Photoinduced, family-specific, site-selective cleavage of TIM-barrel proteins

Article abstract of DOI:10.1021/ja9026105

(Chemical Equation Presented) Nonenzymatic, chemical methods for the controlled cleavage of proteins at predictable sites in a site-specific manner are rare and of strong potential utility in clean, post-translational manipulation of protein structure for

Full text of DOI:10.1021/ja9026105

Published on Web 08/13/2009
Photoinduced, Family-Specific, Site-Selective Cleavage of TIM-Barrel Proteins
Nicola Floyd,† Neil J. Oldham,‡ Christopher J. Eyles,§ Stephen Taylor,| Dmitry A. Filatov,⊥
Mark Brouard,§ and Benjamin G. Davis*,†
Department of Chemistry, UniVersity of Oxford, Chemistry Research Laboratory, Mansfield Road, Oxford, OX1
3TA, U.K., Department of Chemistry, UniVersity of Oxford, Physical and Theoretical Chemistry Laboratory, South
Parks Road, OX1 3QZ, U.K., The School of Chemistry, UniVersity of Nottingham, UniVersity Park, Nottingham,
NG7 2RD, U.K., Computational Biology Research Group, UniVersity of Oxford, Oxford, OX1 3RE, U.K., and
Department of Plant Sciences, UniVersity of Oxford, South Parks Road, Oxford, OX1 3RB, U.K.
Received April 1, 2009; E-mail: ben.davis@chem.ox.ac.uk
Proteins have long been known to be affected by UV irradiation in
Scheme 1. Photoinduced Site Selective Cleavage of TIM-Barrel
GH-1 Proteinsa
processes ranging from amino acid side chain degradation, the
reduction of disulfide bridges, aggregation, and even to consequent
loss of biological activity.1-3 However, natural examples of clean,
site-selective photocleavage are rare. To the best of our knowledge,
the only reported examples are fluorescent proteins Kaede and EosFP
that experience cleavage (between Phe61,His62 at 400 nm) and require
prior formation of a specific chromophore.4,5
A more general, clean, photoinduced cleavage mechanism, with
potential biotechnological utility in protein preparation6 or proteomic
analyses,7,8 is therefore also of fundamental interest. Indeed, while a
number of site-selective photolytic methods requiring the addition of
extraneous cleavage agents have been investigated,9-14 these are to
date poorly selective and low yielding. Use of unnatural amino acids,
such as o-nitro-Phe,15,16 only allows photocleavage in up to ∼30%.17
Unexpectedly, while investigating the photochemical modification of
proteins, we have discovered a clean, photocleavage reaction localized
to TIM-barrel proteins from family 1 of glycosylhydrolases (GH-1).
UV irradiation of archetypal GH-1 protein from the hyperthermo-
philic archaeon Sulfolobus solfataricus (SsꢀG)18 gave fragment product
profiles that were sharp and clean (Scheme 1), suggesting protein
cleavage at a single site.19 To probe reactive transitions precise
frequency light (193-355 nm) was generated using (Nd:YAG)-
pumped tunable dye and ArF-excimer (193 nm) lasers. No cleavage
was observed in the absence of UV light or at wavelengths outside
the range ~240-310 nm.
Cleanly formed fragment masses (18.2, 39.1 kDa) were determined
by ESI-MS (Scheme 1). N-Terminal sequencing revealed the latter to
be “blocked” by an unnatural, unreactive residue and the former to
share the sequence found at the N-terminus of SsꢀG, indicating a
cleavage site between residues His150 and Trp151. SsꢀG mutants
containing alterations at this junction (H150A, W151A, H150F)
remained intact and did not fragment, giving additional evidence for
this specific cleavage site and highlighting the need for specific residues
and not simply for potential functional (e.g., aromatic His150fPhe)
group equivalents.
a (a) ESI-MS after 1 h irradiation; (b) SDS-PAGE time course analysis
1-6 h; (c) densitometry determined reaction course.
H-LNMYHWPLPL-NH2, representing residues 150-152 and 146-155
of SsꢀG, respectively. Neither fragmented under UV light.
These results implicated conformation and secondary/tertiary struc-
ture as key determinants in fragmentation. Ala-scanning was used to
probe a 3.5 Å sphere of residues around the cleavage site: SsꢀG
mutations Y149A, P152A, F222A, and W433A prevented fragmenta-
tion, while alteration of more remote residues (Q18A, R79A, N81A,
L153A, N205A, E206A, V210A, and E387A) had no effect. Modeling
of the cleavage site (Figure 1) reveals that essential residue groups
Y149, F222, W433, and P152 cradle residue W151 in an hydrophobic,
π-rich environment.
Bioinformatic analysis reveals that the His-Trp (HW) diad is
widespread and, indeed, is repeated at positions His424Trp425 in SsꢀG.
Since this site did not dissociate on exposure to UV light, we considered
that an extended series of residues bordering His150 and Trp151 must
contribute to the remarkable regioselectivity of this cleavage reaction.
To probe the underlying molecular contributions, short peptide
fragments of SsꢀG, expected to have little inherent conformation but
correct primary sequence, were synthesized: H-HWP-NH2 and
† CRL, University of Oxford.
‡ University of Nottingham.
§ PTCL, University of Oxford.
| Computational Biology, University of Oxford.
⊥ Plant Sciences, University of Oxford.
Figure 1. SsꢀG’s photocleavage site (blue) and associated essential (red)
and nonessential (green) residues. Model based on PDB 1gow.
9
12518 J. AM. CHEM. SOC. 2009, 131, 12518–12519
10.1021/ja9026105 CCC: $40.75  2009 American Chemical Society

C O M M U N I C A T I O N S
Scheme 2. Putative Photoinduced Cleavage Mechanism
In conclusion, we have discovered an efficient photocleavage
reaction present in a natural motif of GH1 proteins that can be used to
create a photocleavable tag. The biological function, if any, of this
photochemical post-translational processing is not clear. However, early
phylogenetic analyses revealed a cluster of photocleavable proteins
that are products of isolated genes or genes in operons associated with
peptide transport.36
Acknowledgment. We thank Drs. J. McCullagh, R. Procter for
technical assistance; Prof. P. Ogilby for helpful discussions; Prof.
J. Noel, Drs. M. Kertesz, M. Moracci for plasmids encoding
chalcone reductase, arylsulfatase, and TaꢀG, respectively; Dr. S.
Kengen for PfꢀG; Prozomix for ReꢀG, CtꢀG, SpꢀG; and the
International AIDS Vaccine Initiative (IAVI) for funding.
Digestion-MS/MS analysis allowed accurate mass and formulae
determination of the C-terminal amino acid of the 18.2 kDa fragment
and N-terminal amino acid of the 39.1 kDa fragment giving results
consistent with His-amide 1 and enamide 2 (Scheme 2). Lack of an
N-terminal amine in 2 is also consistent with its lack of reactivity under
Edman degradation. MS analyses of the b1 ions of 2 and 3-indole-
acrylic acid were identical; both showed characteristic loss of CO
followed by loss of HCN from the indole ring. The presence of
indolylenamide in 2 was also confirmed by UV spectroscopy (λmax
340 nm).20-22
Evidence to suggest the involvement of singlet oxygen in the
cleavage reaction was not found. Exclusion of oxygen did not alter
reaction rate. After bubbling with argon or addition of scavenging
reagent trolox,23 protein cleavage occurred at the same rate had oxygen
not been removed. Reaction rate was also unaffected by D2O.24
Furthermore, one-photon photosensitized production of singlet oxygen
using TMPyP25 at 390 nm, a wavelength known not to cause protein
fragmentation, did not result in the cleavage of WT SsꢀG. Together
these suggested a mechanism other than one involving singlet oxygen.
A plausible sequence consistent with 240-308 nm light-induced
cleavage is shown in Scheme 2: suprafacial 1,3-ꢀ-H sigmatropic shift,26
R-elimination of fragment 1, and rearomatization. Nearby critical
residues might serve to lower the transition state of energy of the shift
and/or provide general acid/base assistance in the elimination. Other
mechanisms, e.g., photoactivated electron transfer-elimination,27
cannot be excluded.28
Supporting Information Available: Experimental procedures,
bioinformatics and mass spectrometry data. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) Permyakov, E. A.; Permyakov, S. E.; Deikus, G. Y.; Morozova-Roche,
L. A.; Grishchenko, V. M.; Kalinichenko, L. P.; Uversky, V. N. Proteins:
Struct., Funct., Genet. 2003, 51, 498–503.
(2) Clymer, D. J.; Geren, C. R.; Ebner, K. E. Biochemistry 1976, 15, 1093–7.
(3) Artyukhov, V. G.; Gusinskaya, V. V.; Dvurekova, E. A.; Rubtsov, M. P.
Biofizika 2007, 52, 1001–1007.
(4) Mizuno, H.; Mal, T. K.; Tong, K. I.; Ando, R.; Furuta, T.; Ikura, M.;
Miyawaki, A. Mol. Cell 2003, 12, 1051–1058.
(5) Wiedenmann, J.; Ivanchenko, S.; Oswald, F.; Schmitt, F.; Roecker, C.;
Salih, A.; Spindler, K.-D.; Nienhaus, G. U. Proc. Natl. Acad. Sci. U.S.A.
2004, 101, 15905–15910.
(6) Arnau, J.; Lauritzen, C.; Petersen, G. E.; Pedersen, J. Protein Expression
Purif. 2006, 48, 1–13.
(7) Gevaert, K.; Vandekerckhove, J. Electrophoresis 2000, 21, 1145–1154.
(8) Gasteiger, E.; Gattiker, A.; Hoogland, C.; Ivanyi, I.; Appel, R. D.; Bairoch,
A. Nucleic Acids Res. 2003, 31, 3784–3788.
(9) Jones, B. J.; Vergne, M. J.; Bunk, D. M.; Locascio, L. E.; Hayes, M. A.
Anal. Chem. 2007, 79, 1327–1332.
(10) Kumar, C. V.; Buranaprapuk, A. Angew. Chem., Int. Ed. Engl. 1997, 36,
2085–7.
(11) Kumar, C. V.; Thota, J. Inorg. Chem. 2005, 44, 825–827.
(12) Buranaprapuk, A.; Kumar, C. V.; Jockusch, S.; Turro, N. J. Tetrahedron
2000, 56, 7019–7025.
(13) Hoyer, D.; Cho, H.; Schultz, P. G. J. Am. Chem. Soc. 1990, 112, 3249–50.
(14) Suzuki, A.; Hasegawa, M.; Ishii, M.; Matsumura, S.; Toshima, K. Bioorg.
Med. Chem. Lett. 2005, 15, 4624–4627.
(15) England, P. M.; Lester, H. A.; Davidson, N.; Dougherty, D. A. Proc. Natl.
Acad. Sci. U.S.A. 1997, 94, 11025–11030.
(16) Endo, M.; Nakayama, K.; Kaida, Y.; Majima, T. Angew. Chem., Int. Ed.
2004, 43, 5643–5645.
(17) Peters, F; B, B., A.; Wang, J.; Schultz, P. G. Chem. Biol 2009, 16, 148–52.
(18) Aguilar, C. F.; Sanderson, I.; Moracci, M.; Ciaramella, M.; Nucci, R.; Rossi,
M.; Pearl, L. H. J. Mol. Biol. 1997, 271, 789–802.
(19) Indiscriminate cleavage results in diffuse PAGE smears.
(20) Takai, K.; Sasai, Y.; Morimoto, H.; Yamazaki, H.; Yoshii, H.; Inoue, S.
J. Biol. Chem. 1984, 259, 4452–7.
(21) Such electron-rich indolylenamides are poor conjugate electrophiles (see
ref22):the39kDafragmentofC344SM21IM73IM148IM206IM236IM275IM-
280IM383IM439I mutant did not react with ethanethiol, thiophenol,
1-pyrenebutanoic hydrazide, or fluorescein hydrazide.
(22) Soloshonoka, V. A.; Uekia, H.; Ellisa, T. K.; Yamadaa, T.; Ohfune, Y.
Tetrahedron Lett. 2005, 46, 1107–1110.
(23) Rasnik, I.; McKinney, S. A.; Ha, T. Nat. Methods 2006, 3, 891–893.
(24) Merkel, P. B.; Nilsson, R.; Kearns, D. R. J. Am. Chem. Soc. 1972, 94,
1030–1.
(25) Snyder, J. W.; Lambert, J. D. C.; Ogilby, P. R. Photochem. Photobiol.
2006, 82, 177–184.
(26) Kropp, P. J.; Fravel, H. G.; Fields, T. R. J. Am. Chem. Soc. 1976, 98,
840–1.
(27) Bent, D. V.; Hayon, E. J. Am. Chem. Soc. 1975, 97, 2612–2619.
(28) The reaction was unaffected by metals, EDTA, or substrate.
(29) Kleywegt, G. J. J. Mol. Biol. 1999, 285, 1887–1897.
(30) Beil, S.; Kehrli, H.; James, P.; Staudenmann, W.; Cook, A. M.; Leisinger,
T.; Kertesz, M. A. Eur. J. Biochem. 1995, 229, 38–94.
(31) Bomati, E. K.; Austin, M. B.; Bowman, M. E.; Dixon, R. A.; Noel, J. P.
J. Biol. Chem. 2005, 280, 30496–30503.
(32) Chi, Y. I.; Martinez-Cruz, L. A.; Jancarik, J.; Swanson, R. V.; Robertson,
D. E.; Kim, S. H. FEBS Lett. 1999, 445, 375–83.
(33) Degradation of noncleavable over 10 h. Optimal cleavable reaction times
2-4 h give background degradation < ∼20% (see SI); LASER light allows
cleaner 15 min reactions.
(34) Edgar, R. C. Nucleic Acids Res. 2004, 32, 1792–1797.
(35) Cantarel, B.; Coutinho, P.; Rancurel, C.; Bernard, T.; Lombard, V.;
Henrissat, B. Nucleic Acids Res. 2009, 37, D233–238.
(36) Alm, E. J.; Huang, K. H.; Price, M. N.; Koche, R. P.; Keller, K.; Dubchak,
I. L.; Arkin, A. P. Genome Res. 2005, 15, 1015–1022.
The generality of the photocleaved His-Trp motif was investigated
through various methods. Using the motif’s atomic coordinates, all
proteins with known 3D structure containing similarly spatially
arranged HWP-triads were identified (using SPASM29). The most
similar, the arylsulfatase from Pseudomonas aeruginosa,30 the chalcone
reductase from Medicago satiVa,31 and the ꢀ-glycosidase from
Thermosphaera aggregans (TaꢀG),32 were expressed and purified;
only the closest TaꢀG (<0.5 Å rmsd, 54% sequence similarity) was
cleaved by UV, whereas other proteins were degraded by UV
radiation.33 Daughter fragments (17.2 and 38.1 kDa) revealed a
homologous scission between His and Trp. Next, motif-pattern analysis
(using MEME with HMM) and sequence alignments (using
MUSCLE34 and CAZy35) revealed ∼1800 proteins with the highest
similarity across a variety of species. Phylogenetic analysis of this entire
grouping (using Quick Neighbor Joining algorithm) based on clustering
of the full protein sequence revealed branches focused according to
the tetrapeptide region that corresponds to Tyr149-Pro152 in WT-SsꢀG.
Representative proteins (10) from these branches were evaluated for
photocleavage activity. Together these analyses correctly predicted
other cleaving proteins (Y/FHWP, e.g., PfꢀG, Y149F-SsꢀG) and
revealed branches and motifs with weak (YHWD, e.g., ReꢀG, CtꢀG)
and no cleavage activity (HHFD/FHWD, e.g., SpꢀG, SaM). Finally,
to explore transplantation of the photocleavable motif, we created a
GFP-SsꢀG-(His6)tag fusion protein; Ni-Chromatography allowed af-
finity purification and then “photorelease” of GFP (see Supporting
Information (SI)).
JA9026105
9
J. AM. CHEM. SOC. VOL. 131, NO. 35, 2009 12519