New phototriggers 9: p-Hydroxyphenacyl as a C-terminal photoremovable protecting group for oligopeptides

Article abstract of DOI:10.1021/ja991014b

In our search for a more versatile protecting group that would exhibit fast release rates for peptides, we have designed and developed the p- hydroxyphenacyl (pHP) group as a new photoremovable protecting group. We report the application of this protecting group for the dipeptide Ala-Ala (1) and for the nonapeptide bradykinin (2), two representative peptides that demonstrate C-terminus 'caging' and photorelease. The synthesis of these p- hydroxyphenacyl esters was accomplished in good yields by DBU-catalyzed displacement of bromide from p-hydroxyphenacyl bromide. As in the case of caged γ-amino acids 11 (pHP glu) and 12 (pHP GABA) and caged nucleotide 17 (pHP ATP) reported earlier, irradiations of the p-hydroxyphenacyl esters of 1 and 2 actuate the release of the peptides with rate constants that are consistently greater than 10⁸ s^-1 and appearance efficiencies (Φ(app)) that range from 0.1 to 0.3. Release of the substrate is accompanied by a deep-seated rearrangement of the protecting group into the near-UV silent p- hydroxyphenylacetic acid (6). Quenching studies of pHP Ala-Ala (7) with either sodium 2-naphthalenesulfonate or potassium sorbate gave good Stern- Volmer kinetics yielding a rate constant for release of 1.82 x 10⁸ s^-1. Quenching of the phosphorescence emission from pHP Ala-Ala (7, E(T) = 70.1 kcal/mol) and pHP GABA (12, E(T) = 68.9 kcal/mol) were also observed. The biological efficacy of bradykinin released from pHP bradykinin (9) was examined on single rat sensory neurons grown in tissue culture. A single 337 nm flash (<1 ns) released sufficient bradykinin from the p-hydroxyphenacyl protected nonapeptide to activate cell-surface bradykinin receptors as indicated by a rapid increase in the intracellular calcium concentration. A selective antagonist of type 2 bradykinin receptors blocked the biological response. From these results, it is apparent that flash photolysis of p- hydroxyphenacyl protected peptides provides a powerful tool for the rapid and localized activation of biological receptors.

Full text of DOI:10.1021/ja991014b

J. Am. Chem. Soc. 2000, 122, 2687-2697
2687
New Phototriggers 9: p-Hydroxyphenacyl as a C-Terminal
Photoremovable Protecting Group for Oligopeptides
Richard S. Givens,*,† Jo1rg F. W. Weber,† Peter G. Conrad, II,† Gyo1rgy Orosz,‡
Sarah L. Donahue,§ and Stanley A. Thayer§
Contribution from the Department of Chemistry, UniVersity of Kansas, Lawrence, Kansas 66045,
Eo¨tVo¨s Lora´nd UniVersity, Budapest 112, H-1518, Hungary, and Department of Pharmacology,
UniVersity of Minnesota Medical School, Minneapolis, Minnesota 55455
ReceiVed March 29, 1999. ReVised Manuscript ReceiVed January 21, 2000
Abstract: In our search for a more versatile protecting group that would exhibit fast release rates for peptides,
we have designed and developed the p-hydroxyphenacyl (pHP) group as a new photoremovable protecting
group. We report the application of this protecting group for the dipeptide Ala-Ala (1) and for the nonapeptide
bradykinin (2), two representative peptides that demonstrate C-terminus “caging” and photorelease. The synthesis
of these p-hydroxyphenacyl esters was accomplished in good yields by DBU-catalyzed displacement of bromide
from p-hydroxyphenacyl bromide. As in the case of caged γ-amino acids 11 (pHP glu) and 12 (pHP GABA)
and caged nucleotide 17 (pHP ATP) reported earlier,1,2 irradiations of the p-hydroxyphenacyl esters of 1 and
2 actuate the release of the peptides with rate constants that are consistently greater than 108 s-1 and appearance
efficiencies (Φapp) that range from 0.1 to 0.3. Release of the substrate is accompanied by a deep-seated
rearrangement of the protecting group into the near-UV silent p-hydroxyphenylacetic acid (6). Quenching
studies of pHP Ala-Ala (7) with either sodium 2-naphthalenesulfonate or potassium sorbate gave good Stern-
Volmer kinetics yielding a rate constant for release of 1.82 × 108 s-1. Quenching of the phosphorescence
emission from pHP Ala-Ala (7, ET ) 70.1 kcal/mol) and pHP GABA (12, ET ) 68.9 kcal/mol) were also
observed. The biological efficacy of bradykinin released from pHP bradykinin (9) was examined on single rat
sensory neurons grown in tissue culture. A single 337 nm flash (<1 ns) released sufficient bradykinin from
the p-hydroxyphenacyl protected nonapeptide to activate cell-surface bradykinin receptors as indicated by a
rapid increase in the intracellular calcium concentration. A selective antagonist of type 2 bradykinin receptors
blocked the biological response. From these results, it is apparent that flash photolysis of p-hydroxyphenacyl
protected peptides provides a powerful tool for the rapid and localized activation of biological receptors.
Introduction
We report here the efficacy of a photorelease strategy for
two peptides, the dipeptide Ala-Ala (1) and the nonapeptide
bradykinin (2), employing p-hydroxyphenacyl (pHP) as a
photoremovable protecting group for C-terminus carboxylate
groups (eqs 1 and 2). Studies with the triplet quenchers sodium
2-naphthalenesulfonate or potassium sorbate gave good linear
Stern-Volmer quenching from which a triplet lifetime of 5.5
ns was derived. Phosphorescence studies indicated a triplet
energy for the p-hydroxyphenacyl chromophore of 69-71 kcal/
mol. To demonstrate the physiological efficacy of the pHP
phototrigger, we photolyzed p-hydroxyphenacyl bradykinin (9)
in dorsal root ganglion (DRG) preparations producing an in vitro
increase of intracellular calcium, that is,[Ca2+]i.
(1)
Earlier, we had reported1,2 that p-hydroxyphenacyl serves as
an efficient phototrigger for the release of the nucleotide ATP
(5) from 17 and of two γ-amino acids, the γ-carboxyl group of
glutamic acid (3) and γ-aminobutyric acid (GABA, 4) from 11
and 12, respectively (eqs 3 and 4). However, the initial attempts
(2)
* Corresponding author. E-mail: rgivens@ukans.edu.
† University of Kansas.
‡ Eo¨tvo¨s Lora´nd University.
§ University of Minnesota Medical School.
(1) (a) Givens, R. S.; Park, C.-H. Tetrahedron Lett. 1996, 37, 6259-
6262. (b) Park, C.-H.; Givens, R. S. J. Am. Chem. Soc. 1997, 119, 2453-
2463.
(2) Givens, R. S.; Jung, A.; Park, C.-H.; Weber, J.; Bartlett, W. J. Am.
Chem. Soc. 1997, 119, 8369-8370.
10.1021/ja991014b CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/10/2000

2688 J. Am. Chem. Soc., Vol. 122, No. 12, 2000
GiVens et al.
of techniques such as NMR and RP-HPLC would be relatively
straightforward with this model.
Similarly, bradykinin (2) was selected because it represents
a more challenging target in our program to demonstrate the
efficacy and versatility of p-hydroxyphenacyl for a protection-
deprotection sequence. This nonapeptide is representative of
many less accessible, yet biologically significant oligopeptides.
Bradykinin, a naturally occurring, potent modulator of pain,
inflammation, and vasodilator responses,4,5 is a neuroactive
peptide produced after local tissue injury by proteolytic process-
ing of plasma kininogen.6 Bradykinin is one of the most active
pain-producing chemical mediators liberated during tissue
damage. It activates cell surface B2 receptors located on sensory
nerve endings, producing a direct excitation of nociceptors as
well as a secondary release of other chemical mediators that
sensitize nociceptors and produce vasodilation.7,8 Thus, the
photorelease of bradykinin from the protected oligopeptide 9
will enable precise temporal and spatial activation of bradykinin
receptors, which in turn would facilitate studies on the trans-
duction of pain.
(3)
(4)
Results
p-Hydroxyphenacyl Ala-Ala (7) was synthesized as shown
in Scheme 1. Protection of the N terminus by standard Boc
chemistry9 was followed by treatment of the N-protected
dipeptide 14 with 2-bromo-4′-hydroxyacetophenone (8). Re-
moval of the Boc group with TFA afforded the C-terminal-
protected Ala-Ala 7 in an overall unoptimized yield of 46%
for the three steps.
to extend this strategy to R-amino acids failed because of the
hydrolytic instability in aqueous or buffered media2 of the
corresponding p-hydroxyphenacyl esters. Our studies on the
more stable γ-amino acids, phosphates, and nucleotide deriva-
tives did establish, however, that the photochemical deprotection
in aqueous media was accompanied by a concomitant rear-
rangement of the protecting group to p-hydroxyphenylacetic acid
(6). The rearrangement was proposed to occur via the p-
hydroxyphenacyl triplet excited state, suggesting a very rapid
release rate estimated to be at least 107-108 s-1 as derived from
Stern-Volmer quenching results with sodium 2-naphthalene-
sulfonate. A recent report3 challenged our mechanism for the
rearrangement, posing instead that the rearrangement occurs
directly from the excited singlet state by an initial transfer of
the phenolic proton to the ketone carbonyl followed by expulsion
of the ester with concomitant reorganization of the p-quinone
methide (either as its ground state or excited singlet state?) to
the spirodienedione. Our results are not in accord with this
mechanism.
Synthesis of the “caged” bradykinin 9 began with the solid-
phase synthesis of protected bradykinin 10 employing 2-chlo-
rotrityl resin10 using an Fmoc strategy. The peptide was released
from the resin by hydrolysis at the C-terminus attachment using
dichloromethane-methanol-acetic acid (8:1:1), freeing the
carboxylic acid 16 while preserving the protecting groups on
the remaining functional groups.11 The C-terminal carboxylate
was subsequently converted to the fully protected p-hydroxy-
phenacyl ester 10 by reaction with 2-bromo-4′-hydroxyac-
etophenone (8) and DBU in dry DMF, conditions similar to
those employed for Ala-Ala (1, see Scheme 1). The fully
protected p-hydroxyphenacyl bradykinin (10) was then treated
with a deprotection cocktail of 88% TFA, 7% thioanisole, and
5% H2O to remove the N-Boc-, tert-butyl-, and N-Pbf-protecting
groups, leaving intact the p-hydroxyphenacyl ester. The overall
yield of 9 from the esterification-deprotection steps followed
by purification by RP-HPLC was 60%.
Photolyses of the caged peptides were conducted in either
aqueous (H2O or D2O) or buffered media. For the model
p-hydroxyphenacyl Ala-Ala (7), the photolysis reaction was
followed by RP-HPLC and by 1H NMR. The 1H NMR spectrum
initially displayed only the starting ester that, after periodic
irradiations of the sample, eventually dissolved into a spectrum
To gain a better understanding of the mechanism and to
explore the application of the p-hydroxyphenacyl photoprotec-
tion of peptides and oligopeptides, we selected the dipeptide
Ala-Ala (1) as a model, chosen for the following reasons: First,
we sought to overcome the hydrolytic instability of p-hydroxy-
phenacyl esters of R-amino acids, which we reasoned was due
to the enhanced electrophilicity of the ester carbonyl caused by
the adjacent protonated R-amino group. The peptide functional-
ity would not be extensively protonated at physiological pH
and thus would lack this driving force for hydrolysis. Second,
the stability of the chiral center of the amino acid to the
protection-deprotection sequence could be evaluated with Ala-
Ala. Finally, the ease of synthesis, product identification, and
the ready analysis of the progress of “uncaging” by a variety
(4) Dray, A. Can. J. Physiol. Pharmacol. 1997, 75, 704-712.
(5) Hornig, B.; Drexler, H. Drugs 1997, 54, 42-47.
(6) Raidoo, D. M.; Bhoola, K. D. Pharmacol. Ther. 1998, 79 105-127.
(7) Cesare, P.; McNaughton, P. Curr. Opin. Neurobiol. 1997, 7, 493-
499.
(8) Hall, J. M. Gen. Pharmacol. 1997, 28, 1-6.
(9) Song, Y.; Niederer, D.; Lane-Bell, P. M.; Lam, L. K. P.; Crawley,
S.; Palcic, M. M.; Pickard, M. A.; Pruess, D. L.; Vederas, J. C. J. Org.
Chem. 1994, 59, 5784-5793.
(10) Barlos, K.; Gatos, D.; Stauros, K.; Papaphotiv, G.; Scha¨fer, W.;
Wenquing, Y. Tetrahedron Lett. 1989, 30, 3947-3950.
(11) Bo´di, J.; Su¨li-Vargha, H.; Luda´nyi, K.; Ve´key, K.; Orosz, G.
Tetrahedron Lett. 1997, 38, 3293-3296.
(3) Zhang, K.; Corrie, J. E. T.; Munasinghe, V. R. N.; Wan, P. J. Am.
Chem. Soc. 1999, 121, 5625-5632.

New Phototriggers 9
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2689
Scheme 1a
a (a) (t-Boc)2O, THF 0 °C, 97%; (b) p-hydroxyphenacyl bromide (8), DBU, 15 °C, 1,4-dioxane, 62%; (c) TFA, 4 h, 0 °C, 78%.
Table 1. Quantum Efficiencies for the Photolysis of
O-(4-hydroxyphenacyl)-Ala-Ala, Trifuoroacetate Salt (7) in D2O,
Determined by NMR and HPLC
Phosphorescence emission spectra were measured for p-
hydroxyphenacyl Ala-Ala (7) and GABA (12), and a series of
p-hydroxyacetophenone derivatives including the parent p-
hydroxyacetophenone (17),15 R,p-dihydroxyacetophenone (13),
the R,R-dimethyl-R-acetate ester (18), the R-phenylacetate ester
(19), and diethyl p-hydroxyphenacyl phosphate (20).16 The
triplet energies ranged from 68.9 to 71 kcal/mol for the series
whether measured in diethyl ether:isopentane:ethanol (EPA; 5:5:
2) or ethylene glycol:water (EG:W; 2:1) glasses (Table 3).
Quenching of the phosphorescence emission of 7 and 12 by
0.005-0.07 M 2-naphthalenesulfonate (ET ) 60.5 kcal/mol (EG:
W; 2:1)) or of 7 by 0.001 M piperylene (ET ) 59 kcal/mol
(CHCl3)) was observed.16
In a separate experiment, the hydrolytic stability of the
p-hydroxyphenacyl Ala-Ala (7) was determined in aqueous
buffer (Table 4). The ester showed no measurable hydrolysis
(by RP-HPLC) in aqueous media in the absence of buffer or in
Ringer’s solution I (pH 6.5); however, 7 did slowly hydrolyze
over a 24 h period in TRIS buffer (pH 7.3) to give 2,4′-
dihydroxyacetophenone (13).
φrear
[4-hydroxyphenylacetic
acid (6)]
φapp
[Ala-Ala (1)]
entry methoda
φdis (7)
0.25
0.33
0.23
1
2
3
4
NMR
NMR
NMRb
0.24
0.31
0.21
0.21
0.29
0.20
0.27
HPLCc 0.26
averaged 0.267 (0.043) 0.253 (0.051)
0.24 (0.044)
1
a The H NMR analyses were by integration of the absorptions at
7.2 (6), 5.4 (7), and 4.2 ppm (1). The NMR analyses were performed
at 2, 4, 6, 8, 10, and 12 min. A single run of 18 min was included in
this study. The slope of the concentrated vs time was used to determine
the quantum efficiency.2 Each entry in the table represents at least four
and as many as eight determinations. For the HPLC analysis, aliquots
were removed at 2, 4, 6 and 8 min to determine the slope. Each aliquot
was run in triplicate (see Experimental for additional details). b DMF
(2.9 ppm) and acetonitrile (1.9 ppm) used as internal standards.
c Anthranilic acid was used as an internal standard. d Standard deviations
are given in parentheses.
The photorelease of bradykinin (2) was demonstrated by
monitoring the photolysis of a 3 mM solution of p-hydroxy-
phenacyl bradykinin (9) in D2O in a Pyrex vessel at 300 nm by
RP-HPLC as shown in Figure 1 and eq 2. Under these
conditions, bradykinin release was complete in less than 8 min,
yielding bradykinin (2) and the rearranged p-hydroxyphenyl-
acetic acid (6) along with a trace of 13 (<10% of 13 was
detected at 98% conversion of 9). The released oligopeptide
was shown to be identical to bradykinin by FAB-MS, exact
mass, CD, and comparison of spectral data and RP-HPLC
retention volumes when compared with an authentic sample of
bradykinin.17 As shown in Figure 2, the retention of the chirality
of the bradykinin obtained from the photolysis reaction was
established by a CD comparison with an authentic sample.17
Quantum efficiencies were also measured for the photorelease
of p-hydroxyphenacyl bradykinin (Φdis, 9), for the appearance
of bradykinin (Φbrad, 2), p-hydroxyphenylacetic acid (Φpa, 6)
and the hydrolysis product (Φhydr, 13) (Table 5).18
of the two photoproducts, Ala-Ala (1) and p-hydroxyphenyl-
acetic acid (6).12 RP-HPLC also revealed another product, 2,4′-
dihydroxyacetophenone (13) that was formed in low yield
(<10%) even at 100% conversion. The released dipeptide was
shown to be identical to and gave the same optical rotation as
the Ala-Ala employed in the synthesis ([R]D ) -14.77° vs [R]D
) -14.75° (obtained from Sigma13)). Quantum efficiencies were
1
measured by both H NMR (for Φ6, Φ1, and Φ7) and by RP-
HPLC (for Φ6 and Φ7, only) and are given in Table 1.
Quenching studies with 7, employing either sodium 2-naph-
thalenesulfonate or potassium sorbate, gave essentially identical
Stern-Volmer quenching results (Table 2). A triplet lifetime14
of 5.5 ns was determined based on the diffusion-controlled
quenching rate of the disappearance of the pHP ester (7) and
the appearance of Ala-Ala (1), a value consistent with our earlier
studies on the γ-amino acid esters of L-glutamate and GABA.2
Assuming that the dipeptide is released directly from the triplet
3
(i.e., from 7), the rate constant for Ala-Ala formation can be
estimated to be 1.82 ((1) × 108 s-1, which is the rate of decay
(15) The parent p-hydroxyacetophenone has an ET ) 70.5 kcal/mol
measured by the phosphorescence S f T excitation and 0, 0 emission bands:
Kearns, D. R.; Case, W. A. J. Am. Chem. Soc. 1966, 88, 5087-5097.
(16) The phosphorescence emission spectra of 20 and Na+ naphthalene-
2-sulfonate (2) in alcohol water glasses were reported earlier: Givens, R.
S.; Athey, P. S.; Matuszewski, B.; Kueper, L. W., III; Xue, J-y.; Fister, T.
J. Am. Chem. Soc. 1993, 115, 6001-6012. The ET of 2 is reported to be 60
kcal/mol from the S f T absorption spectra (Treinin, A.; Hayon, E. J. Am.
Chem. Soc. 1976, 98, 3884-3891) and that of (E)-piperylene is reported
to be ET ) 59 kcal/mol from its S f T absorption spectrum in CHCl3
(Kellogg, R. E.; Simpson, W. T. J. Am. Chem. Soc. 1965, 87, 4230-4234).
(17) Obtained from Fluka Chemical Co.
3
of 7.
(12) Givens, R. S.; Weber, J. F. W.; Jung, A. H.; Park, C.-H. Caged
Compounds. In Methods in Enzymology; Marriott, G., Ed.; Academic
Press: New York, 1998; Vol. 291, pp 1-29. See Figure 1.
(13) Obtained from Sigma Chemical Co, St. Louis, MO.
(14) The value of 41 M-1 from the Stern-Volmer slope (Table 2) was
divided by 7.4 × 109 M-1 s-1, the calculated rate of diffusion in H2O
(Simons, J. P. Photochemistry and Spectroscopy; Wiley-Interscience: New
York, 1971; pp 212-213) to give the triplet lifetime.

2690 J. Am. Chem. Soc., Vol. 122, No. 12, 2000
GiVens et al.
Table 2. Stern-Volmer Quenching Data for the Photolysis of O-(4-hydroxyphenacyl)-Ala-Ala (7) by Sodium 2-Naphthalenesulfonate and
Potassium Sorbate
(A) [Na+ 2-naphthalenesulfonate]
[M]
Φdis
Φacid
(6)
ΦAla-Ala
(1)
(pHP Ala-Ala, 7)
0
0.267 (0.043)a
0.174 (0.044)
0.13 (0.026)
0.098 (0.04)
0.24 (0.10)
0.253 (0.051)
0.151 (0.03)
0.11 (0.01)
0.0197
0.0364
0.0450
0.133 (0.10)
0.09 (0.003)
0.076 (0.009)
0.099 (0.005)
b
slope (KSV)[M-1
τ (ns)c
]
36 (2.0)
54 (3.7)
32 (1.8)
4.3 (0.2)
4.9 (0.3)
7.3 (0.5)
τ ) 5.5 ns (av)
kr, 108 (s-1
)
2.0 (1.7)
1.4 (0.7)
2.3 (1.0)
d
kr ) 1.82 ((1.1) × 108 s-1 (av)
(B) [K+ sorbate]
[M]
Φdis
Φacid
(6)
ΦAla-Ala
(1)
(pHP-Ala-Ala, 7)
0.00
0.27 (0.02)e
0.12 (0.01)
0.066 (0.016)
0.038 (0.01)
0.24 (0.016)
0.11 (0.007)
0.0663 (0.007)
0.0531(0.005)
0.25 (0.01)
0.025
0.050
0.125
0.11 (0.006)
0.053 (0.005)
0.042 (0.004)
b
slope (KSV)[M-1
τ (ns)c
]
49 (4)
36 (8)
38 (11)
6.6 (0.5)
4.9 (1.1)
5.1 (1.5)
τ ) 5.5 ns (av)
kr, 108 (s-1
)
1.5 (0.2)
2.0 (0.9)
2.0 (0.7)
d
kr ) 1.82 (( 0.6) × 108 s-1 (ave.)
a Standard deviations are given in parentheses. See Table 1 and experimental for details. b The Stern-Volmer slope (KSV) was determined from
a plot of the reciprocal of the quantum efficiency vs the 2-naphthalenesulfonate concentration. c The lifetime (τ) of the triplet was determined from
τ ) KSV/kdiff, where kdiff is 7.4 × 109 M-1 s-1 (ref 14). d The rate constant (kr) was determined from kr ) τ-1, which assumes that dissociation is
the major pathway out of the triplet. e Slopes for the concentrations vs time were converted to relative quantum efficiencies using the values from
Table 1. Standard deviations are given in parentheses. See Table 1 and experimental for details.
shifts from 405 to 490 nm and thus, [Ca2+]i is related to the
ratio of these two signals in manner that is independent of dye
concentration, photobleaching, and optical path length (see
Table 3. Phosphorescence Emission for a Series of
p-Hydroxyacetophenones in EPA, Water:Ethanol:Methanol and
Ethylene Glycol:Water Glasses at 77 K
λex λem λ0.0
ET
Experimental Section for calculations). As shown in Figure 3A,
a 60 s application of 100 nM bradykinin evoked a rapid and
transient increase in [Ca2+]i as indicated by the simultaneous
decrease in fluorescence detected at 405 nm and increase at
490 nm. Approximately one-third of the DRG neurons are
known to express bradykinin receptors, consistent with our
finding that 37% (n ) 40) of the cells responded. The 50%
effective concentration (EC50) for bradykinin activation of
sensory neurons is 6 nM, and thus, the response to 100 nM
bradykinin defines the maximal response for a given cell.20 The
[Ca2+]i rose from a basal level of 38 ( 11 nM to peak at 282
( 99 nM (n ) 14). Cells with robust responses to bradykinin
were allowed to recover for 20 min, and then 10 nM p-
hydroxyphenacyl bradykinin was perfused into the bath. To
trigger the photolysis reaction, light from a nitrogen laser was
directed through a single optical fiber positioned approximately
50 µm from the soma. A single flash (<1 ns) of UV light (337
nm) evoked a [Ca2+]i transient in 8 of 9 cells with a mean peak
amplitude of 115 ( 45 nM. The absence of detectable bleaching
of the indo-1 (absorbance maximum at 350 nm), as indicated
by the raw fluorescence intensity values reported in Figure 3A
(lower panel), suggests that significant uncaging occurred with
flash intensities well tolerated by biological systems. Flashes
applied to cells in the absence of pHP bradykinin failed to elicit
a detectable change in [Ca2+]i. The selectivity of the flash-
induced response was further tested in the experiment shown
in Figure 3B. Following an initial control response to bath-
applied bradykinin (100 nM), the cell was washed for 20 min,
and then, 10 nM pHP bradykinin and the bradykinin receptor
antagonist HOE 14021 were added to the bath. In the presence
of HOE 140, flash photolysis of pHP bradykinin failed to
increase [Ca2+]i. In this cell the flash produced a slight bleaching
ketone
solventa
EPA
concn (M)
(0.003)
(nm) (nm) (est) (kcal/mol)
7
302 435 408 70.1
345 445 415 68.9
350 435 407 70.2
324 425 405 70.6 (70.5)b
309 429 403 71.0
305 435 415 68.9
320 428 405 70.6
274 510 485 60.5 (60)c
302 435d 410 - - - - -
320 440e 415 - - - - -
12
13
17
18
19
20
EG:W (0.05)
EPA
EPA
EPA
EPA
EPA
(0.03)
(0.02)
(0.01)
(0.02)
(0.02)
2-NS
7, 2-NS
7, piperylene EPA
12, 2-NS
WAM (0.012)
EG:W (0.003, 0.005)
(0.003, 0.001)
EG:W (0.1-0.05, 0.07) 316 440f 415 - - - - - -
a EPA ) 5:5:2 ether:isopentane:ethanol; EG:W ) 2:1 ethylene
glycol:water; WAM ) 1:2:2 water:ethanol:methanol. b See ref 15. c See
ref 16. d The intensity was reduced by 50%. e The intensity was reduced
by 50%. f The intensity was reduced by 75%
Table 4. Half-life of O-(p-Hydroxyphenacyl)-Ala-Ala (7) in H2O,
D2O, and Buffered Media
medium
pH
τ2
khydr
Tris buffer
Ringer Ia
H2O
7.3
6.5
7.0
7.0
214 min
>24 h
5.39 × 10-5 s-1
>24 h
D2O
>24 h
a Ringers Solution I: 600 mg of NaCl, 30 m of KCl, and 20 mg of
CaCl2 were dissolved in 100 mL of H2O, resulting in a pH of 6.5.
To test the efficacy of the release of bradykinin in vitro, we
chose to examine bradykinin-evoked excitatory responses in rat
dorsal root ganglion neurons grown in primary tissue culture.
Excitation of sensory neurons produces an increase in [Ca2+]i.19
[Ca2+]i was measured in single neurons using the Ca2+-sensitive
dye indo-1 as described in the Experimental Section. Upon
binding Ca2+, the fluorescence emission maximum for indo-1
(19) Stucky, C. L.; Thayer, S. A.; Seybold, V. S. Neuroscience 1996,
74, 1111-1123.
(20) Burgess, G. M.; Mullaney, I.; McNeill, M.; Dunn, P. M.; Rang, H.
P. J. Neurosci. 1989, 9, 3314-3325.
(18) The recent report by Zhang et al.3 also reports trace of photohy-
drolysis reactions competing with the rearrangement for a series of simple
ester analogues of these p-hydroxyphenacyl peptides.

New Phototriggers 9
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2691
Figure 1. Sequential recordings of the RP-HPLC analysis of the irradiation of p-hydroxyphenacyl bradykinin (9) in D2O. Irradiation was conducted
in a Pyrex tube with 4 RPR-300 nm lamps, aliquots removed at 0.5, 1.0, 1.5, 3.5, 5.5, and 8.0 min which were analyzed by RP-HPLC on a 1.0 cm
Vydac C-18 protein/peptide column using a gradient of 0.1% TFA (5-50% over 45 min) in acetonitrile. Hemimellitic acid (0.85 mg, 2.93 mM)
was added as an internal standard. Retention times were 16.3 min for 13, 17.7 min for hemimellitic acid (i.s.), 19.2 min for 6, 24.7 min for 2 and
27.9 for 9. Additional details are given in the Experimental Section.
of the indo-1 as indicated by the decrease in fluorescence
intensity detected at both 405 and 490 nm (Figure 3b lower
traces).
In fact, the Ala-Ala ester (7) proved sufficiently stable to
satisfy our requirements for a model application of the p-
hydroxyphenacyl group as a phototrigger for the C terminus of
peptides. As shown in Table 4, the rate of hydrolysis was slow,
even in TRIS buffer. This observation and the efficient
photorelease reaction allowed us to extend our studies to
bradykinin, an oligopeptide that plays a major role in the
nociceptive and vascular response to tissue injury.
Discussion
In our earlier reports on the applications of p-hydroxyphenacyl
as a photoremovable protecting group,1,2 we noted that simple
R-amino acids were not amenable to this approach because the
protected esters readily hydrolyzed in aqueous buffer. We further
speculated that the pHP ester hydrolysis may have been
enhanced by the protonated amine group at the pHs employed
(6.5-7.5) in our photochemical studies. To prevent or minimize
this factor, we examined two peptides, that is, the dipeptide
Ala-Ala and the nonapeptide bradykinin, in which the R-amino
group would be less basic and therefore would be far less likely
to be protonated.
The syntheses of these and other p-hydroxyphenacyl-protected
carboxylates are generally accomplished through an SN2 dis-
placement of bromide from 2-bromo-4′-hydroxyacetophenone
catalyzed by DBU. For synthetic oligopeptides, the incorporation
of the photoremovable group can be administered on the C
terminus of the otherwise fully protected oligopeptide after
release of the carboxylate terminus from the solid support,
controlled-pore glass beads, or resin employed in the synthe-
sis.9,10 The resulting p-hydroxyphenacyl esters are stable to the
standard TFA deprotection cocktails employed to remove the
common side chain and amine-protecting groups employed in
peptide synthesis as demonstrated in the synthesis of pHP Ala-
Ala and bradykinin shown in Schemes1 and 2.
(21) Hock, F. J.; Wirth, G. Henke, S. Breipohl, G.; Konig, W.; Knolle,
J.; Schoolkens, B. A. Br. J. Pharmacol. 1991, 102, 769-773. HOE 140
(D-Arg-Arg-Pro-Hyp-Gly-Thi-Ser-D-Tic-Oic-Arg) is a selective agonist for
the B2 bradykinin receptors.

2692 J. Am. Chem. Soc., Vol. 122, No. 12, 2000
GiVens et al.
Figure 2. Comparison of the CD spectra of bradykinin (2) isolated
by HPLC from the photolysis of p-hydroxyphenacyl bradykinin (9)
with authentic bradykinin obtained from Fluka.
Table 5. Quantum Efficienciesa for the Photolysis of
p-Hydroxyphenacyl Bradykinin (9) in D2O
Figure 3. Photolysis of p-hydroxyphenacyl bradykinin (9) evoked a
ædis
æapp
ærear
(6)
æhydr
(13)
neuronal response. Bradykinin-induced increases in [Ca2+
] were
i
(pHP-bradykinin, 9) (bradykinin, 2)
measured in single rat DRG neurons with indo-1-based photometry as
described in the Experimental Section. Indo-1 fluorescence at 490 nm
(unbound) and 405 nm (calcium-bound) is plotted in the lower panels.
Drugs were applied by superfusion at the times indicated by the
horizontal bars. Cells were not stimulated during the 20 min gaps in
the recordings. (A) Bradykinin (100 nM, 60 s) produced an increase
in [Ca2+]i that rapidly desensitized. Following a 20 min recovery period,
p-hydroxyphenacyl bradykinin was applied and at the time indicated
by the arrow, a single pulse of UV light (1 ns, 337 nm) from a nitrogen
laser was delivered to the cell by a single optical fiber positioned
approximately 50 µm from the cell soma. The flash evoked a second
[Ca2+]i response. A representative recording from 8 experiments is
0.205 (0.021)
0.219 (0.022) 0.193 (0.013) 0.021 (0.003)
a Standard deviations in parentheses. See Experimental Section and
footnote a, Table 1, for additional details on the HPLC analysis.
Hemimellitic acid was used as an internal standard.
The photorelease of the two peptides was examined in
aqueous and in buffered media. Both gave release and disap-
pearance efficiencies of slightly greater than 20% along with
the appearance of the rearranged p-hydroxyphenylacetic acid
(6) as the only other major photoproduct (eqs 1 and 2; Tables
1 and 5). A minor amount of photohydrolysis was also observed,
but this amounted to less than 10% of the released peptide.
Quenching studies with sodium 2-naphthalenesulfonate and
potassium sorbate (Table 2) demonstrated the intermediacy of
the triplet of pHP Ala-Ala (37) in the photodeprotection process.
The rate constant, 1.82 × 108 s-1, derived from the Stern-
Volmer slope is essentially the same as the release rate obtained
in our earlier studies on p-hydroxyphenacyl ATP (17)1.
Phosphorescence emission studies (Table 3) on a series of
hydroxyphenacyl esters including pHP Ala-Ala (7), pHP GABA
(12), pHP acetate (18), pHP phenylacetate (19) and pHP diethyl
phosphate (20) showed triplet energies of 68.9-70.6 kcal/mol
in EPA (5:5:2) and ethylene glycol:H2O (2:1) glasses at 77 K,
a value that is >3 kcal/mol above those of the triplet quenchers
employed here (56-60 kcal/mol) under conditions that assured
efficient triplet energy transfer to the quencher.
shown. (B) Bradykinin (100 nM, 60 s) produced an increase in [Ca2+
]
i
that rapidly desensitized. Following a 20 min recovery period, p-
hydroxyphenacyl bradykinin in combination with 300 nM HOE 140,20
a B2 type bradykinin receptor antagonist, were applied and at the times
indicated by the arrows, 1, 2, or 4 pulses of UV light (1 ns, 337 nm)
were delivered to the cell. The flashes failed to elicit a second [Ca2+
]
i
response in the presence of the receptor antagonist. A representative
recording from four experiments is shown.
studies coupled with laser flash photolyses by Andrieux, Save´ant
et al.,23a and Wayner23b,c of several substituted phenacyl
derivatives led Andrieux and Save´ant to suggest that the bond
fragmentation is assisted by an electron transfer from the π*
orbital to the leaving group’s antibonding σ* orbital followed
by bond cleavage and release of a carboxylate ion. We observe
no decarboxylation in our studies with C-terminal-protected
peptides nor with our previously reported deprotection reactions
with pHP glutamate and pHP GABA.2,12 The absence of
decarboxylation products in these studies militates against
; α, α-dimethyl
(22) Banerjee, A.; Falvey, D. E. J. Am. Chem. Soc. 1998, 120, 2965-
2966.
(23) (a) Andrieux, C. P.; Save´ant, J.-M.; Tallec, A.; Tradivel, R.; Tardy,
C. J. Am. Chem. Soc. 1996, 118, 9788-9789. (b) Andersen, M. L.; Long,
W.; Wayner, D. D. M.; J. Am. Chem. Soc. 1997, 119, 6590-6595. (c)
Mathivanan, N.; Johnston, L. J.; Wayner, D. D. J. Chem. Phys. 1995, 99,
8190-8195.
These and earlier results of Falvey,22 Save´ant,23a and
Wayner23b,c suggest a mechanism involving release of the
carboxylate group from the phenacyl triplet. Electrochemical

New Phototriggers 9
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2693
Scheme 2a
a (a) CH2Cl2, CH3OH, CH3CO2H, (8:1:1), 2 h (quantitative); (b) p-hydroxyphenacyl bromide (8), DBU, DMF, rt, 24 h (72%); (c) 88% TFA, 7%
thioanisole, 5% H2O, 2.5 h (84%).
Scheme 3
hexadienoate), the latter quencher has an even lower triplet
energy and does not absorb above 280 nm. The small competi-
tive absorption by 2-naphthalenesulfonate, which we estimate
is 6% at equimolar concentrations of quencher and pHP Ala-
Ala and only 16% at the highest concentration of quencher
employed in this study, does not account for the >60%
quenching we observed. The quenching of both the disappear-
ance of 7 and the appearance of the two products by triplet-
triplet energy transfer clearly demonstrates that the triplet of 7
is the intermediate excited state leading to release of the
dipeptide and rearrangement of the phenacyl nucleus.
The biological utility of pHP bradykinin was demonstrated
in an experiment in which a single nanosecond pulse at 337
nm released sufficient bradykinin to excite sensory neurons. It
was not possible to precisely calculate the amount of energy
delivered to the vicinity of the cell in this system, although
uncaging occurred with laser pulses that did not appear to
produce any observable damage to the cell. Furthermore, our
optical method for [Ca2+]i measurement relied on the fluorescent
indicator indo-1, a dye that absorbs at 337 nm, but was not
significantly bleached at pulses that efficiently uncaged pHP
bradykinin. Clearly, pHP bradykinin is sufficiently stable that
it can be introduced into biological systems without degradation
and yet retains adequate photoreactivity to uncage at UV
energies that spare the cells from toxicity. The actual efficiency
of release can only be approximated using [Ca2+]i as the
bioassay because the amplitude of bradykinin-induced response
is dependent on factors in addition to concentration. The
concentration response relationship for bradykinin-induced
responses in these cells is quite steep because of a threshold
that results from amplification of signaling through the second
messenger cascade. Furthermore, the response depends not only
on bradykinin concentration but also the rate of its application
because the receptor rapidly desensitizes. We can estimate that
at least 30% of the 10 nM pHP bradykinin present was released
in biologically active form because the threshold to elicit a
response by application of bradykinin to the bath is 3 nM.
formation of a free carboxyl radical intermediate, possibly
formed from direct homolysis of the CH2-O bond. These results
are also consistent with the studies reported earlier by us1,2 and
recently by Falvey22 but contrast with an earlier result of
Sheehan and Umezawa,24 where partial decarboxylation was
observed for some p-methoxyphenacyl esters, especially for the
phthaloyl protected R-amino group on glycine. We have
suggested the mechanism paralleling that of Andrieux and
Save´ant shown in Scheme 3 which involves participation of
the CH2-O σ* orbital followed by carboxylate disengagement.12
A recent mechanistic proposal by Zhang, Corrie, Munasinghe,
and Wan3 suggested that a proton transfer occurs from the
phenolic hydroxy group to the carbonyl from the initial singlet
state of the pHP ester and that it is this intermediate p-quinone
methide that releases the carboxylate through a cyclic transfer
of the proton to the ester carbonyl with concomitant participation
of the phenoxy ligand to produce the diene-dione (Scheme 4).
The electronic state of the p-quinone methide (excited or ground
state??) was left undefined.3
In fact, our mechanism has not addressed the state of
protonation of the phenolic group, which would be expected to
be altered in the excited-state singlet or triplet. A proton transfer
to solvent may precede or occur in concert with the migration
of the aryl moiety. We have determined, however, that the
reaction does proceed from a quenchable triplet as shown now
with two different quenchers. Our observed quenching of the
reaction and of the phosphorescence emission by sodium
2-naphthalenesulfonate and by potassium sorbate (potassium 2,4-
Conclusions
In summary, pHP bradykinin is a useful reagent for studying
the kinetics and localization of bradykinin-mediated signaling
processes. Both the efficiency of the release and the lack of
side reactions suggest that the p-hydroxyphenacyl group is ideal
for biochemical kinetic studies at the molecular and physiologi-
(24) Sheehan, J. C.; Umezawa, K. J. Org. Chem. 1973, 38, 3771-3774.

2694 J. Am. Chem. Soc., Vol. 122, No. 12, 2000
GiVens et al.
Scheme 4
was concentrated to ∼10 mL to give white crystals of N-Boc-O-(4-
hydroxyphenacyl)-Ala-Ala (15) (374 mg, 62%): mp 150.5 °C; [R]D )
-35.2° (c ) 0.6, ethyl acetate); IR (KBr) 3420, 3371, 3333, 3260,
2977, 3937, 1733, 1714, 1686, 1608, 1587, 1518, 1459, 1450, 1429,
1367,1331, 1280, 1250, 1212, 1165, 1094, 1070, 1022, 969, 842, 828,
cal level. In fact, the chemical properties of p-hydroxyphenacyl
bradykinin (9) indicate that such compounds are well suited
for activating biological processes by flash photolytic release
of the substrate.
The results reported here and earlier show that the p-
hydroxyphenacyl group can be confidently employed as a
C-terminal photoremovable protecting group for oligopeptides
and provides a general approach to developing reagents for
studying the interactions of oligopeptides with biological
systems.
1
802, 565 cm-1; H NMR (300 MHz, acetone-d6) δ ) 9.39 (s, 1H),
7.86 (d, 2H, J ) 8.8 Hz), 7.50 (m, 1H), 6.91 (d, 2H, J ) 8.8 Hz), 6.04
(m, 1H), AB (δA ) 5.46, δB ) 5.29, |JAB| ) 16.3 Hz, CH2), 4.56 (m,
1H), 4.14 (m, 1H), 1.44 (d, 3H, J ) 7.3 Hz), 1.35 (s, 9H), 1.26 (d, 3H,
7.1 Hz); 13C NMR (100 MHz, acetone-d6) δ ) 190.9, 173.3, 173.0,
163.4, 156.2, 130.6, 126.8, 115.8, 78.7, 66.5, 50.1, 48.2, 28.0, 18.3,
17.7; FAB-MS m/z (relative intensity) 395 (M + 1, 14), 339 (18), 295
(100), 224 (58), 143 (43), 121 (30). Exact mass Calcd for C19H27O7N2
(M + H): 395.1818. Found: 395.1801.
O-(4-Hydroxyphenacyl)-Ala-Ala, Trifluoroacetate Salt‚H2O (7).
A solution of N-Boc-O-(4-hydroxyphenacyl)-Ala-Ala (15) (389 mg,
1.0 mmol) in 15 mL of trifluoroacetic acid (TFA) was cooled to 0 °C
and reacted for 4 h with stirring. The resulting solution was concentrated
by a rotary evaporator and residual solvent removal with a high vacuum
pump. The crude product was extracted with the solution of H2O/EtOAc.
The aqueous layer was collected, and the water was removed by
lyophilization to give a white solid of O-(4-hydroxyphenacyl)-Ala-Ala,
Finally, the photochemical reaction proceeds through a short-
lived, quenchable triplet (3τ ≈ 5.5 ns) with a rate constant of
1.8 × 108 s-1, sufficiently rapid that studies of most fast
biological processes can be initiated with this phototrigger.
Experimental Section
General Procedures. All NMR spectra are reported in ppm (δ) from
tetramethylsilane (1H and 13C). Melting points are uncorrected. TEA
was dried by refluxing over KOH and distilled under Ar, collecting
the middle fraction and storing in the dark. Pyridine and DMF were
dried over phosphorus pentoxide and distilled under reduced pressure,
collecting the middle fraction, which was stored over NaOH or
molecular sieves, respectively. Benzene was distilled from sodium/
benzophenone ketyl prior to use. Technical grade diethyl ether, MeOH,
CH2Cl2, and THF were distilled from CaH2.
Synthesis of 2-Bromo-4′-hydroxyacetophenone (8). 2-Bromo-4′-
hydroxyacetophenone (8) was prepared by the modification of the
method of Durden and Juorio25 which we reported earlier.12
Synthesis of p-Hydroxyphenacyl-Ala-Ala (7). Boc-Ala-Ala (14).
Boc-Ala-Ala (14) was prepared by the method of Song et al.9 Ala-Ala
(1, 300 mg, 1.87 mmol) was converted to Boc-Ala-Ala (14, 470 mg,
97% yield): mp 98 °C; 1H NMR (300 MHz, acetone-d6) δ ) 7.40 (s,
1H), 6.07 (s, 1H), 4.40 (m, 1H), 4.11 (m, 1H), 1.37 (s, 9H), 1.32 (d,
3H, J ) 6.5 Hz), 1.27 (d, 3H, J ) 6.5 Hz).
trifluoroacetate salt‚H2O (7) (326 mg, 77.6%): mp 143 °C; [R]D
)
-39.7° (c ) 0.59, H2O); IR (KBr) 3429, 3357, 2996, 2954, 1738, 1685,
1673, 1608, 1586, 1543, 1520, 1432, 1375, 1280, 1252, 1210, 1186,
1171, 1141, 1060, 970, 846, 828, 806, 728 cm-1; UV-vis (H2O) λmax
(ꢀ) 282 nm (13 323); 1H NMR (300 MHz, D2O) δ ) 7.82 (d, 2H, J )
8.4 Hz), 6.9 (d, 2H, J ) 8.4 Hz), 5.43 (s, 2H), 4.57 (m, 1H), 4.04 (m,
1H), 1.47 (two superimposed d ) s, 6H; J ) 6 Hz); 19F NMR (300
MHz, D2O) δ ) 111.3 (s, 3F); 13C NMR (100 MHz, D2O) δ ) 194.4,
174.0, 171.0, 162.3, 131.3, 126.1, 116.1, 67.5, 49.2, 49.0, 16.8, 16.3;
FAB-MS (free amine) m/z (relative intensity) 295 (M + 1, 35), 277
(12), 185 (100), 161 (8). Exact mass Calcd for C14H19N2O5 (free amine,
M + H): 295.1294. Found: 295.1308. Anal. Calcd for C16H21N2O8F3:
C, 45.07; H, 4.96; N, 6.57. Found: C, 45.09; H, 4.89; N, 6.80.
Synthesis of p-Hydroxyphenacyl Bradykinin (9). N(r-Boc)-N-
(ω-Pbf)-L-arginyl-L-prolyl-L-prolyl-glycyl-L-phenylalanyl-(O-tert-
butyl)-L-seryl-L-prolyl-L-phenylalanyl-Nω-Pbf-L-arginine (16). The
partially protected bradykinin (16) was synthesized by a solid-phase
Fmoc strategy.10,11 The side chain functional groups on bradykinin were
introduced as the protected amino acids, i.e., arginine was protected as
its 2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl (Pbf) derivative
and serine as its tertiary butyl ether. The coupling of the protected amino
acids was accomplished using standard Fmoc protocol on 2-chlorotrityl
resin (nominal capacity: 1.2 mmol/g). The loading of the resin was
0.6 mmol/g initial capacity as determined by the Fmoc assay. The
couplings were carried out by using two equiv of the protected amino
acid, 1.9 equiv of O-benzotriazol-1-yl-N,N,N′,N′-tetramethyluronium
hexafluorophosphate (HBTU) in 1-methyl-2-pyrrolidinone (NMP) or
N-Boc-O-(4-hydroxyphenacyl)-Ala-Ala (15). To a solution of
2-bromo-4′-hydroxyacetophenone (8) (328 mg, 1.5 mmol) dissolved
in 3 mL of 1,4-dioxane and cooled to 15 °C was added 397 mg (1.5
mmol) of Boc-Ala-Ala (14) dissolved in 17 mL of 1,4-dioxane. This
was followed by dropwise addition (10 min) of 1,8-diazabicyclo[5.4.0]-
undec-7-ene (DBU, 269 mg, 1.8 mmol). The reaction mixture was
allowed to warm to room temperature, and the mixture was stirred
overnight. Thin-layer chromatography indicated that the reaction was
complete. The solvent was removed in vacuo, and the crude product
was purified by silica gel column chromatography (EtOAc:hexane )
1:1). After collection of appropriate fractions for the product, the solvent
(25) Durden, D. A.; Juorio, A. V.; Davis, B. A. Anal. Chem. 1980, 52,
1815-1820.

New Phototriggers 9
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2695
dimethylformamide (DMF) in the presence of N,N-diisopropylethyl-
amine (DIEA). When the Kaiser test was positive (usually after 30
min), the resin was washed and coupled again until no free amino
groups were detected. The Fmoc group was then cleaved with 30%
piperidine in DMF. This process was repeated for each amino acid.
The N-terminal arginine was incorporated as the Fmoc-protected
derivative, and the Fmoc group was cleaved as described above. The
N-terminal amino group was acylated with Boc-pyrocarbonate in the
presence of DIEA. Finally, the fully protected oligopeptide was cleaved
from the resin by treatment with a mixture of CH2Cl2:MeOH:CH3CO2H
) 8:1:1 which gave complete conversion to 16 within 2 h.
The protected bradykinin carboxylic acid (16) was purified by
precipitation from CHCl3 with MeOH, and the purity was established
by RP-HPLC and TLC (RF in CHCl3:CH3OH:CH3CO2H ) 8:1:1) )
0.6; (RF in ethyl acetate:pyridine:CH3CO2H:H2O ) 60:20:6:11) ) 0.6
on air-dried Mack Kieselgel 60 F254 glass plates. The purity of the
protected bradykinin carboxylic acid was greater than 90% based on
RP-HPLC analysis. The amino acid composition was confirmed by a
FAB-MS spectrum.
was introduced 2.5 mg (0.006 mmol) of Ala-Ala p-hydroxyphenacyl
ester and trifluoroacetate salt‚H2O (7) containing 3 mg (0.02 mmol) of
p-hydroxyphenylacetic acid (6) as an internal standard. Aliquots (2 µL)
were taken after 10, 30, 60, and 80 min and analyzed by HPLC. The
buffer system for analytical HPLC was 90% ammonium acetate buffer
(50 mM, pH 4.5) and 10% acetonitrile at a flow rate of 1 mL/min
through a Bondesil C-18 3 µm 4.6 mm × 5 cm column. The products
and disappearance of the ester were monitored at 280 nm. The data
were fit to a linear least-squares regression line using a Microcal Origin
software program to obtain the half-life and the rate constant for the
first-order hydrolysis in tris buffer. The reactions were too slow to
monitor in Ringer’s solution I, D2O or in H2O to obtain rate constants
under these conditions. The results are given in Table 4.
General Procedure for Photolysis. All water was distilled and
passed through a Nanopure deionizing system. Ammonium acetate (for
Ala-Ala) and H2O:TFA:CH3CN buffers (bradykinin) were employed
for analytical HPLC. The HPLC system employed a UV-vis detector
set at 254 or 280 nm, and the separations were carried out on a reverse
phase column. All analytical HPLC analyses employed either a solvent
gradient or an isocratic elution with a flow rate of 1.0 mL/min.
Photolysis was performed on a merry-go-round apparatus equipped with
4 × 300 nm lamps. The light output for the determination of quantum
efficiencies was measured by using the potassium ferrioxalate method.26
All determinations of quantitative measurements were performed at
least in triplicate. The data were submitted to linear least-squares
regression analysis using a Microcal Origin software program. Standard
deviations are given in parentheses, and r values were 0.97512 to
0.99999.
N(r-Boc)-N(ω-Pbf)-L-arginyl-L-prolyl-L-prolyl-glycyl-L-phenyl-
alanyl-(O-tert-butyl)-L-seryl-L-prolyl-L-phenylalanyl-N(ω-Pbf)-L-arginine
4-hydroxyphenacyl ester. (10). To a solution of 1,8-diazabicyclo[4.3.0]-
undec-7-ene (DBU, 23.1 mg, 0.15 mmol) dissolved in 2 mL of dry
DMF and cooled to 0 °C with ice was added 100 mg (0.058 mmol) of
N(R-Boc)-N(ω-Pbf)-L-arginyl-L-prolyl-L-prolyl-glycyl-L-phenylalanyl-
(O-tert-butyl)-L-seryl-L-prolyl-L-phenylalanyl-N(ω-Pbf)-L-arginine (16)
under Ar. 2-Bromo-4′-hydroxyacetophenone (8, 37.5 mg; 0.17 mmol)
was then added, and the mixture was cooled to 0 °C. The reaction
mixture was allowed to reach room temperature and was stirred
overnight. TLC indicated complete reaction. The solvent was removed
in vacuo, and the crude product was first purified by silica gel column
chromatography (CHCl3/MeOH/AcOH ) 90:8:2) to give a light oil
which was further purified by HPLC on a 2.2 cm Vydac -C18 peptide/
protein column using a gradient of 0.1% TFA in acetonitrile (50 to
100% over 90 min) at a flow rate of 8 mL/min (product retention time
) 35.6 min). The solvent was concentrated to give a white crystalline
solid of N(R-Boc)-N(ω-Pbf)-L-arginyl-L-prolyl-L-prolyl-glycyl-L-phen-
ylalanyl-(O-tert-butyl)-L-seryl-L-prolyl-L-phenylalanyl-N(ω-Pbf)-L-arginine
p-hydroxyphenacyl ester (10, 78 mg, 72%): 1H NMR (400 MHz,
CDCl3) (only signals of the 4-hydroxyphenacyl- and the two phenyl
groups are assigned) δ 7.84 (m, 2H), 7.27 (m, 10H), 6.93 (d, 2H, J )
8.6 Hz), 5.33 (m, 2H); FAB-MS m/z (relative intensity) 1856.1 (M +
1, 100). Exact mass Calcd for C93H128N15O21S2 (M + H): 1854.8850.
Found: 1854.8810.
L-Arginyl-L-prolyl-L-prolyl-glycyl-L-phenylalanyl-L-seryl-L-pro-
lyl-L-phenylalanyl-L-arginine 4-Hydroxyphenacyl Ester (Bradykinin
4-Hydroxyphenacyl Ester) (9). To a 25 mL conical glass flask were
added 54 mg (0.0291 mmol) of the dry, protected bradykinin 4-hy-
droxyphenylacetic ester 16, followed by 2 mL of the deprotection
cocktail of 88% TFA, 7% thioanisole, 5% H2O. The solution was stirred
with a glass rod every 15 min for 2.5 h to effect the deprotection. The
resulting mixture was concentrated at 30 °C on a rotary evaporator to
yield a light yellow syrup which was dissolved in 400 µL of TFA and
the resulting solution was taken up into a Pasteur pipet and then squirted
forcibly into 12 mL of methyl tert-butyl ether (-75 °C) in a 12 mL
centrifuge tube. A fine white precipitate formed immediately. After 5
min of centrifugation, the ether was decanted, and the crude product
was dispersed thoroughly with a glass rod in 10 mL of fresh ether and
centrifuged once again. This process was repeated six times. The crude
product was dried overnight to give 35.5 mg as a white pellet. The
product was further purified by HPLC on a 2.2 cm Vydac C-18 peptide/
protein column using a gradient of 0.1% TFA in acetonitrile (5 to 50%
over 45 min) at a flow rate of 8 mL/min (product retention time )
33.6 min). The solvent was concentrated on a rotary evaporator to
remove the CH3CN and the remaining material was lyophilized to give
white crystals of bradykinin p-hydroxyphenacyl ester (9, 29.1 mg,
84%): FAB-MS m/z (relative intensity) 1194.6 (M + H, 43). Exact
mass Calcd for C58H80N15O13 (M + H): 1194.6060. Found: 1194.6016.
UV-vis (H2O) λmax (ꢀ) 191 nm (117 490), 282 (11 853).
Photolysis of O-(4-Hydroxyphenacyl)-Ala-Ala, Trifluoroacetate
Salt (7). (A) Quantum Efficiencies by NMR. Into three NMR tubes
containing 2 mL of D2O were introduced 17.6 mg of O-(4-hydroxy-
phenacyl)-Ala-Ala and trifluoroacetate salt (7) (0.0394 mmol) contain-
ing 0.00296 mmol Ala-Ala as an impurity as determined by 1H NMR.
Dimethylformamide (DMF) (8 µl, 0.1034 mmol) was added as an
internal standard. The solutions were deaerated for 10 min with argon
and irradiated at 300 nm for 8 min with 4 RPR-3000 lamps. 1H NMR
spectra (400 MHz) were measured at 2 min intervals during irradiation.
The rearrangement product, 4-hydroxyphenylacetic acid (6), the free
dipeptide Ala-Ala (1), and the disappearance of O-(4-hydroxyphenacyl)-
Ala-Ala (7) were identified by the characteristic 400 MHz-1H NMR-
signals. Quantitative measurement of the products and starting ester
were performed with an internal standard by RP-HPLC (see below)
and by integration of characteristic 1H NMR signals for each compound.
A very small amount of a photohydrolysis product, 2,4′-dihydroxyac-
etophenone (13), was also formed. Linear least squares regression
analyses (vide supra) of the integrated 1H NMR absorptions at 4.2 (1),
7.2 ppm (6), and 5.4 (7), respectively, were carried out for the
appearance of Ala-Ala (1), p-hydroxyphenylacetic acid (6) and the
disappearance of starting ester 7 (relative to the internal standards DMF
and acetonitrile) as a function of time to determine the quantum
efficiencies. The results are shown in Table 1.
(B) Test for Chirality Stability. Into an NMR tube containing 1
mL of D2O was introduced 10 mg (0.0235 mmol) of O-(4-hydroxy-
phenacyl)-Ala-Ala, trifluorocetate salt (7), which was synthesized as
described above from optically active Ala-Ala ([R]D ) -35.9° (c ) 2,
6 N HCl) Ala-Ala (Aldrich)). The solution was deaerated for 10 min
with argon and irradiated at 300 nm for 18 min with 16 RPR-300 lamps.
Spectra (300 MHz 1H NMR) were measured after 0, 2, 4, 6, 8, 12, and
18 min of irradiation. The conversion of the phototrigger 7 to the
rearrangement product, 4-hydroxyphenylacetic acid (6) and the free
dipeptide Ala-Ala, trifluoroacetate salt (1) was complete (see Figure
1). The rotation of the released Ala-Ala was measured directly from
the photolysis solution. The measured rotation of Ala-Ala as its
trifluoroacetate salt was [R]D ) -14.75° (c ) 0.13, H2O). At 100%
conversion as shown in Figure 1, the 0.0235 mmol of Ala-Ala produced
from the photolysis of 0.0235 mmol of 7 gave a specific rotation of
[R]D ) -14.77° (c ) 0.13, D2O) for Ala-Ala as the trifluoroacetate
salt, equivalent to the value obtained from the precursor Ala-Ala.
Hydrolytic Stability of Ala-Ala p-Hydroxyphenacyl Ester (7). Into
a test tube, containing 2 mL of Tris buffer solution (50 mM, pH 7.3)
(26) Hatchard, C. G.; Parker, C. A. Proc. R. Soc. London 1956, A-235,
518-522.

2696 J. Am. Chem. Soc., Vol. 122, No. 12, 2000
GiVens et al.
(C) Quantum Efficiencies by HPLC. Into a Pyrex tube containing
5 mL of D2O were placed 17.6 mg of O-(4-hydroxyphenacyl) Ala-
Ala, trifluoroacetate salt (7) (0.0394 mmol) which contained a trace
(0.00296 mmol) of Ala-Ala (1) as an impurity and 5.7 mg (0.04256
mmol) anthranilic acid as the internal standard. The solution was
deaerated for 10 min with Ar and irradiated at 300 nm. Three 5 µL
aliquots at a time were taken at 2 min intervals (for a total of 8 min)
and injected on HPLC on a Bondesil C-18 3 µm × 4.6 mm × 5 cm
column. The buffer system for analytical HPLC was 90% ammonium
acetate buffer (50 mM, pH 4.5) and 10% acetonitrile and the effluent
was monitored at 280 nm. All analytical HPLC analyses employed an
isocratic flow rate of 1 mL/min for 10 min. The rearrangement product
4-hydroxyphenylacetic acid (6) was identified by injection on HPLC
with authentic sample. In addition a small amount (<10%) of the
hydrolysis product 2,4′-dihydroxyacetophenone (13) was also detected.
The quantum efficiencies for the disappearance of O-(4-hydroxyphena-
cyl)-Ala-Ala, trifluoroacetate salt (7) and the appearance of the
4-hydroxyphenylacetic acid (6) were obtained by linear squares
regression analyses. The data were submitted to linear least-squares
regression analysis using a Microcal Origin software program and the
results are shown in Table 1.
(D) Quenching Studies of Ala-Ala p-Hydroxyphenacyl Ester,
Trifluoroacetate Salt‚H2O (7) with Sodium 2-Naphthalene Sul-
fonate. Into four NMR tubes containing 2 mL of D2O were introduced
17.6 mg of Ala-Ala p-hydroxyphenacyl ester and trifluoroacetate salt
(7) (0.0394 mmol) containing 0.00296 mmol of Ala-Ala (1) as an
impurity. Dimethylformamide (DMF) (8 µL, 0.1034 mmol) was added
as an internal standard. To three of these tubes was added sodium
2-naphthalene sulfonate, 9.1 mg (39 µmol), 16.8 mg (73 µmol), or 20.9
mg (91 µmol), respectively. The tubes were deaerated for 10 min with
Ar at room temperature and irradiated for a total of 8 min with 4 RPR-
3000 Å lamps. 1H NMR spectra (400 MHz) were determined at 2 min
intervals during irradiation. The appearance of the rearrangement
product, p-hydroxyphenylacetic acid (6), the free dipeptide Ala-Ala
(1) and the disappearance of p-hydroxyphenacyl Ala-Ala (7) were
identified and quantitatively determined by integration of characteristic
peaks in their 1H NMR spectra. The data were submitted to linear least-
squares regression analysis using a Microcal Origin software program,
and the results are given in Table 2.
The competitive absorption of sodium 2-naphthalenesulfonate in this
quenching study was determined to be relatively minor in the region
between 300 to ∼320 nm where it overlaps with the absorption of 7.
The emission range of the 300 nm lamps extends to about 335-340
nm. Pyrex test tubes were used, thus filtering out the incident radiation
below 295 nm, making the dynamic range essentially ∼295-335 nm.
In this region, the p-hydroxyphenacyl ester has a ꢀ that ranges from
7700 at 300 nm to 4000 at 310 nm and then tails to 0 at ∼340 nm. The
2-NS absorption spectrum consists of three small absorption λmax’s
between 300 and 320 nm, none with ꢀ greater than 550. Thus, from
the spectral data, the calculated competitive absorption by 2-NS at the
highest concentration used was less than 16% that of 7.
(E) Quenching Studies of Ala-Ala p-Hydroxyphenacyl Ester with
Potassium Sorbate. Into each of four NMR tubes containing 2 mL of
D2O were introduced Ala-Ala p-hydroxyphenacyl ester (15 mg, 0.05
mmol), DMF (5 µL, 0.065 µmol), and increasing amounts of potassium
sorbate (0 mg, 0 µmol; 7.6 mg, 50 µmol; 15.3 mg, 100 µmol, 38.0 mg,
250 µmol, respectively). The NMR tubes were placed inside a Pyrex
test tube in the merry-go-round apparatus and were irradiated with four
RPR-3000 Å lamps. The samples were analyzed at 2 min intervals
during the irradiation by 400 MHz 1H NMR using DMF as the internal
standard. The disappearance of p-hydroxyphenacyl Ala-Ala (7) and the
appearance of p-hydroxyphenylacetic acid (6) and Ala-Ala (1) were
determined (see A, vide supra). Stern-Volmer slopes were determined
as indicated above and the results are given in Table 2.
C-18 peptide/protein column using a gradient of 0.1% TFA (5 to 50%
over 45 min) in acetonitrile at a flow rate of 2 mL/min. The
rearrangement product, 4-hydroxyphenylacetic acid (6) was identified
by co-injection on HPLC with authentic sample. In addition a small
amount of the photohydrolysis product 2,4′-dihydroxyacetophenone (13)
was also observed. The photoreleased bradykinin (2) was identified
by co-injection with an authentic sample, by FAB-MS and by exact
mass of the collected fraction (retention time ) 24.8 min): FAB-MS
m/z (relative intensity) 1060.5 (M, 100). Exact mass calcd for
C50H74N15O11 (M + H): 1060.5692. Found: 1060.5720. CD spectro-
scopy (H2O) (Figure 2): Θmax ) 918 deg cm2 dmol-1 (222 nm), Θmin
) -6965 deg cm2 dmol-1 (202 nm), for bradykinin (Fluka): Θmax
)
954 deg cm2 dmol-1 (222 nm), Θmin ) -6705 deg cm2 dmol-1 (202
nm).
At 100% conversion, the product yields of 6 and 13 were 0.00263
and 0.00035 mmol, respectively, in good agreement with the total
conversion of 9 to bradykinin and the two byproducts. This also
corroborates the high degree of purity of pHP bradykinin (9). The
quantum efficiencies for the disappearance of bradykinin 4-hydroxy-
phenacyl ester (9) and for the appearance of bradykinin (2), 4-hydroxy-
phenylacetic acid (6), and 2,4′-dihydroxyacetophenone (13) were
obtained by linear least-squares regression analyses using Microcal
Origin software as discussed above. These results are shown in Table
4.
Intracellular Ca+2 Determinations in Rat Dorsal Root Gan-
glion: General Procedures for Flash Photolysis of p-Hydroxy-
phenacyl Bradykinin (9). Rat dorsal root ganglion (DRG) neurons
were grown in culture as previously described.19 The intracellular Ca2+
concentration ([Ca2+]i) was measured in single cells with the Ca2+
-
sensitive dye indo-1 using previously published procedures and
instrumentation.19 Briefly, cells were loaded with dye by incubation
with indo-1 acetoxymethyl ester, which is membrane permanent. De-
esterification by cytosolic esterases traps the Ca2+-sensitive, free acid
form of indo-1 within the cell. The cells were then placed in a recording
chamber on the stage of an epifluorescence microscope and excited
with UV light (350 ( 5 nm). The emitted fluorescence was detected
at 405 ( 20 and 490 ( 20 nm. The ratio (R) of the 405:490 signals
was converted to [Ca2+]i by the equation [Ca2+]i ) Kâ(Rmax - R)/(R
- Rmin). The constants K, â, Rmax, and Rmin were derived as previously
described.19
Flash photolysis was accomplished with light from a pulsed nitrogen
laser (Photon Technologies Inc., GL-2300). Ultraviolet light (337 nm)
was focused (plano convex f ) 10 cm) onto the end of a single optical
fiber (polyimide, core diameter ) 50 mm; Ceram Optec Inc.), the
terminating end of which was placed approximately 50 µm from the
cell soma with a micromanipulator. A single pulse of approximately 1
ns in duration (approximately 2.3 mJ) was sufficient for photolysis.
Phosphorescence Emission Studies. The phosphorescence measure-
ments were performed in either ethylene glycol:H2O (EG:W; 2:1) or
ether:isopentane:ethanol (EPA; 5:5:2) glasses in a rotating cam double
monochromator phosphoroscope illuminated with a 200 W xenon-
mercury lamp and detected with an IP28 photomultiplier. The solvents
were HPLC or spectral grade. The samples were placed in a 0.1 mm
quartz tube and cooled to 77 K.
In a typical study, the phosphorescence spectrum of p-hydroxy-
phenacyl γ-aminobutyrate (18-36 mg, 0.05-0.10 mmol) was dissolved
in 1.0 mL of EPA or EG:W and cooled to 77 K to form a clear glass.
The phosphorescence emission and excitation spectra were measured
for each sample. A blank of each solvent was checked for impurities.
In the quenching experiments, a 0.05-0.10 M solution of 7 and 17
mg (0.04 M) of Na+ 2-naphthalenesulfonate16 or piperylene27 in EW
was cooled to 77 K and the phosphorescence emission and excitation
measured. Table 3 shows the results for 7, 12, 13, p-hydroxyacetophen-
one (17),15 R,R-dimethyl-p-hydroxyphenacyl acetate (18), p-hydrox-
yphenacyl phenylacetate (19), diethyl p-hydroxyphenacyl phosphate
(20), and mixtures of Na+ naphthalene-2-sulfonate with 7 and 12 and
mixtures of 7 with piperylene.
Photolysis of Bradykinin 4-Hydroxyphenacyl Ester (9): Quan-
tum Efficiency Determinations. Into a Pyrex tube containing 5 mL
of D2O was dissolved 3.5 mg (0.00293 mmol) bradykinin 4-hydroxy-
phenacyl ester (9) and hemimellitic acid (0.85 mg, 0.00405 mmol) as
the internal standard. The solution was deaerated for 10 min with Ar
and irradiated at 300 nm for 8 min. Aliquots (100 µL) were taken at
0.5, 1.0, 2.0, and 2.5 min and analyzed by HPLC on a 1.0 cm Vydac
(27) The triplet energy of (E)-piperylene is reported to be ET ) 60 kcal/
mol from the S f T 0, 0 band: Kellogg, R. E.; Simpson, W. T. J. Am.
Chem. Soc. 1965, 87, 4230-4234.

New Phototriggers 9
J. Am. Chem. Soc., Vol. 122, No. 12, 2000 2697
Acknowledgment. We thank the NSF (NSF/O5R-9255223;
R.S.G. and IBN9723796; S.A.T.), NIH (DA07340, DA092393,
DA11806; S.A.T.) and the Minnesota Medical Foundation
(S.A.T.) for partial support of this research. G.O. thanks the
Council for International Exchange of Scholars for a Fulbright
Fellowship and J.F.W.W. thanks the University of Kansas
College of Liberal Arts and Sciences for partial support. We
thank Professor Michael Doughty, University of Kansas, for
suggesting the bradykinin problem and D. Williamson for
helpful advice.
JA991014B