reaction condition to cyclize 6, 8, and 9 in toluene to obtain
Sortin1 methyl ester 10 gave highly complex mixture of
products. Therefore, one pot cyclization of 6, 8, and 9 was
conducted under a brief refluxing condition in acetic acid to give
10 (Scheme 2). The reaction was stopped shortly after being
brought to reflux despite resulting in low yield due to the
formation of increasing amounts of multiple byproducts with
longer reaction time. We found that C7”-methyl ester of 10
could be selectively hydrolyzed over C10-methyl ester to give
Sortin1 with LiOH-EtOH. The structures of Sortin1 and the
analog 3 were characterized by MS and NMR analysis. An
analysis of HMBC spectra for Sortin1 methyl ester 10 allowed
unambiguous assignment of the carbon peaks corresponding to
carbonyl groups at C10 and C7” (DMSO-d6, δ = 167.1, 165.9
ppm) and methyl groups at C10 and C7” (DMSO-d6, δ = 52.1,
51.2 ppm). The C10 gives rise to crosspeaks with C11 methyl
proton peak as well as with H4 proton peak. Similarly,
crosspeaks were observed between C7” peak and C9” methyl
proton and H3”/ H5” proton peaks. 13C NMR spectrum of
Sortin1 in comparison with that of 10 revealed a characteristic
downfield shift of the C7” peak (DMSO-d6, δ = 167.0 ppm, Δδ =
1.1 ppm). Therefore, we were able to verify that selective ester
hydrolysis occurred at C7” methyl ester moiety. Next, Sortin1
and the analog 3 were derivatized with propargyl group by amide
coupling to give alkyne-tagged probes 1 and 4. Alkyne tagged
probes were further conjugated with biotin-PEG azide 11 by
CUAAC reaction to provide biotin-tagged probes of Sortin1 (2)
and the analog (5) with 50% and 90% respectively.
UV absorption spectra of compounds 1 and 4 were almost
identical, which were characterized by two broad bands at around
342 nm and 474 nm (Fig. 2). It was thus verified that Sortin1
and its analog can be exited at around 365 nm, at which
benzophenone group is typically irradiated for photoaffinity
labeling experiments.12
Figure 2. UV-VIS spectra of probe 1 (500 μM, ―) and 4 (
1 mM, ―) in DMSO.
We next evaluated the photocrosslinking reactivity of PAL
probes 1 and 4. As a model binding protein, we employed
bovine serum albumin (BSA) as it is known to bind hydrophobic
small molecules such as sterols and aromatic dyes.13 100 μM
probes were first allowed to bind to BSA at 4 °C for 1 h then was
irradiated with a UV lamp at 365 nm at 0 °C for 30 min. To
detect the probe-crosslinked BSA, the unreacted probes were
removed by acetone precipitation then was conjugated to
fluorescein-PEG-azide 12 (Fig. 3a) by CUAAC reaction.7 The
resultant mixture was partially purified by acetone precipitation
to remove unreacted 12 then was resolved by SDS-PAGE. In-gel
fluorescence imaging analysis showed that a fluorescent band
approximately corresponding to the molecular weight of BSA
was detected only under the UV irradiation condition (Fig. 3b).
It therefore demonstrated that Sortin1 effectively served as
photoreactive groups. Similarly, probe 4 was also found to be
capable of photocrosslinking BSA albeit with slightly lower
efficiency (Fig. 3b). This observation may be explained by the
fact that the Sortin1 analog moiety of probe 4 lacks a furan ring
rending it less hydrophobic than Sortin1 in probe 1, and hence is
a weaker BSA binder. Previous structure-activity relationship
studies have shown that removal of the furan moiety in Sortin1
decreases its vacuolar biogenesis inhibitory activity, which raises
a possibility that it may provide an additional binding site to its
target receptor to augment the binding affinity. While probe 1
and 4, which share the common oxyindanodihydropyridine core,
are both photoreactive, we anticipate that there will be
differences in their photocrosslinking efficiency based on the
different target binding affinity of Sortin1 photoaffinity probe 1
and its analog 4. Ultimately, the specificity of photocrosslinking
a target protein by probe 1 should be verified by performing a
control experiment where probe 1 will be competitively
photoreacted with cellular proteins in the presence of excess
amounts of Sortin1. Probe 1-crosslinked protein bands, which
disappear under such a competitive reaction condition should
represent specific Sortin1 binding proteins. Time course analysis
of PAL experiments using probe 1 furthermore revealed that the
photocrosslinked products gradually increased and saturated
beyond 30 min (Fig. 3c). These results are consistent with the
reactivity of a ketobiradical intermediate, which is reversibly
photoactivated in aqueous media.12
Scheme 1. Synthesis of Sortin1 analog and their probes.
Reagents and conditions: (a) toluene (0.25 M), reflux, 4 h, 9%;
(b) propargyl amine·HCl, HBTU, DIPEA, DMF, 1.5 h, 80%; (c)
biotin-PEG-azide 11, CuSO4·5H2O, sodium ascorbate, TBTA,
DMSO/tBuOH/H2O = 1/1/1, 65 °C, 1.5 h, 90%.
Scheme 2. Synthesis of Sortin1 and their probes.
Reagents and conditions: (a) AcOH (0.25 M), 110 °C, 1 min,
10%; (b) LiOH·H2O, MeOH/H2O=2/1, 40 °C, 1 h, 72%; (c)
propargyl amine·HCl, COMU, collidine, DMF, 3 h, 61%; (d)
In conclusions, we have developed Sortin1-based
photoaffinity probes bearing alkyne tag or biotin tag toward
exploration of its target proteins. Based on the structural feature
of Sortin1, we predicted its potential utility as a photoreactive
group. Sortin1 and its analog lacking the furan unit were
constructed respectively by one-step Hantzsch cyclocondensation
reaction. Installation of propargyl group to Sortin1 and the
analog provided alkyne-tagged probes 1 and 4 and the
subsequent CUAAC reaction furnished the desired biotin-tagged
probes 2 and 5 in good yields. We found that Sortin1 is capable
of photocrosslinking a model binding protein by UV-irradiation
at 365 nm. To the best of our knowledge, Sortin1 as an
oxoindanodihydropyridine derivative was exploited as a
photoaffinity reagent for the first time in this present study.
Efforts to identify Sortin1 binding proteins in yeast and plant
cells are underway.
biotin-PEG-azide 11, CuSO4·5H2O, sodium ascorbate, TBTA,
tBuOH/H2O=1/1, 65 °C, 50%.
0.6
Probe 1
0.5
Probe 4
0.4
0.3
0.2
0.1
0
300
350
400
450
500
550
600
Wave length (nm)