a BF3·OEt2-mediated reaction, giving exclusively the b-anomer
of the glycosylated dipeptide with a free C-terminus. Peptide
synthesis was accomplished by fragment condensation on
solid phase using Fmoc-Ser(b-Ac4Gal)-Gly-OH as the building
block followed by purification by HPLC (see Supporting Infor-
mation for details). All peptides were characterized by reverse-
phase liquid chromatography and mass spectrometry.
For direct comparison of glycosylated and non-glycosylated
peptides, we investigated MR121-[Ser(b-Gal)-Gly-]8Trp and
compared the results to previous studies and new control
measurements of MR121-[Gly-Ser-]8Trp. We measured an over-
all steady-state fluorescence quantum yield QYss [that is the
ratio of fluorescence emission from a peptide with (I) and with-
out (I0) quencher involved] of QYss =I/I0 =0.54Æ0.04 for
MR121-[Ser(b-Gal)-Gly-]8Trp (at 208C) as compared to QYss =
0.44Æ0.05 for MR121-[Gly-Ser-]8Trp (interpolated from two
measurements for MR121-[Gly-Ser-]7Trp and MR121-[Gly-Ser-
]9Trp). Fluorescence lifetime measurements by time-correlated
single-photon counting (TCSPC) were performed to estimate
the dynamic quantum yield QYdyn, which reports on collision-
induced fluorescence quenching,[10] and the static quantum
yield QYst =QYss/QYdyn, which reports on fluorescence quench-
ing due to complex formation. The dynamic quantum yield
QYdyn =t/t0 =0.96Æ0.02 for MR121-[Ser(b-Gal)-Gly-]8Trp is
much larger than QYst =0.56Æ0.06 and similar to those mea-
sured for MR121-[Gly-Ser-]8Trp (QYdyn =0.88Æ0.02 and QYst =
0.50Æ0.06), thus confirming that quenching is mostly due to
static quenching from formation of stacked complexes as de-
scribed previously.[9a,b]
Figure 2. a) FCS data for MR121-[Ser(b-Gal)-Gly-]8Trp measured at 208C in
aqueous solution. The data fits well to a single-exponential decay on ns
times and a diffusion decay on ms times [Eq. (S2), Supporting Information].
Residuals are shown as inset. b) FCS data measured at temperatures of 10,
20, 30, and 408C plotted after normalization to the same diffusion-term am-
plitude.
Fluorescence fluctuations due to translational diffusion
through the confocal observation volume (on millisecond time
scales) and due to contact formation and opening processes of
the reporter system (on nano- to microsecond time scales)
were recorded and analyzed by FCS (Figure 2). FCS data could
all be fitted with a simple diffusion model and a single or
double exponential decay function for the fast phase of the
correlation decay [Eq. (S2), Supporting Information]. A small-
amplitude phase on microsecond time scales with an ampli-
tude of <0.05 that results from intersystem crossing of MR121
does not have a significant influence on all other fit parame-
ters.
The amplitude of the fast phase of the correlation decay K=
cc/co (with concentration cc/o of peptides in their closed/open
state) results from formation of quenched MR121-Trp com-
plexes reporting on the end-to-end contact formation and is a
measure of the static quantum yield QYst =1/(K+1). For exam-
ple, at 208C we measure K=0.61Æ0.01, which corresponds to
an ensemble estimate of QYst =0.62Æ0.01. The complex forms
due to short-range hydrophobic interactions and is not rate-
limiting for end-to-end contact formation rates, as was demon-
strated before with unmodified MR121-[Gly-Ser-]xTrp.[9d,11]
In aqueous solution all fast correlation decays are well de-
scribed by a single exponential decay corresponding to a two-
state system (Figure 2). We derived end-to-end contact forma-
tion and opening rate constants kc and ko from amplitude K
and relaxation time trel according to K=kc/ko and trel =
1/(kc+ko). Assuming the power law dependence of contact
rates as described previously,[9d,11] we estimated contact rate
constants for MR121-[Gly-Ser-]8Trp of kc =(2.9Æ0.5)ꢁ107 sÀ1 at
208C (which were confirmed on the current setup). For
MR121-[Ser(b-Gal)-Gly-]8Trp we measured a contact rate con-
stant of kc =(0.8Æ0.1)ꢁ107 sÀ1 which is smaller by a factor be-
tween three and four. In order to illuminate how glycans slow
down end-to-end contact formation, we performed tempera-
ture- and viscosity-dependent measurements.
Focusing on the millisecond phase of the correlation decay
we estimated the translational diffusion constant D of MR121-
[Gly-Ser-]8Trp and MR121-[Ser(b-Gal)-Gly-]8Trp from FCS meas-
urements in a temperature range from 5 to 408C and found
significant smaller D values for glycosylated peptides. Accord-
ing to the Stokes–Einstein equation D is related to an average
hydrodynamic radius Rh as D=kBT/6phRh (with Boltzmann con-
stant kB, temperature T, viscosity h). In Figure 3 diffusion con-
stants for MR121-[Ser(b-Gal)-Gly-]8Trp and MR121-[Gly-Ser-]8Trp
are plotted as function of T/h(T) and fitted by linear regression
confirming the Stokes–Einstein relation. From the fit we esti-
mated the relation of Rh of non-glycosylated to glycosylated
peptides Rh(ÀGal)/Rh(+Gal) to be 0.74Æ0.03. The number con-
firms that the glycopeptide is slightly expanded spending less
time in the closed conformation with the reporter system in
contact.
2908
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 2011, 12, 2907 – 2911