 |
Description  |
|
|
1. Field of the Invention
The present invention relates to improved deprotection reagents used in
peptide synthesis, especially in automated solid-phase peptide synthesis.
2. Brief Description of the Prior Art
The synthesis of peptides is generally carried out through the condensation
of the carboxyl group of an amino acid, and the amino group of another
amino acid, to form a peptide bond. A sequence can be constructed by
repeating the condensation of individual amino acid residues in stepwise
elongation, or, in some cases, by condensation between two preformed
peptide fragments (fragment condensation). In such condensations, the
amino and carboxy groups that are not to participate in the reaction must
be blocked with protecting groups which should be readily introduced, be
stable to the condensation reactions and be removed selectively from the
completed peptide. If a peptide involves amino acids with side chains that
may react during condensation, the problem of protection becomes
increasingly difficult. A great range of reactive groups and side
chains--amino, carboxy, thiol, hydroxy, and so on--must be adequately
blocked. Their blocking must be stable to unmasking of the .alpha.-amino
or .alpha.-carboxy block for stepwise condensation, and must be readily
removed at the final stage, leaving the completed peptide moiety intact.
A successful synthesis for a large peptide by a linear strategy must
achieve nearly quantitative recoveries for each chemical step. This
demanding requirement has been met by solid-phase peptide synthesis,
pioneered by R. B. Merrifield. In such a synthesis, the peptide chain is
normally attached by a benzyl-type carboxy-protective group to a
polystyrene resin. A first amino acid is attached to the resin by a benzyl
linkage, deblocked at its amino side, and coupled with a second amino
acid, carrying a protected .alpha.-amino group. The resulting protected
dipeptide ester is deblocked with trifluoroacetic acid, converted into the
free amine with base, and coupled to a second N-protected amino acid.
After many repetitions of these steps, the complete peptide is cleaved
from the resin with acid treatment. By using the insoluble resin support
it is possible to isolate the product of each coupling reaction simply by
filtering the resin and washing it free of by-products and excess starting
materials. In fact, the synthetic processes are so simplified and the time
required for one cycle is so shortened that in recent years it has become
quite common to use automated peptide synthesizers. (See for example
Barany, G. and Merrifield, R. B., "The Peptides, Vol. 2": Academic Press,
Inc., New York, 1979, pp. 1-284; or Kemp-Vellaccio, "Organic Chemistry",
pp. 1030-1032 (1980)).
Normally, the synthesis of peptides, either in solution or in solid phase,
culminates in a final strong acid step in which all the protecting groups
and the polymeric support are removed. For this purpose, acids with strong
protonating properties such as hydrogen fluoride (Lenard, J. et al,
Journal of the American Chemical Society, 89:181-182 (1967)), hydrogen
bromide (Merrifield, R. B., Biochemistry, Vol. 3, 1385-1390 (1964)) or
sulfonic acids (Yajima, H. et al, Chem. Pharm. Bull., 22:1087-1094 (1974))
have been used. However, several serious side reactions are known to be
associated with these strong acids. For example, alkylation of
nucleophilic side chains of tyrosine, methionine, tryptophan and cysteine,
by carbocations generated from the alcohol component of the protecting
groups (benzyl, tertiary butyl, and the like), is normally observed (See
for example, Martinez, J. et al, Synthesis: 1981, 333-356). Another
problem is dehydration of the protonated side chain carboxylic acid of
aspartic and glutamic acids, followed by acylation reactions of the
resulting acylium ions (Feinberg, R. S. and Merrifield, R. B., J. Amer.
Chem. Soc. 97:3485-3494 (1975)). For example, the design of phenolic OH
tyrosine protecting groups has been a challenging problem for peptide
synthesis. The difficulty has been the tendency of the alkylation of the
phenolic ring of tyrosine during strong acid removal of the
O-alkyl-tyrosine protecting group. Among the commonly used protecting
groups for tyrosine are R.dbd.benzyl, 2,6-dichlorobenzyl,
2-bromobenzoxycarbonyl and cyclohexyl. Of these, O-benzyl tyrosine gives
the most alkylated product during strong acid treatment. This is shown in
Scheme I:
##STR1##
It has long been recognized that the 3-alkylated product (2) is the result
of a cleavage mechanism in which the generated alkylcarbocation leads to
the C-alkylation of the nucleophilic phenyl ring at the 3-position.
Sequential acid treatment during peptide synthesis, also often leads to
electrophilic alkylation of unprotected tryptophan residues. This
alkylation has been recognized to be more serious than the final strong
acid cleavage step (Chino N. et al, Peptide Chemistry 1977 Protein
Research Foundation, pp 27-32, Osaka, Japan). For example, the acidolytic
removal of N.sup..alpha. -tertbutoxycarbonyl group by trifluoroacetic acid
results in about 30% of t-butylated tryptophan side products. Although the
seriousness of the alkylation can be reduced in the presence of
appropriate scavengers during the acid treatment, it is best to prevent
the alkylation side reaction by protecting the indole moiety against
electrophiles during the synthesis. This has been achieved by using
N.sup.i -formyl protecting groups, which effectively decrease the
electrophilicity of the indole tryptophan ring and prevent alkylation side
reactions. N.sup.i -formyl tryptophan is stable to strong acid, including
HF and sulfonic acids, but is usually removed after the strong acid
deprotection of side chain protecting groups and removal from the resin
support, by nucleophiles or aqueous bases, thus requiring an additional
step, and resulting in further side reactions.
Still another amino acid residue which is the subject of problems during
peptide synthesis is methionine. The incorporation of unprotected
methionine into a synthesis faces side reactions of the thioether side
chain being S-alkylated to sulfonium, or oxidized to sulfoxide groups.
Since this side reaction is often left to be corrected during the
purification step, not all S-alkylation is reversible (see for example
Hofmann, K., Journal of the American Chemical Society 87: 631 (1965)).
Alternatively, the use of methionine sulfoxide (met (O)) in the synthesis
avoids the S-alkylation side reaction due to reduced nucleophilicity of
the thioether side chain. However, both strategies result in the similar
uncertainty of converting met(O) to met at the purification step. In
general, methionine sulfoxide is stable to HF cleavage conditions and
requires subsequent thiolytic reductions to methionine, after the removal
of all the protecting groups. This again necessitates an extra step, and
is, in addition, always slow, and often accompanied by side reactions
(Houghton, R. A. Analytical Biochemistry 98: 36 (1979)).
The art has, for a long time, sought final step deprotection conditions
which will overcome one or more, or possibly all, of the aforementioned
problems. Thus, several studies on the effect of nucleophilic sulfur
scavenger systems during acidolytic deprotection have appeared in the
literature on peptide chemistry. In 1977 it was observed by Brady et al
(Journal of Organic Chemistry 42: 143 (1977)) that dimethyl sulfide
enhanced the rate of cleavage of the benzoxycarbonyl group ("Z group") in
trifluoroacetic acid. Kiso et al (Chem Pharm Bull 28: 673 (1980)) found
that the Z group could be completely removed by trifluoroacetic acid at
room temperature when thioanisole, instead of anisole, was employed as a
cation scavenger. These authors also stated that removal of the benzyl
group (Bzl) from Tyr(Bzl), Thr(Bzl) and Ser(Bzl) by trifluoroacetic acid
was accelerated by various sulfur compounds, but to a different extent.
Yajima et al (Chem Pharm Bull 28: 1214-1218 (1980)) developed a system of
using trifluoroacetic acid-thioanisole-m-cresol for the removal of
benzyloxycarbonyl groups from N-.omega.-amino group of lysine. Kiso et al
(Chem Pharm Bull 26: 2563-2564 (1978)) described the use of thioanisole as
a scavenger for the deprotection of O-protected tyrosine under acidic
conditions. Mixtures of anisole and HF were also tested in their work and
the extent of rearrangement of tyrosine protected with 2,6-dichlorobenzyl,
or tyrosine protected with benzyl, were examined. Finally, Node et al have
suggested the use of mixtures of dimethylsulfide and aluminum trichloride
for dealkylation of esters, such as aliphatic or aromatic carboxylic acid
esters.
All of the aforementioned research, however, is based on the use of acids
such as HF or trifluoroacetic (TFA) in high concentrations (>60-65%).
Under these conditions carbocations are generated due to an S.sub.N 1 type
of mechanism, and scavengers are therefore necessary to remove these
carbocations. In some instances it has been proven useful to work with an
acid such as HBr which operates partly on a S.sub.N 2 mechanism, and may
thus prevent side reactions (Homer, R. B. et al Proc. Chem. Soc., 1963,
367). Hydrogen fluoride, however, which is nonoxidizing and much more
volatile, is usually much preferred over HBr. No successful method using
HF, which prevents the serious side reactions described previously, has
yet been designed by the art.
A need therefore continues to exist for a deprotection reagent for peptide
synthesis, especially for solid phase peptide synthesis, which will not
suffer from the severe side reactions, rearrangements, and the like which
characterize the reagents of the prior art, and which may also serve as a
deprotection reagent for a variety of amino acid residues.
SUMMARY OF THE INVENTION
It is therefore an object of the invention to provide new reagents for
acidolytic deprotection of amino acid residues in peptide synthesis.
It is another object of the invention to provide a method for the synthesis
of peptides.
These and other objects of the invention as will hereinafter become more
readily apparent have been attained by providing:
A method of releasing a functional group present in an amino acid or amino
acyl residue from a resin or protecting residue which is bonded to said
functional group by a linkage having proton affinity, which comprises:
reacting said functional group bonded to said resin or protecting residue
with a mixture of HF and a base, for a time and under conditions effective
to produce said release;
wherein the amounts of HF and base in said mixture are adjusted so as to
cause said release to occur substantially by an S.sub.N 2 mechanism.
Another object of the invention has been attained by providing:
In a method of synthesizing peptides which includes protecting amino acid
or amino acyl side residues with acid labile protecting groups through a
linkage having proton affinity, the improvement wherein said protecting
groups are removed with a mixture of HF and a base, wherein the amounts of
HF and base in the mixture are adjusted so as to cause the removal of said
protecting group to occur substantially by an S.sub.N 2 mechanism.
Yet another object of the invention has been attained by providing:
In a method of synthesizing peptides by the solid-phase methodology,
wherein the growing peptide chain is covalently attached by a functional
group through a linkage having proton affinity, to the organic residue of
an insoluble resin, the improvement wherein said peptide is detached from
said resin using a mixture of HF and a base, wherein the amounts of HF and
base in said mixture are adjusted so that said detachment occurs
substantially by an S.sub.N 2 mechanism.
BRIEF DESCRIPTION OF THE DRAWINGS
A more complete appreciation of the invention and many of the attendant
advantages thereof will be readily obtained as the same becomes better
understood by reference to the following detailed description when
considered in connection with the accompanying drawings, wherein:
FIG. 1 demonstrates the composition of a generalized ternary system
(HF-solvent-base) in the S.sub.N 1 and S.sub.N 2 cleavages of Tyr(Bzl).
The shaded area is of the S.sub.N 2 type. The smooth curve dividing the
equilateral triangle is made up of points where transition from S.sub.N 2
to S.sub.N 1 is observed;
FIG. 2 demonstrates the different behavior of weak bases such as sulfides
in a ternary system (HF-trifluoroacetic acid-sulfide) as represented by
the equilateral triangle diagram. The smooth curves dividing the triangle
show the change-over points from S.sub.N 2 to S.sub.N 1 mechanism when a
given sulfide is used as base;
FIG. 3 demonstrates the effect of potentiating and deactivating solvents in
a ternary system: HF-solvent-dimethylsulfide as represented by the
equilateral triangle diagram;
FIG. 4 demonstrates the dependence of the rate of deprotection of Ser(Bzl)
on the concentration of HF in dimethylsulfide;
FIG. 5 demonstrates the dependence of deprotection of O-Benzyltyrosine on
the concentration of HF in dimethylsulfide, showing both the cleavage
yield of O-Benzyltyrosine (--.quadrature.--), the formation of
3-Benzyltyrosine by the alkylation side reaction (--.DELTA.--) and of the
benzyldimethylsulfonium ion ( );
FIG. 6 demonstrates the dependence of deprotection of N.sup.i -formyl
tryptophan by various concentrations of HF in dimethylsulfide, in the
presence of p-thiocresol, maintained at 5% by volume;
FIG. 7 demonstrates the dependence of methionine sulfoxide reduction, on
the percent volume of HF in dimethyl sulfide;
FIG. 8 demonstrates the high pressure liquid chromatographic analysis of
crude pentagastrin amide on a reverse phase .mu.Bondapak C-18 column
(4.times.300 mm), linear gradient of 2 to 98% solution B (H.sub.2
O:CH.sub.3 CN:H.sub.3 PO.sub.4 50:50:0.1, v/v/v) into solution A (H.sub.2
O:CH.sub.3 CN:H.sub.3 PO.sub.4 95:5:0.1, v/v/v), 2 ml/min. Detection at
280 nm, 0.1 absorbance full scale;
FIG. 9A demonstrates the high pressure liquid chromatographic analysis of
crude Cecropin A(1-33) on a reverse phase .mu.Bondapak C-18 column
(4.times.300 mm), linear gradient of 10 to 50% solution B (H.sub.2
O:CH.sub.3 CN:H.sub.3 PO.sub.4 2:8:0.01, v/v/v) into solution A (H.sub.2
O:CH.sub.3 CN:H.sub.3 PO.sub.4 9:1:0.01, v/v/v), 2 ml/min. Detection at
225 nm, 0.1 absorbance full scale;
FIG. 9B demonstrates the high pressure liquid chromatographic analysis of
purified Cecropin A(1-33) on a reverse phase .mu.Bondapak C-18 column
(4.times.300 mm). Conditions of the analysis are similar to those in FIG.
9A;
FIG. 10 demonstrates the high pressure liquid chromatographic analysis of
crude (Tyr.sup.22) glucagon on a reverse phase .mu.Bondapak C-18 column
(4.times.300 mm), linear gradient of 30 to 50% solution B (H.sub.2
O:CH.sub.3 CN:H.sub.3 PO.sub.4 2:8:0.01, v/v/v) into Solution A (H.sub.2
O:CH.sub.3 CN:H.sub.3 PO.sub.4 9:1:0.01, v/v/v), 2 ml/min, Detection at
280 nm, 0.1 absorbance full scale;
FIG. 11 demonstrates the DEAE-Sephacel chromatographic analysis of crude
(Tyr.sup.22) glucagon after HF cleavage by the improved method of the
invention. Elution with a linear gradient of 0.01M Tris in 6M urea, 0.3N
NaCl (Buffer B) to 0.01M Tris in 6M urea (Buffer A). 0.3 ml/mn. Detection
by radioactivity (--.DELTA.--) (10.sup.3 cpm) and absorbance at 280 nm
(--0--);
FIG. 12 demonstrates the CM-52 chromatographic analysis of crude
(Tyr.sup.22) glucagon after HF cleavage by the standard method
(HF-p-cresol 9:1, v/v). Elution with a linear gradient of 0.1M ammonium
acetate in 6M urea (Buffer B) to 0.01M ammonium acetate in 6M urea (Buffer
A). Flow rate 0.3 ml/min, detection by radioactivity (--0--), (10.sup.3
cpm) and absorbance at 280 nm (--.DELTA.--); and
FIG. 13 demonstrates the high pressure liquid chromatographic analysis of
purified (Tyr.sup.22) glucagon on a reverse phase .mu.Bondapak column
(4.times.300 mm). Conditions similar to those in FIG. 11
DESCRIPTION OF THE PREFERRED EMBODIMENTS
In their search for a reagent which would be useful for carrying out the
final step in peptide synthesis, (i.e., the deprotection of all amino acid
functional groups which are protected with acid labile protecting groups,
and, when working in solid phase, the detachment of the synthetic peptide
from the solid phase), the present inventors discovered limiting
conditions within which HF could be used substantially without any of the
serious side reactions of the prior art, and could be used for an expanded
number of deprotections. Although in the prior art it has generally been
believed that HF is, in fact, useful as a final step reagent in peptide
synthesis, the prior art has had to accept the fact that severe side
reactions occur with this reagent. The present inventors, however, have
discovered that, under certain controlled circumstances, HF can be used
without the side reactions.
The controlled circumstances or limiting conditions for final step
reaction, according to the present invention, are those wherein HF is used
at much lower concentrations than was previously customary in the prior
art. In addition, the reaction is carried out in the presence of a base.
The amount of HF is adjusted in the reagent mixture so that it will be
sufficient to at least partially protonate the linkage joining the
functional group present in the amino acid or amino acyl residue with the
residue to be cleaved therefrom, yet insufficient to fully protonate the
base present in the mixture.
Without being bound by any particular theory, applicants believe that,
under these conditions, the fraction of unprotonated base present in the
mixture enters--as a nucleophile--into a bimolecular nucleophilic
substitution reaction (S.sub.N 2) with the detachable organic residue
involved, (e.g. the protecting group), thereby releasing the amino acid or
amino acyl functional group in protonated form. In other words, applicants
have found conditions wherein an HF-based, final step reagent is fully
capable of utilizing an S.sub.N 2 type of mechanism, instead of the
S.sub.N 1 cleavage mechanism seen in the prior art. The first of these two
alternatives is schematically shown in Scheme II:
Scheme II (S.sub.N 2 mechanism):
##STR2##
In this scheme, P is an insoluble polymeric resin used in solid phase
synthesis. --Q.sub.r1 --CH.sub.2, --Q.sub.r2 --CH.sub.2 -- and --Q.sub.r3
are organic residues which are released from the amino acid functional
groups during the reaction. Thus, --Q.sub.r1 --CH.sub.2 -- is attached to
the polymeric resin and serves as the bridging group between the polymeric
resin and the peptide (in this case a single amino acid, shown for
illustration only). Q.sub.r2 --CH.sub.2 -- is an organic residue which is
utilized as a protecting group for the amino acid functional group
R.sub.1. Q.sub.r3 is a protecting group for the terminal amino group. B is
a base. Scheme II shows the probable mechanism by which reactions occur in
the present invention. First, protonation occurs at the residue R.sub.1
having proton affinity, and at the carbonyl oxygen of the terminal carboxy
group of the amino acid. In a second step, the base attacks the CH.sub.2
carbon of the organic residue, releasing the protonated functional groups,
and yielding detached and deprotected amino acid, as well as resin, and
protecting groups covalently attached to the base.
In comparison, at high concentrations of hydrogen fluoride, such as those
of the prior art, it is believed that the mechanism is as shown in Scheme
III.
Scheme III (S.sub.N 1 mechanism):
##STR3##
It is shown in Scheme III that, at high HF concentration, the mechanism
proceeds via free carbocations P--Q.sub.r1 --CH.sub.2.sup.+, Q.sub.r2
--CH.sub.2.sup.+ and Q.sub.r3.sup.+. In the prior art, these carbocations
have been scavenged with cationic scavengers such as bases. However, in
the present invention, the primary function of the base is to serve as the
nucleophilic displacement reagent, although it is envisioned that some
cationic scavenging would also occur by such base. Because of the nature
of the mechanism, and the absence of free cationic intermediates (which
would otherwise undergo acylation or alkylation reactions), such side
reactions are severely suppressed in the present invention.
In essence, the invention is based on the discovery that a changeover in
mechanism from S.sub.N 1 to S.sub.N 2 occurs in going from high HF
concentrations to low HF concentrations. Although such change in
mechanistic behavior has been observed with other acids (see for example
Yates, K. and McClelland, R. A. Journal of the American Chemical Society
89: 2686-2692, 1967), the present inventors were the first ones to
recognize the existence of such changeover in HF solutions, and to
successfully utilize this changeover in formulating a reagent for
final-step reaction in peptide synthesis.
A great variety of functional groups present in amino acids or amino acid
residues can be involved in deprotection reactions. In addition, the term
"functional group" as used in the present invention and claims also
includes the amino acid group utilized to bind the peptide chain to a
solid polymeric resin, in the case of solid phase synthesis. Normally,
this bridging functional group is a carboxy group. The most common
functional groups which are protected during peptide syntheses are those
present in the side chains of the amino acids, and include such functional
groups as OH (both aromatic and aliphatic); NH.sub.2 ; COOH; nitrogen
atoms of imidazole rings, of tryptophan rings, or of arginine; and SH or
--S--CH.sub.3. When one includes the resin-bridging amino acid functional
group used during solid phase synthesis, the amino acids involved in the
method of the present invention thus comprise the aliphatic amino acids,
such as glycine, alanine, valine, leucine, and isoleucine; the hydroxy
amino acids such as serine and threonine; the dicarboxylic amino acids and
amides such as aspartic acid, asparagine, glutamic acid and glutamine; the
amino acids having basic functions such as lysine, hydroxylysine,
histidine, and arginine; the aromatic amino acids such as phenylalanine,
tyrosine, tryptophan, and thyroxine; the sulfur containing amino acids
such as cysteine and methionine; and other amino acids such as proline or
hydroxyproline. Any other amino acids, natural or non-natural, are of
course also included in the present invention. The amino acids may be
present by themselves, at the end, or along the chain of a synthetic
peptide.
By the term "releasing" as used in the invention and in the claims is meant
the splitting or cleavage of the covalent chemical bond which serves to
unite the amino acid functional groups to their respective protecting
groups or resin-bridging residues. In the process of the present
invention, i.e. in the final step of peptide synthesis, the functional
groups are freed from whatever protecting group or resin they may have
been covalently attached to during the previous process of synthesis.
The terms "resin or protecting residue" as used in the present invention
and in the claims are meant to include both the normally used protecting
groups, as well as the organic functional groups which link a growing
peptide chain to a polymeric resin. In the most general sense, these
residues should be able to undergo bimolecular nucleophilic substitution
by a base, i.e. they are the electrophilic acceptors in the S.sub.N 2
mechanism. A wide range of protecting residues are used and are well known
to the art. They include for example, benzyl, 2-chlorobenzyl,
2,6-dichlorobenzyl, bromobenzyl, carbobenzoxy, methylbenzyl,
methoxybenzyl, benzhydryl, substituted carbobenzoxy, and the like. In the
case where the residue is a bridging residue between the peptide and a
polymeric resin, it has been customary to use the benzyl group in that
role, since it is derived from attachment of the first amino acid in the
peptide to chloromethylated residues on a polystyrene resin. The
invention, however, is not only limited to that particular embodiment.
Polyacrylamide, substituted polyethylenes, polyethyleneglycol, phenolic
resins, polysaccharides (e.g. Sephadex, cellulose) soluble,
non-crosslinked polystyrene, etc. can also be used as resins.
The linkage linking the functional group and the resin or protecting
residue is, by the nature of the atoms present therein (O, N or S in most
cases) basic, that is, it has proton affinity.
Any bond cleavable by a bimolecular nucleophilic substitution reaction of
the base on the protecting or resin residue can be used. Preferably, the
bonds are acidolytically labile, and include all those which, in the past,
have been deprotected with mixtures containing high concentrations of HF.
The most common linkages as used in the invention are ethers or thioethers
(R.sup.1 --CH.sub.2 --O (or S) --CH.sub.2 --R.sup.2), esters
##STR4##
carbamates
##STR5##
benzhydrylamide, sulfenylamine, or sulfonylamides, wherein R.sup.1 is the
radical of an amino acid or aminoacyl moiety and R.sup.2 is the radical of
a protecting group therefor, or R.sup.2 is a polymeric resin. In the most
preferred mode, R.sup.2 is a phenyl or halo-substituted phenyl group which
may or may not be attached to a resin.
The bases useful in the present invention are normally weak bases since,
given their crucial role as nucleophiles in the reaction, they have to
comply with the condition that they not be fully protonated by the HF. The
most preferred bases are those having pK.sub.a values ranging between -2
and -11. In addition, the bases should be nucleophilic. Among the
preferred bases are sulfur containing bases, such as thiols or sulfides.
Among the best are di-C.sub.1 -C.sub.10 -alkylsulfides (linear or cyclic)
wherein the alkyls are the same or different, alkyl phenyl sulfides,
thiophenols, p-thiocresol, 3,4-dithiotoluene, 1,2-dithioethane,
ethanethiol, thioanisole, tetrahydrothiophene and the like. Oxygen bases
such as phenol, m-cresol, p-cresol or anisole can also be used. Because of
their nucleophilicity, pKa and volatility properties, the di-lower alkyl
sulfides are the most preferred and dimethylsulfide is, at the present
time, the best of all.
The concentration of HF is to be adjusted according to the proton affinity
(strength) of the accompanying base and according to the proton affinity
of the linkage between the functional group and the protecting or resin
residue, in order to cause the mechanism to be predominantly S.sub.N 2 in
character, rather than S.sub.N 1 in character, i.e., fall within the range
wherein the linkage is partially protonated and the base is not fully
protonated. Given the availability of "indicator systems," see infra and
of pKa values for bases and proton affinity values of the linkage
functional groups, e.g. ethers, thioethers, carbamates, amides, acyl
guanidine, acylimidazole and the like, it is a matter of testing or
calculation to define this concentration. The concentration will also
depend on the polarity of the solvent mixture, the hydrogen-bonding
ability of the cosolvent and the dissociation of HF itself. Therefore, in
the more polar solvents the effective HF concentrations may be smaller
than in less polar solvents. In general, the concentration of HF is
adjusted so that the acidity of the solution, as measured by the Hammett
acidity function HO (see for example Hammett, Chemical Reviews 16: 67
(1935); "Physical Organic Chemistry" McGraw-Hill Book Company, New York,
N.Y. 1940; as well as Hyman and Garber "The Hammett Acidity Function Ho
for Trifluoroacetic Acid Solutions of Sulfuric and Hydrofluoric Acids,"
Journal of the American Chemical Society, Volume 81, 1847, 1849 (1959),
both of which are herein incorporated by reference), is at most within
plus or minus one whole logarithmic unit of the pKa of the base. Thus, the
acidity of the mixture may be in the range of from -1 to -11. In general
terms, the percent by volume of HF in the solution can be between 0.1% and
60%. More particularly, in using lower dialkyl sulfides as the bases and
as the diluents, the concentration of HF may range from about 15% to 50%,
most preferably 20-40%.
In order for the method of the invention to proceed effectively, the
linkage between the functional group and the protecting or resin residue
has to be at least partially protonated. By "partially protonated" is
meant that at least 1% of the linkage is in the protonated form and 99% is
in the unprotonated form. This ratio can be calculated from the proton
affinity of the linkage functional group. Also, the base should be
sufficiently unprotonated. By the terms "sufficiently unprotonated" is
meant that at least 1% should not be protonated and be free to function as
a nucleophile.
The release reaction carried out in the final step of peptide synthesis
according to the present invention can be performed over a wide range of
temperature and time conditions, which conditions are only limited by
experimental expediency and can be readily determined by those of ordinary
skill in the art without undue experimentation. Normally, the reaction can
be carried out at from about -10.degree. C. to room temperature
(.about.25.degree. C.), preferably at 0.degree. C. The time can be
ascertained by following the yield of free peptide, and is normally within
the range of from 5 minutes to 12 hours. The reagent of the present
invention is used in molar excess over the molar amount of protected and
attached functional groups in the peptide.
The reagent can be used as such (HF plus base, binary systems) or diluted
in a solvent, preferably in polar protic solvents such as trifluoroacetic
acid, acetic acid, and other protic acids (ternary systems). The reagent
can also be used in the presence of additional cationic scavenging
compounds such as 1-20% v/v of phenols, indoles, ethers, thioethers,
thiophenols, sulfides. Sufficient amounts of resin swelling compounds
preferably 1-50%, especially 1-20%, of compounds such as aromatics,
chlorinated hydrocarbons, ethers, esters and the like can also be added.
In the present invention the question of whether any given cleavage
reaction is occurring under S.sub.N 2 or S.sub.N 1 mechanism conditions,
can be readily determined (infra). The reaction can then be done under HF,
solvent and base conditions which will fall within the (desired) S.sub.N 2
region.
The change-over point can be readily measured by a number of reactions
("indicator systems"). Some "indicator systems" that show the change of
S.sub.N 2 to S.sub.N 1 mechanism are:
(a) Deprotection of Boc-Tyr(Bzl). (See Example 2 and FIG. 5). The S.sub.N 2
reaction in this indicator system is characterized (at 0.degree. C., 1
hour) by a reduction in side product, 3-benzyl-tyrosine, and a
corresponding quantitative increase in production of sulfonium salts. The
S.sub.N 1 reaction, on the other hand, gives significant amounts of
3-benzyl-tyrosine as side product and, furthermore, it produces little
sulfonium salts, but significant amounts of aromatic side products, due to
the dimerization or polymerization of the benzylic cation. This pattern is
repeatedly seen when such reaction product mixtures are analyzed by
reverse phase high performance liquid chromatography.
Any system (binary or ternary or higher) wherein the amount of
3-benzyltyrosine is lower than about 15%, preferably lower than about 5%,
and/or wherein the amount of sulfonium salts is higher than about 85%,
preferably higher than about 95% in this indicator system, can be
considered to be in the S.sub.N 2 region.
(b) Rate of deprotection of Ser(Bzl). The S.sub.N 2 reaction in this
indicator system (at 0.degree. C.) is characterized by a slow rate of
benzyl protection group removal and a slow change in rate with the
addition of HF, while the S.sub.N 1 reaction produces faster rate changes
with the addition of HF. The slope change for the rate of reaction by the
S.sub.N 2 mechanism is less than 1, while the slope change for the S.sub.N
1 type is greater than +2;
(c) In addition, the S.sub.N 2 type of reaction causes a chemical reduction
of Met(O) to Met in the presence of the appropriate sulfide, and
deprotects Trp(For) to Trp in the presence of a thiol. Both of these
reactions are slow or do not occur significantly in an S.sub.N 1 type of
reaction.
Therefore using any of these "indicator systems" (a, b or c) one skilled in
the art can determine the changeover for any HF/base system, whether the
system is binary or higher.
In binary HF/base mixtures, with phenolic or thiophenolic additives, there
are many combinations that will be effective in the present invention,
where the cleavage mechanism is largely an S.sub.N 2 type (as long as the
sulfide is not totally protonated). For example, in systems with
HF-sulfide as the binary mixture, the change-over of mechanism from
S.sub.N 2 to S.sub.N 1 with respect to HF concentration (in vol %) is
about 55% for dimethylsulfide, and about 65% for thioanisole. However, the
addition of phenolic and thiophenolic compounds affects this change-over
point. The effect depends on the solvent system. In general, the pKa of
the weak base, such as sulfide, and the potentiating or deactivating
effect of the added solvent will shift the change-over point. Usually, the
weaker the base, the higher the concentration of the HF required for the
change-over in mechanism. If thioanisole is further substituted with
electron withdrawing groups such as C1, Br, NO.sub.2 etc., the pKa of the
sulfide will be lower. Accordingly, the change-over point will be moved
towards higher concentrations of HF.
Since the ability of the reagent of the invention to remove protecting
groups or resin groups by an S.sub.N 2 mechanism depends on the H.sub.o of
the acid system, it is possible either to potentiate or to deactivate the
HF by the addition of various solvents. For example, trifluoroacetic acid
(TFA) has long been known to potentiate the H.sub.o of HF. The H.sub.o
value for 100% HF is -11 and for 50% HF in an inert diluent is -6. 100%
TFA has a pKa of -3.3. However, the addition of 0.9% vol % of HF to TFA
will bring the H.sub.o of the HF-TFA mixture to -6.0. Thus, when various
concentrations of HF in TFA, containing 10% of dimethylsulfide, are used
to test the change-over point from an S.sub.N 2 to an S.sub.N 1 reaction
according to the tyrosine alkylation indicator system, it is found that
the mechanism changes at only about 5% by volume of HF. This is in sharp
contrast to the HF-dimethylsulfide binary system where the change-over
point is at about 55% by volume.
The variation in the change-over point of a three-component system with
changes in the composition can be conveniently represented on an
equilateral triangle graph. FIGS. 1-3 illustrate the application of the
invention results on a series of equilateral triangle plots. The three
illustrated components e.g. solvent, HF and base occupy the apices of the
triangle, which represent 100% of each component. Any point along the line
opposite to the apex represents 0% of that component. The composition of
any point in the system can be represented by the length of a
perpendicular to the given side for each component. If the change-over
points from S.sub.N 2 to S.sub.N 1 reaction given by the solvent, HF and
base system are plotted as in FIG. 1, then a smooth curve can be drawn
through these points, dividing the equilateral triangle into two parts.
The shaded area represents the compositions of solvent, HF and base that
operate through an S.sub.N 2 mechanism, while the clear part represents
the compositions where the ternary mixture operates predominately through
an S.sub.N 1 mechanism. The S.sub.N 2 area can be roughly divided into 3
regions. In part A, where the concentration of solvent is high (>65%), the
HF concentration is rather low (<20%) and can still maintain an S.sub.N 2
character. In part B where the concentration of solvent is intermediate
(30 to 65%) the HF concentration changes accordingly so as to give both an
ionic and acidic condition. In part C, where the solvent concentration is
low and the base increases >35%, the HF concentration has to change to
accomodate for the solvent change. Thus, in the final analysis, the HF
concentration in such system can vary from 0.1 to 60% and still be
effective as the deprotecting reagent.
When trifluoroacetic acid, HF and thioanisole (pKa--7) are plotted in the
diagram, the S.sub.N 2 area becomes larger than the corresponding system
where dimethylsulfide (pKa--5) is used (FIG. 2). This clearly reflects
that thioanisole is a weaker base than dimethylsulfide and requires a
stronger HF-TFA acid system to tie up all the sulfide and cause the change
from an S.sub.N 2 to an S.sub.N 1 mechanism.
In practice there are many solvent systems that can potentiate the acidity
function of HF and produce the just described effects. Solvents such as
formic, acetic and other C.sub.1 -C.sub.10 branched or unbranched
aliphatic acids are common solvents in peptide synthesis, and can be used
accordingly to substitute for trifluoroacetic acid.
A ternary system has an advantage over the HF-sulfide binary system in that
it does not necessarily require a complicated HF apparatus. To achieve the
effect of an S.sub.N 2 deprotection reagent, a stable HF-DMS (1:1 v/v)
reagent can simply be diluted according to a three-component diagram with
trifluoroacetic acid and additional sulfide, and used accordingly in
plasticware. (An example of a two component system that does not require
an HF apparatus is known, although it has several deficiencies and has not
been used with a potentiating third component. Thus, an HF-pyridine
complex (3:1, v/v) can be obtained commercially (Aldrich Chemical Co) as a
stable compound, and can be diluted with TFA and sulfide to obtain a
cleavage reagent within this invention.)
Solvent systems that form strong hydrogen bonds with HF will also
deactivate and reduce the H.sub.o of HF. Thus the change-over point of the
S.sub.N 2 to S.sub.N 1 mechanism will be shifted toward higher
concentrations of HF. Solvents such as H.sub.2 O, alcohols, phenols,
thiols, or thiophenols fall into this category. An example of this system
is represented in FIG. 3 in which HF, H.sub.2 O and dimethylsulfide are
plotted on the equilateral diagram. Since an addition of 10% of H.sub.2 O
will lower the H.sub.o to -8.10, the effect of even small amounts of
H.sub.2 O will be large. Accordingly the diagram in FIG. 3 shows a large
area where the S.sub.N 2 mechanism predominates.
Finally, addition of certain salts can also affect the H.sub.o of HF. For
example, the presence of Lewis acids such as BF.sub.3, SbF.sub.5, or
TaF.sub.5 raises the H.sub.o of HF to so called "super acid" strength,
i.e. H.sub.o <-11. Thus, in such mixtures as HF-BF.sub.3 -DMS or
HF-BF.sub.3 -TFA-DMS the mechanism change-over will be shifted to a much
lower concentration of HF. On the other hand, addition of NaF, KF of LiF
will change the H.sub.o of HF to a more positive value, making it a weaker
acid system. In such system of HF-KF-DMS, or HF-KF-DMS-TFA, the change
over point will shift to a higher concentration of HF.
In addition to being useful for the deprotection of acid labile protecting
groups of such functional groups as carboxylic or hydroxy groups, it has
been discovered that the final step reagent of the present invention is
also capable of deprotecting formylated tryptophan (Trp(For)) in one
single manipulation. In this particular embodiment, it is preferred to add
between 1 and 10% by volume of a C.sub.1 -C.sub.10 alkylthiol or arylthiol
scavenger to the low HF concentration reagent, wherein aryl can be phenyl
or lower alkyl- or halo substituted phenyl. Side products which are
observed in high concentration HF/thiol mixtures are not found at low HF
concentrations. These by-products are thiol addition side reactions to
Trp(For), and occur in high concentrations of HF. Other scavengers can of
course also be used, such as phenols, substituted phenols and the like, in
addition to the HF/base/thiol mixture. It is believed that a low
HF-base-thiol condition, where the HF concentration is maintained within
the limitations described previously, is optimal. When HF concentrations
are above or below these limits, deprotection of the formyl groups is
either too slow or side reactions occur. If a weaker acid such as
trifluoroacetic acid is substituted for HF at the same concentration, the
deprotection of formyl tryptophan is found to be too slow (t 1/2 greater
than 24 hours).
Another discovery made by the present inventors is that when the
accompanying base in the low HF reagent in the present invention is a
reductant, capable of reducing a sulfoxide group, the reagent is fully
capable of reducing methionine sulfoxide to methionine. Thus, for example,
when methionine sulfoxide is treated with HF and a dialkysulfide as the
base, reduction is essentially quantitative.
Given these results, the reagent of the present invention can be used in
one step to mildly and without side reactions deprotect most functional
groups present in peptides in the final step, detach the peptide from the
polymeric resin, remove formyl groups from tryptophan residues and reduce
methionine sulfoxide to methionine.
The regeant can also be used in multi step deprotection procedures, wherein
the first step is deprotection/detachment with the reagent of the peptide
synthesis and, of course, the most commonly utilized, solid phase peptide
synthesis, most especially in the automated solid phase peptide synthesis.
Thus, the reagent can be provided pre-made for direct utilization in
plastic vessels or in automated peptide synthesizers.
In this respect, compositions which are useful for the methodology of the
present invention are also covered by the present application. These
compositions encompass those comprising HF solutions plus C.sub.1
-C.sub.10 dialkyl sulfide or alkyl phenyl sulfide, especially
dimethysulfide, wherein the amounts of HF and sulfide are adjusted so that
when said reagent is used with any of the "indicator systems", supra, the
reagent will result in the operation of an S.sub.N 2 mechanism. The
compositions can be binary, ternary or higher and contain solvents,
scavengers, swelling compounds, etc.
Having now generally described this invention, the same will be better
understood by reference to certain specific examples, which are included
herein for purposes of illustration only and are not intended to be
limiting unless otherwise specified.
EXAMPLE 1
Deprotection of Ser(Bzl).
The kinetic response for the deprotection of Ser(Bzl) to changes in acidity
of HF and dimethysulfide (DMS) was studied. (This is one of the so called
"indicator systems", supra). The deprotection mechanism of HF-DMS mixtures
can basically be accommodated by two distinct acid-rate profiles (see FIG.
4). At high HF concentration (greater than 50% of HF) the rate of removal
of Ser(Bzl) increased rapidly with rising concentration of HF. The rate of
deprotection of Ser(Bzl) at 50% HF concentration was found to be
1.35.times.10.sup.-3 sec.sup.-1 but increased to 3.01.times.10.sup.-3
sec.sup.-1 at 60%. The rate was too fast to be measured accurately in
concentrations above 60% but was estimated to be 23.0.times.10.sup.-3
sec.sup.-1 at 75%, and 60.0.times.10.sup.-3 sec.sup.-1 at 90% HF. Thus,
for practical purposes, most Ser(Bzl) residues were deprotected at
0.degree. C. in the first two minutes when exposed to the high
concentration of HF used in normal deprotecting procedures. However, when
the HF concentration was below 50%, the rate of change was much less
dramatic and slower. With a slope of 0.13 the rate increased only 3-fold
from 10% to 40% of HF concentration. For the same amount of increase of HF
concentration from 60% to 90% of HF, the increase was over one 20-fold
with a slope of 2. Such a sudden break in the rate-acidity profile is
indicative of a change-over from an S.sub.N 2 to an S.sub.N 1 mechanism,
as is shown in FIG. 4. The role of dimethylsulfide in this example serves
not only as diluent but also as nucleophile. Thus, the HF-dimethylsulfide
binary mixture resembles a nucleophilic acid such as HBr, which is known
to remove N.alpha.- benzoxycarbonyl protecting groups by an S.sub.N 2
mechanism, rather than by an S.sub.N 1 mechanism in sulfuric acid (Homer,
R. B. et al Proceedings of the Chemical Society 1963, 367).
The change-over in mechanisms of HF-dimethysulfide can also be explained by
the acidity function of HF-dimethylsulfide mixtures at various
concentrations. Since dimethylsulfide has a pKa of -5.3, the effective
acidity function for this binary mixture should be near the pKa of
dimethylsulfide so t | | |
