Design, syntheses, and evaluations of bicyclomycin-based rho inactivators

Article abstract of DOI:10.1021/jo970575n

The commercial antibiotic bicyclomycin (1) has been shown to target the essential transcriptional termination factor rho in Escherichia coli. Little is known, however, about the bicyclomycin binding site in rho. A recent structure-activity relationship study permitted us to design modified bicyclomycins that may irreversibly inactivate rho. The four compounds selected were C(5a)-(4-azidoanilino)dihydrobicyclomycin (3), C(5a)-(3- formylanilino)dihydrobicyclomycin (4), C(5)-norbicyclomycin C(5)-O-(4- azidobenzoate) (5), and C(5)-norbicyclomycin C(5)-O-(3-formylbenzoate) (6). In each of these compounds the inactivating unit was placed at the C(5)- C(5a) site in bicyclomycin. In compounds 3 and 5 an aryl azide moiety was used as a photoaffinity label whereas in 4 and 6 an aryl aldehyde group was employed as a reductive amination probe. The synthesis and spectral properties of 3-6 are described. Chemical studies demonstrated that 3 and 4 were stable in D₂O and CD₃OD (room temperature, 7 d), while 5 and 6 underwent significant change within 1 d. Biochemical investigations showed that 3 and 4 retained appreciable inhibitory activities in rho-dependent ATPase and transcription termination assays. In the ATPase assay, I₅₀ values for 1, 3, and 4 were 60, 135, and 70 μM, respectively. Correspondingly, the I₅₀ values for 5 and 6 were >400 and 225 μM, respectively. In the transcription termination assay, compounds 1, 3, and 4 all prevented (≤97%) the production of rho-dependent transcripts at 40 μM, whereas little (≤15%) inhibition of transcription termination was observed for 5 and 6 at this concentration. Antimicrobial evaluation of 3-6 showed that none of the four compounds exhibited antibiotic activity at 32 mg/mL or less against W3350 E. coli. The combined chemical and biochemical studies led to our further evaluation of 3 and 4. Photochemical irradiation (254 nm) of 3 in the presence of rho led to a 29-32% loss of rho ATPase activity. Attempts to confirm the irreversible adduction of 3 to rho by electrospray mass spectrometry were unsuccessful. No higher molecular weight adducts were detected. Incubation of rho with 4 at room temperature (4 h) followed by the addition of NaBH₄ led to significant losses (>62%) of rho ATPase activity. Analyses of the 4-rho modified adduct showed appreciable levels of adduction (~40%). Mass spectrometric analyses indicated a molecular weight for the adduct of approximately 47 410, consistent with a modification of a rho lysine residue, by 4. Compound 4 was selected for additional studies.

Full text of DOI:10.1021/jo970575n

5432
J . Org. Chem. 1997, 62, 5432-5440
Design , Syn th eses, a n d Eva lu a tion s of Bicyclom ycin -Ba sed Rh o
In a ctiva tor s
Hangjin Cho,† Hyeung-geun Park,† Xiangdong Zhang,† Isabel Riba,§ Simon J . Gaskell,§
William R. Widger,*,‡ and Harold Kohn*,†
Department of Chemistry, University of Houston, Houston, Texas 77204-5641, Department of Biochemical
and Biophysical Sciences, University of Houston, Houston, Texas 77204-5934, and Michael Barber Centre
for Mass Spectrometry, UMIST, P.O. Box 88, Manchester M60 1QD, U.K.
Received March 31, 1997X
The commercial antibiotic bicyclomycin (1) has been shown to target the essential transcriptional
termination factor rho in Escherichia coli. Little is known, however, about the bicyclomycin binding
site in rho. A recent structure-activity relationship study permitted us to design modified
bicyclomycins that may irreversibly inactivate rho. The four compounds selected were C(5a)-(4-
azidoanilino)dihydrobicyclomycin (3), C(5a)-(3-formylanilino)dihydrobicyclomycin (4), C(5)-norbi-
cyclomycin C(5)-O-(4-azidobenzoate) (5), and C(5)-norbicyclomycin C(5)-O-(3-formylbenzoate) (6).
In each of these compounds the inactivating unit was placed at the C(5)-C(5a) site in bicyclomycin.
In compounds 3 and 5 an aryl azide moiety was used as a photoaffinity label whereas in 4 and 6
an aryl aldehyde group was employed as a reductive amination probe. The synthesis and spectral
properties of 3-6 are described. Chemical studies demonstrated that 3 and 4 were stable in D2O
and CD3OD (room temperature, 7 d), while 5 and 6 underwent significant change within 1 d.
Biochemical investigations showed that 3 and 4 retained appreciable inhibitory activities in rho-
dependent ATPase and transcription termination assays. In the ATPase assay, I50 values for 1, 3,
and 4 were 60, 135, and 70 µM, respectively. Correspondingly, the I50 values for 5 and 6 were
>400 and 225 µM, respectively. In the transcription termination assay, compounds 1, 3, and 4 all
prevented (g97%) the production of rho-dependent transcripts at 40 µM, whereas little (e15%)
inhibition of transcription termination was observed for 5 and 6 at this concentration. Antimicrobial
evaluation of 3-6 showed that none of the four compounds exhibited antibiotic activity at 32 mg/
mL or less against W3350 E. coli. The combined chemical and biochemical studies led to our further
evaluation of 3 and 4. Photochemical irradiation (254 nm) of 3 in the presence of rho led to a
29-32% loss of rho ATPase activity. Attempts to confirm the irreversible adduction of 3 to rho by
electrospray mass spectrometry were unsuccessful. No higher molecular weight adducts were
detected. Incubation of rho with 4 at room temperature (4 h) followed by the addition of NaBH4
led to significant losses (>62%) of rho ATPase activity. Analyses of the 4-rho modified adduct
showed appreciable levels of adduction (∼40%). Mass spectrometric analyses indicated a molecular
weight for the adduct of approximately 47 410, consistent with a modification of a rho lysine residue
by 4. Compound 4 was selected for additional studies.
Bicyclomycin (1) is a structurally unique antibiotic1-4
prolonged incubation with excess drug, multiple covalent
modification of the protein occurs.8
whose primary target in Escherichia coli is the transcrip-
tion termination factor rho.5 We have demonstrated that
1 reversibly binds to rho and as a result inhibits both
ATPase and transcription termination activities.6,7 Upon
† Department of Chemistry.
‡ Department of Biochemical and Biophysical Sciences.
§ UMIST.
X Abstract published in Advance ACS Abstracts, J uly 15, 1997.
(1) Tanaka, N. Antibiotics (N.Y.) 1979, 5, 18-25.
(2) (a) Miyoshi, T.; Miyari, N.; Aoki, H.; Kohsaka, M.; Sakai, H.;
Imanaka, H. J . Antibiot. 1972, 25, 569-575. (b) Kamiya, T.; Maeno,
S.; Hashimoto, M.; Mine, Y. J . Antibiot. 1972, 25, 576-581. (c)
Nishida, M.; Mine, Y.; Matsubara, T. J . Antibiot. 1972, 25, 582-593.
(d) Nishida, M.; Mine, Y.; Matsubara, T.; Goto, S.; Kuwahara, S. J .
Antibiot. 1972, 25, 594-601.
(3) (a) Miyamura, S.; Ogasawara, N.; Otsuka, H.; Niwayama, S.;
Takana, H.; Take, T.; Uchiyama, T.; Ochiai, H.; Abe, K.; Koizumi, K.;
Asao, K.; Matuski, K.; Hoshino, T. J . Antibiot. 1972, 25, 610-612. (b)
Miyamura, S.; Ogasawara, N.; Otsuka, H.; Niwayama, S.; Tanaka, H.;
Take, T.; Uchiyama, T.; Ochiai, H. J . Antibiot. 1973, 26, 479-484.
(4) Williams, R. M.; Durham, C. A. Chem. Rev. 1988, 88, 511-540
and references cited therein.
(5) Zweifka, A.; Kohn, H.; Widger, W. R. Biochemistry 1993, 32,
3564-3570.
(6) Park, H.-g.; Zhang, X.; Moon, H.-s.; Zwiefka, A.; Cox, K.; Gaskell,
S. J .; Widger, W. R.; Kohn, H. Arch. Biochem. Biophys. 1995, 323, 447-
454.
(7) Magyar, A.; Zhang, X.; Kohn, H.; Widger, W. R. J . Biol. Chem.
1996, 271, 25369-25374.
The exact binding site of bicyclomycin in rho is
unknown, although several studies have provided struc-
tural information concerning rho. Rho protein is com-
posed of six identical 47-kDa proteins of 419 amino acids9
each in a proposed planar, hexagonal D3 symmetry.10-12
(8) Riba, I.; Gaskell, S. J .; Park, H.-g.; Zhang, X.; Widger, W. R.;
Kohn, H. Proceedings of the 44th ASMS Conference on Mass Spec-
trometry and Allied Topics, Portland, OR, May 1996; p 1416.
(9) Finger, L. R.; Richardson, J . P. J . Mol. Biol. 1982, 156, 203-
219.
(10) Bear, D. G.; Hicks, P. S.; Escudero, K. W.; Andrews, C. L.;
McSwiggen, J . A.; von Hippel, P. H. J . Mol. Biol. 1988, 199, 623-635.
(11) Geiselmann, J .; Yager, T. D.; Gill, S. C.; Calmettes, P.; von
Hippel, P. H. Biochemistry 1992, 31, 111-121.
(12) Geiselmann, J .; Seifried, S. E.; Yager, T. D.; Liang, C.; von
Hippel, P. H. Biochemistry 1992, 31, 121-132.
S0022-3263(97)00575-6 CCC: $14.00 © 1997 American Chemical Society

Bicyclomycin-Based F Inactivators
J . Org. Chem., Vol. 62, No. 16, 1997 5433
Investigations show that rho possesses primary RNA,
secondary RNA, and ATP binding domains.13-18 The
N-terminal 151 amino acids have been assigned to the
primary RNA binding site19 whereas the ATP binding
domain is believed to extend from residues 160 to 340
and contain sequence similarity to the E. coli â subunit
of F1 ATPase and adenylate kinase.15 The secondary
RNA binding site has not been identified. We showed
that bicyclomycin binding affects the secondary RNA
binding site,7 and mutations conferring bicyclomycin
resistance (M218K, S266A, G337S) were located at or
near the ATP domain.5 These studies suggest that part
of the bicyclomycin, ATP, and secondary RNA binding
domains are close to each other. We report herein the
design and evaluation of a modified bicyclomycin deriva-
tive that efficiently modifies a lysine residue(s) in rho
and inactivates the protein.
over, the high extinction coefficient (ꢀ) and photochemical
quantum yield (φ) at 250-280 nm of aryl azides allow
photolysis to occur rapidly with minimal ultraviolet
damage to the protein.
Reductive amination probes require a lysine group to
be present at or near the receptor binding site close to
the reactive aldehyde group to permit imine formation
and subsequent conversion to the amine by reduction.
We chose the aryl aldehyde moiety as the reductive
amination probe because this group provides an attrac-
tive balance between stability and reactivity. In a
relevant study, Platt and co-workers utilized the adenine
nucleotide analogue 2 to identify the ATP binding domain
in rho.14
Resu lts a n d Discu ssion
A. Ch oice of P r obes. In a recent structure-activity
relationship (SAR) study, we demonstrated that modifi-
cation of either the C(1) triol group20 or the piperazinedi-
one group within the [4.2.2] bicyclic ring system21 led to
a significant loss in inhibitory activity of the bicyclomycin
analogue in rho-dependent ATPase22 and in transcription
termination assays.23 These findings showed that both
structural sites played important roles in the antibiotic’s
binding to rho. Correspondingly, selective modification
of the C(5)-C(5a) site in 1 led to bicyclomycin analogues
exhibiting excellent inhibitory activities in these assays.24
Accordingly, we installed a rho irreversible inactivating
unit at the C(5)-C(5a) locus in bicyclomycin.
Numerous examples have been reported of functional
groups which affect enzyme inactivation.25 We focused
in this study on photoaffinity and reductive amination
probes since they have been used extensively and suc-
cessfully to target receptor sites within proteins. The
phenylazido group was selected as the photoaffinity
unit.26 Irradiation of this moiety generates nitrenes,
which can undergo insertion reactions in amino acid
residues surrounding the receptor binding site. More-
Compounds 3-6 were selected for synthesis and evalu-
ation. Our SAR studies demonstrated that the C(5a) aryl
dihydrobicyclomycin derivatives bound tightly to rho.24
In particular, we observed C(5a)-(anilino)dihydrobicyclo-
mycin (7) possessed an I50 value of 120 µM in the ATPase
assay, comparing favorably to bicyclomycin (I50 ) 60 µM).
This finding, coupled with the commercial availability of
aniline derivatives, prompted us to choose 3 and 4 as
potential photoaffinity and reductive amination probes,
respectively. Our high-yield synthesis of norbicyclomy-
cin-5-ol C(2′),C(3′)-acetonide (8)27 led us also to synthesize
the photoaffinity probe 5 and reductive amination agent
6.
(13) Dombroski, A. J .; Brennan, C. A.; Spear, P.; Platt, T. J . Biol.
Chem. 1988, 263, 18802-18809.
(14) Dombroski, A. J .; La Dine, J . R.; Cross, R. L.; Platt, T. J . Biol.
Chem. 1988, 263, 18810-18815.
(15) Dombroski, A. J .; Platt, T. Proc. Natl. Acad. Sci. U.S.A. 1988,
85, 2538-2542.
(16) Dolan, J . W.; Marshall, N. F.; Richardson, J . P. J . Biol. Chem.
1990, 265, 5747-5754.
(17) Platt, T.; Richardson, J . P. In Transcriptional Regulation;
McKnight, S. L., Yamamoto, K. R., Eds.; Cold Spring Harbor Press:
New York, 1992; pp 365-388.
(18) Platt, T. Mol. Microbiol. 1994, 11, 983-990.
(19) Modrak, D.; Richardson, J . P. Biochemistry 1994, 33, 8292-
8299.
(20) Park, H.-g.; Zhang, X.; Widger, W. R.; Kohn, H. J . Org. Chem.
1996, 61, 7750-7755.
(21) Santilla´n, A., J r.; Park, H.-g.; Zhang, X.; Lee, O.-S.; Widger,
W. R.; Kohn, H. J . Org. Chem. 1996, 61, 7756-7763.
(22) Sharp, J . A.; Galloway, J . L.; Platt, T. J . Biol. Chem. 1983, 258,
3482-3486.
(23) Wu, A. M.; Christie, G. E.; Platt, T. Proc. Natl. Acad. Sci. U.S.A.
1981, 78, 2913-2917.
(24) Park, H.-g.; Zhang, Z.; Zhang, X.; Widger, W. R.; Kohn, H. J .
Org. Chem. 1996, 61, 7764-7776.
B. Syn th esis a n d Str u ctu r a l Ch a r a cter iza tion of
Bicyclom ycin P r obes. Compounds 3 and 4 were
prepared using the method employed for 724 (Scheme 1).
Dissolution of bicyclomycin C(2′),C(3′)-acetonide28,29 (9)
(25) Silverman, R. B. The Organic Chemistry of Drug Design and
Drug Action; Academic Press: New York, 1992; Chapter 5.
(26) (a) Bayley, H.; Staros, J . V. Photoaffinity Labeling and Related
Techniques. In Azides and Nitrenes: Reactivity and Utility; Scriven,
E. F. V., Ed.; Academic Press: Orlando, 1984; pp 433-490. (b)
Fleming, S. A. Tetrahedron 1995, 51, 12479-12520. (c) Kotzyba-
Hibert, F.; Kapfer, I.; Goeldner, M. Angew. Chem., Int. Ed. Engl. 1995,
34, 1296-1312.
(27) Zhang, Z.; Park, H.-g.; Kohn, H. J . Org. Chem. 1995, 60, 5346-
5351.
(28) Kamiya, T.; Maeno, S.; Kitaura, Y. Belgium Patent 847 475.

5434 J . Org. Chem., Vol. 62, No. 16, 1997
Sch em e 1. P r ep a r a tion of Com p ou n d s 3 a n d 4
Cho et al.
Ta ble 1. Select 1H NMR a n d 13C NMR Sp ectr a l P r op er ties for Rh o P h otoa ffin ity a n d Red u ctive Am in a tion Rea gen ts
1H NMRa
13C NMRb
C(5)
compd
C(4)HH′
C(4)HH′
C(5)H
C(5a)HH′
C(5a)HH′
C(3)
C(4)
C(5a)
C(6)
3
1.80-2.10 (m)
2.28-2.40 (m)
2.30-2.45 (m)
5.06 (t, 3.9)
2.94 (dd, 9.3, 13.8)
3.02 (dd, 9.6, 13.8)
3.50-3.65 (m)
3.62-3.85 (m)
62.6
62.6
57.3
59.4
29.9
29.9
31.1
31.9
50.6
50.7
78.7
80.3
44.2
43.6
83.9
83.8
80.8
82.4
4
1.85-2.10 (m)
1.85-2.00 (m)
2.10-2.35 (m)
5c
6d
5.47 (t, 3.9)
a
The number in each entry is the chemical shift value (δ) observed in ppm relative to Me4Si, followed by the multiplicity of the signal
and the coupling constant(s) in hertz. All spectra were recorded at 300 MHz, and the solvent used was CD3OD unless otherwise indicated.
b
The number in each entry is the chemical shift value (δ) observed in ppm relative to Me4Si. All spectra were recorded at 75 MHz, and
the solvent used was CD3OD unless otherwise indicated. c The solvent used was DMSO-d6. The solvent used was CD3OD:D2O (1:1).
d
and 4-azidoaniline‚HCl in 50% aqueous methanol ad-
justed to “pH” 10.5 yielded the Michael addition adduct
10 as a mixture of diastereomers (>10:1, NMR analysis).
Deprotection of its acetonide group in 10 at room tem-
perature (2 h) with trifluoroacetic acid furnished photo-
affinity probe 3. A similar procedure was employed for
the preparation of a precursor to 4. Dissolution of 9 and
3-(hydroxymethyl)aniline in 50% aqueous methanol (“pH”
10.5) warmed to 50 °C gave the addition adduct 11.
C(5) as (R).27 Treatment of a tetrahydrofuran solution
of 8 with either 4-azidobenzoic acid or 3-formylbenzoic
acid, using dicyclohexylcarbodiimide, (dimethylamino)-
pyridine, and triethylamine, gave 13 and 14, respec-
tively. Deprotection of 13 and 14 furnished 5 and 6,
respectively.
Satisfactory spectroscopic data (IR, 1H NMR, 13C NMR,
low- and high-resolution mass) were obtained for com-
pounds 3-6. Table 1 contains key 1H NMR and 13C NMR
chemical shift values for these bicyclomycin derivatives.
Similar NMR patterns were observed for 3 and 4, as well
as for 5 and 6.
30
Oxidation of the benzyl alcohol group in 11 with MnO2
produced 12, which upon deprotection of its acetonide
group yielded reductive amination probe 4. The major
diastereomer in 3 (10) and 4 (11, 12) has been assigned
C. Sta bility Stu d ies of Bicyclom ycin P r obes. An
important requirement for the use of probes 3-6 in
bicyclomycin-rho studies is that these reagents do not
undergo fragmentation during rho binding, protein modi-
fication, and protein digestion. If this requirement is met
amino acid residue(s) surrounding the bicyclomycin bind-
ing domain may be identified. Accordingly, we deter-
mined the stability of these compounds in CD3OD and
1
as C(5)-R on the basis of the finding that the H and 13C
NMR spectral data for 3 and 4 were similar to that
reported for 7.24 The stereochemical identity of 7 was
determined from X-ray crystallographic analysis of the
corresponding acetonide derivative.24
Both 5 and 6 were prepared by coupling norbicyclo-
mycin-5-ol C(2′),C(3′)-acetonide27 (8) with the requisite
carboxylic acid (Scheme 2). Ozonolysis of a methanolic
solution of 9 followed by catalytic reduction (10% Pd/C)
gave 8 as a single diastereomer. X-ray crystallographic
analysis of 8 revealed the stereochemical orientation of
1
D2O by H NMR spectroscopy and TLC analyses (data
not shown). Compounds 3 and 4 did not undergo change
in either solvent at room temperature for 1 week.
Dissolution of 5 and 6 in either CD3OD or D2O led to the
gradual hydrolysis of these bicyclomycin derivatives to
15, as evidenced by the upfield shift of the C(5) methine
signal to δ 3.82-3.90.27 The approximate t1/2 value for 5
(29) Zhang, Z.; Kohn, H. J . Chem. Soc., Chem. Commun. 1994,
1343-1344.
(30) Gritter, R. J .; Wallace, T. J . Org. Chem. 1959, 24, 1051-1056.

Bicyclomycin-Based F Inactivators
J . Org. Chem., Vol. 62, No. 16, 1997 5435
Sch em e 2. P r ep a r a tion of Com p ou n d s 5 a n d 6
F igu r e 1. A Plot of the relative percent ATPase activity of
rho in the presence of 1: (O), 3 (b), 4 (1), 5 (3), 6 (0).
C-dependent ATPase processes than the corresponding
aryl azide adducts 3 and 5. The I50 value for 4 was 70
µM, while that for 3 was 135 µM, and the I50 value for 6
(I50 ) 225 µM) was lower than that observed for 5 (I50
>
400 µM). Evaluation of 3-6 in the rho-dependent
transcription termination assay led to similar results
(Table 2). Bicyclomycin, 3, and 4 all prevented the
production of rho-dependent transcripts at 40 µM (g97%),
while little inhibition of transcription termination was
observed for 5 and 6 at this concentration (e15%).
Accordingly, the I50 values for 3 and 4 were >10, <40,
and ∼10 µM, respectively, while the I50 values for 5 and
6 exceeded 100 µM. The corresponding I50 value for
bicyclomycin was 5 µM.
These results demonstrated that bicyclomycin C(5a)
aryl amine derivatives 3 and 4 inhibit rho more efficiently
than bicyclomycin C(5) benzoate derivatives 5 and 6 and
were consistent with the previous SAR findings.24 We
have attributed the increased inhibitory properties of the
aryl-substituted aldehyde derivatives 4 and 6 over the
aryl azide adducts 3 and 5 to the adverse binding
interactions accompanying the placement of the polar
azide residue on the aromatic ring.24
and 6 in CD3OD (22 °C) was 12 h and in D2O (22 °C)
was 24 h (1H NMR analyses).
The antimicrobial properties of 3-6 are listed in Table
2. None of the four compounds exhibited antibiotic
activity (MIC > 32 mg/mL) against W3350 E. coli in the
filter disc assay.31 These results paralleled the result
previously observed for 7.24
The enhanced binding of 3 and 4 with rho compared
to that of 5 and 6, along with the ease with which 5 and
6 underwent hydrolysis, led us to select 3 and 4 for
further study.
These studies indicated that compounds 3 and 4 were
stable in protic solvents in the absence of added reactants
and that probes 5 and 6 underwent cleavage of the
bicyclomycin unit in CD3OD and D2O. This finding
indicated that the use of 5 and 6 to identify the bicyclo-
mycin binding site could be hampered by hydrolysis of
the C(5) aryl ester bond.
D. Bioch em ica l a n d Biologica l P r op er ties of Bi-
cyclom ycin P r obes. The inhibitory properties of com-
pounds 1 and 3-6 in the poly C-dependent rho ATPase
assay22 are provided in Figure 1, and these data are
summarized along with the transcription termination
assay23 results in Table 2. The I50 values for 3 and 4 were
135 and 70 µM, respectively, and they compared favor-
ably to bicyclomycin (60 µM).6 Compound 6 only mod-
erately inhibited rho hydrolysis of ATP (I50 ) 225 µM),
and 5 was a weak inhibitor (I50 > 400 µM). Two patterns
emerged from this limited data set. First, the aniline-
based Michael addition adducts 3 and 4 inhibited rho-
dependent ATPase activity to a greater extent than their
bicyclomycin ester counterparts, 5 and 6. The I50 values
for 3 and 4 were 135 and 70 µM, respectively, and the
I50 values for 5 and 6 were >400 and 225 µM, respec-
tively. Second, the aryl aldehyde bicyclomycin deriva-
tives 4 and 6 were more potent inhibitors of poly
E. Stu d ies on th e La belin g of Rh o by P h otoa ffin -
it y Bicyclom ycin P r ob e 3. The effectiveness of bi-
cyclomycin irreversible inactivating probe 3 to covalently
modify the drug-binding pocket in rho was assessed by
the poly C-dependent ATPase assay.22 Solutions (pH 7.9)
containing 3 (1 mM) and rho with ATP were irradiated
(254 nm, 15 min) and then extensively dialyzed to remove
any unbound 3. A similar experiment was done with
4-azidoaniline (16) in place of 3. Comparison of the
results obtained from 3 with 16 permitted us to assess
the effect of random rho modification on poly C-depend-
ent ATPase activity and provided a preliminary meas-
urement of the site specificity of 3. Efforts to increase
the extent of protein modification by increasing the
concentration of 3 (5 mM) and by the incremental
(31) Ericsson, H. M.; Sherris, J . C. Acta Pathol. Microbiol. Scand.
1971, Suppl. 217, 1-90.

5436 J . Org. Chem., Vol. 62, No. 16, 1997
Cho et al.
Ta ble 2. Bioch em ica l a n d Biologica l Activities of P h otoa ffin ity a n d Red u ctive Am in a tion Rea gen ts
inhibition of ATPase activitya
I50 (µM) (BCM)b 400 µM (%) (BCM)c
60
TT activityd
compd
X
Y
I50 (µM)e
40 µM (%)f
MICg (mg/mL)
1
3
4
5
6
CH2
5
∼5
100
97
100
0
0.25
H
H
H
H
p-N3C6H4N(H)CH2
m-(C(O)H)C6H4N(H)CH2
p-N3C6H4C(O)O
135 (60)
70 (60)
78 (95)
89 (95)
22 (95)
61 (95)
>10, <40
∼10
>32 (0.25)
>32 (0.25)
>32 (0.25)
>32 (0.25)
>400 (60)
>100
>100
m-(C(O)H)C6H4C(O)O
225 (60)
15
a
b
Activity measured using the ATPase assay (ref 22) and described in an earlier paper (ref 6). The I50 value is the average 50%
inhibition concentration determined from duplicate tests. The corresponding value obtained from bicyclomycin in a concurrently run
experiment is provided in parentheses. c The percent inhibition of ATPase activity at 400 µM. The corresponding value obtained from
d
bicyclomycin in a concurrently run experiment is provided in parentheses. Activity in the transcription termination assay was determined
using the method of T. Platt and co-workers (ref 23) and described in an earlier paper (ref 7). e The I50 value is the average concentration
for termination of rho-dependent transcripts determined from duplicate tests. f The estimated amount of transcription termination at 40
g
µM. MIC value is the average minimum inhibitory concentration of the test compound determined from duplicate tests against W3350
E. coli using a filter disc assay (ref 31). The number in parentheses is the corresponding value obtained from bicyclomycin in a concurrently
run experiment.
addition of 3 (2 × 1 mM) over a 15 min time period did
not produce any appreciable changes in the level of
protein inactivation (data not shown). We included DTT
(1 mM) and mercaptoethanol (1 mM) in the reactions to
minimize random protein modification.26b Use of higher
concentrations of mercaptoethanol (20 mM) led to little
probe-rho bonding (data not shown).
led to partial protein degradation without formation of
an adduct with the intact protein.
The results of the experiments are presented in Figure
2. UV irradiation of rho solutions in the absence of an
aryl azide did not produce any detectable loss of rho
ATPase activity (Figure 2, lanes 1 and 2). A comparable
result was observed when aryl azide 16 (1 mM) was
added to the reaction solution (Figure 2, lanes 2 and 4).
Correspondingly, irradiation of a rho solution containing
bicyclomycin photoaffinity probe 3 (1 mM, 5 mM) led to
a 29-32% loss of ATPase activity (Figure 2, lanes 6 and
8). Significantly, lower levels of rho ATPase inactivation
(10%) were observed when rho was treated with 3 in the
dark (Figure 2, lanes 5 and 7). The observed losses in
ATPase activity were consistent with levels typically
reported in aryl azide-protein photolabeling experi-
ments.26b Attempts to verify that the loss of rho ATPase
activity was associated with covalent modification were
unsuccessful. Photolysis of 3 was expected to initially
give the singlet aryl nitrene 17 and then the ring-
expanded ketene imine azepine 18.26b Insertion reactions
stemming from 17 or nucleophilic addition reactions from
18 would lead to an increase in the molecular weight of
rho by 408. We observed no evidence by electrospray
mass spectrometry that the rho-3 photolysis experi-
ments led to protein adduction compared with controls
(data not shown). We concluded that either the photo-
product was unstable to the conditions of storage and
analysis26b or that photolysis of the mixture of rho and 3
Attempts to induce 3 modification of rho by including
poly C in the reaction proved unsuccessful (data not
shown). Poly C is known to bind tightly to rho10,15,32 and
induce a protein conformational change.16,33 We observed
that irradiation of rho solutions containing poly C (24
µM) without 3 led to noticeable losses of ATPase activity
(50-60%). This result was consistent with an earlier
study showing that poly C is cross-linked to rho by UV
light.16 Inclusion of 3 with poly C in the photolysis
experiments did not lead to a significant further loss of
rho ATPase activity.
F . Stu d ies on th e La belin g of Rh o by Red u ctive
Am in a tion w ith Bicyclom ycin P r obe 4. The ef-
fectiveness of reductive amination probe 4 to selectively
modify rho was determined by comparing the ATPase
activity after dialysis of NaBH4 treated solutions34 con-
taining rho and 4 with those of rho and 3-carboxybenz-
aldehyde (19).
Figure 3 shows the effect of reductive amination of rho
with either 4 or 19 in the absence of poly C and the
(32) (a) Galluppi, G. R.; Richardson, J . P. J . Mol. Biol. 1980, 138,
513-539. (b) McSwiggen, J . A.; Bear, D. G.; von Hippel, P. H. J . Mol.
Biol. 1988, 199, 609-622.
(33) Bear, D. G.; Andrews, C. L.; Singer, J . D.; Morgan, W. D.; Grant,
R. A.; von Hippel, P. H.; Platt, T. Proc. Natl. Acad. Sci. U.S.A. 1985,
82, 1911-1915.
(34) Means, G. E.; Feeney, R. E. Anal. Biochem. 1995, 224, 1-16.

Bicyclomycin-Based F Inactivators
J . Org. Chem., Vol. 62, No. 16, 1997 5437
presence of ATP. We adopted the following protocol. The
rho solution containing the aromatic aldehyde was first
incubated in the dark at room temperature for 4 h
followed by the addition of excess NaBH4. The solution
was allowed to stand at room temperature for 20 min
and was dialyzed overnight (5 °C) to remove any unbound
4. The extent of protein modification was determined by
the ATPase activity assay,22 and these results were
compared with control experiments conducted without
NaBH4. We observed no appreciable losses in poly
C-dependent ATPase activity for rho solutions treated
with NaBH4 compared with untreated NaBH4 solutions
(Figure 3, lanes 2 and 3). No detectable protein inactiva-
tion was found when 19 (1 mM) was included in the
reaction with NaBH4 (Figure 3, lane 6). Use of 4 in place
of 19 led to significant decreases (72%) in rho ATPase
activity (Figure 3, lanes 3 and 9). Moreover, significantly
decreased levels of inactivation were observed for 4 when
NaBH4 was excluded from the reaction (Figure 3, lane
8). These results indicated that 4-mediated rho inactiva-
tion proceeds by NaBH4 reduction of an intermediate
imine. Further evidence in support of imine formation
was the reduced levels of ATPase activity in the absence
of NaBH4 (Figure 3, lane 8).
F igu r e 2. Histogram depicting the percentages of rho ATPase
activity after photochemical activation in the presence and
absence of 3 and 16. The solutions were successively irradi-
ated with UV light (254 nm) for 0.25 h (25 °C) and then
dialyzed. The reactions were (1) no antibiotic and no UV; (2)
no antibiotic and UV; (3) 16 and no UV; (4) 16 and UV; (5) 3
(1 mM) and no UV; (6) 3 (1 mM) and UV; (7) 3 (5 mM) and no
UV; (8) 3 (5 mM) and UV. Figure code: open, no UV; shaded,
UV.
An NaBH4-treated sample obtained from rho and 4 (1
mM) in the absence of ATP, found to have 26% rho
ATPase activity, was examined by electrospray mass
spectrometry. We observed two proteins in an ap-
proximate 1:0.7 ratio that corresponded to approximately
47 008 and 47 410 Da (Figure 4). We have assigned these
to rho and the 4-rho reduced imine product since
reductive amination of rho by 4 leads to an increase in
the molecular weight of rho by 407.
The detection of the 47 410 Da signal does not permit
us to conclude that bonding proceeded at a specific site
in rho. Information concerning the specificity and site
selectivity of reductive amination probe 4 is being
obtained by mass spectral analysis of the tryptic digest
of 4-modified rho samples.35 Complementary information
can be obtained by radiodetection of the HPLC tryptic
digest of radiolabeled 4-rho. Our syntheses of C(5a)-
(anilino)dihydrobicyclomycin enzyme inactivators (e.g., 3
and 4) provide an attractive opportunity to introduce a
radiolabel into the compound, and we have tested this
possibility. These compounds are formed by Michael
addition of a substituted aniline to the ring-opened form
of bicyclomycin C(2′),C(3′)-acetonide (9) followed by trap-
ping of a proton from the solvent (CH3OH:H2O (1:1)) at
C(5) and deprotection of the acetonide group. Accord-
ingly, use of CH3OT and T2O in place of CH3OH and H2O,
respectively, would allow generation of the corresponding
C(5) radiolabeled probes. The feasibility of this approach
has been examined by treating 9 with 4-azidoaniline in
F igu r e 3. Histogram depicting the percentages of rho ATPase
activity after reductive amination with either 4 or 19 in the
presence of ATP (1 mM) using NaBH4. The solutions were
incubated in select cases for 4 h at room temperature, treated
with NaBH4 in select cases, and dialyzed. The reactions
were: (1) no antibiotic, no incubation, and no NaBH4; (2) no
antibiotic, with incubation, and no NaBH4; (3) no antibiotic,
with incubation and NaBH4; (4) 19 (1 mM), no incubation, and
no NaBH4; (5) 19 (1 mM), with incubation, and no NaBH4; (6)
19 (1 mM), with incubation and NaBH4; (7) 4 (5 mM), no
incubation, and no NaBH4; (8) 4 (5 mM), with incubation, and
no NaBH4; (9) 4 (5 mM), with incubation and NaBH4. Figure
code: open, no incubation, no NaBH4; shaded, with incubation
and no NaBH4; solid, with incubation and NaBH4.
1
CD3OD:D2O (1:1) to give 10-d. In the H NMR spectrum
of 10-d the C(5a) protons appeared as two doublets (δ
2.92, J ) 13.8 Hz; δ 3.61, J ) 13.8 Hz) rather than the
pair of doublets of doublets found for 10. Dissolution of
10-d in a buffered aqueous solution (pH 8.0) at room
temperature for 2 d followed by recovery of the starting
material did not produce detectable C-D f C-H exchange
at C(5) (data not shown).
(35) Riba, I.; Gaskell, S. J .; Cho, H.; Park, H.; Widger, W. R.; Kohn,
H. Unpublished results.

5438 J . Org. Chem., Vol. 62, No. 16, 1997
Cho et al.
F igu r e 4. Electrospray mass spectrometric analysis of an NaBH4-treated sample containing rho and 4 (1 mM). Upper panel:
mass/charge ratio spectrum as recorded, showing multiple charge states of each component. Lower panel: spectrum obtained by
maximum entropy processing (Ferrige, A. G.; Seddon, M. J .; J arvis, S. Rapid Commun. Mass Spectrom. 1991, 5, 374-377) of the
acquired data to achieve deconvolution and transformation to a mass scale.
structural information concerning both the sites of anti-
biotic and secondary RNA binding.17,18
Exp er im en ta l Section 36
P r ep a r a tion of C(5a )-(4-Azid oa n ilin o)d ih yd r obicyclo-
m ycin C(2′),C(3′)-Aceton id e (10). To a 50% methanolic
aqueous solution (2 mL) of 928,29 (26 mg, 0.076 mmol) was
added 4-azidoaniline‚HCl (19 mg, 0.11 mmol). The “pH” was
Con clu sion s
adjusted to 10.5 with an aqueous 0.1 N NaOH solution. The
solution was stirred at room temperature (24 h), and the “pH”
was adjusted to 7.0 with an aqueous 5% HCl solution, and then
the solvent was removed in vacuo. The residue was purified
by column chromatography (SiO2, 10% MeOH-CHCl3) to give
10 as a mixture of diastereomers (>10:1, NMR analysis): 13
mg (33%); Rf 0.60 (10% MeOH-CHCl3); mp 120 °C dec; FT-IR
These studies document that probes 3 and 4 retained
excellent inhibitory activity in rho-dependent assays,
providing preliminary evidence that these agents bind
to the bicyclomycin binding pocket in rho. Photochemical
activation of 3 in the presence of rho and ATP led to
reduced rho-ATPase activity after dialysis compared
with controls. Efforts to corroborate 3 adduction by mass
spectrometry were unsuccessful. Correspondingly, treat-
ment of a rho solution containing 4 with NaBH4 led to
an appreciable loss of rho-dependent ATPase activity
after dialysis. This finding coupled with the detection
of a higher-molecular-weight adduct by mass spectrom-
etry provided preliminary information that 4 binds to the
bicyclomycin binding pocket close to a lysine residue.
Future studies will be directed toward determining the
site and specificity of rho bonding and documenting that
4 bonds at the bicyclomycin binding domain in the
protein. Identification of the lysine residue would provide
(KBr) 3385 (br), 3313 (br), 2111, 1688, 1511, 1044 cm-1 1H
;
NMR (CD3OD) δ 1.36 (s, 3 H), 1.46 (s, 6 H), 1.80-2.10 (m, 2
H), 2.28-2.40 (m, 1 H), 2.93 (dd, J ) 9.0, 14.1 Hz, 1 H), 3.61
(dd, J ) 4.5, 14.1 Hz, 1 H), 3.71 (d, J ) 8.4 Hz, 1 H), 3.81 (dd,
J ) 8.4, 13.8 Hz, 1 H), 4.00-4.10 (m, 2 H), 4.56 (d, J ) 8.4
Hz, 1 H), 6.68 (d, J ) 9.0 Hz, 2 H), 6.81 (d, J ) 9.0 Hz, 2 H);
13C NMR (CD3OD) 24.9, 26.8, 28.3, 30.4, 44.6, 50.7, 64.1, 73.2,
73.4, 84.0, 86.4, 88.8, 111.7, 115.3, 120.8, 129.7, 147.5, 167.9,
171.9 ppm; MS (+CI) 477 [M + 1]+; Mr (+CI) 477.208 13 [M +
1]+ (calcd for C21H29N6O7 477.209 77).
P r ep a r a t ion of C(5)-Deu t er a t ed C(5a )-(4-Azid oa n i-
lin o)d ih yd r obicyclom ycin C(2′),C(3′)-Aceton id e (10-d ).
Using the preceding procedure and utilizing a CD3OD-D2O
(36) For the general experimental procedures employed, see ref 20.

Bicyclomycin-Based F Inactivators
J . Org. Chem., Vol. 62, No. 16, 1997 5439
(1:1) solution (1 mL), 928,29 (5 mg, 0.015 mmol) and 4-azido-
aniline hydrochloride (5 mg, 0.029 mmol) gave 10-d as a
mixture of diastereomers (>10:1, NMR analysis). The “pD”
was adjusted to 10.9 with a 0.1 N NaOD D2O solution. The
“pD” of the solution was determined from the observed pH
meter reading by using the relationship pD ) pH meter
reading + 0.4.37 The isolated yield for 10-d was 1 mg (13%):
solvent was removed in vacuo, and the residue was purified
by preparative TLC (10% MeOH-CHCl3) to give 3 as a
mixture of diastereomers (>10:1, NMR analysis): yield, 7 mg
(76%); Rf 0.20 (10% MeOH-CHCl3); mp 190 °C dec; FT-IR
(KBr) 3384 (br), 3275 (br), 2112, 1686, 1512, 1405 cm-1 1H
;
NMR (CD3OD) δ 1.33 (s, 3 H), 1.80-2.10 (m, 2 H), 2.28-2.40
(m, 1 H), 2.94 (dd, J ) 9.3, 13.8 Hz, 1 H), 3.51-3.85 (m, 4 H),
4.00-4.10 (m, 2 H), 6.68 (d, J ) 9.0 Hz, 2 H), 6.81 (d, J ) 9.0
Hz, 2 H); 13C NMR (CD3OD) 24.2, 29.9, 44.2, 50.6, 62.6, 68.5,
72.2, 78.1, 83.9, 89.5, 115.3, 120.8, 129.6, 147.5, 168.5, 172.4
ppm; MS (+CI) 437 [M + 1]+; Mr (+CI) 437.178 40 [M + 1]+
(calcd for C18H25N6O7 437.178 47).
P r ep a r a t ion of C(5a )-(3-F or m yla n ilin o)d ih yd r ob i-
cyclom ycin (4). To an acetone solution (1 mL) of 11 (15 mg,
0.032 mmol) was added MnO2 (10 mg), and the reaction
mixture was stirred at 40 °C (2 h). The solvent was removed
in vacuo, the crude product 12 was dissolved in a 50% aqueous
methanolic solution (1 mL), and then 1 drop of trifluoroacetic
acid was added. The solution was stirred at 40 °C (1 h), and
then the solvent was removed in vacuo. The residue was
1
Rf 0.60 (10% MeOH-CHCl3); H NMR (CD3OD) δ 1.37 (s, 3
H), 1.46 (s, 6 H), 1.86 (dd, J ) 8.4, 16.5 Hz, 1 H), 2.04 (dd, J
) 8.1, 16.5 Hz, 1 H), 2.92 (d, J ) 13.8 Hz, 1 H), 3.61 (d, J )
13.8 Hz, 1 H), 3.72 (d, J ) 8.4 Hz, 1 H), 3.81 (dd, J ) 8.1, 13.5
Hz, 1 H), 4.02-4.10 (m, 2 H), 4.46 (d, J ) 8.4 Hz, 1 H), 5.34
(d, J ) 8.7 Hz, 2 H), 6.82 (d, J ) 8.7 Hz, 2 H); MS (+CI) 478
[M + 1]+; Mr (+CI) 477.208 79 [M]+ (calcd for C21H27DN6O7
477.208 22).
P r ep a r a tion of C(5a )-(3-(Hyd r oxym eth yl)a n ilin o)d i-
h yd r obicyclom ycin C(2′),C(3′)-Aceton id e (11). Using the
procedure described for 10 and utilizing 928,29 (100 mg, 0.29
mmol) and 3-(hydroxymethyl)aniline (72 mg, 0.44 mmol) gave
11 as a single diastereomer (NMR analysis) after treatment
at 50 °C (5 h): yield, 105 mg (78%); mp 150-180 °C; FT-IR
purified by preparative TLC (10% MeOH-CHCl3
) to give pale
(KBr) 3390 (br), 3302 (br), 1687, 1608, 1044 cm-1
;
1H NMR
yellow 4 as a mixture of diastereomers (>10:1, NMR analy-
(CD3OD) δ 1.37 (s, 3 H), 1.46 (s, 6 H), 1.80-2.10 (m, 2 H),
2.32-2.47 (m, 1 H), 2.95 (dd, J ) 9.0, 14.1 Hz, 1 H), 3.60-
3.73 (m, 2 H), 3.81 (dd, J ) 8.1, 13.5 Hz, 1 H), 4.00-4.10 (m,
2 H), 4.45-4.48 (m, 3 H), 6.55-6.70 (m, 3 H), 7.07 (t, J ) 7.0
Hz, 1 H); 13C NMR (CD3OD) 24.8, 26.8, 28.2, 30.4, 44.6, 50.7,
64.2, 65.6, 73.2, 73.5, 84.1, 86.3, 88.8, 111.7, 112.9, 113.4, 117.0,
130.1, 143.5, 149.8, 167.8, 171.9 ppm; MS (+CI) 466 [M + 1]+;
Mr (+CI) 466.217 97 [M + 1]+ (calcd for C22H32N3O8 466.218 94).
P r ep a r a t ion of C(5)-Nor b icyclom ycin C(2′),C(3′)-
Aceton id e C(5)-O-(4-Azid oben zoa te) (13). To an anhy-
drous THF solution (1 mL) of 827 (5 mg, 0.015 mmol) were
added 4-azidobenzoic acid (3.5 mg, 0.019 mmol), dicyclohexyl-
carbodiimide (7 mg, 0.034 mmol), (dimethylamino)pyridine (2
mg, 0.016 mmol), and triethylamine (10 µL, 0.071 mmol), and
the solution was stirred at room temperature (5 h). The
solvent was removed in vacuo, and the residue was purified
by preparative TLC (10% MeOH-CHCl3) to give 13: yield, 3
mg (41%); mp 125-130 °C; Rf 0.60 (10% MeOH-CHCl3); FT-
IR (KBr) 3452 (br), 3320 (br), 2128, 1716, 1603, 1268 cm-1; 1H
NMR (DMSO-d6) δ 1.25 (s, 3 H), 1.35 (s, 3 H), 1.40 (s, 3 H),
1.95-2.10 (m, 2 H), 3.60-3.85 (m, 2 H), 3.90-4.07 (m, 2 H),
4.31 (d, J ) 8.1 Hz, 1 H), 5.14 (t, J ) 4.2 Hz, 1 H), 5.86 (d, J
) 8.1 Hz, 1 H), 7.03 (s, 1 H), 7.26 (d, J ) 8.4 Hz, 2 H), 7.97 (d,
J ) 8.4 Hz, 2 H), 8.12 (s, 1 H), 9.00 (s, 1 H); 13C NMR (DMSO-
d6) 24.6, 26.1, 27.8, 30.9, 58.6, 70.6, 71.2, 78.4, 80.8, 85.0, 87.3,
109.3, 119.2, 126.1, 131.3, 144.6, 163.7, 167.0, 167.4 ppm; MS
(+CI) 492 [M + 1]+; Mr (+CI) 492.172 00 [M + 1]+ (calcd for
C21H26N5O9 492.173 05).
P r ep a r a t ion of C(5)-Nor b icyclom ycin C(2′),C(3′)-
Acet on id e C(5)-O-(3-F or m ylb en zoa t e) (14). Using the
preceding procedure and utilizing 827 (50 mg, 0.15 mmol),
3-carboxybenzaldehyde (35 mg, 0.23 mmol), dicyclohexylcar-
bodiimide (50 mg, 0.24 mmol), (dimethylamino)pyridine (10
mg, 0.08 mmol), and triethylamine (100 µL, 0.71 mmol) gave
14 as a single diastereomer (NMR analysis) after treatment
at room temperature (3 h): yield, 30 mg (43%); mp 152-157
°C; Rf 0.50 (10% MeOH-CHCl3); FT-IR (KBr) 3447 (br), 1699,
1603, 1384, 1271, 1190, 1073 cm-1; 1H NMR (DMSO-d6) δ 1.26
(s, 3 H), 1.37 (s, 3 H), 1.40 (s, 3 H), 2.00-2.10 (m, 2 H), 3.65-
3.80 (m, 2 H), 3.90-4.05 (m, 2 H), 4.32 (d, J ) 8.1 Hz, 1 H),
5.21 (t, J ) 4.5 Hz, 1 H), 5.86 (d, J ) 8.1 Hz, 1 H), 7.07 (s, 1
H), 7.78 (t, J ) 7.8 Hz, 1 H), 8.10-8.70 (m, 3 H), 8.43 (s, 1 H),
9.03 (s, 1 H), 10.07 (s, 1 H); 13C NMR (DMSO-d6) 24.7, 26.1,
27.9, 31.0, 58.9, 70.6, 71.2, 78.8, 80.9, 85.1, 87.5, 109.4, 129.7,
130.1, 130.6, 134.1, 134.9, 136.4, 163.7, 166.9, 167.5, 195.5
ppm; MS (+CI) 479 [M + 1]+; Mr (+CI) 479.165 40 [M + 1]+
(calcd for C22H27N2O10 479.166 57).
sis): yield, 5 mg (36%); mp 134-136 °C; Rf 0.16 (10% MeOH-
CHCl3); FT-IR (KBr) 3433 (br), 1685, 1436, 1207, 1141 cm-1
;
1H NMR (CD3OD) δ 1.34 (s, 3 H), 1.85-2.10 (m, 2 H), 2.30-
2.45 (m, 1 H), 3.02 (dd, J ) 9.6, 13.8 Hz, 1 H), 3.54 (d, J )
11.4 Hz, 1 H), 3.62-3.85 (m, 3 H), 4.00-4.15 (m, 2 H), 6.90-
7.38 (m, 4 H), 9.84 (s, 1 H);
50.7, 62.6, 68.5, 72.3, 78.7, 83.8, 89.5, 113.2, 119.7, 120.3, 130.7,
-; Mr
13C NMR (CD3OD) 24.2, 29.9, 43.6,
139.0, 150.6, 168.5, 172.4, 195.0 ppm; MS (-CI) 423 [M]
(-CI) 423.166 46 [M]- (calcd for C19H25N3O8 423.164 17).
P r ep a r a tion of C(5)-Nor bicyclom ycin C(5)-O-4-(Azid o-
ben zoa te) (5). Using the procedure for 3 and utilizing 13 (15
mg, 0.03 mmol) gave 5 as a single diastereomer (NMR
analysis): yield, 7 mg (50%); mp 152-160 °C; Rf 0.50 (20%
MeOH-CHCl3); FT-IR (KBr) 3436 (br), 2129, 1700, 1603, 1273
cm-1; 1H NMR (DMSO-d6) δ 1.08 (s, 3 H), 1.85-2.00 (m, 2 H),
3.20-3.40 (m, 2 H), 3.50-3.65 (m, 1 H), 3.75 (m, 2 H), 4.39 (t,
J ) 5.7 Hz, 1 H), 5.06 (t, J ) 3.9 Hz, 1 H), 5.16 (s, 1 H), 5.23
(d, J ) 7.5 Hz, 1 H), 6.73 (s, 1 H), 7.16 (d, J ) 6.0 Hz, 2 H),
7.89 (d, J ) 6.0 Hz, 2 H), 8.68 (s, 1 H), 9.02 (s, 1 H); 13C NMR
(DMSO-d6) 23.8, 31.1, 57.3, 66.5, 70.3, 77.0, 78.7, 80.8, 87.6,
119.1, 126.1, 131.4, 144.5, 163.7, 167.4, 167.9 ppm; MS (+CI)
452 [M + 1]+; Mr (+CI) 452.143 30 [M + 1]+ (calcd for
C18H22N5O9 452.141 75).
P r epar ation of C(5)-Nor bicyclom ycin C(5)-O-(3-For m yl-
ben zoa te) (6). Using the procedure for 3 and utilizing 14 (25
mg, 0.05 mmol) gave 6 as a single diastereomer (NMR
analysis): yield, 13 mg (43%); mp 138-145 °C; Rf 0.20 (10%
MeOH-CHCl3); FT-IR (KBr) 3423 (br), 3273 (br), 1698, 1407,
1
1274, 1193 cm-1; H NMR (CD3OD:D2O (1:2)) δ 1.36 (s, 3 H),
2.10-2.35 (m, 2 H), 3.58 (d, J ) 11.4 Hz, 1 H), 3.70 (d, J )
11.4 Hz, 1 H), 3.80-3.95 (m, 1 H), 4.08 (s, 1 H), 4.15-4.30 (m,
1 H), 5.47 (t, J ) 3.9 Hz, 1 H), 7.70-8.52 (m, 4 H), 10.03 (s, 1
H); 13C NMR (CD3OD:D2O (1:2)) 23.9, 31.9, 59.4, 67.7, 71.6,
78.2, 80.3, 82.4, 88.8, 130.8, 131.1, 132.0, 135.6, 136.7, 137.3,
166.2, 170.2, 170.7, 195.6 ppm; MS (-CI) 437 [M - 1]-; Mr
(-CI) 437.120 53 [M - 1]- (calcd for C19H21N2O10 437.119 62).
Stu d ies on th e P h otoa ffin ity La belin g of Rh o by
Bicyclom ycin P r obe 3. A solution (0.2 mL) containing the
buffer (40 mM Tris‚HCl, pH 7.9, 50 mM KCl, 12 mM MgCl2,
0.1 mM EDTA, 0.1 mM DTT), rho (10 µg, 1 µM), ATP (1 mM),
3 (1 mM), DTT (1 mM), and mercaptoethanol (1 mM) was
irradiated (254 nm) using a Mineralight UVG-54 lamp (UVP
Inc.) at 25 °C (0.25 h). The reaction was then dialyzed (5 °C,
20 h) against 100 mM NaCl, 10 mM Tris‚HCl (pH 7.6), 5%
glycerol, 0.1 mM EDTA, and 0.1 mM DTT and then assayed.
The percentage inactivation of rho was determined by measur-
ing the initial velocity of ATPase activity using the previously
described protocol20,22 with rho (1 µg), poly C (24 µM), ATP
(250 µM), and 0.5 µCi [γ-32P]ATP. A second reaction was
conducted using 16 in place of 3. The results are presented
in Figure 2.
P r ep a r a tion of C(5a )-(4-Azid oa n ilin o)d ih yd r obicyclo-
m ycin (3). To a 50% aqueous methanolic solution (2 mL) of
10 (10 mg, 0.02 mmol) was added trifluoroacetic acid (1 drop),
and the solution was stirred at room temperature (2 h). The
Stu d ies on th e Red u ctive Am in a tion of Rh o by Bicy-
clom ycin P r obe 4. A solution (0.2 mL) containing the buffer
(40 mM Tris‚HCl, pH 7.9, 50 mM KCl, 12 mM MgCl2, 0.1 mM
(37) Bates, R. G. Determination of pH: Theory and Practice, 2nd
ed.; Wiley: New York, 1973; pp 375-376.

5440 J . Org. Chem., Vol. 62, No. 16, 1997
Cho et al.
EDTA, 0.1 mM DTT), rho (10 µg), ATP (1 mM), and 4 (5 mM)
was incubated at 25 °C (4 h). An aqueous solution (20 µL) of
NaBH4 (600 mM) was added, and the reaction was allowed to
stand at 25 °C (20 min). The reaction was dialyzed (5 °C, 20
h) against 100 mM NaCl, 10 mM Tris‚HCl (pH 7.6), 5%
glycerol, 0.1 mM EDTA, and 0.1 mM DTT and assayed. The
percentage inactivation of rho was determined by measuring
the initial velocity of ATPase activity using the previously
described protocol20,22 with rho (1 µg), poly C (24 µM), ATP
(250 µM), and 0.5 µCi [γ-32P]ATP. A second reaction was
conducted using 19 in place of 4. The results presented in
Figure 4 were done using 4 (1 mM) in the absence of ATP.
Electrospray mass spectra were recorded using a Quattro
tandem quadrupole mass spectrometer upgraded to Quattro
II specifications (Micromass UK Ltd., Altrincham, UK). The
electrospray capillary potential was 3.0-3.9 kV, and the cone
was held at 20-30 V. The resolution was set to give a peak
width at half height of 0.4 m/z unit for the monoisotopic peak
of a singly charged ion. Analytes were introduced at a 5 µL/
min flow rate using a syringe driver (Harvard Apparatus,
South Natick, MA). Data were collected and processed using
Micromass MassLynx software running on a Celebris XL 590
personal computer (Digital Equipment Corporation, Maynard,
MA).
Ack n ow led gm en t. This research was supported by
National Institutes of Health Grant GM37934 and
Robert A. Welch Foundation Grant E-607. Work at
UMIST is supported by EPSRC Grant GR/K18658 (to
S.J .G). I.R. is supported by a research studentship from
the U.K. Engineering and Physical Sciences Research
Council (EPSRC). We thank Dr. M. Kawamura and the
Fujisawa Pharmaceutical Co., Ltd., J apan, for the gift
of bicyclomycin, Dr. T. Platt (University of Rochester)
for the overproducing strain of rho, and Dr. A. Magyar
(University of Houston) for his help in the transcription
termination assays.
Su p p or tin g In for m a tion Ava ila ble: 1H and 13C NMR
spectra of all new compounds prepared in this study (17 pages).
This material is contained in libraries on microfiche, im-
mediately follows this article in the microfilm version of the
journal, and can be ordered from the ACS; see any current
masthead page for ordering information.
J O970575N