Cu-free cycloaddition for identifying catalytic active adenylation domains of nonribosomal peptide synthetases by phage display

Article abstract of DOI:10.1016/j.bmcl.2008.08.085

To engineer the substrate specificities of nonribosomal peptide synthetases (NRPS), we developed a method to display NRPS modules on M13 phages and select catalytically active adenylation (A) domains that would load azide functionalized substrate analogs to the neighboring peptidyl carrier protein (PCP) domains. Biotin conjugated difluorinated cyclooctyne was used for copper free cycloaddition with an azide substituted substrate attached to PCP. Biotin-labeled phages were selected by binding to streptavidin.

Full text of DOI:10.1016/j.bmcl.2008.08.085

Bioorganic & Medicinal Chemistry Letters 18 (2008) 5664–5667
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Cu-free cycloaddition for identifying catalytic active adenylation domains of
nonribosomal peptide synthetases by phage display
*
Yekui Zou, Jun Yin
Department of Chemistry, The University of Chicago, 929 E. 57th Street, GCIS, E505A, Chicago, IL 60637, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
To engineer the substrate specificities of nonribosomal peptide synthetases (NRPS), we developed a
method to display NRPS modules on M13 phages and select catalytically active adenylation (A) domains
that would load azide functionalized substrate analogs to the neighboring peptidyl carrier protein (PCP)
domains. Biotin conjugated difluorinated cyclooctyne was used for copper free cycloaddition with an
azide substituted substrate attached to PCP. Biotin-labeled phages were selected by binding to
streptavidin.
Received 18 July 2008
Revised 22 August 2008
Accepted 22 August 2008
Available online 28 August 2008
Keywords:
Phage display
Ó 2008 Elsevier Ltd. All rights reserved.
Substrate specificity
Nonribosomal peptide synthetase
Cu-free click chemistry
Adenylation domain
Polyketides and nonribosomal peptides are two large classes of
pharmaceutically important natural products constituting a rich
source of anticancer, antifungal, antibiotic, and immunosuppressant
reagents.1 The biosynthesis of these natural products shares the
same logic by stepwise chain elongation on multimodular polyke-
tide synthases (PKS) or nonribosomal peptide synthetases (NRPS)
as the enzymatic assembly lines of simple building blocks—carbox-
ylic acids for PKS and amino acids for NRPS.2–5 The acyltransferase
(AT) domain in PKS is responsible for selecting a specific acyl-CoA
as the substrate and attaching the acyl group to the free thiol at
the end of the phosphopantetheinyl prosthetic group (Ppant) of
the acyl carrier protein (ACP) domain.6 Similarly the adenylation do-
main in NRPS is responsible for selecting a specific amino acid to be
attached to the Ppant arm on the peptidyl carrier protein (PCP) do-
mainembedded withintheNRPSenzymes(Fig. 1A).7,8 Subsequently,
substrates loaded on the carrier proteins by AT and A domains are
incorporated into the elongating polyketide or nonribosomal pep-
tide chains by stepwise condensation reactions.
To overcome the difficulties in organic synthesis, several strat-
egies have been developed for the biosynthesis of structural ana-
logs of polyketides and nonribosomal peptides. One method,
known as ‘combinatorial biosynthesis’, combines genes originated
from different PKS and NRPS biosynthetic pathways in one bacte-
rial host for the assembly of natural product molecules of new
structures.9–13 In another strategy called ‘precursor directed bio-
synthesis’, nonnative building block molecules are supplied to
the culture media for their incorporation in the biosynthetic reac-
tions catalyzed by PKS and NRPS enzymes.14 Often the substrate
tolerance of the biosynthetic enzymes limits the structural diver-
sity of the natural product analogs that can be produced by biosyn-
thesis. For examples, the AT domain in a PKS module and the A
domain in a NRPS module act as the gate-keeping domains exhib-
iting strict specificity for their cognate acyl-CoA or amino acid sub-
strates, respectively.6–8 To achieve greater diversity for substrate
loading in NRPS enzymes, we here report a phage selection method
for the identification of the A domains with altered substrate spec-
ificity. This method can be potentially used for directed evolution
of the AT and A gate-keeping domains in the PKS and NRPS en-
zymes to reprogram their substrate specificities.
Phage display establishes a direct linkage between the polypep-
tide entities displayed on the phage surface and the encoding gene
encapsulated inside the phage particle.15 In this way, phage display
provides a general platform for quickly sorting through large li-
braries of peptides or proteins for the selection of desired bind-
ing16,17 or catalytic activities.18–22 To engineer the substrate
specificity of the NRPS A domains by phage selection, we displayed
the A-PCP didomain on the surface of M13 phages and used alkyne
or azide functionalized substrate analogs for A domain catalyzed
substrate loading on the neighboring PCP domain (Fig. 1B). Sub-
strate analogs covalently attached to PCP can then be conjugated
to biotin by Huisgen’s 1,3-cycloaddion reaction between the azide
and alkyne functionalities.23,24 Such strategy directly couples sub-
strate loading catalyzed by the A domain on the phage surface with
biotin conjugation to the corresponding phage particles. Biotin-la-
beled phages can then be selected by binding to streptavidin.
* Corresponding author. Tel.: +1 773 8345881; fax: +1 773 7020805.
E-mail address: junyin@uchicago.edu (J. Yin).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.08.085

Y. Zou, J. Yin / Bioorg. Med. Chem. Lett. 18 (2008) 5664–5667
5665
Figure 1. (A) Loading of AHB 1 on the A-PCP didomain of RifA LM after the PCP domain being activated by Sfp catalyzed Ppant modification. (B) Phage selection strategy for
engineering the substrate specificity of the A domain of RifA LM.
To test this selection method, we cloned the gene encoding A-
PCP didomain of RifA LM into phagemid vector pCom3H for the
display of A-PCP on the surface of M13 phages as the N-terminal
fusion to truncated phage capsid protein pIII.25 RifA is the N-termi-
nal component of rifamycin synthetase from Amycolatopsis medi-
terranei that is responsible for the biosynthesis of proansamycin
X, a precursor to antibiotics rifamycin B.26 RifA LM was previously
cloned by Khosla and colleagues for expression and enzymatic
studies.27 To assay the display efficiency of the A-PCP didomain,
we used Sfp phosphopantetheinyl transferase to label the PCP do-
main on the phage surface with biotin-CoA 6 for the attachment of
biotin-Ppant group to a specific Ser residue on PCP.28,29 Western
blot of the labeling reaction mixture was probed with streptavi-
din–horseradish peroxidase (HRP) conjugate and showed a band
of 84 kDa in size corresponding to A-PCP fused to pIII (Supplemen-
tal Fig. 1). This suggests that the A-PCP didomain from RifA LM can
be displayed on the surface of M13 phages with good efficiency.
The A domain of RifA LM is specific for 3-amino-5-hydroxyben-
zoate (AHB) 1 and has been shown to have substrate tolerance for
substituted benzoates at the 3 and 5 positions (Fig. 1A).27,30 We
thus synthesized AHB analogs 2 and 4 with alkyne or azide func-
tionalities at the 3 position for their conjugation with biotin linked
cycloaddition partners 3 and 5 after substrate loading on the A-PCP
didomain displayed on phage surface (Fig. 1B). The reaction be-
tween 2 and 3 needs to be catalyzed by Cu(I) salt34 while the reac-
tion between 4 and 5 does not need the copper catalyst following
the method developed by Bertozzi and colleagues using a strained
alkyne functionality in cyclooctyne for azide conjugation.31,32
To assay the activities of 2 and 4 for substrate loading, the A-
PCP didomain of RifA LM was expressed as C-terminal fusions to
6Â His tag. ATP-pyrophosphate (PPi) exchange assay showed that
both 2 and 4 can be taken as the substrates of the A domain with
the alkyne analog 2 retaining 77% of the activity of the native sub-
strate AHB 1 and the azide analog 4 5% of the AHB activity (Supple-
mental Fig. 2). Covalent loading of 2 and 4 on PCP was also
confirmed by Western blot (Supplemental Fig. 3). In this assay,
A-PCP didomain was first activated by Sfp catalyzed Ppant modifi-
cation of PCP for the formation of holo-A-PCP. After incubation
with 2 and 4, substrate loading on holo-A-PCP was detected by
reacting biotin functionalized 3 and 5 with substrate analogs 2
and 4 attached to PCP, respectively. Western blot of the reaction
mixture was probed with streptavidin–HRP and biotin conjugation

5666
Y. Zou, J. Yin / Bioorg. Med. Chem. Lett. 18 (2008) 5664–5667
with the A-PCP didomain was confirmed by detecting a band of
67 kDa, matching the size of A-PCP of RifA LM (Supplemental
Fig. 3). These results suggest that although substrate analogs 2
and 4 have lower activity then the native substrate 1, both sub-
strates can be loaded on PCP by A domain catalysis. We thus
decided to test if 2 and 4 could be loaded on the A-PCP didomain
displayed on phage surface for the selection of catalytic active A
domains.
face. Biotin conjugation to phage displayed A-PCP was dependent
on substrate loading with 2 since no biotin conjugation was found
when 2 was not included in the loading reaction or phage dis-
played A-PCP was not activated for Ppant modification (Fig. 2A,
lanes 1 and 2). We then assayed if biotin-labeled phages from cop-
per catalyzed cycloaddition reactions can be bound to streptavidin
surface for affinity enrichment. Unfortunately we found that the
efficiency of phage infectivity was decreased by more than 100-
Compound 2 was incubated with phages displaying A-PCP dido-
main after Sfp catalyzed Ppant modification of PCP. After substrate
loading, phage particles were precipitated by polyethylene glycol
(PEG) followed by conjugation with biotin-azide 3 in the presence
of CuBr (Fig. 1B). Western blot of the phage loading reaction
showed a band corresponding to the size of A-PCP didomain fusion
with the phage capsid protein pIII (Fig. 2A, lane 3). This suggests
that biotin-azide 3 can be specifically conjugated with alkyne func-
tionalized substrate 2 loaded on A-PCP didomain on the phage sur-
fold upon the phages being exposed to 10 lM CuBr or CuSO4 for
30 min at room temperature. Under these conditions, phages
would not be able to infect the Escherichia coli cells for phage
amplification and subsequent rounds of selection after the cycload-
dition reaction catalyzed by the copper salt.
We then switched to azide substrate 4 for A-PCP loading on
phage surface since the cycloaddition reaction between 4 and bio-
tin-DIFO 5 does not require copper catalysts. Following the previ-
ous procedure, phage displayed holo-A-PCP of RifA was first
loaded with 4 in the presence of ATP. After PEG precipitation to re-
move the free substrate analog in the reaction, 5 was added to the
resuspended phage particles in copper free buffer and the coupling
reaction was allowed to proceed for 4 h. Western blot of the reac-
tion mixture showed biotin attachment to the A-PCP didomain
fused to phage capsid protein pIII (Fig. 2B, lane 1). In contrast, con-
trol reactions without Ppant activation or without substrate load-
ing with 4 did not show biotin attachment (Fig. 2B, lanes 2 and
3). This suggests copper free cycloaddition can be used for biotin
conjugation to azide functionalized substrate covalently attached
to the A-PCP didomain displayed on the phage surface. We also
found phages retained full infectivity after substrate loading with
4 and biotin conjugation with 5 without copper in the reaction.
We then tested if biotin-DIFO 5 can be used for specific selec-
tion of phages displaying A-PCP didomain of RifA with the loading
of azide substrate 4. A-PCP didomain displayed phages were acti-
vated by Ppant modification in the presence of Sfp and CoA fol-
lowed by substrate loading with 4. In parallel a control reaction
was set up in which same number of phages displaying a single
PCP domain of the NRPS module GrsA33 was modified by Ppant
and incubated with 4. After PEG precipitation, the resuspended
phage particles were reacted with 5. Reaction mixtures were PEG
precipitated again to remove unreacted 5 before they were added
to a 96-well plate coated with streptavidin for binding the biotin
conjugated phages. Plates were washed and E. coli cells were added
to the plate for direct infection by phages bound to the plate. Phage
infected cells were then titered and the results were shown in
Fig. 3. Affinity selection with streptavidin binding gave more than
a 10-fold increase in phage enrichment for phages displaying A-
PCP didomain after substrate loading with 4 comparing to the con-
trol phages displaying just the PCP domain after incubating with
the same substrate (Fig. 3, lanes 1 and 5). Also A-PCP displayed
phages without Ppant activation did not give any phage enrich-
ment over the control phages displaying the single PCP domain
(Fig. 3, lanes 4 and 5). The enrichment of A-PCP displayed phages
also relied on substrate loading with 4 and biotin conjugation with
5 (Fig. 3, lanes 2 and 3). These experiments demonstrate that cop-
per free cycloaddition between an azide functionalized substrate
loaded on phage displayed A-PCP and biotin-DIFO 5 can be used
for the enrichment of A domain displayed phages based on cata-
lytic turnover. Based on this method, substrate analogs of diverse
structures with a common azide substitute can be used to identify
matching mutants of the A domains that load nonnative substrates
on the NRPS assembly lines. We are currently using this method to
carry out selections on A domain libraries of RifA LM to identify A
domain mutants with increased catalytic activities with 4.
Figure 2. (A) Western blot of phage displayed A-PCP didomain of RifA LM after
substrate loading with 2 and conjugation with biotin-azide 3. (B) Western blot of
phage displayed A-PCP didomain of RifA LM after substrate loading with 4 and
conjugation with biotin-DIFO 5. The Western blots were probed with streptavidin–
HRP.
In summary we have developed a phage selection strategy to
engineer the substrate specificities of the A domains of NRPS based
on enzymatic loading of azide functionalized substrate analogs on

Y. Zou, J. Yin / Bioorg. Med. Chem. Lett. 18 (2008) 5664–5667
5667
Figure 3. Phage titers for the enrichment of catalytic active RifA LM. Phages displaying the A-PCP didomain of RifA LM were loaded with substrate 4 and conjugated with
biotin-DIFO 5. Biotin-labeled phages were enriched by binding to a streptavidin plate and titered. Phages displaying the single PCP domain were used as a control.
6. Liou, G. F.; Khosla, C. Curr. Opin. Chem. Biol. 2003, 7, 279.
the A-PCP didomain displayed on phage surface. We also applied
7. Challis, G. L.; Ravel, J.; Townsend, C. A. Chem. Biol. 2000, 7, 211.
copper free cycloaddition31,32 for biotin attachment to the phage
8. Stachelhaus, T.; Mootz, H. D.; Marahiel, M. A. Chem. Biol. 1999, 6, 493.
particles displaying catalytically active A-PCP didomains. We ex-
pect the same phage selection method can be used for displaying
AT-ACP didomains from PKS modules on phage surface and select-
ing for substrate loading with azide functionalized acyl-CoA ana-
logs. In this way, the substrate specificities of PKS and NRPS
enzymatic assembly lines can be reprogramed for the incorpora-
tion of nonnative building block molecules into natural product
biosynthesis. This method should be combined with existing meth-
ods of combinatorial biosynthesis9 and precursor directed biosyn-
thesis14 to afford natural product analogs of diverse structures.
9. Cane, D. E.; Walsh, C. T.; Khosla, C. Science 1998, 282, 63.
10. Marsden, A. F.; Wilkinson, B.; Cortes, J.; Dunster, N. J.; Staunton, J.; Leadlay, P. F.
Science 1998, 279, 199.
11. Menzella, H. G.; Reid, R.; Carney, J. R.; Chandran, S. S.; Reisinger, S. J.; Patel, K.
G.; Hopwood, D. A.; Santi, D. V. Nat. Biotechnol. 2005, 23, 1171.
12. Pfeifer, B. A.; Admiraal, S. J.; Gramajo, H.; Cane, D. E.; Khosla, C. Science 2001,
291, 1790.
13. Stachelhaus, T.; Schneider, A.; Marahiel, M. A. Science 1995, 269, 69.
14. Jacobsen, J. R.; Hutchinson, C. R.; Cane, D. E.; Khosla, C. Science 1997, 277, 367.
15. Smith, G. P.; Petrenko, V. A. Chem. Rev. 1997, 97, 391.
16. Avrantinis, S. K.; Stafford, R. L.; Tian, X.; Weiss, G. A. Chembiochem 2002, 3,
1229.
17. Baca, M.; Scanlan, T. S.; Stephenson, R. C.; Wells, J. A. Proc. Natl. Acad. Sci. U.S.A.
1997, 94, 10063.
18. Cesaro-Tadic, S.; Lagos, D.; Honegger, A.; Rickard, J. H.; Partridge, L. J.;
Blackburn, G. M.; Pluckthun, A. Nat. Biotechnol. 2003, 21, 679.
19. Damoiseaux, R.; Schultz, P. G.; Johnsson, K. Chembiochem 2002, 3, 573.
20. Janda, K. D.; Lo, L. C.; Lo, C. H.; Sim, M. M.; Wang, R.; Wong, C. H.; Lerner, R. A.
Science 1997, 275, 945.
21. Xia, G.; Chen, L.; Sera, T.; Fa, M.; Schultz, P. G.; Romesberg, F. E. Proc. Natl. Acad.
Sci. U.S.A. 2002, 99, 6597.
22. Yin, J.; Mills, J. H.; Schultz, P. G. J. Am. Chem. Soc. 2004, 126, 3006.
23. Agard, N. J.; Baskin, J. M.; Prescher, J. A.; Lo, A.; Bertozzi, C. R. ACS Chem. Biol.
2006, 1, 644.
Acknowledgments
This work was supported by a lab startup grant from the Uni-
versity of Chicago, the New Faculty Award from the Camille and
Henry Dreyfus Foundation and the NSF-MRSEC program (DMR-
0213745) (J.Y.) We thank P.G. Schultz (Scripps) for providing
pcomb3H-7G12 plasmid and C.T. Walsh (Harvard) for providing
PSsA8 plasmid.
24. Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem. Int. Ed.
Engl. 2002, 41, 2596.
25. Barbas, C. F., 3rd; Kang, A. S.; Lerner, R. A.; Benkovic, S. J. Proc. Natl. Acad. Sci.
U.S.A. 1991, 88, 7978.
26. August, P. R.; Tang, L.; Yoon, Y. J.; Ning, S.; Muller, R.; Yu, T. W.; Taylor, M.;
Hoffmann, D.; Kim, C. G.; Zhang, X.; Hutchinson, C. R.; Floss, H. G. Chem. Biol.
1998, 5, 69.
27. Admiraal, S. J.; Walsh, C. T.; Khosla, C. Biochemistry 2001, 40, 6116.
28. La Clair, J. J.; Foley, T. L.; Schegg, T. R.; Regan, C. M.; Burkart, M. D. Chem. Biol.
2004, 11, 195.
29. Yin, J.; Liu, F.; Li, X.; Walsh, C. T. J. Am. Chem. Soc. 2004, 126, 7754.
30. Admiraal, S. J.; Khosla, C.; Walsh, C. T. Biochemistry 2002, 41, 5313.
31. Agard, N. J.; Prescher, J. A.; Bertozzi, C. R. J. Am. Chem. Soc. 2004, 126, 15046.
32. Baskin, J. M.; Prescher, J. A.; Laughlin, S. T.; Agard, N. J.; Chang, P. V.; Miller, I. A.;
Lo, A.; Codelli, J. A.; Bertozzi, C. R. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 16793.
33. Yin, J.; Liu, F.; Schinke, M.; Daly, C.; Walsh, C. T. J. Am. Chem. Soc. 2004, 126, 13570.
34. Zou, Y.; Yin, J. Chembiochem 2008, in press.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2008.08.085.
References and notes
1. Clardy, J.; Walsh, C. Nature 2004, 432, 829.
2. Khosla, C. Chem. Rev. 1997, 97, 2577.
3. Marahiel, M. A.; Stachelhaus, T.; Mootz, H. D. Chem. Rev. 1997, 97, 2651.
4. Staunton, J.; Weissman, K. J. Nat. Prod. Rep. 2001, 18, 380.
5. Walsh, C. T. Science 2004, 303, 1805.