peptide macrocyclization catalyzed by a prolyl oligopeptidase involved in α-amanitin biosynthesis

Please cite this article in press as: Luo et al., Peptide Macrocyclization Catalyzed by a Prolyl Oligopeptidase Involved in a-Amanitin Biosynthesis,Chemistry & Biology (2014), http://dx.doi.org/10.1016/j.chembiol.2014.10.015

Chemistry & Biology

Brief Communication

Peptide Macrocyclization Catalyzedby a Prolyl Oligopeptidase Involvedin a-Amanitin BiosynthesisHong Luo,1,3 Sung-Yong Hong,1,3 R. Michael Sgambelluri,1,2 Evan Angelos,1 Xuan Li,1,4 and Jonathan D. Walton1,*1Department of Energy Plant Research Laboratory2Department of Biochemistry and Molecular Biology

Michigan State University, East Lansing, MI 48824, USA3Co-first author4Present address: Faculty of Environmental Science and Engineering, Kunming University of Science and Technology, Kunming 650091,

Yunnan, People’s Republic of China

*Correspondence: [email protected]://dx.doi.org/10.1016/j.chembiol.2014.10.015

SUMMARY

Amatoxins are ribosomally encoded and posttrans-lationally modified peptides that account for themajority of fatal mushroom poisonings of humans.A representative amatoxin is the bicyclic octapep-tide a-amanitin, formed via head-to-tail macrocyc-lization, which is ribosomally biosynthesized as a35-amino acid propeptide in Amanita bisporigeraand in the distantly related mushroom Galerina mar-ginata. Although members of the prolyl oligopepti-dase (POP) family of serine proteases have been pro-posed to play a role in a-amanitin posttranslationalprocessing, the exact mechanistic details are notknown. Here, we show that a specificPOP (GmPOPB)is required for toxin maturation in G. marginata. Re-combinant GmPOPB catalyzed two nonprocessivereactions: hydrolysis at an internal Pro to releasethe C-terminal 25-mer from the 35-mer propeptideand transpeptidation at the second Pro to producethe cyclic octamer. On the other hand, we showthat GmPOPA, the putative housekeeping POP ofG. marginata, behaves like a conventional POP.

INTRODUCTION

Ribosomally encoded and posttranslationally modified peptides

(RiPPs) have been found in plants, bacteria, fungi, and animals

(Arnison et al., 2013; Umemura et al., 2014). A common

posttranslational modification of RiPPs is backbone macrocycli-

zation. Known natural macrocyclized peptides include the cya-

nobactins from marine bacteria (Donia et al., 2006) and the orbi-

tides and cyclotides from plants (Craik and Malik, 2013).

Cyclization confers several critical attributes to peptides, in-

cluding increased rigidity, enhanced membrane permeability,

and resistance to proteolysis and denaturation (Craik and Alle-

well, 2012). As a result, many cyclic peptides have potent biolog-

ical activities against a wide range of medically and ecologically

Chemistry & Biology 21

significant targets (Bockus et al., 2013; Giordanetto and Kihl-

berg, 2014; Walton, 1996).

The amatoxins, such as a-amanitin, and the phallotoxins, such

as phalloidin, are among the most notorious of known naturally

occurring cyclic peptides. Amatoxins account for the majority

of fatal mushroom poisonings of humans and other sensitive

mammals (Bresinsky and Besl, 1990; Enjalbert et al., 2002; Ka-

neko et al., 2001; Sgambelluri et al., 2014; Wieland, 1986). Ama-

toxins are defining inhibitors of eukaryotic RNA polymerase II

(Bushnell et al., 2002).

Structurally, the amatoxins are bicyclic octapeptides with C-

to-N (head-to-tail) condensation of the peptide backbone and

a cross-bridge between Trp and Cys, which dipeptide has the

trivial name tryptathionine (May and Perrin, 2007) (Figure 1).

Three of the amino acids (Trp, Pro, and Ile) are hydroxylated.

Phallotoxins also contain tryptathionine but their macrocycles

comprise only seven amino acids. Amatoxins are resistant to

all forms of cooking, are not inactivated in the mammalian

digestive tract, and are rapidly absorbed into the bloodstream

and across the plasma membrane. Because these traits are

also desirable properties of beneficial drugs, the amatoxin scaf-

fold could provide the basis for new pharmaceuticals (Craik and

Allewell, 2012; Giordanetto and Kihlberg, 2014; Wong et al.,

2012).

a-Amanitin is biosynthesized ribosomally as a 35-amino acid

propeptide in Amanita bisporigera and in the distantly related

mushroom Galerina marginata (Hallen et al., 2007; Luo et al.,

2012; Riley et al., 2014). The genomes of A. bisporigera and

other amatoxin-producing species of Amanita species contain

>50 sequences homologous to the core gene, AMA1, called

the MSDIN family (Hallen et al., 2007; Li et al., 2013, 2014;

Zhou et al., 2013).

The amino acids of the mature toxins are located within the

propeptides between two conserved Pro residues. An enzyme

capable of cleaving the propeptide at both Pro residues was

purified from the phallotoxin-producing mushroom Conocybe

albipes and identified as a member of the prolyl oligopepti-

dase (POP) family of serine proteases (Fulop et al., 1998; Luo

et al., 2009; Szeltner and Polgar, 2008). However, the POP of

C. albipes produced only linear peptide, leaving the mechanism

of head-to-tail macrocyclization an open question.

, 1–8, December 18, 2014 ª2014 Elsevier Ltd All rights reserved 1

mailto:[email protected]

http://dx.doi.org/10.1016/j.chembiol.2014.10.015

Figure 1. Structure of a-Amanitin

The macrocyclic outer ring has the sequence IWGIGCNP in three-letter amino

acid code.

Chemistry & Biology

Peptide Cyclization by Prolyl Oligopeptidase


The enzymological basis of macrocyclization has been eluci-

dated for several other RiPPs. A protease that cuts at a con-

served Asn residue in the precursor to the cyclotide kalata B1

is necessary for cyclization (Saska et al., 2007). Cyclization of

patellamide, a cyanobactin, is catalyzed by PatG (Agarwal

et al., 2012; Koehnke et al., 2012; Lee et al., 2009). Cyclization

of the orbitide segetalin A is catalyzed by PCY1 (Barber et al.,

2013). PCY1 is, like POPB, a member of the S9A family of

serine proteases. Butelase 1 is a cyclizing enzyme that can

cyclize a variety of peptides of 14 to 58 residues (Nguyen

et al., 2014).

A. bisporigera and G. marginata each have two POP genes,

known as POPA and POPB. POPB of G. marginata (GmPOPB)

and POPB of A. bisporigera (AbPOPB) are present only in am-

atoxin-producing species in each genus, whereas the POPA

genes are present in all species of Amanita and Galerina as

well as other agarics (Luo et al., 2010, 2012). GmPOPB is

clustered with one of two copies of GmAMA1 (Riley et al.,

2014). Collectively, these observations suggested that POPB

is involved in, and dedicated to, the biosynthesis of, a-amanitin

and that it catalyzes cleavage of the toxin propeptides to

release the mature octa- or heptapeptides. However, direct

proof of a role of POPB in amatoxin and phallotoxin biosyn-

thesis is lacking, and no evidence addresses its possible role

in macrocyclization.

RESULTS

Targeted Mutation of GmPOPBA transformation system using Agrobacterium tumefaciens and

hygromycin selection was developed based on a method devel-

oped for Laccaria bicolor (Kemppainen and Pardo, 2010, 2011).

The POPB gene was disrupted by targeted gene replacement

in a monokaryotic strain of G. marginata. Four independent

POPB-knockout mutants generated by homologous integration

of the transforming DNA were shown by Southern blotting to

be deleted for GmPOPB (Figures S1A and S1B available online).

All four mutants lost the ability to produce amanitin, as shown for

transformant 1 (Figure S1C). The GmPOPBmutant had no other

discernible phenotype such as a change in growth rate or

pigmentation.

2 Chemistry & Biology 21, 1–8, December 18, 2014 ª2014 Elsevier L

Expression andAssay ofGmPOPAandGmPOPB in YeastGmPOPB protein was expressed in Saccharomyces cerevisiae

and purified on an anti-c-myc peptide antibody affinity col-

umn (Figure S1D). GmPOPA protein was also expressed in

S. cerevisiae (Figure S1D). GmPOPA showed good activity

against the chromogenic substrate Z-Gly-Pro-pNA (176 ±

5 mmol p-nitroaniline/min/mg protein), which is comparable

with published values for POP from lamb kidney (Koida and

Walter, 1976), but GmPOPB showed much weaker activity

(10 ± 2 mmol/min/mg). Versions of GmPOPB expressed without

the c-myc tag or with an N-terminal c-myc tag were also ex-

pressed in yeast, but they had the same poor activity (data

not shown).

GmPOPA and GmPOPB were also assayed using the 35-

amino acid a-amanitin precursor (GmAMA1) as substrate.

GmPOPB converted GmAMA1 to a series of products consistent

with cleavage at either one or both flanking Pro residues (Figures

2A and 2B). Another major product had a mass and retention

time identical to synthetic cyclo(IWGIGCNP) (Figure 2C). In a

time-course experiment, conversion of the 35-mer to the puta-

tive cyclic product progressed to near completion in 50 min

(Figure 3A). The 25-mer intermediate accumulated transiently

(Figure 3B). This indicates that GmPOPB catalyzed two nonpro-

cessive reactions: hydrolysis at the carboxyl side of the first Pro

and transpeptidation of the new N-terminal residue (Ile) to the

second Pro, with release of the C-terminal 17-amino acid pep-

tide (Figure 3C).

POPA hydrolyzed the 25-amino acid intermediate but not

the full-length 35-amino acid propeptide (Figures S2A–S2D).

In neither case was any cyclo(IWGIGCNP) produced. Mamma-

lian POP, obtained commercially, could also neither cleave

GmAMA1 nor cyclize either GmAMA1 or the 25-amino acid inter-

mediate (data not shown).

Proof of the Structure of the Cyclic Product by NuclearMagnetic Resonance SpectroscopyA large-scale reaction using purified GmPOPB and 10 mg of

GmAMA1 was used to produce 1.5 mg of the putative cyclic

product. All 1H and 13C atoms in the product were assigned

using correlation spectroscopy (COSY), total correlation spec-

troscopy (TOCSY), heteronuclear single-quantum correlation

(HSQC) spectroscopy, and heteronuclear multiple-bond correla-

tion (HMBC) NMR spectroscopy (Figures S3A and S3B). A signal

from the free thiol proton of Cys indicated that the product did

not contain an internal thioester (Figure S3C). Rotating-frame

nuclear Overhauser effect correlation spectroscopy (ROESY)

indicated that the Ile1-HN and Pro8-Ha atoms were positioned

within �5 A of each other (Figure S3D), and HMBC NMR

spectroscopy indicated through-bond coupling of Ile-1HN to

Pro8-13CO (i.e., the presence of a new amide bond) (Figure S3E).

Taken together, the structural evidence proved that the com-

pound produced by the action of pureGmPOPBon the 35-amino

acid GmAMA1 propeptide was cyclo(IWGIGCNP).

Substrate Requirements of GmPOPBSynthetic peptides were tested with pure recombinant GmPOPB

to explore its substrate structural requirements (Table 1). Consis-

tent with the fact that GmPOPB also catalyzes removal of the

N-terminal 10 amino acids in a nonprocessive manner, the

td All rights reserved

Figure 2. HPLC Fractionation of Reaction

Products Produced by GmPOPB from

GmAMA1

The reaction was stopped before completion to

show the intermediates.

(A) Mass spectrometric identification of peaks.

(B) UV trace (optical density at 280 nm [OD280]).

(C) Chromatogram of chemically synthesized

cyclo(IWGIGCNP).

The compound eluting at 8.7min in (B) is unknown.

Chemistry & Biology



resulting 25-amino acid product of this first reaction (Table 1,

substrate 2) was cyclized to completion.

Truncating the N terminus by four amino acids led to a sharp

decrease in efficiency of cyclization due to failure to cut at

the first Pro (Table 1, substrate 3). Therefore, the sequence

and/or length of the N terminus is important for recognition by

GmPOPB. Five amino acids (ATRLP) immediately upstream of

the toxin region are shared in common between AbAMA1 and

GmAMA1 (Figure S4A). Changing this region to either AAAAP

or ATAAP was sufficient to reduce, but not completely abolish,

cyclization (Table 1, substrates 4 and 5).

To investigate the role of the C-terminal region in macrocycli-

zation, two peptides truncated in the C terminus were synthe-

sized. They were both cut by GmPOPB at the first Pro, but

neither was cut at the second Pro or cyclized, indicating that

the C terminus is important for complete processing (Table 1,

substrates 6 and 7). Because the primary sequences of the

C-terminal regions of GmAMA1 and AbAMA1 are quite different

(Figure S4A), we investigated their secondary structures. Both

C domains are predicted to form an a helix (Figures S4A and

Chemistry & Biology 21, 1–8, December 18, 20

S4B). Consistent with secondary struc-

ture rather than primary structure being

important, GmPOPB efficiently cyclized

AbAMA1, indicating that the critical fea-

tures that are necessary for cyclization

are conserved in AbAMA1 (Table 1, sub-

strate 12) (Figures S2E and S2F). To test

specifically whether the putative a helix

is critical, selected residues within the he-

lix region were changed to Gly, an amino

acid that tends to disrupt a helix forma-

tion (Table 1, substrates 8–10). All three

modifications (i.e., introduction of one,

two, or three Gly residues) resulted in

drastic reduction of cyclization efficiency.

The terminal Cys is conserved in

GmAMA1 and AbAMA1 (Figure S4A).

Changing this amino acid to Ala resulted

in a product that was cleaved at the first

Pro but not at the second Pro or cyclized

(Table 1, substrate 11).

A. bisporigera, but not G. marginata,

also biosynthesizes phallotoxins from

genes in the MSDIN family. GmPOPB

did not efficiently cleave the phallacidin

propeptide from A. bisporigera, even at

the first Pro (Table 1, substrate 13). This

suggests that the ‘‘toxin’’ region (i.e., the region between the

two Pro residues) might be involved in binding to the active

site of GmPOPB, but more peptides will need to be tested to

confirm this. A carboxy-truncated version of AbPHA1 was cut

effectively at both Pro residues, but no cyclization occurred

(Table 1, substrate 14). This result was unexpected, because

GmPOPB did not cut two carboxy-truncated versions of

GmAMA1 (substrates 6 and 7) at the second Pro. Perhaps the

precise length of the C-terminal overhang is critical for the pro-

teolytic cleavage and/or cyclization at the second Pro.

GmPOPB Enzyme KineticsThe enzyme kinetic constants of GmPOPB with both the 35-mer

and the 25-mer as substrates were determined bymeasuring the

rate of cyclic product formation by high-performance liquid chro-

matography (HPLC) (Figures S4C and S4D). The Km and kcat for

the 35-mer were 25.5 mM and 5.6 s�1, respectively. The Km and

kcat for the 25-mer were 23.2 mMand 5.7 s�1. The similarity of the

values of the kinetic constants for the two substrates implies that

the overall conversion of the 35-mer to the cyclic octapeptide is

14 ª2014 Elsevier Ltd All rights reserved 3

Figure 3. Time Course of Conversion of GmAMA1 Propeptide to cyclo(IWGIGCNP)

(A) HPLC analysis of products at different time points.

(B) Graph of time-course results. The quantitation based on OD280 was normalized to the Trp content of the products.

(C) Diagram of the reactions catalyzed by GmPOPB.

Chemistry & Biology



limited by the second (cyclization) reaction, consistent with the

transient accumulation of the 25-mer intermediate (Figure 3).

Sequence Comparisons of POPA and POPB fromG. marginata and A. bisporigera

The POPB enzymes of G. marginata and A. bisporigera are

75.5% identical, the two POPAs are 65.7% identical, and the

average identity between the POPAs and the POPBs of the

two fungi is 57.6% (Figure S4E). The homology among the four

POPs is roughly distributed throughout the proteins, as are the

differences between the POPAs and the POPBs. GmPOPB is


36% identical to PCY1 and 37% identical to porcine POP (Fulop

et al., 2001). All four of the mushroom POP proteins have the b

propeller domain and the peptidase domain with an a/b hydro-

lase fold typical of POPs. On the basis of alignment with porcine

POP, the catalytic domain of GmPOPB is predicted to comprise

amino acids 1 to 80 and 451 to 730. POPA and POPB of

A. bisporigera and G. marginata have the conserved catalytic

triad typical of serine proteases in family S9A (in GmPOPB, these

are Ser577, Asp661, and His698). Both POPA and POPB also

have the Trp residue (Trp619 in GmPOPB) that stacks with the

Pro of the substrate (Fulop et al., 2001; Li et al., 2010). Unlike


Table 1. Synthetic Peptides Tested with GmPOPB

Substrate No. Substrate Major Product Conclusion

1 MFDTNATRLPIWGIGCNPWTAEHV

DQTLASGNDIC (full-length GmAMA1)

cyclo(IWGIGCNP) efficiently cyclized

2 IWGIGCNPWTAEHVDQTLASGNDIC

(N-truncated 25-mer, equivalent to

native intermediate)

cyclo(IWGIGCNP) cyclized as efficiently as full length

3 NATRLPIWGIGCNPWTAEHVDQ

TLASGNDIC (N-truncated 31-mer)

none N-terminal four amino acids

important for first cut

4 MFDTAAAAAPIWGIGCNPWTAEHVDQT

LASGNDIC (modification of amino acids

conserved in the N termini of GmAMA1

and AbAMA1)

cyclo(IWGIGCNP) (<2%) N-terminal amino acid composition

important for efficient processing

5 MFDTNATAAPIWGIGCNPWTAEHVDQTL

ASGNDIC (modification of amino acids

conserved in the N termini of GmAMA1

and AbAMA1)

cyclo(IWGIGCNP) (<2%) N-terminal amino acid composition

important for efficient processing

6 MFDTNATRLPIWGIGCNPWTAEHVD

(C-truncated 25-mer)

IWGIGCNPWTAEHVD C terminus important for cyclization

7 MFDTNATRLPIWGIGCNPWTAEHVDQ

TLAS (C-truncated 30-mer)

IWGIGCNPWTAEHVDQTLAS C terminus important for cyclization

8 IWGIGCNPWTAEGVGQGLASGNDIC

(Native 25-mer intermediate altered

in a helix region)

cyclo(IWIGCNP) (<1%) C-terminal a helix necessary

for efficient cyclization

9 IWGIGCNPWTAEHGDQGLASGNDIC


in a helix region)

cyclo(IWIGCNP) (�2%) C-terminal a helix necessary


10 IWGIGCNPWTAEHVDQTGASGNDIC


in a helix region)

cyclo(IWIGCNP) (�5%) C-terminal a helix necessary


11 MFDTNATRLPIWGIGCNPWTAEHVD

QTLASGNDIA (ultimate C changed to A)

IWGIGCNPWTAEHVDQTLASGNDIA

(partial) cyclo(IWGIGCNP) (trace)

conserved C-terminal Cys

important

12 MSDINATRLPIWGIGCNPCVGDDVTTL

LTRGEALC (AbAMA1)

cyclo(IWGIGCNP) (efficient;

Figure S2)

AbAMA1 recognized and cyclized

by GmPOPB despite low sequence

identity

13 MSDINATRLPAWLVDCPCVGDDVTTLL

TRGEALC (phallacidin gene from

A. bisporigera (AbPHA1)

AWLVDCPCVGDDVTTLLTRGEALC

(partial)

GmPOPB cannot cyclize

phallacidin

14 MSDINATRLPAWLATCPCAGDD

(C-truncated phalloidin propeptide

from A. bisporigera)

AWLATCP GmPOPB cannot cyclize phalloidin

lacking C terminus

Structural assignments of the products were based on their m/z values, OD280 (for Trp-containing peptides), and retention times compared with stan-

dards, when available. The toxin regions in the substrates are underlined. Altered amino acids are double-underlined.

Chemistry & Biology



PatG, GmPOPB and AbPOPB lack any apparent sequence or

structural motifs that could account for their unique enzymatic

properties (Figure S4E).

DISCUSSION

A dedicated POP, GmPOPB, cleaves the 35-amino acid

amanitin propeptide at two internal Pro residues, and the second

cleavage is a transpeptidation reaction between Ile and Pro to

produce the cyclic octapeptide found in the mature toxin.

Themushroom system has similarities and differences to other

peptide macrocyclases (Agarwal et al., 2012; Barber et al.,

2013; Nguyen et al., 2014). One common theme is that the pre-

cursors include leader or follower amino acids that are required


for cyclization (Oman and van der Donk, 2010). A second com-

mon theme is that the propeptides in whose processing POPB,

PCY1, and PatG are involved require two proteolytic reactions

in going from the primary translation products to the final cyclic

peptide product. However, in the case of PCY1 and PatG, two

separate enzymes catalyze the two proteolytic reactions,

whereas POPB catalyzes both in a-amanitin biosynthesis. The

comparison of POPB with PCY1 is especially intriguing because

both enzymes are serine proteases in family S9A (Fulop et al.,

2001).

GmPOPB is unlike other described POPs in several regards. It

accommodates peptides as large as 35 amino acids, it does not

efficiently use the synthetic chromogenic substrate Z-Gly-Pro-

pNA, and it catalyzes macrocyclization. GmPOPA, on the other


Chemistry & Biology



hand, is a more typical POP; it can cut a 25-mer but not a 35-mer

(Koida and Walter, 1976; Li et al., 2010), it efficiently uses the

chromogenic substrate, and it does not catalyze cyclization.

While the current paper was in preparation, butelase 1 was

shown to be a peptide cyclase involved in cyclotide biosynthesis

(Nguyen et al., 2014). GmPOPB and butelase 1 have comparable

turnover rates (5.6 versus 0.6–17 s�1), both of which are much

higher than PCY1 or PatG. The Km values of butelase 1 and

GmPOPB are also similar (31–213 and 25 mM, respectively).

Butelase 1 cyclizes not only derivatives of kalata B1, but also

sequence-unrelated peptides (Nguyen et al., 2014). However,

unlike GmPOPB, butelase 1was not successfully overexpressed

in a heterologous system, which limits its availability for further

studies. Insofar as POPB might have biotechnological applica-

tions, AbPOPB could prove more versatile than GmPOPB,

because Amanita mushrooms make many more types of small

cyclic peptides thanGalerina (Hallen et al., 2007; Wieland, 1986).

SIGNIFICANCE

Ribosomally synthesized cyclic peptides comprise a large

class of biologically active natural products. Here we show

that the cyclic peptide a-amanitin is processed from its pro-

peptide gene product by an enzyme in the proline oligopep-

tidase family of serine proteases. In two steps, the enzyme

converts the primary propeptide to the cyclic octapeptide.

EXPERIMENTAL PROCEDURES

Transformation of G. marginata

Monokaryotic strain CBS 339.88 of G. marginata strain was obtained from the

Centraalbureau voor Schimmelcultures, Utrecht, the Netherlands, and main-

tained on potato dextrose agar (PDA) (Luo et al., 2012). Fresh cultures were

used for transformation.

A. tumefaciens strain LBA1100, a generous gift of Alejandro Pardo, Uni-

versidad Nacional de Quilmes, Buenos Aires, Argentina, was transformed

by electroporation (Weigel and Glazebrook, 2006) with vector pHg (Kemp-

painen and Pardo, 2011). Fresh colonies from glycerol stocks were used

to inoculate seed cultures of Agrobacterium grown in 1 ml Luria broth

with 100 mg/ml kanamycin in 10 ml test tubes at 28�C with 200 rpm agitation

for 24 hr. This seed culture was used for inoculating 5 ml of minimal medium

(Kemppainen and Pardo, 2011) containing 100 mg/ml kanamycin in loosely

capped 50 ml plastic centrifuge tubes (Sarstedt) and grown overnight at

28�C and 200 rpm. The cells were pelleted at 9,500 3 g for 10 min at

4�C, the supernatant removed, and the bacteria resuspended in 5 ml of in-

duction medium buffered at pH 5.3 with 40 mM MES and supplemented

with 100 mg/ml kanamycin and 200 mM acetosyringone. Bacteria were

grown for 6 hr at 28�C with shaking at 200 rpm, at which point the optical

density at 600 nm of the culture was typically 0.4 to 0.5 and presented a

slightly flocculent appearance.

The fungal transformation protocol was a modification of the method

developed for L. bicolor (Kemppainen and Pardo, 2011). Round cello-

phane membranes (8 cm in diameter) were autoclaved and placed on 9-

cm plates of PDA. Small fragments (1 3 2 mm) of fresh G. marginata

mycelium were cut from the edge of a colony with a scalpel and placed

on the cellophane membrane discs, spaced �0.5 cm apart (�15–20 col-

onies per 9 cm plate). The plates were incubated at room temperature

for 3 days, at which point the colonies were �0.3 to 0.5 cm in diameter.

The discs plus fungus were then transferred to agar plates containing

P0.2% medium (Kemppainen and Pardo, 2011) supplemented with

200 mM acetosyringone. Each fungal colony was inoculated with 5 to

10 ml of induced Agrobacterium culture by pipetting the bacteria directly

onto the colonies. The plates were sealed with Parafilm (Bemis), covered

in aluminum foil, and placed in darkness.


After 3 to 4 days, the cellophane discs with the fungal colonies were trans-

ferred to selection plates containing HSV-2Cmedium (Luo et al., 2012) in 1.5%

agar supplemented with 30 mg/ml hygromycin B. The colonies were resealed

and further incubated at room temperature in light. After 2 weeks, pieces

of growing mycelium were transferred to fresh HSV-2C agar containing

30 mg/ml hygromycin B. When the colonies reached �3 cm in diameter (10–

20 days) they were transferred again to fresh HSV-2C plus hygromycin. This

process was repeated two to three times. Transformants were confirmed by

colony PCR (Alshahni et al., 2009) and Southern blotting at standard stringency

(Luo et al., 2012).

a-Amanitin was extracted from mycelium of G. marginata using 50% meth-

anol and analyzed by reverse-phase HPLC as previously described (Luo et al.,

2012; Sgambelluri et al., 2014).

Yeast Expression of POP Genes

Subcloning in Escherichia coli DH5a (Invitrogen) used standard procedures.

S. cerevisiae YPH501 (Agilent Technologies) was cultured on YPAD medium

(1% yeast extract, 2% peptone, 2% dextrose, 0.0075% adenine hemisulfate

salt) for competent cell preparation. SD-HIS medium (0.67% yeast nitrogen

base without amino acids, 2% glucose, 0.13% amino acid mixture minus

histidine) was used as a selective medium for isolation of yeast transformants

and SG-HIS medium (galactose instead of glucose in SD-HIS) for total sol-

uble protein extraction. For protein production, yeast cells were grown in

100 ml cultures of SG-HIS with shaking at 200 rpm and 30�C according to

the manufacturer’s instructions (pESC Yeast Epitope Tagging Vectors, Agilent

Technologies).

For expression of GmPOP, yeast expression vector pESC-HIS (Agilent

Technologies) was used. A 2.19-kb cDNA of POPB (GenBank AEX26938.2)

was synthesized by GENEART (Life Technologies), using S. cerevisiae codon

bias, and cloned into the ApaI/SalI sites of pESC-HIS. The resulting plasmid,

pGM-POPB, had a c-myc affinity tag fused in-frame to the C terminus of

GmPOPB. To construct the plasmid for expression of GmPOPA (GenBank

accession number AEX26937.2), a 2.25 kb GmPOPA cDNA was synthesized

and cloned into the ApaI and SalI sites of pESC-HIS. The resulting plasmid

(pGM-POPA) had a c-myc epitope tag fused at the C terminus of GmPOPA.

Yeast transformation was performed by the lithium acetate method according

to the manufacturer’s instructions (pESC Yeast Epitope Tagging Vectors, Agi-

lent Technologies).

GmPOPB and GmPOPA were purified from total soluble protein extracted

from transformants carrying pGM-POPB using anti-c-Myc agarose affinity

gel (Thermo Scientific) according to the manufacturer’s instruction. Yields

were �5 mg/l.

POP Assays

The chromogenic POP substrate Z-Gly-Pro-pNA (Sigma-Aldrich) was used at

1 mM in a reaction volume of 100 ml 50 mM Tris-HCl (pH 7.5). After 4 hr incu-

bation at 37�C, the released p-nitroanilide was measured at 405 nm. Each

assay typically contained 2 mg enzyme.

For assay of POP activity on peptides, the peptides were dissolved at 1 mg/

ml in water, 20 mM Tris-HCl (pH 7.5), containing 1 mM dithiothreitol, or 0.1%

trifluoroacetic acid. Reactions were performed at 37�C in a volume of 50 ml.

The temperature optimum was between 35�C and 45�C, and the pH optimum

was 7.5. At the end of the incubation, 50 ml of methanol was added and the

samples were centrifuged for 5 min. The supernatant was analyzed by liquid

chromatography (LC)/mass spectrometry (MS) using an Agilent 1100 pump

system and Agilent 6120 quadrupole mass spectrometer equipped with auto-

sampler and multiwavelength UV detector. The column was a reverse-phase

C18 column (RS-2546-W185, Higgins Analytical). Solution A was 0.02 M

ammonium acetate (pH 5), and solution B was 100% acetonitrile. Flow rate

was 1 ml/min, with a linear gradient from 100% A solution to 100% B solution

in 20 min. UV absorbance was monitored at 220, 250, and 280 nm.

For the kinetic studies, the reaction conditions were 20 mM Tris (pH 7.5),

10 mM dithiothreitol, and 30 min at 37�C. Each reaction contained 15 ng

(0.18 pmol) of enzyme in 50 ml. Reactions were stopped with methanol and

analyzed as above by HPLC. Kinetic constants were calculated using

nonlinear curve fitting with GraphPad Prism (GraphPad Software).

Chemically synthesized peptides were supplied by Bachem or Elim

Biopharmaceuticals. Peptide structures were predicted using PEP-FOLD


Chemistry & Biology



(Thevenet et al., 2012). Recombinant mammalian POP was purchased from

Sigma-Aldrich. Two micrograms were incubated with 10 mg GmAMA1 for

3 hr at pH 7.5 and 37�C. Protein concentrations were measured using BCA

(Pierce Biotechnology) with BSA as standard.

Large-Scale Purification of Cyclic Product, cyclo(IWGIGCNP)

GmPOPB protein (5 mg) purified on a c-myc affinity column was incubated with

10 mg GmAMA1 for 15 hr at 37�C. Progress of the reaction was monitored by

HPLC as described elsewhere. The protein was removed by precipitation with

50% v/v methanol followed by centrifugation. The supernatant was dried un-

der vacuum and redissolved in 4 ml 20 mM ammonium acetate (pH 5). The

product was purified on a preparative C18 column. Drying under vacuum

yielded 1.25 mg of white powder (57% yield). LC/MS analysis showed the

same mass ([M+H+] 841 m/z) and HPLC retention time as synthetic cyclo(IW-

GIGCNP). The purity was �88% on the basis of integration of absorbance at

280 nm.

NMR Spectroscopy

NMR analysis of the putative cyclic peptide product cyclo(IWGIGCNP) was

done at 5 mM peptide in DMSO-d6 and a temperature of 25�C. 1H atoms

were assigned with COSY and TOCSY. TOCSY spectra were acquired with

an MLEV17 mixing sequence with a mixing time (tm) of 80 ms. 13C atoms (nat-

ural abundance) were assigned with HSQC and HMBC. ROESY was used to

determine the proximity of the Ile1-HN to Pro8-Ha atoms. The tm was

200ms. NMR spectra were collected on a Varian 600MHz instrument. Spectra

were referenced using 1H impurities in the DMSO solvent.

SUPPLEMENTAL INFORMATION

Supplemental Information includes four figures and can be found with this

article online at http://dx.doi.org/10.1016/j.chembiol.2014.10.015.

AUTHOR CONTRIBUTIONS

H.L. developed the transformation system, cloned and characterized AbPOPB

and GmPOPB, and made the GmPOPB knockout lines. S.-Y.H. expressed

GmPOPA and GmPOPB in yeast and performed cyclase assays. R.M.S. puri-

fied and determined the structure of the cyclo product. E.A. did cyclase as-

says. X.L. assisted with the transformation experiments. J.D.W. supervised

the work and wrote the manuscript.

ACKNOWLEDGMENTS

This research was supported by grant 1R01-GM088274 from the NIH General

Medical Sciences. Additional support came from grant DE-FG02-91ER20021

to the Plant Research Laboratory from the Division of Chemical Sciences, Ge-

osciences and Biosciences, Office of Basic Energy Sciences, Office of Sci-

ence, U.S. Department of Energy. We thank the Michigan State University

(MSU) Mass Spectrometry Core Facility for high-resolution MS, the MSU

Research Technology Support Facility for DNA sequencing, and the MSU

Max T. Rogers NMR facility for NMR analysis. Dan Holmes and Kermit John-

son (MSU Department of Chemistry) gave invaluable advice on interpretation

of NMR spectra. We thank Alejandro Pardo (Universidad Nacional de Quilmes

and Consejo Nacional de Investigaciones Cientıficas y Tecnicas, Buenos

Aires, Argentina) for plasmids and bacterial strains and John Scott-Craig

(MSU) and Gary Foster (University of Bristol, UK) for discussions and advice

about fungal transformation.

Received: September 9, 2014

Revised: October 8, 2014

Accepted: October 22, 2014

Published: December 4, 2014

REFERENCES

Agarwal, V., Pierce, E., McIntosh, J., Schmidt, E.W., and Nair, S.K. (2012).

Structures of cyanobactin maturation enzymes define a family of transamidat-

ing proteases. Chem. Biol. 19, 1411–1422.


Alshahni, M.M., Makimura, K., Yamada, T., Satoh, K., Ishihara, Y., Takatori, K.,

and Sawada, T. (2009). Direct colony PCR of several medically important fungi

using Ampdirect plus. Jpn. J. Infect. Dis. 62, 164–167.

Arnison, P.G., Bibb, M.J., Bierbaum, G., Bowers, A.A., Bugni, T.S., Bulaj, G.,

Camarero, J.A., Campopiano, D.J., Challis, G.L., Clardy, J., et al. (2013).

Ribosomally synthesized and post-translationally modified peptide natural

products: overview and recommendations for a universal nomenclature. Nat.

Prod. Rep. 30, 108–160.

Barber, C.J.S., Pujara, P.T., Reed, D.W., Chiwocha, S., Zhang, H., and

Covello, P.S. (2013). The two-step biosynthesis of cyclic peptides from linear

precursors in a member of the plant family Caryophyllaceae involves cycliza-

tion by a serine protease-like enzyme. J. Biol. Chem. 288, 12500–12510.

Bockus, A.T., McEwen, C.M., and Lokey, R.S. (2013). Form and function in

cyclic peptide natural products: a pharmacokinetic perspective. Curr. Top.

Med. Chem. 13, 821–836.

Bresinsky, A., and Besl, H. (1990). A Colour Atlas of Poisonous Fungi. (London,

United Kingdom: Wolfe).

Bushnell, D.A., Cramer, P., and Kornberg, R.D. (2002). Structural basis of tran-

scription: alpha-amanitin-RNA polymerase II cocrystal at 2.8 A resolution.

Proc. Natl. Acad. Sci. U S A 99, 1218–1222.

Craik, D.J., and Allewell, N.M. (2012). Thematic minireview series on circular

proteins. J. Biol. Chem. 287, 26999–27000.

Craik, D.J., and Malik, U. (2013). Cyclotide biosynthesis. Curr. Opin. Chem.

Biol. 17, 546–554.

Donia, M.S., Hathaway, B.J., Sudek, S., Haygood,M.G., Rosovitz, M.J., Ravel,

J., and Schmidt, E.W. (2006). Natural combinatorial peptide libraries in cyano-

bacterial symbionts of marine ascidians. Nat. Chem. Biol. 2, 729–735.

Enjalbert, F., Rapior, S., Nouguier-Soule, J., Guillon, S., Amouroux, N., and

Cabot, C. (2002). Treatment of amatoxin poisoning: 20-year retrospective

analysis. J. Toxicol. Clin. Toxicol. 40, 715–757.

Fulop, V., Bocskei, Z., and Polgar, L. (1998). Prolyl oligopeptidase: an unusual

b-propeller domain regulates proteolysis. Cell 94, 161–170.

Fulop, V., Szeltner, Z., Renner, V., and Polgar, L. (2001). Structures of prolyl oli-

gopeptidase substrate/inhibitor complexes. Use of inhibitor binding for titra-

tion of the catalytic histidine residue. J. Biol. Chem. 276, 1262–1266.

Giordanetto, F., and Kihlberg, J. (2014). Macrocyclic drugs and clinical candi-

dates: what canmedicinal chemists learn from their properties? J.Med. Chem.

57, 278–295.

Hallen, H.E., Luo, H., Scott-Craig, J.S., and Walton, J.D. (2007). Gene family

encoding the major toxins of lethal Amanita mushrooms. Proc. Natl. Acad.

Sci. U S A 104, 19097–19101.

Kaneko, H., Tomomasa, T., Inoue, Y., Kunimoto, F., Fukusato, T., Muraoka, S.,

Gonmori, K., Matsumoto, T., and Morikawa, A. (2001). Amatoxin poisoning

from ingestion of Japanese Galerina mushrooms. J. Toxicol. Clin. Toxicol.

39, 413–416.

Kemppainen, M.J., and Pardo, A.G. (2010). Gene knockdown by ihpRNA-trig-

gering in the ectomycorrhizal basidiomycete fungus Laccaria bicolor. Bioeng.

Bugs 1, 354–358.

Kemppainen, M.J., and Pardo, A.G. (2011). Transformation of the mycorrhizal

fungus Laccaria bicolor using Agrobacterium tumefaciens. Bioeng. Bugs 2,

38–44.

Koehnke, J., Bent, A., Houssen, W.E., Zollman, D., Morawitz, F., Shirran, S.,

Vendome, J., Nneoyiegbe, A.F., Trembleau, L., Botting, C.H., et al. (2012).

The mechanism of patellamide macrocyclization revealed by the characteriza-

tion of the PatG macrocyclase domain. Nat. Struct. Mol. Biol. 19, 767–772.

Koida, M., and Walter, R. (1976). Post-proline cleaving enzyme. Purification of

this endopeptidase by affinity chromatography. J. Biol. Chem. 251, 7593–

7599.

Lee, J., McIntosh, J., Hathaway, B.J., and Schmidt, E.W. (2009). Using marine

natural products to discover a protease that catalyzes peptide macrocycliza-

tion of diverse substrates. J. Am. Chem. Soc. 131, 2122–2124.

Li, M., Chen, C., Davies, D.R., and Chiu, T.K. (2010). Induced-fit mechanism for

prolyl endopeptidase. J. Biol. Chem. 285, 21487–21495.


http://dx.doi.org/10.1016/j.chembiol.2014.10.015

Chemistry & Biology



Li, P., Deng, W.Q., Li, T.H., Song, B., and Shen, Y.H. (2013). Illumina-based de

novo transcriptome sequencing and analysis of Amanita exitialis basidiocarps.

Gene 532, 63–71.

Li, P., Deng, W., and Li, T. (2014). The molecular diversity of toxin gene families

in lethal Amanita mushrooms. Toxicon 83, 59–68.

Luo, H., Hallen-Adams, H.E., and Walton, J.D. (2009). Processing of the

phalloidin proprotein by prolyl oligopeptidase from the mushroom Conocybe

albipes. J. Biol. Chem. 284, 18070–18077.

Luo, H., Hallen-Adams, H.E., Scott-Craig, J.S., and Walton, J.D. (2010).

Colocalization of amanitin and a candidate toxin-processing prolyl oligopepti-

dase in Amanita basidiocarps. Eukaryot. Cell 9, 1891–1900.

Luo, H., Hallen-Adams, H.E., Scott-Craig, J.S., and Walton, J.D. (2012).

Ribosomal biosynthesis of a-amanitin in Galerina marginata. Fungal Genet.

Biol. 49, 123–129.

May, J.P., and Perrin, D.M. (2007). Tryptathionine bridges in peptide synthesis.

Biopolymers 88, 714–724.

Nguyen, G.K.T., Wang, S., Qiu, Y., Hemu, X., Lian, Y., and Tam, J.P. (2014).

Butelase 1 is an Asx-specific ligase enabling peptide macrocyclization and

synthesis. Nat. Chem. Biol. 10, 732–738.

Oman, T.J., and van der Donk, W.A. (2010). Follow the leader: the use of leader

peptides to guide natural product biosynthesis. Nat. Chem. Biol. 6, 9–18.

Riley, R., Salamov, A.A., Brown, D.W., Nagy, L.G., Floudas, D., Held, B.W.,

Levasseur, A., Lombard, V., Morin, E., Otillar, R., et al. (2014). Extensive sam-

pling of basidiomycete genomes demonstrates inadequacy of the white-rot/

brown-rot paradigm for wood decay fungi. Proc. Natl. Acad. Sci. U S A 111,

9923–9928.

Saska, I., Gillon, A.D., Hatsugai, N., Dietzgen, R.G., Hara-Nishimura, I.,

Anderson, M.A., and Craik, D.J. (2007). An asparaginyl endopeptidase medi-

ates in vivo protein backbone cyclization. J. Biol. Chem. 282, 29721–29728.


Sgambelluri, R.M., Epis, S., Sassera, D., Luo, H., Angelos, E.R., and Walton,

J.D. (2014). Profiling of amatoxins and phallotoxins in the genus Lepiota by

liquid chromatography combined with UV absorbance and mass spectrom-

etry. Toxins (Basel) 6, 2336–2347.

Szeltner, Z., and Polgar, L. (2008). Structure, function and biological relevance

of prolyl oligopeptidase. Curr. Protein Pept. Sci. 9, 96–107.

Thevenet, P., Shen, Y., Maupetit, J., Guyon, F., Derreumaux, P., and Tuffery, P.

(2012). PEP-FOLD: an updated de novo structure prediction server for both

linear and disulfide bonded cyclic peptides. Nucleic Acids Res. 40, W288–

W293.

Umemura, M., Nagano, N., Koike, H., Kawano, J., Ishii, T., Miyamura, Y.,

Kikuchi, M., Tamano, K., Yu, J., Shin-ya, K., and Machida, M. (2014).

Characterization of the biosynthetic gene cluster for the ribosomally synthe-

sized cyclic peptide ustiloxin B in Aspergillus flavus. Fungal Genet. Biol. 68,

23–30.

Walton, J.D. (1996). Host-selective toxins: agents of compatibility. Plant Cell 8,

1723–1733.

Weigel, D., and Glazebrook, J. (2006). Transformation of Agrobacterium Using

Electroporation. (Cold Spring Harbor: Cold Spring Harbor Protocols).

Wieland, T. (1986). Peptides of Poisonous Amanita Mushrooms. (New York:

Springer).

Wong, C.T., Rowlands, D.K., Wong, C.H., Lo, T.W., Nguyen, G.K., Li, H.Y., and

Tam, J.P. (2012). Orally active peptidic bradykinin B1 receptor antagonists en-

gineered from a cyclotide scaffold for inflammatory pain treatment. Angew.

Chem. Int. Ed. Engl. 51, 5620–5624.

Zhou, P., Silverstein, K.A.T., Gao, L., Walton, J.D., Nallu, S., Guhlin, J., and

Young, N.D. (2013). Detecting small plant peptides using SPADA (Small

Peptide Alignment Discovery Application). BMC Bioinformatics 14, 335.


peptide macrocyclization catalyzed by a prolyl oligopeptidase involved in α-amanitin biosynthesis

Documents