Deciphering the late steps of rifamycin biosynthesis

Article abstract of DOI:10.1038/s41467-018-04772-x

Rifamycin-derived drugs, including rifampin, rifabutin, rifapentine, and rifaximin, have long been used as first-line therapies for the treatment of tuberculosis and other deadly infections. However, the late steps leading to the biosynthesis of the industrially important rifamycin SV and B remain largely unknown. Here, we characterize a network of reactions underlying the biosynthesis of rifamycin SV, S, L, O, and B. The two-subunit transketolase Rif15 and the cytochrome P450 enzyme Rif16 are found to mediate, respectively, a unique C-O bond formation in rifamycin L and an atypical P450 ester-to-ether transformation from rifamycin L to B. Both reactions showcase interesting chemistries for these two widespread and well-studied enzyme families.

Full text of DOI:10.1038/s41467-018-04772-x

A
R
T
I
C
L
E
DOI: 10.1038/s41467-018-04772-x
OPEN
D
e
o
c
i
y
p
n
h
t
e
h
r
i
n
g
s
t
h
e
l
a
t
e
s
t
e
p
s
o
f
r
i
f
a
m
y
c
i
n
b
i
s
e
s
i
1
2
1
1
2
1
2
2
F
e
i
f
e
i
Q
i
,
C
h
a
o
L
e
i
,
F
e
n
g
w
e
i
L
i
,
X
i
n
g
w
a
n
g
Z
h
a
n
g
,
J
i
n
W
a
n
g
,
W
e
i
Z
h
a
n
g
,
Z
h
e
n
F
a
n
,
W
e
i
c
h
a
o
L
i
,
Go n g - Li Tan g 3, Y ou l i X ia o
2, 4
, G uo p i ng Zhao 2, 4 & Sh engyi n g L i 1,4
Ri fam yc in - d er i ve d d ru g s, in cl udi ng ri fam pi n, ri fabu ti n , r ifa pe nt in e , an d r ifa xim in , h av e lo n g
b e e n u s e d a s fir st- li n e t he ra pie s fo r t he t re at m e nt of t ub er cu lo sis an d o t he r d ead ly in fe ct io n s.
Ho wev e r, th e l ate s te p s le a din g t o t he bi os yn th e si s of t h e in d us tr ial ly im po r ta nt r ifa myci n SV
an d B r e ma i n l ar g e ly un kn ow n. H e re , we c har act e ri ze a n et w or k o f r eac ti on s u nd e rly in g t h e
b io sy nt he s is o f r if amy ci n S V, S , L , O, an d B. T he t wo -su b un it t ra ns ket o las e Ri f15 an d t h e
cy to ch ro m e P 45 0 e nz ym e R if1 6 ar e fo und t o m e di at e, r es pe ct iv ely , a u ni qu e C –O b on d
f or mati o n i n r if am y cin L an d an at yp ica l P4 5 0 e s te r -t o -et h er t ra ns fo r mat io n f ro m r ifa myci n L
to B . Bo th r e act io n s s ho wc as e in te r e st in g c he m is t rie s f or t h ese t wo w id esp r ead an d w ell-
st u di ed e n zy me fa m il ie s .
1
Shandong Provincial Key Laboratory of Synthetic Biology, CAS Key Laboratory of Biofuels, Qingdao Institute of Bioenergy and Bioprocess Technology,
Chinese Academy of Sciences, Qingdao, Shandong 266101, China. 2 CAS Key Laboratory of Synthetic Biology, Institute of Plant Physiology and Ecology,
Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 200032 Shanghai, China. State Key Laboratory of Bio-Organic and Natural
Products Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, 200032 Shanghai, China. University of Chinese Academy of
3
4
Sciences, 100049 Beijing, China. These authors contributed equally: Feifei Qi, Chao Lei. Correspondence and requests for materials should be addressed to
Y.X. (email: ylxiao@sibs.ac.cn) or to S.L. (email: lishengying@qibebt.ac.cn)
N A TU RE C OMM U N IC AT I ON S | ( 2 0 1 8) 9 :2 3 4 2 | D OI : 1 0 . 1 03 8 / s 41 4 67 - 0 1 8- 0 4 7 7 2 - x | w ww . na tu r e . co m /n a tu r e co m m u n ic ati o n s
1

A
R
T
I
C
L
E
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
|
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
8
-
0
4
7
7
2
-
x
ifamycins are ansamycin antibiotics that show a wide biotransformation process that yield multiple highly potent
2
spectrum of antimicrobial activities against both Gram- clinical drugs . Consequently, strains that produce high levels of
positive and Gram-negative bacteria . Their semisynthetic R-SV are now preferred by industry (e.g., the well-studied mutant
derivatives such as rifampin, rifabutin, rifapentine, and rifaximin strain A. mediterranei U32). Nonetheless, the strain improvement
1
R
have been used for decades in the clinic for the treatment of for R-SV high-producers has not been as successful as that for R-
2
2
tuberculosis, leprosy, and AIDS-related mycobacterial infections , B producing strains , so both kinds of strains are still required by
and their recognized pharmacological mode-of-action is the industry. Notably, rifamycin S (R-S), rifamycin L (R-L), and
specific inhibition of prokaryotic DNA-dependent RNA rifamycin O (R-O) are also important intermediates that can be
synthesis2
,3
.
prepared from A. mediterranei fermentation cultures
11–14
.
Floss and co-workers discovered the first rifamycin biosyn- However, the biosynthetic relationship between these late rifa-
thetic gene cluster comprised of 34 genes in the bacterium mycin derivatives remains elusive, despite previous research
4
Amycolatopsis mediterranei S699 . Five type I polyketide syn- attempts based on extensive isotopic feeding and mutagenesis
2
,13,15
thases (PKSs: RifA-E) encoded by the cluster are responsible for experiments
assembling the first macrocyclic intermediate (proansamycin X) Recently, comparative analysis of the rifamycin biosynthetic
using 3-amino-5-hydroxybenzoic acid as the starter unit, and gene clusters of A. mediterranei S699 (an R-B producer) and U32
.
5
,6
malonyl-CoA and methylmalonyl-CoA as extender units
.
(an R-SV producer), together with corresponding genetic com-
Further post-PKS modifications, including the dehydrogenation plementation testing, strongly suggested that the cytochrome
of C-8 and the hydroxylation of C-34, lead to the intermediate P450 enzyme Rif16 is involved in the conversion of R-SV to R-
7
16
rifamycin W . Subsequently, Rif5 converts the Δ12,29 olefinic B . Furthermore, gene inactivation and complementation
bond of rifamycin W into a ketal moiety, followed by Rif20- experiments showed that Rif16 and the two-subunit transketolase
mediated acetylation of the hydroxyl group at C-25 and Rif14- Rif15 (encoded by rif15a and rif15b, two overlapping genes,
mediated O-methylation at C-27, producing rifamycin SV (R-SV) Supplementary Table 1) are essential and sufficient for this
2
,8–10
17
(
Supplementary Fig. 1)
from R-SV to the end product rifamycin B (R-B) (Fig. 1) have not mechanisms underlying the unusual ether bond formation
been biochemically characterized. between the C-4 phenolic hydroxyl group of R-SV and a glycolic
. The later biosynthetic steps leading transformation . However, the biochemical and chemical
During fermentation, R-B is the predominant rifamycin pro- acid moiety leading to R-B remain unknown. Here, by recon-
duct accumulated by wild type A. mediterranei and by the first stituting the in vitro activity of Rif15 and Rif16, we reveal a
11
industrial strains . However, as the antibacterial activity of R-B is biosynthetic network for the inter-conversion of R-SV, R-S, R-L,
modest, R-B needs to be transformed back to the more bioactive R-O, and R-B (Fig. 1), finally elucidating the mechanisms for the
R-SV before being subjected to the chemical, enzymatic, or late reactions of rifamycin biosynthesis.
H3CCOO
H3CO
H3CCOO
H3CO
OH
OH
OH
OH
OH
OH
OH
O
O
OH
1
X
1
NH
NH
O
O
O
4
4
O
OH
O
O
38
O
Hydrolysis
O
39 OH
Rifamycin SV (R–SV)
Rifamycin B (R–B)
H3CCOO
H3CO
OH
OH
OH
OH
NADP+
O
NADP+
1
NH
O2, M2+
[O]
O
4
O
NADPH
39
O
38
O
NADPH
OH
O
Fdx
Rifamycin L (R–L)
NADPH
FdR
Rif16
H3CCOO
H3CO
H3CCOO
H3CO
NADP+
OH
OH
OH
OH
OH
OH
O
2-ketose
O
O
O
1
NH
Hydrolysis
NH
O
4
O
O
O
O
O
O
38
O
O
39
O
Rifamycin S (R–S)
Rifamycin O (R–O)
Fig. 1 Th e bi os y nt h et ic ne tw or k of l at e r i fa myc in de r i va t i ve s . R- SV c an be oxi diz e d t o R-S spon ta ne ously in t he pr es e nce of dio xyg e n and div a le nt met a l
i o ns . Th e t r an s ke t ol as e R if 1 5 i s r e spo nsibl e fo r t r ans fe r r i ng a C ke t o- cont aini ng f rag me nt f rom a 2 -k e t ose t o R-S , g iv ing ri se t o R-L . T he P 4 5 0 en z y me Ri f 16
2
c a ta l y z es t he t r an s fo r ma t io n fr o m R- L t o R- O i n t he pr e se nc e o f N AD P H, f e rr ed oxin (F dx), and f e rr ed oxin re du ct a se (F dR). Fi nall y, R-O is no n-
en z y m a ti ca ll y r e duc e d t o R- B by N ADP H
2
N A TU RE CO MMUN I C A TI O N S | ( 2 0 1 8) 9: 2 3 4 2 | D OI : 1 0 . 1 0 3 8/ s 4 1 46 7 - 0 1 8- 0 4 7 7 2 - x | w ww . na tu r e . c o m /n atu r e c o m m u n ic ati o n s

NA TUR E COMM UNICAT IO NS | D OI: 1 0.1 0 38/ s4 1 467 - 01 8- 0 47 7 2 -x
A R T I C L E
a
b
R–L
R–B R–S R–SV
R–L
R–B R–S R–SV
R–O
ꢀ
= 425 nm
xi
ꢀ = 425 nm
i
xii
ii
iii
xiii
xiv
iv
v
xv
xvi
vi
vii
xvii
xviii
viii
ix
xix
xx
x
5
10
15
20
25
5
10
15
20
25
Time (min)
Time (min)
Fig. 2 HP L C a na l y si s of t he r e ac t io ns ca t al y ze d by Ri f 15 an d Ri f 16. a Re ac t ion s c at al yz e d by t he t ra nsk e to lase Ri f1 5 . (i ) T he mix e d R-L , R-B, R-S , a nd R- SV
2
+
s ta n d ar d s; ( ii ) R-S wi th R if 15 i n t he pr e se nc e of F -6 -P , T hDP an d Mg ; (i ii–vi ) t he ne g at iv e c ont rol s of (i i) wi t h t he omis sion of Ri f1 5 a (i ii), Ri f1 5 b ( iv) , T hD P
2
+
2 +
(
v ), o r Mg ( vi ); ( vi i) R -S V w it h Ri f 15 i n t he pr e se nc e of F -6 -P , ThD P and M g ; (v iii ) R-S V in re a ct i on buf f e r; (i x) R-S V in re a ct i on buf f e r wi t h t h e a dd it i o n
2+
o f 2 m M a s c or bi c a c id ; a n d (x ) R- SV w it h Ri f 15 i n t he pr e se nc e o f F- 6-P , T hD P , M g , and 2 mM asc or bic ac id. b Re ac t ion s c at al yz e d by t he cy t o chr ome
P 45 0 e nz y me R if 16 . ( xi ) The mi x ed R- L, R- B, R- S , an d R- SV st and ards; (x ii) T he f re shl y pr ep ared R-O aut he nt i c st and ard ; (x iii ) R-L wi t h Ri f16 i n t h e
p re s en c e of seF dx , seF dR , an d N ADP H ; (x i v) t he ne g at i ve co nt r ol of (x iii ) wi t h t he omis sion of N AD P H; (x v) c o-in je c t ion of (x iii ) wi t h 5 0 μM R- B; ( xvi )
R-L wi t h R if 16 in t he pr e s enc e of 2 0 mM H
2 2 2 2
O ; (x v i i) R- O i n r e act i on buf f e r; (x vii i) R-L wi t h Ri f1 6 in t he pr es e nce of 2 0 mM H O and 1 mM N AD P H ; ( xix )
R-L wi th 2 0 mM H ; ( xx ) R- O an d 1 mM N ADP H i n r e ac t io n buf f e r
2 2
O
(e.g., Cu2 , Mn , etc.) (Supplementary Fig. 6), similar to
+
2+
Results and Discussion
Initial examination of the enzymatic activity of Rif16. Gen- previously reported findings . However, we cannot exclude the
erally, P450 enzymes catalyze oxidative reactions
2
3
1
8,19
and trans- possibility that an oxidase might be responsible for enzymatic
ketolases can, in the presence of the essential cofactor thiamine oxidation of R-SV into R-S in vivo. Taken together, our results
diphosphate (ThDP), transfer a C2 keto-containing fragment suggest that Rif16, rather than performing a normal bio-
from a 2-ketose (e.g., fructose-6-phosphate (F-6-P), xylulose-5- oxidation, may catalyze an atypical P450 reaction in rifamycin
phosphate
(Xu-5-P),
ribulose-5-phosphate
(Ru-5-P), biosynthesis.
sedoheptulose-7-phosphate (S-7-P), dihydroxyacetone (DHA),
etc.) to the first carbon atom of an aldose (e.g., ribose 5-phos-
2
0
phate, glyceraldehyde 3-phosphate, etc.) . Based on the previous Functional characterization of Rif15. We next evaluated the
2
proposal that R-SV could be a biosynthetic precursor of R-B , we in vitro activity of Rif15a/Rif15b at a 1:1 ratio (i.e., the recon-
2
1
first surmised that Rif16 (CYP105G1 ) may oxidize R-SV to R-S stituted Rif15 transketolase) in the presence of R-S and F-6-P as
and that R-S (containing a C-4 keto group) may be a substrate of the potential C keto acceptor and donor, respectively, with ThDP
2
Rif15. To test these enzymatic hypotheses, we sub-cloned rif15a, and MgCl as cofactors. As predicted, Rif15 converted R-S into a
2
rif15b, and rif16 (Supplementary Fig. 2) and heterologously different product with higher polarity than R-B, while single
expressed these genes in Escherichia coli Codon Plus (DE3)-RIPL. subunits (that is, either Rif15a or Rif15b alone) were not able to
We then used Ni-NTA chromatography to purify the N-term- catalyze the same transformation. Additionally, we found that
2
+
inally His -tagged Rif15a, Rif15b, and Rif16 to homogeneity both ThDP and Mg were required for the catalytic activity of
6
(
Supplementary Fig. 3). Notably, the N-His -tagged Rif15a and Rif15 (Fig. 2a, trace i–vi). This is unsurprising since the dipho-
6
the non-tagged Rif15b were able to be co-expressed and co- sphate moiety of ThDP is bound to the transketolase through a
purified, suggesting a strong interaction between these two bivalent cation to form the catalytically active holo-enzyme from
2
4,25
Rif15 subunits (Supplementary Fig. 3).
the apo-protein
. Protein sequence alignment of multiple
Purified Rif16 appeared to be a functional P450 enzyme, as it transketolases shows that the residues involved in the interactions
had the expected red color and showed a signature peak at 450 with ThDP and the metal ion are highly conserved regardless
nm in its CO-reduced difference spectrum (Supplementary their origins and subunit organization modes (Supplementary
Fig. 4). To reconstitute the in vitro activity of Rif16, we used Fig. 7).
two surrogate redox partner proteins to shuttle electrons from
High performance liquid chromatography-high resolution
NADPH to the heme-iron reactive center for P450 catalysis: the mass spectrometry (HPLC-HRMS) analysis revealed that the m/
–
ferredoxin seFdx (SynPcc7942_1499) and the ferredoxin reduc- z value of the product was 754.3069 ([M-H] , deduced to be
–
tase seFdR (SynPcc7942_0978), both of which are from the [C H NO ] ) (Supplementary Fig. 8), which is consistent with
39
48
14
cyanobacterium strain Synechococcus elongatus PCC 7942 and that of R-L or R-B in negative ion mode (calc. 754.3069). Since
2
2
were here expressed heterologously in E. coli and purified . the retention time of this product was distinct from that of R-B,
1
13
Against our expectations, Rif16 was not able to catalyze the we suspected that the product here is R-L. Both the 1D ( H, C,
1
1
conversion from R-SV to R-S, while R-S was readily reduced to R- DEPT135) and 2D ( H- H COSY, HSQC, HMBC) NMR spectra
SV by addition of NADPH alone (Supplementary Fig. 5). (Supplementary Figs. 9–13, Supplementary Table 2) of the
Importantly, the hydroquinone R-SV was spontaneously oxidized purified product were acquired, and spectral comparisons of the
to the quinone R-S by ambient O , and this transformation was proton NMR data obtained from the product and the substrate R-
2
dramatically accelerated by the presence of divalent metal ions S (Supplementary Fig. 9) revealed a new set of CH signals (δ
2
H-39
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
|
(
2
0
1
8
)
9
:
2
3
4
2
|
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
8
-
0
4
7
7
2
-
x
|
w
w
w
.
n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
r
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
3

A
R
T
I
C
L
E
N AT URE CO M MU N ICAT I ON S | D O I: 10 . 10 38/ s414 67 -01 8- 047 7 2 - x
a
1
1
OH
O
O
1
*
CH2OH
C
H
H
*
OH
OH
H
H
H
C
OH
OH
H
*
NH
C
C
C
C
O
C
C
C
C
C
C
C
C
G-6-P isomerase
Hexokinase
HO
H
H
O
O
HO
H
HO
H
O
OH
OH
OH
OH
H
O
H
H
O
CH OPO 2–
2 3
CH OPO32–
2
CH2OH
R–S
F-6-P
Glucose
G-6-P
ThDP
Rif15a Rif15b
OH
E-4-P
OH
O
O
C
C
C
H
O
O
O
H
OH
O
OH
OH
O
NH
O
H
OH
NH
NH
CH2OPO32-
O
-O
O
R–S
O
O
N
1
*
CH2OH
N
*
CH2OH
N
*
CH2OH
N
N
38
O
*
C
C
OH
C
C
OH
C
C
OH
O
3
9
OH
C
S
C
C
C
C
O
C
C
C
C
C
OH
H
2
3
4
5
H+
S
S
S
S
O
HO
H
H
HO
CH2OH
CH2OH
R–L
ThDP
OH
H
H
OH
OH
*
*
H
OH
CH2OPO32–
6
CH2OPO32–
Fdx
FdR Rif16
(
CpdII)
Fe
IV
OH
H
O
H
O
OH
OH
OH
OH
O
OH
OH
OH
NH
O
NH
NH
NH
NH
O
NADPH
H
O
O
O
O
O
O
O
O
O
O
O
O
O
38
38
O
O
O
O
O
* O
O
Fe
39
O
O
39
O
O
O
O
IV
*
O
*
O
*
O
*
R–B
(CpdI)
R–O
b
ii
i
ii′
i′
iv
ii
ii
ii′
ii′
v
iii
97.0
95.0
93.0
f1 (ppm)
91.0
97
94
91
88
f1 (ppm)
65
62
71.0
65.5
f1 (ppm)
61.0
Fig. 3 P ut a t iv e me c ha n isms fo r Ri f 15 an d Ri f 16 su ppor te d by C -t ra ce r N M R re s ult s. a T he pr opose d c at al yt ic me c hani sms f or Ri f1 5 and Ri f 16. T he 13C
1
3
1
3
1 3
1 3
1 3
l
a
b
e
l
e
d
c
a
r
b
o
n
s
m
a
r
k
e
d
b
y
a
s
t
e
r
i
s
k
s
o
r
i
g
i
n
a
t
e
f
r
o
m
[
1
-
C
]
g
l
u
c
o
s
e
.
b
T
h
e
C
N
M
R
s
p
e
c
t
r
a
o
f
[
1
-
C
]
(
±
)
-
g
l
u
c
o
s
e
(
i
,
i
′
)
,
[
1
-
C
]
(
±
)
-
G
-
6
-
P
(
i
i
,
i
i
′
)
,
1
3
1 3
1
3
[1- C] F- 6-P ( ii i) in D
2 3
O, an d [39- C]R- L (i v ) an d [38- C]R- B (v) in CD O D . T he t ri ang le s ind ic at e t he c arb on si gna ls of re s idual g lyc e ro l d eri ve d f ro m
en z y m a ti c r e ac t i on bu f f e r
4
4
.63, d, J = 17.1 Hz, 1 H, δH-39 4.56, d, J = 17.1 Hz, 1 H vs δH-38 reaction product (Fig. 2a, trace x), establishing that the previously
.72, s, 2 H in R-B; δC-39 62.3 vs δC-38 67.8 in R-B) from the detected R-L was actually derived from the spontaneously formed
product. Further extensive analyses of 2D NMR correlations of R-S.
this methylene group confirmed it to be R-L. The steady-state
Collectively, our results from these in vitro assays demonstrate
kinetic constants of Rif15 were determined with K = 8.8 ± 2.4 μM that Rif15 is a two-component transketolase that transforms a
m
−2
−1
and kcat = (2.2 ± 0.1) × 10 min . Moreover, Xu-5-P, Ru-5-P, S- ketone (in R-S) into an ester (in R-L), a reaction that has not been
-P, and DHA were also able to serve as alternative C keto found previously in natural product synthesis, to the best of our
7
2
donors for Rif15, with Xu-5-P being optimal in terms of knowledge. Mechanistically, the deprotonation of ThDP at the
conversion ratios under the same conditions (Supplementary thiazolium ring generates a carbanion, which is responsible for
Fig. 14).
cleaving the C-2/C-3 bond in the 2-ketose. The resultant ThDP-
To examine whether R-SV is also a direct precursor of R-L as bound dihydroxyethyl group then undertakes nucleophilic attack
2
previously proposed , similar Rif15 reactions were performed of the C-4 carbonyl carbon of R-S, which is followed by bond
using R-SV as substrate, and, interestingly, we detected a small rearrangements and re-aromatization, ultimately yielding R-L
amount of R-L as a product (Fig. 2a, trace vii). In light of the (Fig. 3a).
spontaneous oxidation from R-SV to R-S by O that we had
2
observed in earlier assays (in aqueous solution in the presence of
2
+
Mg , Supplementary Fig. 6), 2 mM ascorbic acid was added to Biochemical, structural, and mechanistic characterization of
the Rif15 reactions to protect R-SV from oxidation (Fig. 2a, trace Rif16. Having characterized the R-SV→R-S→R-L transformation,
26
viii, ix) . Upon this addition, we no longer detected R-L as a we next sought to resolve the conversion of R-L into R-B. These
4
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
|
(
2
0
1
8
)
9
:
2
3
4
2
|
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
8
-
0
4
7
7
2
-
x
|
w
w
w
.
n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
r
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s

NA TUR E COMM UNICAT IO NS | D OI: 1 0.1 0 38/ s4 1 467 - 01 8- 0 47 7 2 -x
A R T I C L E
a
b
B′
B′
C
C
B
B
H
H
A
A
G
G
I
I
F
F
E
E
D
D
c
d
B′
P197
A196
R-L
L406
F308 F310
I255
G
T407
V303
N108
I107 I307
F
4.5 Å
T260
S76
F/G loop
Heme
C366
Fig. 4 S t ru c tu re s of R if 16 . a S ubs t ra t e -f re e Ri f 16. T he di so rd e re d B ′ loo p (wi t h 83- 97 re s idue s mis sing as show n wi t h t he das he d re d li ne ) t hat r ep la ce s t h e
t yp i c a l B ′ he li x is s ho wn in br ow n. T he F an d G he l ic e s ar e sh own in mag e nt a and or ang e , re s pe ct i ve ly . T he dis orde r e d FG loo p (19 2 -2 0 5 re si d ue s) l ac ks
el e ct ro n de ns i t y . Th e he me g ro up i s sh own as a st i c k i n r e d. b S ub st ra te -bou nd Ri f1 6. T he B ′ loo p in br own re ma ins dis orde r e d. T he FG loo p (he re o rde red )
i s s h own in bl ue . c S up er i mpos it i on of t he su bst r a te -fr e e (g r ay) and t he subs t rat e -b ound (l ime g re e n) Ri f1 6. T he bla ck ar row s poi nt out t he re g io ns t h at
u n de rg o c on f or mat io na l ch ang e s upo n R- L bi ndi ng . d T he ac t i ve si t e of Ri f1 6. T he subs t rat e R-L and he me ar e show n as st i ck s in ye l low and re d,
r es p ect i ve ly, wi th t he he m e- i r on de pi ct e d as a sp he r e. T he ke y r es idue s f ormi ng hydr og e n bonds (bl ac k das he d li ne s) wi t h R-L ar e show n as st i c ks i n g re en .
T he r e si du es wi th in 5 Å a r oun d t he su bst r a te t ha t co nst i t ute a la rg e hydr ophobic poc k e t ar e show n as li ne s in c yan. T he c onse r ve d T 2 60 an d C 366 a re
s h own a s s t ic k s in s il ve r . T he di st a nce s (i n an gs tr o ms) ar e i ndi cat e d by t he das he d ye l low li ne
two rifamycin derivatives have the same oxidation state, but we substrates (Supplementary Fig. 16). The missing electron density
still chose to test the activity of Rif16 against R-L, since this P450 of this region in both structures suggests the great structural
enzyme was previously shown to be required for R-B biosynth- flexibility. Both findings might help explain how Rif16 is able to
1
7
esis . Indeed, we found that R-L was significantly converted into accommodate its bulky substrate R-L, which represents one of the
R-B by Rif16 in the presence of seFdx, seFdR, and NADPH largest substrates for a P450 enzyme with the substrate-bound
2
8
(
Fig. 2b, trace xiii, xiv); the structure of the product was con- crystal structure available . In the absence of substrate, Rif16
firmed by the identical retention time of the product and the adopts an open conformation characterized by retraction of the F
R-B authentic standard, co-elution with R-B in a co-injection and G helices, loss of order in the B′ helix, and missing electron
experiment (Fig. 2b, trace xv), and the consistently observed density for the B′C and FG loops (Fig. 4a). A water molecule that
¯
m/z value of 754.3069 ([M-H] , calc.754.3069) (Supplementary is 2.5 Å away from the heme-iron forms the sixth axial ligand of
3
+
Fig. 15). These results clearly establish that Rif16 is the long- Fe (Supplementary Fig. 17a). Upon binding with R-L, Rif16’s
sought R-B synthase of rifamycin biosynthesis.
FG loop becomes ordered but the B’ helix and the B′C loop
To elucidate the catalytic mechanism for this atypical ester-to- remain disordered (Fig. 4b), thereby adopting a partially open
ether transformation, the crystal structures of substrate-free Rif16 conformation rather than the closed conformation observed in
(
PDB ID code: 5YSM, Fig. 4a) and R-L-bound Rif16 (PDB ID many substrate-bound P450 enzymes29,30 (Fig. 4c).
code: 5YSW, Fig. 4b) were solved at 1.90 Å and 2.60 Å resolution,
In the substrate-bound structure (Fig. 4d and Supplementary
respectively. In both of the structures, there was only one typical Fig. 17b), R-L forms hydrogen bonds with residues S76, N108,
cytochrome P450 fold existing in an asymmetric unit. The BB’ F308, and T407, and additionally interacts with residues I107,
loop-B’ helix-B’C loop region, which is known to be important A196, P197, I255, V303, P305, I307, F310, and L406 via
2
7
for substrate specificity determination , is significantly longer hydrophobic interactions (all of these residues are within 5 Å of
than those of many P450 enzymes that recognize smaller R-L). Critically, the axial water ligand is displaced and the
N
A
T
U
R
E
C
O
M
M
U
N
I
C
A
T
I
O
N
S
|
(
2
0
1
8
)
9
:
2
3
4
2
|
D
O
I
:
1
0
.
1
0
3
8
/
s
4
1
4
6
7
-
0
1
8
-
0
4
7
7
2
-
x
|
w
w
w
.
n
a
t
u
r
e
.
c
o
m
/
n
a
t
u
r
e
c
o
m
m
u
n
i
c
a
t
i
o
n
s
5

A
R
T
I
C
L
E
N AT URE CO M MU N ICAT I ON S | D O I: 10 . 10 38/ s414 67 -01 8- 047 7 2 - x
hydroxyl group at C-39 of R-L is closest (4.5 Å) to the heme-iron transformation from R-L to R-B (Supplementary Fig. 22). Since
reactive center. These structural features suggest a possible R84 is located at the B’ loop of Rif16 (Supplementary Fig. 16),
27
mechanism for R-B production (Fig. 3a): the ferryl-oxo species which is an important region for P450 substrate recognition , its
(Compound I) likely abstracts the hydrogen atom of the C-39 replacement by a tryptophan abolishes the productive substrate
hydroxyl group, leading to formation of a substrate radical and binding via a mechanism to be elucidated. Furthermore,
the ferryl-hydroxo Compound II. The resultant oxygen radical according to the biosynthetic network shown in Fig. 1, the U32
can then directly attack the neighboring arene to form a five- mutant should accumulate R-L instead of the observed R-SV and
1
6
membered ring pendant, and the radical would be delocalized to R-S . We reason that the ester R-L might be unstable, which
the aromatic ring. Next, the relocation of the spirocyclic could be hydrolyzed to R-SV either enzymatically or sponta-
intermediate could induce the second hydrogen abstraction from neously (Supplementary Fig. 23).
the C-1 hydroxyl group by Compound II. This diradical
Our elucidation of the network comprising the late steps of
mechanism might result in the formation of R-O. Notably, rifamycin biosynthesis revealed a unique C–O bond formation
similar mechanisms—involving two alternative substrate binding reaction mediated by a transketolase that involves both normal
poses being responsible for hydrogen abstractions from two C–C bond formation and unusual bond rearrangements. Notably,
distant sites—have been proposed for C–O coupling reactions transketolases primarily participate in central metabolic pathways
3
1,32
catalyzed by a number of P450 enzymes
. Finally, the such as pentose phosphate pathway and the Calvin cycle, and
pentabasic cyclic compound R-O could be reduced to R-B there have been few reports on transketolases that are involved in
3
6,37
(rather than R-L) by the NADPH-derived hydride, since the natural product biosynthesis
. The ether bond formation
carboxylic acid is a better leaving group than the alcohol.
derived from the concomitant oxidation-reduction reactions and
This R-L→R-B conversion reaction lacks net oxidation- complex bond rearrangements also represents a highly atypical
reduction. To dissect this unusual P450 reaction experimentally, reaction for a P450 reaction system. The knowledge on the slow
we elected to oxidize R-L by taking advantage of the peroxide kinetics and the optimal C keto donor of Rif15 could also help
2
1
8
shunt pathway of Rif16 , in which H O acts as the sole oxygen direct the future rational strain improvement. Finally, BLAST
2
2
and electron donor of Rif16; this approach allowed us to bypass searches demonstrate that there exist other protein sequences
the dual roles of NADPH from our previous reaction system (its with high similarity to Rif15 and Rif16 (Supplementary Table 3,
roles as an electron donor for the P450 enzyme and as a hydride Supplementary Fig. 24), suggesting that more Rif15-like and
provider for direct reduction of R-O). Interestingly, R-S was the Rif16-like functionality could be further identified. Some of these
dominant product from this reaction (Fig. 2b, trace xvi), which enzymes come from rifamycin producing microorganisms3
8–41
,
likely resulted from the spontaneous hydrolysis of the P450 which may suggest an effective method for discovery of more
product R-O1
5,33
(Fig. 2b, trace xvii). The addition of NADPH rifamycin producers by using Rif15 and Rif16 sequences as
into the Rif16/R-L/H O system led to predominant production probes.
2
2
of R-B (Fig. 2b, trace xviii, xix), as R-O can be reduced to R-B in
3
4,35
the presence of NADPH (Fig. 2b, trace xx)
. Furthermore, the
Methods
unstable compound R-O with the correct m/z value of 752.2920 Chemicals. Rifamycin SV and rifamycin O authentic standards were purchased
[M-H] , calc. 752.2924) was observed in a time-course study
Supplementary Fig. 18). These results strongly suggest that R-O
is the intermediate that enables the conversion of R-L to R-B.
–
from Sigma Aldrich (USA) and Toronto Research Chemicals (Canada), respec-
tively. Rifamycin S and rifamycin B authentic standards were bought from National
Institutes for Food and Drug Control (China).
(
(
To validate our proposed enzymatic reaction mechanisms, we
General DNA manipulation. The E. coli DH5α strain was used for plasmid con-
struction, storage, and isolation. Fast-digest restriction endonucleases (Thermo
1
3
performed a series of C-tracer NMR experiments. First, [39-
1
3
13
C]R-L was prepared by mixing [1- C]glucose, ATP, Fisher Scientific, USA) and T4 DNA ligase (Takara, Japan) were used for con-
2
+
TM
Mg , hexokinase, G-6-P isomerase, Rif15a/Rif15b, ThDP, and struction of vectors. PCR reactions were performed using I-5 2 × High-Fidelity
1
3
Master Mix DNA polymerase (TsingKe Biotech, Beijing, China). Plasmid isolations
from E. coli cells were performed using the Plasmid Miniprep Kit (TsingKe Bio-
tech, Beijing, China). Purification of DNA fragments from agarose gels or PCR
R-S in a one-pot reaction. We observed that [1- C]glucose was
1
3
phosphorylated to [1- C]G-6-P by hexokinase, which was
1
3
subsequently transformed into [1- C]F-6-P by G-6-P isomerase reactions was carried out using Gel Extraction Kit (Omega, USA) and Cycle Pure
1
3
(
Fig. 3). The Rif15-mediated transfer of the C-labeled glycolic Kit (Omega, USA), respectively. Primers were synthesized by TsingKe (China).
1
3
acid C moiety from [1- C]F-6-P to R-S resulted in production
2
1
3
of [39- C]R-L, with an enriched C-39 signal of δ 62.4 (Fig. 3b). Molecular cloning. The DNA sequences that encode the two-subunit transketolase
C
Rif15, and the separated subunits Rif15a and Rif15b were amplified from the
The identity of this product was further confirmed by LC-HRMS
−
genomic DNA of Amycolatopsis mediterranei U32 (A. mediterranei U32 was
deposited in Institute of Microbiology, Chinese Academy of Sciences designated as
CGMCC4.5720) under standard PCR conditions using the primer pairs of rif15-F/
analysis indicating an m/z value of 755.3106 ([M-H] , calc.
7
55.3105, Supplementary Fig. 19), which is ~1 Da greater than
−
that of unlabeled R-L [M-H] = 754.3069). Next, Rif16, seFdx/ rif15-R, rif15a-pSJ2F/rif15a-pSJ2R, and rif15b-F/rif15b-R, respectively (Supple-
mentary Table 4). The P450 gene rif16 and the mutant gene rif16R84W were
amplified from A. mediterranei S699 and U32 (Professor Guoping Zhao’s labora-
tory collection) gDNA, respectively, using a pair of primers rif16-F and rif16-R
seFdR, and NADPH were added into the above one-pot reaction.
1
3
As expected, the P450 enzyme converted [39- C]R-L into [38-
1
3
C]R-B, confirmed by LC-HRMS with an m/z value of 755.3100
(
Supplementary Table 4). The rif15a fragment was double digested by BamHI/
−
(
[M-H] , calc. 755.3105, Supplementary Fig. 19) and by our HindIII and cloned into the expression vector pSJ2 (a derivative of pET21a, which
13
observation that the C-labeled carbon signal shifted downfield is a gift from Professor Jiahai Zhou at Shanghai Institute of Organic Chemistry,
Chinese Academy of Sciences) using standard molecular cloning methods to
generate pSJ2-rif15a (Supplementary Fig. 2). The rif15, rif15b, rif16, and rif16R84W
PCR products were double digested by NdeI/HindIII, NdeI/EcoRI, NdeI/XhoI, and
from δC 62.4 to δC 67.9 (Fig. 3b); both analytical results are
consistent with the conversion of R-L to R-B via R-O (Fig. 3a).
It was previously reported that the R-SV high-producer A. NdeI/XhoI, respectively, and ligated into the corresponding pre-treated pET28b
1
6
mediterranei U32 has an R84W single mutation in Rif16 . The (Novagen, USA) to afford pET28b-rif15, pET28b-rif15b, pET28b-rif16, and
pET28b-rif16R84W (Supplementary Fig. 2). All constructs were confirmed by DNA
sequencing (Genewiz, China), and transformed into E. coli Codon Plus(DE3)-RIPL
for expression of N-terminal His -tagged recombinant proteins.
6
understanding of Rif16 mechanism allowed us to rationalize this
industrially important phenotype. Specifically, the dissociation
constant (K ) of R-L toward Rif16 was determined to be
d
1
.3 ± 0.1 μM (Supplementary Fig. 20), while the purified
Protein expression and purification. The E. coli Codon Plus(DE3)-RIPL trans-
formant carrying pSJ2-rif15a was grown at 37 °C overnight in LB media supple-
Rif16R84W mutant (Supplementary Figs. 3 and 21) showed no
detectable binding of R-L and lost the ability of catalyzing the mented with ampicillin (50 μg/mL), chloramphenicol (34 μg/mL), and
6
N A TU RE CO MMUN I C A TI O N S | ( 2 0 1 8) 9: 2 3 4 2 | D OI : 1 0 . 1 0 3 8/ s 4 1 46 7 - 0 1 8- 0 4 7 7 2 - x | w ww . na tu r e . c o m /n atu r e c o m m u n ic ati o n s

NA TUR E COMM UNICAT IO NS | D OI: 1 0.1 0 38/ s4 1 467 - 01 8- 0 47 7 2 -x
A R T I C L E
streptomycin (50 μg/mL). The transformant carrying pET28b-rif15, pET28b-rif15b,
pET28b-rif16 or pET28b-rif16R84W was cultured at 37 °C overnight in LB media
supplemented with kanamycin (50 μg/mL), chloramphenicol (34 μg/mL), and
streptomycin (50 μg/mL). The overnight seed culture was inoculated (1:100) into 1
L LB broth containing 10% glycerol, appropriate selective antibiotics and rare salt
UV-visible spectrum, and confirmed by co-injection experiments and LC-HRMS
analysis.
Kinetic analysis of Rif15. The kinetic analysis of Rif15 was carried out using 2–10
solution (6.75 mg/L FeCl
μg/l CuSO , and 125 μg/l H
.6–1 (~3–4 h). Next, 0.1 mM isopropyl β-D-thiogalactoside (IPTG, Sigma, USA)
3
, 500 μg/L ZnCl
2
, CoCl
2
, Na
2
MoO
4
, 250 μg/l CaCl
2
, 465
μM Rif15a/Rif15b, 10–100 μM R-S as substrate, 2 mM F-6-P as the C keto donor,
2
4
3
BO ), and then cultured at 37 °C until OD600 reached
3
2.5 mM MgCl , and 0.5 mM ThDP in 100 μL of reaction buffer. The reactions were
2
0
performed 15–30 min at 28 °C and quenched by mixing with the same volume of
was added to induce protein expression, and thiamin (1 mM, Sigma, USA) and δ-
aminolevulinic acid (1 mM, Sigma, USA) were supplemented to support efficient
expression of holo-form P450 enzymes. The cells were grown at 16 °C for 20 h. All
recombinant proteins including Rif15, Rif15a, Rif15b, Rif16, and Rif16R84W were
purified by Ni-NTA affinity chromatography . Briefly, E. coli cells were collected
by centrifugation at 5000 ×g for 5 min, the cell pellet was re-suspended with 20 mL
lysis buffer (20 mM Tris-HCl, 300 mM NaCl, 10% (w/v) glycerol, and 10 mM
imidazole, pH 8.0), and the cells were broken by sonication (5 s on/5 s off) for 30
min on ice. Next, the cellular debris was removed by centrifugation at 12,000 ×g for
methanol. After high-speed centrifugation (20,000 ×g) for 15 min, the supernatants
were analyzed on an Agilent 1260 infinity HPLC system as described above. Each
peak area of R-L was used to calculate the product concentration based on the
standard curve of R-L. The triplicated data were fit to the Michaelis–Menten
equation to determine the kcat and Km values using Origin 9.0.
4
2
Preparation of 13C labeled F-6-P, R-L, and R-B. To prepare [1-13C]-F-6-P, the
1
3
reaction mixture containing 5 mM [1- C] glucose (Sigma, USA), 6.25 mM MgCl2,
12.5 mM ATP, 1 U hexokinase (Sigma, USA), 1 U phosphoglucose isomerase
30 min. To the supernatant 1 mL nickel-nitrilotriacetic acid resin (Qiagen, Ger-
4
6,47
13
many) was added, and each mixture was incubated for 30 min with gentle shaking
at 4 °C. The resin was loaded onto an empty column and washed with wash buffer
(Sigma, USA) was incubated at 28 °C for 16 h
. To prepare the C-labeled R-L,
5 mM [1-1 C]-glucose, 8.75 mM MgCl , 12.5 mM ATP, 1 U hexokinase, 1 U
3
2
(
8
20 mM Tris-HCl, 300 mM NaCl, 10% (w/v) glycerol, and 20 mM imidazole, pH
.0) until no protein could be detected in flow-though. The proteins bound to resin
phosphoglucose isomerase, 10 μM Rif15a, 10 μM Rif15b, 200 μM R-S, and 0.5 mM
13
ThDP were mixed and incubated at 28 °C for 16 h. To prepare the C-labeled
rifamcyin B, [39-1 C] R-L was first prepared as described above; after 8 h, 2 μM
Rif16, 20 μM seFdx, 10 μM seFdR, and 1 mM NADPH were added into the reaction
mixture, and the reaction continued for 16 h at 28 °C. Notably, the one-pot reaction
was unsuccessful because NADPH would reduce R-S to R-SV, thus preventing the
Rif15 catalyzed reaction.
3
were eluted by 10 mL elution buffer (20 mM Tris-HCl, 300 mM NaCl, 10% (w/v)
glycerol, and 250 mM imidazole, pH 8.0). Finally, the eluents were concentrated
and buffer-exchanged to reaction buffer (20 mM Tris-HCl, 10% (w/v) glycerol, pH
7.4) via repeated ultrafiltration using Amicon Ultra-15 centrifugal filter units
(
Millipore, Ireland) with a 10-kDa cutoff. All protein purification steps were per-
formed at 4 °C. The SDS-PAGE analysis (Supplementary Fig. 3) showed that Rif15,
Rif15a, and Rif15b were >95% pure. However, the purity of Rif16 and Rif16R84W
were unsatisfactory. Therefore, the collected Rif16 and Rif16R84W proteins were
buffer-exchanged again to 20 mM Tris-HCl buffer (pH 8.0) by repeated ultra-
filtration using Amicon Ultra-15 centrifugal filter units (Millipore, Ireland) with a
Isolation and purification of R-L and R-B. The scaled-up Rif15 (for R-L) and
Rif16 (for R-B) reactions (20–100 mL) were respectively extracted by the same
volume of ethyl acetate five times. Then, the extracts were dried under nitrogen
3
5
0-kDa cutoff. The buffer-exchanged proteins were further loaded onto a Mono Q
/50 GL column (GE Healthcare, USA) pre-equilibrated with 20 mM Tris-HCl
flow, and re-dissolved in methanol. The purification of R-L or R-B was performed
TM
using semi-preparative HPLC (Waters XBridge
Prep C18 5 μm, 10 × 250 mm)
buffer (pH 8.0), and eluted with a gradient volume of 20 CV and an increasing
ionic strength up to 1 M NaCl at a flow rate of 1 mL/min using ÄKTA purifier-100
with a linear gradient of 40-80% acetonitrile in ddH O + 0.1% trifluoroacetic acid
2
over 20 min, and then 100% acetonitrile for 5 min at a flow rate of 2.5 mL/min. The
(
GE Healthcare, USA). Fractions containing proteins of interest were flash-frozen
collected fractions containing R-L or R-B were combined. The solvents were
removed using a Rotavapor R-3 rotary evaporator (Buchi, Switzerland) and N2
blowing. Finally, 10.0 mg R-L and 3.0 mg R-B with > 98.0% purity were obtained.
by liquid nitrogen and stored at −80 °C for later use.
Protein concentration determination. For Rif16 and Rif16R84W, the UV-visible
spectra were recorded on a DU 800 spectrophotometer (Beckman Coulter, USA).
The CO-bound reduced difference spectrum was employed to determine the
functional concentration of P450 enzymes using the extinction coefficient (ε450nm-
LC-HRMS analysis. The LC-HRMS analysis was performed on a Waters symmetry
column (4.6 × 150 mm, RP18) using the negative-mode electrospray ionization
with a linear gradient of 10–100% acetonitrile in ddH
2
O with 0.1% formic acid over
–
1
–143
4
90nm) of 91,000 M ·cm
2 2 4
. Briefly, CO was slowly bubbled through the Na S O
20 min, and followed by 100% acetonitrile for 5 min at a flow rate of 0.5 mL/min.
reduced P450 enzyme solution using a Pasteur pipette in a fume hood. The spectra
of ferric, CO-bound, and CO-bound reduced forms of the P450 enzyme were
recorded between 250 and 550 nm for generation of the CO-bound reduced dif-
ference spectrum. The protein concentrations of other proteins were determined
The high resolution mass spectra were recorded on a Dionex Ultimate 3000
coupled to a Bruker Maxis Q-TOF.
4
4
using the Bradford assay with bovine serum albumin as standard
.
NMR analysis. The 13C NMR spectra of 13C-labeled glucose, G-6-P, and F-6-P
13
were acquired using D
2
O as solvent. The 1H, C and 2D NMR spectra of rifa-
3
mycin derivatives were obtained using CDCl or MeOD as solvent on a Bruker
Avance III 600 MHz spectrometer with a 5 mm TCI cryoprobe.
Spectral substrate binding assays. Spectral substrate binding assays were carried
out on a UV-visible spectrophotometer 50 Bio (Cary, USA) at room temperature
by titrating 100 μM rifamycin L DMSO solution (blank DMSO for the reference
group) into 1 mL of 1 μM Rif16 or Rif16R84W solution in 1 μL aliquots, leading to
the substrate concentrations ranging from 0.1 to 1.2 μM. The series of Type I
difference spectra were used to deduce ΔA (Apeak(390 nm)–Atrough(420 nm)). Then, the
ΔA data versus substrate concentrations were fit to Michaelis–Menten equation to
Crystallization and structure determination. The crystal of Rif16 alone and the
complex crystal of Rif16 and R-L were obtained at 16 °C by hanging drop vapor
diffusion. The native Rif16 crystal screen droplets consisted of a 1:1 (v/v) protein at
16 mg/mL and the well solution of 100 mM Bis-Tris, pH 6.5, 200 mM magnesium
chloride hexahydrate, and 25% PEG3350. Rif16 and R-L were mixed at a molar
ratio of 1:5 and the co-crystallization was carried out in 200 mM magnesium
chloride hexahydrate, 100 mM Tris, pH 8.5, 30% PEG4000. Crystals appeared after
1 week and were ready for data collection in 20 days. The crystals were flash-cooled
in liquid nitrogen. The diffraction data were collected at 100 K under the syn-
chrotron radiation at beamline BL19U1 of the Shanghai Synchrotron Radiation
Facility (SSRF). The data sets were integrated and scaled with the HKL3000
package4 . The structure is determined by molecular replacement with the struc-
ture of CYP105A (PDB accession code 4OQS) as the initial search model with the
45
d
.
calculate the dissociation constant K
In vitro enzymatic assays of Rif15 and Rif16. The Rif15 reaction mixture con-
tained 10 μM Rif15a, 10 μM Rif15b, 200 μM R-S (or other rifamycin derivatives,
National Institutes for Food and Drug Control, China), 0.5 mM ThDP (Sigma,
2
USA), 2.5 mM MgCl , and 2 mM F-6-P (J&K Scientific, China) (or Xu-5-P, Ru-5-
P, S-7-P, DHA from Sigma), in 100 μL of reaction buffer (20 mM Tris-HCl, 10%
glycerol, pH 7.4). The Rif16 reaction assay was carried out in 100 μL of the same
reaction buffer containing 2 μM Rif16, 200 μM R-L, 20 μM seFdx plus 10 μM
seFdR, and 1 mM NADPH (Solarbio, China) or 2 μM Rif16, 200 μM R-L, and 20
8
program Phaser. The programs Refmac5 and Coot9 were used for the refinement
and model building4
9–52
. Ramachandran plots were generated with Coot9. The
2
mM H O
2
. The Rif15 and Rif16 reactions were incubated at 28 °C for 4 h and 1 h
(
unless otherwise specified), respectively, and quenched by mixing with the same
statistics for data processing and structure refinement are shown in Supplementary
Table 5. The coordinates were deposited to Protein Data Bank and the PDB ID
codes are 5YSM and 5YSW for substrate-free Rif16 and R-L-bound Rif16,
respectively. Figures were prepared using PYMOL (http://www.pymol.org).
volume of methanol. After high-speed centrifugation (20,000 ×g) for 15 min, the
supernatants were analyzed on an Agilent 1260 infinity HPLC system (Agilent
Technologies, USA) equipped with an ultraviolet detector. Compounds were
separated by SB-C18 reverse-phase column (Thermo, 5 μm, 46 mm, USA) in a
2
gradient system consisting of ddH O + 0.1% trifluoroacetic acid as solvent A and
acetonitrile as solvent B. The program of solvent gradient is as follow: 40% B for 5
min, 40-80% B over 15 min, and 80–100% B over 5 min at a flow rate of 1 mL/min.
The wavelength of detection was 425 nm for all rifamycin derivatives except for R-
O (370 nm). The peak identity in each HPLC trace was assigned by comparison
with authentic standards regarding the retention time and the
Data availability. Data that support the findings of this study have been deposited
in Protein Data Bank with the PDB ID codes 5YSM and 5YSW. Data referenced in
this study are available in GenBank with the accession codes AMED_0651,
AMED_0652, and AMED_0653. All other relevant data are available from the
corresponding authors.
N A TU RE C OMM U N IC AT I ON S | ( 2 0 1 8) 9 :2 3 4 2 | D OI : 1 0 . 1 03 8 / s 41 4 67 - 0 1 8- 0 4 7 7 2 - x | w ww . na tu r e . co m /n a tu r e co m m u n ic ati o n s
7

A
R
T
I
C
L
E
N AT URE CO M MU N ICAT I ON S | D O I: 10 . 10 38/ s414 67 -01 8- 047 7 2 - x
Received: 3 April 2018 Accepted: 12 May 2018
28. Rudolf, J. D., Chang, C. Y., Ma, M. & Shen, B. Cytochromes P450 for natural
product biosynthesis in Streptomyces: sequence, structure, and function. Nat.
Prod. Rep. 34, 1141–1172 (2017).
29. Poulos, T. L., Finzel, B. C. & Howard, A. J. High-resolution crystal structure of
cytochrome P450cam. J. Mol. Biol. 195, 687–700 (1987).
3
0. Sherman, D. H. et al. The structural basis for substrate anchoring, active site
selectivity, and product formation by P450 PikC from Streptomyces
venezuelae. J. Biol. Chem. 281, 26289–26297 (2006).
References
1
2
3
4
.
.
.
.
Riva, S. & Silvestri, L. G. Rifamycins: a general view. Ann. Rev. Microbiol. 26,
3
3
3
3
1. Tang, M. C., Zou, Y., Watanabe, K., Walsh, C. T. & Tang, Y. Oxidative
199–224 (1972).
cyclization in natural product biosynthesis. Chem. Rev. 117, 5226–5333
Floss, H. G. & Yu, T. W. Rifamycin-mode of action, resistance, and
biosynthesis. Chem. Rev. 105, 621–632 (2005).
Calvori, C., Frontali, L., Leoni, L. & Tecce, G. Effect of rifamycin on protein
synthesis. Nature 207, 417–418 (1965).
(
2017).
2. Grandner, J. M., Cacho, R. A., Tang, Y. & Houk, K. N. Mechanism of the
P450-catalyzed oxidative cyclization in the biosynthesis of griseofulvin. ACS
Catal. 6, 4506–4511 (2016).
3. Seong, B. L., Son, H. J., Mheen, T. I. & Han, M. H. Microbial transformation of
rifamycin B: a new synthetic approach to rifamycin derivatives. J. Antibiot.36,
August, P. R. et al. Biosynthesis of the ansamycin antibiotic rifamycin:
deductions from the molecular analysis of the rif biosynthetic gene cluster of
Amycolatopsis mediterranei S699. Chem. Biol. 5, 69–79 (1998).
Schupp, T., Toupet, C., Engel, N. & Goff, S. Cloning and sequence analysis of
the putative rifamycin polyketide synthase gene cluster from Amycolatopsis
mediterranei. FEMS Microbiol. Lett. 159, 201–207 (1998).
Tang, L., Yoon, Y. J., Choi, C. Y. & Hutchinson, C. R. Characterization of the
enzymatic domains in the modular polyketide synthase involved in rifamycin
B biosynthesis by Amycolatopsis mediterranei. Gene 216, 255–265 (1998).
White, R. J., Martinelli, E. & Lancini, G. Ansamycin biogenesis: studies on a
novel rifamycin isolated from a mutant strain of Nocardia mediterranei. Proc.
Natl. Acad. Sci. USA 71, 3260–3264 (1974).
1402–1404 (1983).
5
6
7
.
.
.
4. Han, M. H., Seong, B. L., Son, H. J. & Mheen, T. I. Rifamycin B oxidase from
Monocillium spp., a new type of diphenol oxidase. FEBS Lett. 151, 36–40
(
1983).
35. Liu, Z. et al. Reduction of quinones by NADH catalyzed by organoiridium
complexes. Angew. Chem. Int. Ed. 52, 4194–4197 (2013).
36. Peng, C. et al. Hijacking a hydroxyethyl unit from a central metabolic ketose
into a nonribosomal peptide assembly line. Proc. Natl. Acad. Sci. USA 109,
8540–8545 (2012).
3
7. Craig, J. W. & Brady, S. F. Discovery of a metagenome-derived enzyme that
produces branched-chain acyl-(acyl-carrier-protein)s from branched-chain
alpha-keto acids. Chembiochem 12, 1849–1853 (2011).
8
.
.
Floss, H. G. From ergot to ansamycins-45 years in biosynthesis. J. Nat. Prod.
6
9, 158–169 (2006).
9
Xu, J., Wan, E., Kim, C. J., Floss, H. G. & Mahmud, T. Identification of
tailoring genes involved in the modification of the polyketide backbone of
rifamycin B by Amycolatopsis mediterranei S699. Microbiology 151,
3
8. Wilson, M. C., Gulder, T. A. M., Mahmud, T. & Moore, B. S. Shared
biosynthesis of the saliniketals and rifamycins in Salinispora arenicola is
controlled by the sare1259-encoded cytochrome P450. J. Am. Chem. Soc. 132,
2515–2528 (2005).
12757–12765 (2010).
1
0. Xiong, Y., Wu, X. & Mahmud, T. A homologue of the Mycobacterium
tuberculosis PapA5 Protein, Rif-Orf20, is an aetyltransferase involved in the
biosynthesis of antitubercular drug rifamycin B by Amycolatopsis mediterranei
S699. Chembiochem 6, 834–837 (2005).
1. Oppolzer, W., Prelog, V. & Sensi, P. The composition of rifamycin B and
related rifamycins. Experientia 20, 336–339 (1964).
3
9. Saxena, A., Kumari, R., Mukherjee, U., Singh, P. & Lal, R. Draft genome
sequence of the rifamycin producer Amycolatopsis rifamycinica DSM 46095.
Genome Announc. 2, e00662-14 (2014).
4
0. Huang, H. et al. Micromonospora rifamycinica sp. nov., a novel actinomycete
from mangrove sediment. Int. J. Syst. Evol. Microbiol. 58, 17–20 (2008).
1. Promnuan, Y., Kudo, T., Ohkuma, M. & Chantawannakul, P. Actinomadura
apis sp. nov., isolated from a honey bee (Apis mellifera) hive, and the
reclassification of Actinomadura cremea subsp. rifamycini Gauze et al. 1987 as
Actinomadura rifamycini (Gauze et al. 1987) sp. nov., comb. nov. Int. J. Syst.
Evol. Microbiol. 61, 2271–2277 (2011).
1
1
1
4
2. Funayama, S. & Cordell, G. A. in Studies in Natural Products Chemistry Vol.
23 (ed. Rahman Atta, u. r.) 51–106 (Elsevier, 2000).
3. Lancini, G. C., Gallo, G. G., Sartori, G. & Sensi, P. Isolation and structure of
rifamycin L and its biogenetic relationship with other rifamycins. J. Antibiot.
(
Tokyo) 22, 369–377 (1969).
4
4
4
4
4
4
2. Du, L. et al. Characterization of a unique pathway for 4-cresol catabolism
1
4. Sugawara, S., Karasawa, K., Watanabe, M. & Hidaka, T. Production of
rifamycin O by Streptomyces 4107 A2. J. Antibiot. 17, 29–32 (1964).
5. Ghisalba, O., Roos, R., Schupp, T. & Nuesch, J. Transformation of rifamycin S
into rifamycins B and L. A revision of the current biosynthetic hypothesis. J.
Antibiot. 35, 74–80 (1982).
6. Zhao, W. et al. Complete genome sequence of the rifamycin SV-producing
Amycolatopsis mediterranei U32 revealed its genetic characteristics in
phylogeny and metabolism. Cell Res. 20, 1096–1108 (2010).
7. Yuan, H. et al. Two genes, rif15 and rif16, of the rifamycin biosynthetic gene
cluster in Amycolatopsis mediterranei likely encode a transketolase and a P450
monooxygenase, respectively, both essential for the conversion of rifamycin
SV into B. Acta Biochim. Biophys. Sin. 43, 948–956 (2011).
8. Zhang, X. & Li, S. Expansion of chemical space for natural products by
uncommon P450 reactions. Nat. Prod. Rep. 34, 1061–1089 (2017).
9. Guengerich, F. P. Common and uncommon cytochrome P450 reactions related
to metabolism and chemical toxicity. Chem. Res. Toxicol. 14, 611–650 (2001).
0. Kochetov, G. A. & Solovjeva, O. N. Structure and functioning mechanism of
transketolase. Biochim. Biophys. Acta 1844, 1608–1618 (2014).
initiated by phosphorylation in Corynebacterium glutamicum. J. Biol. Chem.
2
91, 6583–6594 (2016).
1
3. Omura, T. & Sato, R. The carbon monoxide-binding pigment of liver
microsomes II. Solubilization, purification, and properties. J. Biol. Chem. 239,
2379–2385 (1964).
1
4. Bradford, M. M. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye binding.
Anal. Biochem. 72, 248–254 (1976).
5. Li, S., Podust, L. M. & Sherman, D. H. Engineering and analysis of a self-
sufficient biosynthetic cytochrome P450 PikC fused to the RhFRED reductase
domain. J. Am. Chem. Soc. 129, 12940–12941 (2007).
1
6. Tsuboi, K. K., Estrada, J. & Hudson, P. B. Enzymes of the human erythrocyte.
IV. Phosphoglucose isomerase, purification and properties. J. Biol. Chem. 231,
9–29 (1958).
7. Hauck, T., Landmann, C., Bruhlmann, F. & Schwab, W. Formation of
-methyl-4-hydroxy-3[2H]-furanone in cytosolic extracts obtained from
1
1
2
2
2
1
5
Zygosaccharomyces rouxii. J. Agric. Food Chem. 51, 1410–1414 (2003).
8. Otwinowski, Z. & Minor, W. Methods in Enzymology. 276, 307–326
Academic Press, 1997).
9. McCoy, A. J. et al. Phaser crystallographic software. J. Appl. Crystallogr. 40,
58–674 (2007).
0. Murshudov, G. N., Vagin, A. A. & Dodson, E. J. Refinement of
macromolecular structures by the maximum-likelihood method. Acta
Crystallogr. D. 53, 240–255 (1997).
4
4
5
1. Nelson, D. R. The cytochrome P450 homepage. Hum. Genomics 4, 59–65
(
(
2009).
2. Ma, L. et al. Reconstitution of the in vitro activity of the cyclosporine-specific
P450 hydroxylase from Sebekia benihana and development of a heterologous
whole-cell biotransformation system. Appl. Environ. Microbiol. 81, 6268–6275
6
(
2015).
2
2
3. Scrutton, M. C. Divalent metal ion catalysis of the oxidation of rifamycin SV
to rifamycin S. FEBS Lett. 78, 216–220 (1977).
51. Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development
of Coot. Acta Crystallogr. D. 66, 486–501 (2010).
4. Lindqvist, Y., Schneider, G., Ermler, U. & Sundström, M. Three-dimensional
structure of transketolase, a thiamine diphosphate dependent enzyme, at 2.5 Å
resolution. EMBO J. 11, 2373–2379 (1992).
52. Collaborative Computational Project, Number 4. The CCP4 suite: programs
for protein crystallography. Acta Crystallogr. D. 50, 760–763 (1994).
2
2
2
5. Kochetov, G. A. Structure of the active center of transketolase*. Ann. N. Y.
Acad. Sci. 378, 306–311 (1982).
Acknowledgements
6. Dampier, M. F., Chen, C. W. & Whitlock, H. W. Jr. Substituent effects on the
solution conformation of rifamycin S. J. Am. Chem. Soc. 98, 7064–7069 (1976).
7. Gotoh, O. Substrate recognition sites in cytochrome P450 family 2 (CYP2)
proteins inferred from comparative analyses of amino acid and coding
nucleotide sequences. J. Biol. Chem. 267, 83–90 (1992).
This work was supported by the Shandong Provincial Natural Science Foundation
(
(
ZR2017ZB0207 to W.Z. and S.L.), the National Natural Science Foundation of China
grants 81741155, 31422002, and 21472204 to S.L., 31600036 to F.Q., 21572243 to Y.X.,
3
1430004 and 31670058 to G.Z.), Chinese Academy of Sciences (grants QYZDB-SSW-
SMC042 to S.L. and XDPB0402 to Y.X.), and the Science and Technology Commission of
8
N A TU RE CO MMUN I C A TI O N S | ( 2 0 1 8) 9: 2 3 4 2 | D OI : 1 0 . 1 0 3 8/ s 4 1 46 7 - 0 1 8- 0 4 7 7 2 - x | w ww . na tu r e . c o m /n atu r e c o m m u n ic ati o n s

NA TUR E COMM UNICAT IO NS | D OI: 1 0.1 0 38/ s4 1 467 - 01 8- 0 47 7 2 -x
A R T I C L E
Shanghai Municipality (grant 15JC1400402 to Y.X.). We thank the staff of BL19U1
beamline at the Shanghai Synchrotron Radiation Facility for assistance during X-ray
diffraction data collection. We are also grateful to Dr. Gong Chen at Nankai University
for helpful discussions about reaction mechanisms.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
Author contributions
F.Q., C.L., J.W., G.-L.T., Y.X., G.Z., and S.L. designed research; F.Q., C.L., F.L., X.Z., J.W.,
W.Z., Z.F., and W.L. performed research; F.Q., C.L., F.L., X.Z., J.W., G.-L.T., Y.X., G.Z.,
and S.L. analyzed results; F.Q., C.L., F.L., X.Z., Y.X., G.Z., and S.L. wrote the manuscript;
All authors read and approved the manuscript.
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. The images or other third party
material in this article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons license and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/.
Additional information
Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467-
0
18-04772-x.
Competing interests: The authors declare no competing interests.
Reprints and permission information is available online at http://npg.nature.com/
© The Author(s) 2018
reprintsandpermissions/
N A TU RE C OMM U N IC AT I ON S | ( 2 0 1 8) 9 :2 3 4 2 | D OI : 1 0 . 1 03 8 / s 41 4 67 - 0 1 8- 0 4 7 7 2 - x | w ww . na tu r e . co m /n a tu r e co m m u n ic ati o n s
9