Journal of the American Chemical Society
Article
predominantly cis peptide bonds while PPII contains
(COD) (L = donor ligand, COD = cyclooctadiene) in
dimethylformamide (DMF). These species are also efficient
initiators for living polymerization of NCAs and allow tuning
18−23
predominantly trans bonds.
Pro is found in high density in the structural protein
2
4
41
collagen, the most abundant protein in mammals, as well as
in mucin proteins which are the essential gel-forming
of polypeptide end groups. Since the Ni amido-amidate is
preformed rather than generated in situ from NCAs, this
method bypasses the NH requirement. The propagation steps
have not been well characterized, and we envisioned possible
mechanisms that would not require an NH proton. Therefore,
we investigated the use of these complexes to polymerize Pro-
based NCAs.
25
component of mucus. Pro plays an important structural
role that contributes to the physical properties of these unique
and essential proteins. Therefore, Pro-based polypeptides are
expected to have useful material properties. In our own work
toward NCA-derived synthetic mucins, we desired the
incorporation of Pro NCA. We were also curious about the
properties of poly(trans-4-hydroxy-L-Pro) (PHP) since this
enzymatically generated derivative of Pro also plays a structural
We prepared Pro and trans-4-acetoxy-Pro (AcOPro) NCAs.
We chose the Ac protecting group since it is easily installed
and removed. Our starting materials were Boc protected AAs
which were treated with triphosgene and triethylamine as
2
4
role in collagen and would have increased aqueous solubility
as compared to Pro.
32,34,35
reported by others.
However, we greatly improved the
Preparation, polymerization, and conformational studies of
purification procedure by use of anhydrous flash chromatog-
8
42
Pro NCAs were first reported in 1948 by Astbury et al., which
raphy. In order to remove unreacted starting material and
were later followed up on in a series of studies led by Fasman,
uncyclized acyl chloride species, a prior procedure reported use
of a cold aqueous workup, followed by precipitations, and
1
9−22,26−29
Katchalski, Blout, Sela, and co-workers.
Subse-
32
quently, acylated 4-OH-Pro NCAs were investigated for
finally crystallization. This procedure resulted in an isolated
yield of 30% and would take an estimated 4−5 days to
accomplish. Our flash chromatography method resulted in 70−
80% yields of analytically pure Pro and AcOPro NCAs and can
be completed in a few hours. Further, in our hands,
crystallization did not fully remove uncyclized species that
still contain the Boc group as evidenced by its persistent ATR-
30,31
potential use in antithrombotic surfaces.
These seminal
studies established PP as a material with unique structural
properties and enabled characterization of PPI and PPII
helices. However, the NCAs were prepared by treatment of
Pro with phosgene followed by addition of silver oxide to
promote cyclization. This method is now known to result in a
significant contamination by uncyclized N-carbamoyl chloride
and diketopiperidine side products that are challenging to
−
1
FTIR carbonyl absorbance at 1660−1720 cm , which was
only fully removed after chromatography (Figure 2D; see SI).
Attempts to drive the cyclization to completion through excess
regents, extended reaction times, and heat did not push
conversion further (see SI).
3
2,33
remove by precipitation or crystallization.
Use of N-tert-
34
33,35
butyloxycarbonyl (N-Boc) Pro and addition of a base
were later shown to improve cyclization, but purification
remained a challenge. Further, the likely impure NCAs were
polymerized using alkoxide and amine bases which are well-
established to be plagued by side-reactions that result in short
To investigate the polymerization of Pro and AcOPro NCAs
with transition metal initiators, we subjected the monomers to
(
PMe ) Co, BiPyNiCOD, or Ni amido-amidate 1. We chose
3 4
11,13,36
oligomeric side chains.
The polypeptides were only
to use an azide species which should install the clickable group
analyzed by intrinsic viscosity which could not have fully
revealed the masses and dispersities. More recently, Gkikas et.
al reported synthesis and polymerization of Pro and 4-BnO-
Pro NCAs using amine initiators and high vacuum
at the chain initiation site for later attachment to fluorophores
3
or other targets. Corroborating the data reported by Deming
40
and Curtin, no polypeptide was formed by treatment of Pro
3
4
3
2,37,38
techniques.
PP chain lengths of ca. 100 residues could
evidenced by ATR-FTIR (Figure 1A, Figure S1A). However,
to our delight, treatment of these NCAs with initiator 1 in
DMF at monomer to initiator (M:I) ratios up to 200:1 resulted
in rapid and quantitative formation of PP and PAcOP (Figure
be obtained (as determined by two-angle light scattering).
However, the low yield of purified NCA (30%), the complex
custom glassware, need for intermittent degassing during
polymerization, and slow reaction times (ca. 7 days) sparked
our interest in improving the process of both purification and
polymerization of Pro-based NCAs. Here, we are pleased to
report a rapid and high-yielding method for preparation and
polymerization of Pro and AcOPro NCAs using anhydrous
chromatography and transition metal catalysts.
1
B, Figure 2A, B, Table 1). The chain lengths of these
materials were readily tunable by initiator stoichiometry, and
we could generate high MW chains. Similar to other NCAs,
Pro polymerizations conducted at M:I ratios of 200:1 took less
than 1 h to observe complete monomer consumption. AcOPro
NCA 200:1 M:I polymerizations completed in 1−2 h. We
characterized the polypeptides by size exclusion chromatog-
raphy (SEC) coupled to 18-angle light scattering (MALS) and
refractive index (RI) detectors and/or via end group analysis
RESULTS AND DISCUSSION
■
In the late 1990s, Deming reported zerovalent Ni, Co, and Fe
catalysts with impressively fast NCA polymerization kinetics
that suppress side-reactions and result in low dispersity (Đ =
1
43,44
by H NMR.
PP had a tendency to gel in our SEC solvent
which impeded obtaining dispersity data on the homopolymer;
however, dispersities of soluble copolymers were low.
<
1.2), high MW (degree of polymerization, DP = 900+)
39
polypeptides. Their careful mechanistic studies revealed that
polymerization initiation results from formation of metal
amido-amidate species that require 3 equiv of NCA per metal
Observed M (number-average molecular weight) increased
n
linearly with increasing M:I ratios and correlated with the
predicted values (Figure 2 A, B; Table 1, entries 1−8).
12,14,40
0
and an NH β-hydride elimination (vide infra).
This
Co amido-amidates can be formed by reaction of
latter step eliminates the possibility of initiation with Pro NCA,
which lacks the NH. However, later work by Deming and
Curtin demonstrated Ni amido-amidates can also be prepared
(PMe ) Co with NH-containing NCAs and should be present
3
4
16
on the resulting polypeptide chain end. Therefore, we
supposed polymerization of Pro NCA should also be possible
using a polypeptide macroinitiator to yield a diblock with PP
by quantitative reaction of alloc-α-aminoamides and L Ni-
2
1
1483
J. Am. Chem. Soc. 2021, 143, 11482−11489