on Web 08/18/2007
Modular Total Chemical Synthesis of a Human
Immunodeficiency Virus Type 1 Protease
Erik C. B. Johnson,† Enrico Malito,‡ Yuequan Shen,‡,§ Dan Rich,|
Wei-Jen Tang,‡ and Stephen B. H. Kent*,†,
Contribution from the Department of Biochemistry and Molecular Biology, Institute for
Biophysical Dynamics, Ben-May Department for Cancer Research, and DDepartment of
Chemistry, The UniVersity of Chicago, Chicago, Illinois 60637, and School of Pharmacy,
The UniVersity of WisconsinsMadison, Madison, Wisconsin 53705
Received April 24, 2007; E-mail: skent@uchicago.edu
Abstract: As part of our ongoing studies of the human immunodeficiency virus type 1 (HIV-1) protease
enzyme, we set out to develop a modular chemical synthesis of the protein from multiple peptide segments.
Our initial attempts were frustrated by the insolubility of intermediate peptide products. To overcome this
problem, we designed a synthetic strategy combining the solubility-enhancing properties of C-terminal (Arg)n
tags and the biological phenomenon of autoprocessing of the Gag-Pol polyprotein that occurs during
maturation of the HIV-1 virus in vivo. Synthesis of a 119-residue peptide chain containing 10 residues of
the reverse transcriptase (RT) open reading frame plus an (Arg)10 tag at the C-terminus was straightforward
by native chemical ligation followed by conversion of the Cys residues to Ala by Raney nickel desulfurization.
The product polypeptide itself completed the final synthetic step by removing the C-terminal modification
under folding conditions, to give the mature 99-residue polypeptide. High-purity homodimeric HIV-1 protease
protein was obtained in excellent yield and had full enzymatic activity; the structure of the synthetic enzyme
was confirmed by X-ray crystallography to a resolution of 1.07 Å. This efficient modular synthesis by a
biomimetic autoprocessing strategy will enable the facile synthesis of unique chemical analogues of the
HIV-1 protease to further elucidate the molecular basis of enzyme catalysis.
Introduction
polyprotein constructs in bacteria and reticulocyte lysate have
suggested that the protease has the ability to fold into an
The HIV-1 protease (HIV-1 PR) is a virally encoded 21.4
kDa aspartyl protease made up of two identical 99-residue
polypeptide chains that together form an enzyme molecule with
a single active site.1 Inhibition of the HIV-1 PR results in
noninfectious viral particles, and this fact has made the protease
a focus of antiviral therapy through development of selective
inhibitors of the enzyme.2 During HIV-1 replication, the HIV-1
PR enzyme is required for processing of the Gag and Gag-Pol
polyprotein precursors into the individual protein components
required for maturation of the virus.3 Because the HIV-1 PR
99-residue polypeptide monomer sequence is itself contained
within the Gag-Pol polyprotein, a conundrum arises: where
does the first molecule of the homodimeric HIV-1 protease come
from?4 Experiments with recombinant expression of Gag-Pol
enzymatically active form while still contained within the Gag-
Pol polyprotein, which allows it to cleave itself from the
polyprotein by an autoproteolytic mechanism.4,5
Although the HIV-1 PR has been well-studied,6 details of its
catalytic mechanism remain to be experimentally determined.
For example, hydrolysis of the substrate peptide bond is thought
to occur by nucleophilic attack of a water molecule activated
by a catalytic aspartate residue located at position 25 in each
monomer.7 Curiously, this water nucleophile has never been
observed in enzyme-inhibitor crystal structures. To test the
validity of this aspect of the mechanism, it would be useful to
systematically vary the electronic properties of the catalytic
aspartate side chains and observe the effects on catalysis.8 Other
aspects of HIV-1 PR catalysis that require further study include
the role of the highly mobile “flaps” that enclose the substrate
within the binding groove of the enzyme.9-12
† Department of Biochemistry and Molecular Biology and Institute for
Biophysical Dynamics, The University of Chicago.
‡ Ben-May Department for Cancer Research, The University of Chicago.
§ Present address: College of Life Science, Nankai University, Tianjin,
China 300071.
(5) Pettit, S. C.; Everitt, L. E.; Choudhury, S.; Dunn, B. M.; Kaplan, A. H. J.
Virol. 2004, 78, 8477-8485.
| The University of WisconsinsMadison.
(6) Wlodawer, A.; Vondrasek, J. Annu. ReV. Biophys. Biomol. Struct. 1998,
27, 249-284.
Department of Chemistry, The University of Chicago.
(1) Miller, M.; Schneider, J.; Sathyanarayana, B. K.; Toth, M. V.; Marshall,
G. R.; Clawson, L.; Selk, L.; Kent, S. B.; Wlodawer, A. Science 1989,
246, 1149-1152.
(2) Swanstrom, R.; Erona, J. Pharmacol. Ther. 2000, 86, 145-170.
(3) Frankel, A. D.; Young, J. A. Annu. ReV. Biochem. 1998, 67, 1-25.
(4) Debouck, C.; Gorniak, J. G.; Strickler, J. E.; Meek, T. D.; Metcalf, B. W.;
Rosenberg, M. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 8903-8906.
(7) Fersht, A. Structure and Mechanism in Protein Science: A Guide to Enzyme
Catalysis and Protein Folding; Freeman: New York, 1999.
(8) Short, G. F., 3rd; Laikhter, A. L.; Lodder, M.; Shayo, Y.; Arslan, T.; Hecht,
S. M. Biochemistry 2000, 39, 8768-8781.
(9) Baca, M.; Kent, S. B. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 11638-
11642.
(10) Baca, M.; Kent, S. B. H. Tetrahedron 2000, 56, 9503-9513.
9
11480
J. AM. CHEM. SOC. 2007, 129, 11480-11490
10.1021/ja072870n CCC: $37.00 © 2007 American Chemical Society

Modular Total Chemical Synthesis of HIV-1 Protease
A R T I C L E S
Scheme 1. Initial Synthetic Strategy for the HIV-1 PR 1-99 Polypeptidea
a Four peptide segments are joined sequentially by native chemical ligation14 (NCL), with conversion of the N-terminal (4R)-1,3-thiazolidine (Thz) into
cysteine after each NCL reaction in order to proceed with the next ligation.16,17 After the third ligation is complete, formyl protection is removed from
tryptophan, and the cysteines are converted into alanine by Raney nickel desulfurization.18
Total synthesis enables precise atom-by-atom control over
the covalent structure of a protein molecule and is thus very
useful for dissecting essential aspects of the molecular basis of
enzyme catalysis.13 Modern chemical protein synthesis is based
on the chemoselective reaction of peptides by native chemical
ligation. In native chemical ligation, an unprotected peptide
thioester reacts with an unprotected Cys peptide in aqueous
solution at neutral pH to give a single product with a native
peptide bond at the ligation site.14 The peptide thioester building
blocks are readily made by optimized Boc chemistry stepwise
solid-phase peptide synthesis (SPPS).15 Because of the statistical
accumulation of resin-bound byproducts in SPPS, yields and
purity are generally better for shorter peptides, and it is therefore
desirable to synthesize proteins in a modular fashion through
ligation of multiple peptide segments. This is of particular
importance when one aims to incorporate nonnatural amino
acids or biophysical probes into the protein molecule, and is
also useful if a large series of analogue proteins is to be
synthesized.
We set out to develop a robust, modular synthetic route to
the HIV-1 PR for our studies of the enzyme. Our initial efforts
to synthesize the HIV-1 PR using multiple peptide segments
were frustrated by the insolubility of intermediate peptide
products. In this paper, we describe how we overcame these
difficulties using state-of-the-art chemical synthesis based on
native chemical ligation (NCL) and Raney nickel desulfuriza-
tion, in a biomimetic synthetic approach inspired by the natural
autoproteolytic maturation of the HIV-1 PR in vivo.
(11) Trylska, J.; Tozzini, V.; Chang, C. E.; McCammon, J. A. Biophys. J. 2007,
92, 4179-4187.
(12) Hornak, V.; Okur, A.; Rizzo, R. C.; Simmerling, C. Proc. Natl. Acad. Sci.
U.S.A. 2006, 103, 915-920.
Results
(13) Kent, S. J. Pept. Sci. 2003, 9, 574-593.
Initial Synthetic Design. Our initial synthetic design is shown
in Scheme 1. The strategy was based on ligation of four peptides,
consisting of residues 1-27 (1-27RCOSR), 28-70 (Thz28-
(14) Dawson, P. E.; Muir, T. W.; Clark-Lewis, I.; Kent, S. B. Science 1994,
266, 776-779.
(15) Schnolzer, M.; Alewood, P.; Jones, A.; Alewood, D.; Kent, S. B. Int. J.
Pept. Protein Res. 1992, 40, 180-193.
9
J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007 11481

A R T I C L E S
Johnson et al.
Scheme 2. (a) Simplified Scheme of Maturation of HIV-1 Gag-Pol Polyproteina and (b) Modification of C-Terminal 95-99 Peptideb
a The protease (PR) catalyzes its own removal from within the polyprotein and then processes the rest of the polyprotein into mature matrix (MA), capsid
(CA), nucleocapsid (NC), p6, reverse transcriptase (RT), and integrase (IN) viral proteins. b To increase the solubility of Cys71-99, 10 residues from the
N-terminus of the reverse transcriptase protein sequence adjacent in the Gag-Pol polyprotein, plus 10 arginine residues, are added to the Cys95-99 peptide.
70RCOSR), 71-94 (Thz71-94RCOSR), and 95-99 (Cys95-
99) of the protease (for peptide sequences, see Table S1,
Supporting Information). The design involved ligation of the
peptides in sequential fashion, starting with the C-terminal
Cys95-99 peptide.
simplify the synthesis, increasing the solubility of Cys71-99
was imperative.
Revised Synthetic Strategy. Our idea was to add a solubi-
lizing sequence of Arg residues at the C-terminus of the first
peptide segment in order to keep all the intermediate peptide
products in solution. The challenge was to devise a way to
remove this Argn tag after assembly of the full-length polypep-
tide chain, to give the mature 99-residue product. Our revised
synthetic design relied on the autoproteolytic biological function
of the protease during HIV-1 maturation (Scheme 2a). On the
basis of this proposed autoprocessing mechanism of HIV-1 PR
maturation, we added 10 residues from the reverse transcriptase
(RT) protein adjacent to the HIV-1 PR sequence in the Gag-
Pol polyprotein to the C-terminus of the 95-99 peptide,
followed by 10 arginine residues to act as a solubilizing “tag”
to keep the polypeptide intermediates in solution (Scheme 2b).19
Ligation of Cys95-99 to Thz71-94RCOSR was rapid and
was complete within 1 h (data not shown). However, the solution
became cloudy as the ligation progressed, suggesting that the
product Thz71-99 peptide was aggregating and precipitating.
In addition to the insolubility of the Thz71-99 peptide, the
synthesis was further complicated by the relative instability of
the Thz28-70RCOSR peptide thioester during the subsequent
ligation (see Figure S2, Supporting Information). Because of
the poor solubility of the Cys71-99 peptide and the poor
stability of the Thz28-70RCOSR peptide thioester, it was
reasoned that ligation between these two peptides could not be
optimized to give satisfactory yields without mutating residues
in Thz28-70RCOSR to stabilize the thioester. In order to
(16) Villain, M.; Vizzavona, J.; Rose, K. Chem. Biol. 2001, 8, 673-679.
(17) Bang, D.; Kent, S. B. Angew. Chem., Int. Ed. 2004, 43, 2534-2538.
(18) Yan, L. Z.; Dawson, P. E. J. Am. Chem. Soc. 2001, 123, 526-533.
9
11482 J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007

Modular Total Chemical Synthesis of HIV-1 Protease
A R T I C L E S
Scheme 3. Biomimetic Synthetic Strategy for HIV-1 PRa
a The protease is synthesized from four peptide segments. The product 1-99RTArg10 polypeptide contains the 1-99 protease polypeptide in addition to
10 residues from the N-terminus of the reverse transcriptase polypeptide adjacent in the Gag-Pol polyprotein and 10 arginine residues to increase the
solubility of intermediate products during the synthesis. After synthesis of the 119-residue polypeptide chain is complete, the polypeptide is dialyzed into
conditions where protease activity is expected, and the protease removes the C-terminal modification through autoprocessing to yield the 2 × 99 residue
homodimeric enzyme.
After assembly of the 119-residue polypeptide was complete,
it was anticipated that the protease would remove the C-terminal
modification after folding through an autoprocessing mechanism
similar to what occurs in vivo, yielding the 2 × 99 residue
homodimeric enzyme. The final biomimetic synthetic strategy
is shown in Scheme 3.
To test this synthetic strategy, the C-terminal peptide was
resynthesized to include the aforementioned modifications, and
the first ligation was performed as before. Satisfyingly, the
Cys71-99RTArg10 peptide product was much more soluble in
ligation buffer than the Cys71-99 peptide, and Thz deprotection
proceeded to completion overnight (Figure 1).
performed with 1-27RCOSR (Figure 3). Ligation was allowed
to proceed for 2 h, after which the ligation solution was made
20% in piperidine to remove formyl protection on tryptophan.
The nonpeptidic materials were removed by SPE, and the
peptide products were treated with Raney nickel to convert Cys28
and Cys71 to the native alanine residues and Cys95 to the mutant
Ala95 (Figure 3). Desulfurization at neutral pH required ap-
proximately 40 h. After desulfurization was complete, the crude
mixture was purified by gradient reverse-phase high-perfor-
mance liquid chromatography (RP-HPLC). The final isolated
yield of 1-99RTArg10, based on the amount of Cys95-
99RTArg10 starting peptide, was 26%.
Methoxylamine was removed from the solution by solid-phase
extraction (SPE), and after lyophilization the peptides were
redissolved in ligation buffer. Thz28-70RCOSR was then added
in order to perform the next NCL reaction, and the progress of
the ligation was monitored by liquid chromatography-mass
spectrometry (LC-MS) (Figure 2). Ligation of Cys71-99RTArg10
to Thz28-70RCOSR went to completion without any of the
complications previously observed.
The purified 119-residue 1-99RTArg10 polypeptide contain-
ing the reverse transcriptase plus arginine tag extension at the
C-terminus was then dissolved in 6 M guanidine hydrochloride
phosphate buffer (0.2 M, pH 7.4) and folded by dialysis against
pH 5.6 acetate buffersconditions expected to result in folding
of the polypeptide chain to form enzymatically active protein.
Analysis by LC-MS and matrix-assisted laser desorption ioniza-
tion time-of-flight (MALDI-TOF) MS of the folding reaction
after overnight incubation at 4 °C showed that quantitative
cleavage of 1-99RTArg10 had occurred between residues 99
and 100 to yield the mature 1-99 polypeptide chain (Figure 4
The Thz group was then removed as before, the methoxy-
lamine was removed by SPE, and the final ligation was
(19) Johnson, E. C.; Kent, S. B. Tetrahedron Lett. 2007, 48, 1795-1799.
9
J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007 11483

A R T I C L E S
Johnson et al.
Figure 1. Ligation of Thz71-94RCOSR to Cys95-99RTArg10 and Thz deprotection. (a) RP-HPLC analysis of the ligation reaction after addition of Thz71-
94RCOSR (t < 1 min). Product has begun to appear before a reaction sample could be quenched with acid and subsequently analyzed. (b) Ligation is
essentially complete within 1 h (expected mass 5633.7 Da, observed mass 5634.9 Da). (c) Thz deprotection after 12 h (expected mass 5621.7, observed mass
5622.2 Da). Reactions were analyzed by analytical RP-HPLC with online ESI-MS detection, using a 5-65% gradient of solvent B (CH3CN + 0.08% TFA)
vs solvent A (H2O + 0.1% TFA). UV detection was at 214 nm. Masses are an average over the principal UV peak during LC-MS analysis, with an
experimental uncertainty of ∼1 part in 5000.
and Figure S3, Supporting Information). Folding yield was 45%,
for an overall synthetic yield of 12% based on the limiting
peptide segment.
after addition of the inhibitor JG-365, a hydroxyethylamine
substrate isostere.22 The HIV1PR‚JG365 crystals diffracted to
1.07 Å (for data collection and refinement statistics, see Table
S3, Supporting Information). Main-chain electron density was
observed to unambiguously end at residue Phe99, showing that
the protease had indeed cut itself from the 1-99RTArg10
sequence at residue 99 (Figure 5c). The atomic resolution of
the electron density map also allowed us to unambiguously
determine maintenance of L stereochemistry at the desulfuriza-
tion sites. The structure with the JG-365 inhibitor showed the
hallmark features present in previously determined structures
of HIV-1 PR inhibitor complexes, including the coplanar
catalytic aspartate side-chain carboxyls, and the enzyme flaps
mediating contact with the inhibitor through a water molecule.
Characterization of Synthetic HIV-1 PR. The mature
enzyme’s polypeptide had an observed mass of 10 704.4 ( 5.4
Da (calculated mass 10 703.5 Da, average isotopes). To verify
that the synthetic HIV-1 PR possessed full activity identical to
the virally encoded enzyme, the synthetic enzyme was assayed
directly from folding buffer by incubation with synthetic
substrate peptides corresponding to the p17/24 (data not shown)
and to the p24/15 Gag-Pol junctions (Figure 5a). The synthetic
enzyme showed the expected substrate specificity by catalyzing
peptide bond hydrolysis only at the expected cleavage sites.
Further kinetic characterization of the synthetic enzyme by use
of a fluorogenic peptide substrate gave steady-state kinetic
parameters kcat and Km of 26.0 s-1 and 25 µM, respectively, in
agreement with previously reported values for the protease under
similar assay conditions (Figure 5b and Table S2, Supporting
Information).10,20,21
Discussion
When nonnatural amino acids or biophysical probes are
incorporated into a protein by total chemical synthesis, it is
desirable to construct the protein in a modular fashion from
shorter peptide segments in order to minimize the difficulty and
expense of synthesizing the peptide that contains the modifica-
tion of interest. However, this approach can occasionally lead
to problems with intermediate polypeptide solubilities during
The structure of the chemically synthesized HIV-1 PR was
characterized by X-ray crystallography. Crystals were grown
from protein material concentrated directly from folding buffer,
(20) Szeltner, Z.; Polgar, L. J. Biol. Chem. 1996, 271, 32180-32184.
(21) Bergman, D. A.; Alewood, D.; Alewood, P. F.; Andrews, J. L.; Brinkworth,
R. I.; Englebretsen, D. R.; Kent, S. B. H. Lett. Pept. Sci. 1995, 2, 99-107.
(22) Rich, D. H.; Green, J.; Toth, M. V.; Marshall, G. R.; Kent, S. B. H. J.
Med. Chem. 1990, 33, 1285-1288.
9
11484 J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007

Modular Total Chemical Synthesis of HIV-1 Protease
A R T I C L E S
Figure 2. Ligation of Thz28-70RCOSR to Cys71-99RTArg10 and Thz deprotection. RP-HPLC analysis of the ligation reaction is shown (a) at t ) 0, (b)
after 1 h, and (c) after 2.5 h. Most of the starting Cys71-99RTArg10 peptide has reacted with Thz28-70RCOSR, which is present in slight molar excess
(Thz28-99RTArg10 expected mass 10 434.2 Da, observed mass 10 437.0). The reaction was allowed to proceed for 5.5 h, at which time methoxylamine
hydrochloride was added to remove Thz protection. (d) Thz deprotection after 15 h (expected mass 10 422.2 Da, observed mass 10 424.8 Da). Reactions
were analyzed as described for Figure 1.
the ligation reactions, even when they are performed in aqueous
denaturant solution at neutral pH. We encountered this difficulty
during our initial attempts at developing a modular synthesis
of the HIV-1 PR, which contains a â-sheet-rich region at the
C-terminus of the protein that has a tendency to form intermo-
lecular H-bonded aggregates, creating problems with chain
assembly during SPPS23 and, in our case, solubility problems
during peptide ligations. While this ordinarily would not be an
insurmountable obstacle to a protein synthesis, as one could
simply extend the reaction time to increase the yield of ligated
product, we were presented with the compounding problem of
reactant instability in the Thz28-70RCOSR peptide thioester.
(23) Henkel, B.; Bayer, E. J. Pept. Sci. 1998, 4, 461-470.
9
J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007 11485

A R T I C L E S
Johnson et al.
Figure 3. Ligation of 1-27RCOSR to Cys28-99RTArg10, formyl deprotection, and desulfurization. (a) RP-HPLC analysis of the ligation reaction at t )
0 after the addition of 1-27RCOSR. A significant amount of ligation has occurred before the reaction could be quenched for analysis. (b) Analysis of the
ligation reaction at 2 h. The ligation was determined to be complete (1-99RTArg10 expected mass 13 449.8 Da, observed mass 13 452.6 Da). (c) RP-HPLC
analysis of the crude products after formyl deprotection, SPE, lyophilization, and redissolution in ligation buffer before addition of Raney nickel (expected
mass 13 393.7 Da, observed mass 13 396.5 Da). (d) Analysis of the desulfurization reaction after 40 h (expected mass 13 296.5 Da, observed mass 13 301.5
Da). Reactions were analyzed as described previously.
Unexpectedly, the Thz28-70RCOSR peptide thioester was
found to be highly labile to hydrolysis and methoxylaminolysis
during the one-pot ligation procedure.17 This reactant thioester
lability, coupled with the poor solubility of reactant cysteine
peptide Cys71-99, created an insurmountable practical barrier
to the attainment of acceptable ligation yields for this reaction.
To overcome this problem, we redesigned the synthesis of
HIV-1 PR, taking inspiration from the natural function of the
protease in vivo. Molecular biology experiments have suggested
that the protease is generated from the Gag-Pol polyprotein
by means of an autoprocessing mechanism, the details of which
are still not completely understood. The protease then functions
9
11486 J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007

Modular Total Chemical Synthesis of HIV-1 Protease
A R T I C L E S
Figure 4. Folding of 1-99RTArg10. (a) RP-HPLC analysis of purified 1-99RTArg10 at t ) 0 in 6 M guanidine hydrochloride phosphate buffer (0.2 M,
pH 7.4). (b) Analysis of folding solution after 20 h. The principal peak has shifted to a slightly later elution time, and a peak corresponding in mass to the
RTArg10 peptide has appeared (for details of mass analyses, see Figure S3, Supporting Information). Analytical RP-HPLC was performed as described
previously.
to process the viral polyproteins into their individual protein
components, including the HIV-1 PR itself, that are required
for viral replication and propagation. We took advantage of this
autoprocessing mechanism in our revised synthetic design by
adding a sequence of the reverse transcriptase polypeptide
adjacent to the protease polypeptide in the Gag-Pol polyprotein
to the C-terminus of the protease, in addition to 10 arginine
residues to increase the solubility of the product Thz71-
99RTArg10 peptide after ligation of Cys95-99RTArg10 to
Thz71-94RCOSR. After all ligations were complete, we
reasoned that the protease would recapitulate in vitro its function
in vivo and remove the RTArg10 modification from its own
C-terminus to give the mature 99-residue polypeptide. To our
satisfaction, not only did the RTArg10 modification increase the
solubility of the Cys71-99RTArg10 peptide to such an extent
where ligation was favored over thioester hydrolysis, but also
the product synthetic protein was able to remove the additional
residues from its own C-terminus upon folding of the 119-
residue polypeptide to yield the desired enzyme molecule, a
homodimer of identical 99-residue polypeptide chains. As
demonstrated by multiple experimental techniques, the final
synthetic protein material was of very high purity and correct
structure, with full catalytic activity.
Efficient autoprocessing of synthetic 1-99RTArg10 provides
strong additional support for the proposed nature of the enzyme
maturation mechanism in vivo: in principle, only one molecule
of HIV-1 PR or another cell-encoded protease would be required
to initiate the autoproteolysis leading to complete polyprotein
maturation. This level of experimental control is difficult to
achieve by the techniques of traditional molecular biology; it
is impossible to definitively rule out a first proteolytic event
by a cell-encoded enzyme. Total chemical protein synthesis is
inherently free from contamination by cellular enzymes and, in
the case of autoproteolytic enzymes, provides the ability to
separate synthesis of the polypeptide from activation of the
enzyme.
This type of “recursive” synthetic design may find other uses
in the synthesis of other enzymes, when one wishes to remove
a functionality useful during the synthesis but not part of the
desired final synthetic product.
Conclusion
We have developed a novel biomimetic approach to the total
chemical synthesis of the HIV-1 protease. A modular synthesis
using native chemical ligation, Raney nickel desulfurization,
and autoprocessing gave good yields of high-purity synthetic
protein, with full enzymatic activity and the correct fold as
determined by atomic-resolution X-ray crystallography. As
exemplified here by the incorporation of Nle36,46 and Abu67,
such a modular synthetic strategy enables the efficient incor-
poration of nonnatural amino acids into the protein in a
straightforward and economical manner. Future structure-
function studies on the HIV-1 protease can now take advantage
of this modular synthetic route to elucidate the details of enzyme
catalysis.
Experimental Procedures
Materials. 2-(1H-Benzotriazol-1-yl)-1,1,3,3-tetramethyluronium
hexafluorophosphate (HBTU), S-tritylmercaptopropionic acid, p-me-
thylbenzhydrylamine (MBHA) resin, Boc-Lys(2ClZ)-OCH2-phenylac-
etamidomethyl (Pam) resin, and tert-butoxycarbonyl- (Boc-) amino
acids were obtained from Peptides International. Boc-Arg(Tos)-OCH2-
Pam resin and Boc-Phe-OCH2-Pam resin were purchased from Applied
Biosystems, Foster City, CA. Side-chain protecting groups used were
Arg (Tos), Cys (4MeBzl), Asp (cHex), Asn (Xan), Glu (cHex), His
(Bom), Lys (2ClZ), Ser (Bzl), Thr (Bzl), Trp (CHO), and Tyr (BrZ).
N,N-Diisopropylethylamine (DIEA) was obtained from Applied Bio-
systems. N,N-Dimethylformamide (DMF), dichloromethane, diethyl
ether, HPLC-grade acetonitrile, and guanidine hydrochloride were
purchased from Fisher. Trifluoroacetic acid (TFA) was obtained from
Halocarbon Products. HF was purchased from Matheson. All other
reagents were purchased from Sigma-Aldrich. All glassware was
thoroughly washed with chromic acid cleaning solution before use.
Peptide Synthesis. The Thz28-70 peptide-aCOSCH2CH2COArg
thioester was synthesized as described previously.24 Peptide-RCOSCH2-
CH2COArg6 thioesters were synthesized by coupling five arginine
This is the first time to our knowledge that a synthetic product
itself completes a synthesis through cleavage of a covalent bond.
9
J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007 11487

A R T I C L E S
Johnson et al.
Figure 5. (a) Cleavage of the p24/p15 substrate peptide by synthetic HIV-1 PR. Substrate peptide (1.8 mg, 0.5 mM) was dissolved at 0.5 mM in acetate
buffer (50 mM, pH 5.6) and incubated with 13 pmol of HIV-1 PR at 37 °C for 9 h. The reaction was then analyzed by MALDI-TOF MS, with 2,5-
dihydroxybenzoic acid as the matrix. All possible peptide cleavage products can be accounted for (p24/p15 expected mass 3582.1 Da, observed mass 3582.6
( 1.8 Da; GHKARVL expected mass 779.9 Da, observed mass 779.5 ( 0.4 Da; AEAMSQVTNSATIM expected mass 1453.6 Da, observed mass 1454.0
( 0.7 Da; MQRGNPRNQRK expected mass 1384.6 Da, observed mass 1385.2 ( 0.7 Da; GHKARVLAEAMSQVTNSATIM expected mass 2215.5 Da,
observed mass 2215.6 ( 1.1 Da; AEAMSQVTNSATIMMQRGNPRNQRK expected mass 2820.2 Da, observed mass 2821.6 ( 1.4 Da). The small peaks
to the right of the principal peaks correspond to sodium adducts. (b) Steady-state kinetics of synthetic HIV-1 PR. The fluorogenic substrate Abz-TI(Nle)F-
(NO2)QR-CO-amide, where Abz ) 2-aminobenzoyl, Nle ) norleucine, and F(NO2) ) p-nitrophenylalanine, was incubated at various concentrations with
the enzyme (100 nM), and initial velocities were determined by the rate of increase in fluorescence upon substrate cleavage. kcat and Km were determined
after fitting of the data points to the Michaelis-Menten equation by use of a nonlinear least-squares fitting program in Origin 7.0. (c) Final σA-weighted 2Fo
- Fc electron density map, contoured at a level of 1σ shown in blue wire mesh, for the C-terminal region of synthetic HIV-1 PR, superimposed on a stick
representation of residues 92-99. Carbon, oxygen, and nitrogen atoms are shown in green, red, and blue, respectively. The map clearly shows the successful
desulfurization of Cys95 to L-Ala,95 as well as the termination of main-chain density at residue Phe.99
9
11488 J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007

Modular Total Chemical Synthesis of HIV-1 Protease
A R T I C L E S
residues to the preloaded Boc-Arg(Tos)-OCH2-Pam resin before
coupling with â-mercaptopropionic acid and continuing with chain
assembly.19 The fluorogenic substrate and Cys95-99RTArg10 peptides
were synthesized on MBHA resin; the p24/p15 substrate peptide was
synthesized on Boc-Lys(2ClZ)-OCH2-Pam resin. Peptides were syn-
thesized on a 0.4 mmol scale by the manual Boc chemistry in situ
neutralization/HBTU protocol described previously,15 with slight
modifications. Briefly, NR-Boc removal was achieved with neat TFA,
via one rapid wash followed by a 2 min batch treatment. The
NR-deprotected peptide-resin was subject to a single flow wash with
DMF. Amino acids (2.2 mmol) were dissolved in 4 mL of 0.5M HBTU
(i.e., 2.0 mmol), were activated by addition of DIEA (1 mL, 6.6 mmol)
for 1 min, and then added to the peptide-resin; coupling was carried
out for 12 min, followed by a single DMF flow wash to remove excess
activated amino acid and soluble coproducts. After chain assembly was
complete, the NR-Boc group was removed as described above, and the
peptide-resin was washed with dichloromethane and dried by aspira-
tion. The peptide was cleaved from the resin and the side-chain
protecting groups were simultaneously removed by treatment with
anhydrous HF at 0 °C for 1 h, with 10% p-cresol added as a scavenger.
After removal of HF by evaporation, the peptide was precipitated,
washed with cold diethyl ether, dissolved in aqueous acetonitrile +
0.1% TFA, and then lyophilized. Isolated yields after HPLC purification,
based on the loading of the resin, were 14.1% for 1-27RCOSR, 13.3%
for Thz28-70RCOSR, 23.7% for Thz71-94RCOSR, and 46% for
Cys95-99.
HPLC. Analytical reversed-phase HPLC was performed on an
Agilent 1100 system with Microsorb C-4 (5 µm, 2.1 × 50 mm) silica
columns packed in-house. Peptides were eluted from the column with
a gradient of 5-65% acetonitrile/0.08% TFA versus water/0.1% TFA
at a flow rate of 0.5 mL/min. Preparative HPLC of crude peptides after
SPPS was performed on a Vydac C-18 (50 × 250 mm) column, and
preparative HPLC of 1-99RTArg10 after desulfurization was performed
on an Agilent C-3 (22 × 250 mm) column. Peptides were eluted from
the column with an appropriate shallow gradient of acetonitrile/0.08%
TFA versus water/0.1% TFA. Fractions containing the desired purified
peptide product were identified by analytical LC-MS, combined, and
lyophilized.
600 mg of NaBH4. The charged nickel was then washed in batch 3×
with 50 mL of water and added to the peptide solution. The reaction
was allowed to proceed under argon until complete as judged by LC-
MS of aliquots, after which the reaction solution was centrifuged and
the supernatant was decanted. The catalyst was washed 3× with 5 mL
of ligation buffer, with the final wash including 1 M imidazole to
displace any peptide adsorbed to the nickel. The supernatants were
pooled and the combined solution was acidified to pH 2 and then
purified by RP-HPLC as described above. The overall yield of purified
119 residue polypeptide for all ligations and other synthetic manipula-
tions was 26% based on the starting peptide segment Cys95-
99RTArg10.
Folding. 1-99RTArg10 polypeptide (15 mg) was dissolved in 12
mL of 6 M guanidine hydrochloride phosphate buffer (0.2 M, pH 7.4)
and folded by a three-step dialysis procedure against sodium acetate
buffer (pH 5.6) by use of a 3500 MW cutoff membrane. The first step
was against 2 L of 50 mM sodium acetate buffer (pH 5.6) for 3 h. The
second step was against 2 L of 25 mM sodium acetate buffer (pH 5.6)
for 2 h. The third step was against 2 L of 10 mM sodium acetate buffer
(pH 5.6) overnight at 4 °C. Final conditions were 10 mM sodium
acetate, pH 5.6. The final concentration of folded protein was
determined by absorbance at 280 nm with a calculated dimer extinction
coefficient of 25 120 M-1 cm-1
.
Enzyme Kinetics. The fluorogenic substrate Abz-TI(Nle)F(NO2)-
QR-CONH, where Abz ) 2-aminobenzoyl, Nle ) norleucine, and
F(NO2) ) p-nitrophenylalanine, was incubated at various concentrations
with the enzyme (100 nM), and initial velocities were determined by
the initial rate of increase in fluorescence after substrate cleavage.25
Final assay conditions were 50 mM sodium acetate buffer, pH 5.6,
and 1% dimethyl sulfoxide (DMSO), 37 °C. Assays were performed
in triplicate for each substrate concentration. kcat and Km were
determined after fitting of the data points to the Michaelis-Menten
equation by use of a nonlinear least-squares fitting program in
Origin 7.0.
Crystallization. Crystals were grown at room temperature by the
hanging drop vapor diffusion method from a well solution consisting
of 0.1 M citrate, 0.2 M sodium phosphate, 30% saturated ammonium
sulfate, and 10% DMSO, pH 6.0. Protein solution (aliquot of
concentrated folding solution) (∼0.5 mM enzyme) was preincubated
with a 20-fold molar excess of (S)-JG-365 and was then mixed in a
2:1 (v/v) ratio with well solution. Crystals grew within 5 days and were
frozen in liquid nitrogen with a cryoprotectant of 30% glycerol.
MS Analysis. Peptide masses were obtained on an Agilent 1100
LC-MS with on-line electrospray ion trap detection. MALDI-TOF MS
analysis of the folding reaction was performed by mixing a sample of
the reaction mixture and a sample of sinapinic acid (saturated in aqueous
acetonitrile + 0.1% TFA) in a 1:1 mixture. MALDI spectra were
acquired on a Perseptive Biosystems Voyager-DE in linear mode.
Structure Determination and Refinement. Diffraction data were
collected at 19-BM station in the Advanced Photon Source (APS) at
Argonne National Laboratory and processed with HKL2000.26 PHASER27
was used to obtain the initial phases, by use of the model of HIV-1
protease in complex with JG-365 (PDB code 7HVP). The starting model
was initially refined by the CNS program28 with data between 50 and
1.5 Å. After free R-factor convergence,29 the model was fed into
SHELX9730 for full anisotropic temperature factor refinement, with
data to the highest resolution shell and with low-resolution cutoff at
10 Å. Iterative model building and refinement were performed with
COOT31 and SHELXL30 until the final Rcryst/Rfree values converged to
14.36/18.37%. Coordinates for the refined model have been deposited
in the Protein Data Bank with accession code 2JE4. Model analysis
Peptide Ligations. Ligations were performed on a 5 µmol scale
(2.5 mM) in 6 M guanidine hydrochloride phosphate buffer (0.2 M) at
pH 7.0, with 200 mM 4-mercaptophenylacetic acid (MPAA) added as
a catalyst and 20 mM Tris(2-carboxyethyl)phosphine hydrochloride
(TCEP) added as a reducing agent. MPAA was repurifed from the
commercial source by HPLC before use. After each ligation, the Thz
was converted into cysteine with 0.2 M methoxylamine HCl at pH
4.0. After Thz conversion was complete, the crude peptide products
were subjected to solid-phase extraction (SPE) on C-4 silica SPE
columns and recovered in 50% aqueous acetonitrile and lyophilized.
After the final ligation was complete, cysteine (1 M) was added and
the solution was made 20% in piperidine for 20 min to remove
formyl protection on tryptophan. The solution was then acidified with
HCl, purified by SPE to remove all nonpeptidic material, and then
lyophilized.
(25) Toth, M. V.; Marshall, G. R. Int. J. Pept. Protein Res. 1990, 36, 544-550.
(26) Otwinowski, Z.; Minor, W. Methods Enzymol. 1997, 276, 307-326.
(27) Read, R. J. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2001, 57, 1373-
1382.
Desulfurization. Lyophilized crude peptide 1-99RTArg10 (67 mg)
was dissolved in 10 mL of 6 M guanidine hydrochloride phosphate
buffer (0.2 M, pH 7.4) with 20 mM TCEP. Raney nickel was prepared
by dissolving 3.6 g of nickel acetate in 24 mL of H2O and slowly adding
(28) Brunger, A. T.; Adams, P. D.; Clore, G. M.; DeLano, W. L.; Gros, P.;
Grosse-Kunstleve, R. W.; Jiang, J. S.; Kuszewski, J.; Nilges, M.; Pannu,
N. S.; Read, R. J.; Rice, L. M.; Simonson, T.; Warren, G. L. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 1998, 54, 905-921.
(29) Brunger, A. T. Nature 1992, 355, 472-475.
(30) Sheldrick, G. M.; Schneider, T. R. Methods Enzymol. 1997, 277, 319-
343.
(24) Hackeng, T. M.; Griffin, J. H.; Dawson, P. E. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 10068-10073.
(31) Emsley, P.; Cowtan, K. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004,
60, 2126-2132.
9
J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007 11489

A R T I C L E S
Johnson et al.
was carried out with programs of the CCP4 package.32 Graphics were
generated by Pymol.33
04ER63786 to S.B.H.K.). Y.S. is funded by an AHA fellowship
(0520123Z), and E.C.B.J. is funded by the Graduate Training
Program in Growth and Development (NIH T32 HD007009).
Acknowledgment. We thank Brad Pentelute for helpful
discussions regarding the desulfurization reaction and Hookang
Im for assistance with X-ray diffraction data collection. This
work was supported by the National Institutes of Health
(GM075993-01 to S.B.H.K., GM62548 to W.-J.T.) and by the
Department of Energy GTL-Genomics Program (DE-FG02-
Supporting Information Available: Synthetic peptide se-
quences, LC-MS of crude polypeptides, ligation data using initial
synthetic strategy, mass analysis of the folding reaction, kinetic
parameters of chemically synthesized HIV-1 PR (PDF); X-ray
crystallography of synthetic HIV-1 PR (CIF). This material is
available free of charge via the Internet at http://pubs.acs.org.
(32) Bailey, S. Acta Crystallogr., Sect. D: Biol. Crystallogr. 1994, 50, 760-
763.
(33) DeLano, W. L. DeLano Scientific LLC, South San Francisco, CA, 2005;
http://www.pymol.org.
JA072870N
9
11490 J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007