Chemoenzymatic synthesis of novel C-ribosylated naphthoquinones

Article abstract of DOI:10.1021/cb400384c

The biological activity of many natural products is dependent on the presence of carbohydrate units, which are usually attached via an O-glycosidic linkage by glycosyltransferases. Recently, an exceptional C-ribosylation event was discovered in the biosynthesis of the polyketide antibiotic alnumycin A. The two-step process involves initial attachment of d-ribose-5-phosphate to the polyaromatic aglycone by the C-glycosynthase AlnA and subsequent dephosphorylation by AlnB, an enzyme of the haloacid dehalogenase family. Here, we tested 23 unnatural substrates to probe the C-ribosylation reaction. The chemoenzymatic synthesis of C-ribosylated juglone, 7-methyl juglone, monomethyl naphthazarin, 8-chloro-7-methyl juglone, and 9-hydroxy-1,4-anthraquinone revealed the importance of a 1,4-quinoid system with an adjacent phenolic ring in order for reaction to occur. To further rationalize the molecular basis for reactivity, factors governing substrate recognition were investigated by NMR binding experiments. Additionally, the suitability of substrates for nucleophilic substitution was assessed by molecular modeling using density functional theory (DFT) calculations.

Full text of DOI:10.1021/cb400384c

Subscriber access provided by University of South Dakota
Letter
Chemoenzymatic Synthesis of Novel C-Ribosylated Naphthoquinones
Bastian Blauenburg, Terhi Oja, Karel D. Klika, and Mikko Metsä-Ketelä
ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/cb400384c • Publication Date (Web): 09 Sep 2013
Downloaded from http://pubs.acs.org on September 15, 2013
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Chemical Biology is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 22
ACS Chemical Biology
1
1
2
Title
3
4
5
6
Chemoenzymatic Synthesis of Novel CꢀRibosylated Naphthoquinones
7
8
9
Authors
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Bastian Blauenburg,†,‡,§ Terhi Oja,† Karel D. Klika,*,‡,∥ and Mikko MetsäꢀKetelä*,†
Affiliations
†Departments of Biochemistry and Food Chemistry and ‡Chemistry, University of Turku,
FINꢀ20014 Turku, Finland
§Department of Chemistry, Biochemistry, PhilippsꢀUniversity Marburg, HansꢀMeerweinꢀ
Straße, Dꢀ35032 Marburg, Germany
∥Molecular Structure Analysis, German Cancer Research Center (DKFZ), Dꢀ69009
Heidelberg, Germany
* Authors for correspondence
Contact
Mikko MetsäꢀKetelä, Eꢀmail: mikko.mk@gmail.com, Tel. +35823336847, Fax.
+35823336860; Karel D. Klika, Eꢀmail: klikakd@yahoo.co.uk, Tel: +35823336804, Fax:
+35823336700.
ACS Paragon Plus Environment

ACS Chemical Biology
Page 2 of 22
2
1
2
Abstract
3
4
5
6
7
8
The biological activity of many natural products is dependent on the presence of carbohydrate
units, which are usually attached via an Oꢀglycosidic linkage by glycosyltransferases.
Recently, an exceptional Cꢀribosylation event was discovered in the biosynthesis of the
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
polyketide antibiotic alnumycin A. The twoꢀstep process involves initial attachment of
riboseꢀ5ꢀphosphate to the polyaromatic aglycone by the Cꢀglycosynthase AlnA and
Dꢀ
subsequent dephosphorylation by AlnB, an enzyme of the haloacid dehalogenase family. Here
we tested 23 unnatural substrates to probe the Cꢀribosylation reaction. The chemoenzymatic
synthesis of Cꢀribosylated juglone, 7ꢀmethyl juglone, monomethyl naphthazarin, 8ꢀchloroꢀ7ꢀ
methyl juglone and 9ꢀhydroxyꢀ1,4ꢀanthraquinone revealed the importance of a 1,4ꢀquinoid
system with an adjacent phenolic ring in order for reaction to occur. To further rationalize the
molecular basis for reactivity, factors governing substrate recognition were investigated by
NMR binding experiments. Additionally, the suitability of substrates for nucleophilic
substitution was assessed by molecular modeling using density functional theory (DFT)
calculations.
ACS Paragon Plus Environment

Page 3 of 22
ACS Chemical Biology
3
1
2
3
4
5
6
7
8
9
Many medically important natural products, such as the anticancer agent doxorubicin,1 the
antibiotic erythromycin2 and the anthelmintic drug avermectin,3 owe their activity to the
carbohydrate moieties attached to the polyketide framework.4 The formation of these units
has been investigated in detail and the studies have clarified that the 6ꢀdeoxy sugars are
modified as nucleoside diphosphates prior to attachment by glycosyltransferases.5,6 The most
common means of attachment is via an Oꢀglycosidic linkage, which unfortunately makes the
metabolites susceptible to hydrolysis.7 Chemically more robust Cꢀglycosylation8 has been
reported, e.g. during the biosynthesis of urdamycin, granaticin and medermycin,9,10 but this
mode of attachment is comparatively rare.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The aromatic polyketide alnumycin A (1, Figure 1) contains an exceptional 4′ꢀhydroxyꢀ5′ꢀ
hydroxymethylꢀ2′,7′ꢀdioxane unit which mimics the Cꢀglycosylated carbohydrates.11,12 In
fact, alnumycin A consists of six stereoisomeric compounds that differ in their configurations
within the dioxane unit.13 The pathway responsible for the biosynthesis of the dioxane moiety
has previously been elucidated in vivo14,15 and reconstructed in vitro,15 which confirmed that
the whole dioxane unit is derived from
Dꢀriboseꢀ5ꢀphosphate. In the initial step, AlnA
catalyzes the attachment of ꢀriboseꢀ5ꢀphosphate to the prealnumycin aglycone (2, Figure 1)
D
via a C8–C1′ bond. The enzyme is homologous only to pseudouridine glycosidases such as
YeiN, which is responsible for the conversion of pseudouridine 5′ꢀphosphate to uracil and
riboseꢀ5ꢀphosphate.16 The AlnA reaction products included an anomeric pair, denoted as
Dꢀ
alnumycin P1 (3, 84%, Figure 1) and P2 (16%), which differed only in the configuration at Cꢀ
1′.17 The overall Cꢀribosylation is completed by dephosphorylation of the intermediates by
AlnB, an enzyme of the haloacid dehalogenase superfamily,18 resulting in the formation of the
anomers alnumycin C1 (4, 76%, Figure 1) and C2 (24%).
ACS Paragon Plus Environment

ACS Chemical Biology
Page 4 of 22
4
1
2
3
Structural and mechanistic studies of AlnA have recently suggested that the attachment of Dꢀ
4
5
6
7
8
9
riboseꢀ5ꢀphosphate proceeds through a Michaelꢀtype 1,4ꢀaddition.17 The spacious active site
of AlnA is composed of two compartments which are separated by a triad of catalytically
important residues: E29, K86 and K159 (Figure 1). Crystallization experiments have further
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
revealed that the phosphosugar coꢀsubstrate binds in a linear configuration in one of the clefts,
while the other compartment, which is predominantly hydrophobic, is the expected binding
site of 2. The reaction is likely to proceed through initial deprotonation of Dꢀriboseꢀ5ꢀ
phosphate by E29 and formation of an enolate, which enables the nucleophilic 1,4ꢀaddition to
2 (Figure S1). The conversion to 3 then occurs via hemiꢀketal formation and restoration of the
chromophore system. In contrast to the mechanistic proposal for AlnA, crystallographic
snapshots of reaction intermediates of the pseudouridine glycosidase YeiN identified a lysine
adduct during catalysis (Figure S1),19 but the 100ꢀfold disparity in activities of lysine mutants
of YeiN19 and AlnA17 alludes that the two enzymes may differ in their respective
mechanisms. The observation that AlnA is able to utilize Dꢀribuloseꢀ5ꢀphosphate as an
alternative substrate (Figure S1) also yields support to the mechanism involving a Michaelꢀ
type 1,4ꢀaddition.
In turn, determination of the crystal structure of AlnB (Figure 1) revealed that the twoꢀdomain
enzyme is composed of a core domain, which provides the catalytically important residues
D15 and D17, and a cap domain that is utilized for substrate recognition.17 Structure/function
studies confirmed that the dephosphorylation reaction is likely to follow a nucleophilic course
in tandem with acid/base catalysis, a fact well established for the HAD enzyme
superfamily.17,18
ACS Paragon Plus Environment

Page 5 of 22
ACS Chemical Biology
5
1
2
3
4
5
In this paper, we have continued the characterization of the unusual twoꢀstep Cꢀribosylation
process by probing the reaction using 23 potential substrates. The results show a strict
6
7
8
9
requirement for a 1,4ꢀquinoid system with an additional adjacent phenolic ring, which was the
common denominator in the five converted substrates. Binding experiments by NMR suggest
that the phenolic hydroxyl group may be required for correct orientation of the substrate in the
active site of AlnA, while computational chemistry calculations highlight the importance of a
suitable charge at the reacting carbon of the substrate amenable to nucleophilic substitution.
Finally, the data demonstrated that AlnB also appeared to exhibit promiscuity towards
alternative substrates as the enzyme was able to convert all of the phosphorylated AlnA
reaction products to the final Cꢀribosylated compounds.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We initially tested various α,βꢀunsaturated carbonyl compounds (Figure 2), which should be
suitable Michael acceptorꢀtype molecules, as substrates. The set included linear molecules
and compounds containing a single sixꢀmembered ring (5–10), naphthalene derivatives (11–
21) with various hydroxy, dehydro, halo and dioxo substituents, 7ꢀethoxy coumarin (22) and
several 9,10ꢀ and 1,4ꢀanthraquinones (23–27). Recombinant AlnA and AlnB were produced
in E. coli and purified as described previously15 and incubated in the presence of the substrate
analogs and Dꢀriboseꢀ5ꢀphosphate. While none of the single ring compounds (5–9), including
uracil (10), were accepted as substrates, analysis by HPLC indicated that juglone (15) was
converted to a product with a distinctly different retention time but a similar UVꢀvis
spectrum. Scaling up the reaction provided sufficient material for structure elucidation by 1H
and 13C NMR and HRꢀESIꢀMS (see Supporting Information). NMR indicated the sample to
actually consist of a pair of novel stereoisomeric compounds differing in their configuration at
Cꢀ1′, denoted as 3ꢀ(βꢀribosyl) juglone (28a, 96.5%, Figure 2) and its α anomer (28b, 3.5%).
These initial experiments revealed the importance of a 1,4ꢀquinone substructure fused
ACS Paragon Plus Environment

ACS Chemical Biology
Page 6 of 22
6
1
2
3
laterally to another ring for reaction to occur since none of the nonꢀquinoid twoꢀring
4
5
6
compounds (11, 12 and 22) nor the 9,10ꢀanthraquinones (23–26) were accepted as substrates.
7
8
9
After these initial experiments, we probed the boundaries of substrate recognition of AlnA
and AlnB in more detail by focusing on compounds resembling 15. Firstly, substitutions at Cꢀ
2 next to the reactive carbon of the polyaromatic substrates appeared to be detrimental to the
activity as neither plumbagin (21) nor 2,3ꢀdichloroꢀ1,4ꢀnaphthoquinone (14) could be
converted to Cꢀribosylates. Secondly, the native substrate 2 contains a heterocyclic 2Hꢀpyran
ring as the third sixꢀmembered ring. In order to test the tolerance of AlnA and AlnB for
variation in this part of the substrate, we tested 7ꢀmethyl juglone (16) and 9ꢀhydroxyꢀ1,4ꢀ
anthraquinone (27). Both of these were found to be suitable substrates and were converted
into the expected novel Cꢀribosylates 29 and 30 (Figure 2), respectively. NMR analysis
(Supporting Information) confirmed that in the case of 29, both Cꢀ1′ anomers were present,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
with the
β
anomer again predominating (96.5%). For 30, the low sample amount precluded
anomer. Thirdly, a free
the observation of a minor anomer, but the one present was the
β
phenolic ring adjacent to the quinone ring was estimated to be requisite for the reaction, since
surprisingly neither 1,4ꢀnaphthoquinone (13) nor Oꢀmethyl juglone (20) were converted. In
addition, naphthazarin (18), possessing a benzeneꢀ1,4ꢀdiol substructure, was also found not to
be a suitable substrate for AlnA/AlnB. However, other compounds with substitutions in the 8ꢀ
position of the phenolic ring appeared to be tolerated to some extent, since new reaction
products could be observed when monomethyl naphthazarin (19) and 8ꢀchloroꢀ7ꢀmethyl
juglone (17) were used as substrates. However, the conversion rate of these two compounds
was drastically reduced in comparison to 15 as estimated by HPLC (Table 1), which limited
characterization of the reaction products to 1H NMR (Supporting Information) and only
ACS Paragon Plus Environment

Page 7 of 22
ACS Chemical Biology
7
1
2
3
tentative assignment of the structures of the corresponding Cꢀribosylates 31 and 32,
4
5
6
respectively, could be conferred.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Thus, the experiments revealed a complex set of rules required for reactivity by AlnA and
AlnB. In order to comprehend the underlying causes of these restrictions, we next
experimentally assessed (i) the relative strength of substrate binding to AlnA by linewidth
broadening (
ν
) of the 1H NMR signals due to the presence of the enzyme20 and (ii) the
½
binding epitope of the substrate to AlnA by waterLOGSY20 NMR experiments. In addition,
we utilized molecular modeling using density functional theory (DFT) calculations to (iii)
quantify the atomic charge at the reactive site Cꢀ3 (QC3) and (iv) evaluate the energy
difference (ꢀꢀG) between the substrate and a postulated enol intermediate (Supporting
Information).
Estimation of the strength of binding (Table 1) revealed that while 15 was bound strongly,
this was not an absolute requirement for reactivity, since 16, which could also be converted
into the corresponding Cꢀribosylate 29, was found to bind only weakly to AlnA. Perhaps more
surprisingly, the nonꢀreacting 18 appeared to bind to AlnA strongly, and even 13 and 20 were
better in this respect than 16. The waterLOGSY results (Table 1) indicated that 15 was wholly
embedded in the active site as indicated by the even intensities for all of the proton
resonances. A similar result was also obtained for two of the other accepted substrates that
were suitable for the binding studies, 16 and 19, although in the case of 19 the signal
originating from the methoxy group was weaker. Of the nonꢀreacting substrates, 21 displayed
even binding for the whole molecule, except for the methyl group, suggesting that the binding
mode of the substrate could be correct, especially since the Hꢀ3 resonance at the reacting
ACS Paragon Plus Environment

ACS Chemical Biology
Page 8 of 22
8
1
2
3
4
5
6
7
8
carbon yielded a clear signal. This was not the case with 13, which gave weak signals
particularly for Hꢀ2 and Hꢀ3 indicating an incorrect mode of binding, and with 20 where the
signal from the methoxy group was weakened.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Since the binding studies could not comprehensively explain the nonꢀreactivity of all of the
compounds (e.g. 18 and to some extent 21), we complemented the experimental results with
DFT calculations in order to elucidate chemical properties that might prohibit the reaction.
The suitability of the molecules to act as Michael acceptors was assessed by the atomic
charge at Cꢀ3 (Table 1). For most of the substrates, QC3 values were similar (−0.17 to −0.18),
but a much higher electron density (−0.25) was obtained for 21 which could explain why the
compound was not converted. In addition, the proposed reaction mechanism assumes the
formation of a highꢀenergy enol intermediate. This prerequisite was simulated in the DFT
calculations in the absence of the enzyme by comparison of the reactant substrate relative to a
Cꢀ3 methylated enol intermediate (see Supporting Information). The differences in the Gibbs'
free energies (ꢀꢀG) for the substrate reactions relative to 2 (ꢀꢀG, Table 1) were then
compared to evaluate the thermodynamic viabilities of each reaction. The calculations
indicated that all reacting substrates had ꢀꢀG values in the range of −0.74–4.62 kcal mol−1, as
did many of the unreactive naphthoquinone substrates except for 18 and 21 which had values
in excess of 7 kcal mol−1. Interestingly, 2,3ꢀdichloronaphthoquinone (14) yielded a reasonable
ꢀꢀG value of 1.28 kcal mol−1.
Hence, it would seem that various naphthoquinones can be utilized as substrates for AlnA and
AlnB but that several conditions must be met for reaction to occur (Figure 3 and Table 1).
Firstly, the substrate must bind to AlnA, and bind in the correct orientation though it appears
that strong binding per se is not an absolute requirement based on the results of 16. For these
ACS Paragon Plus Environment

Page 9 of 22
ACS Chemical Biology
9
1
2
3
4
5
6
7
8
9
conditions to be fulfilled, a hydroxyl group at Cꢀ5 (Cꢀ9 in 27) seems obligatory as this
provides an anchor point and may also be essential for placing the substrate in the correct
orientation. Based on the crystal structure of AlnA and docking experiments with 2, the
contact point may be the backbone amide group of L32, which is within hydrogen bonding
distance to the hydroxyl group and one of the quinone oxygens (Figure 1). A particularly
illustrative example of this is 20, which binds to AlnA and has suitable chemical
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
characteristics for reactivity, but which appears to bind in an incorrect orientation due to the
methoxy group at Cꢀ5. Secondly, the substrate must possess suitable reactivity at Cꢀ3 that is
amenable to nucleophilic substitution, and high electron density may inhibit the reaction. For
example, 21 has a much higher calculated atomic charge on Cꢀ3 than all of the reacting
substrates. Thirdly, it would appear that some of the substrates tested would result in a highꢀ
energy intermediate, which may prohibit the reaction. Thus, substrates that were initially
expected to react but did not despite having a hydroxyl present at Cꢀ5, e.g. 18 and 21 were
found to have much higher ꢀꢀG values compared to substrates that did react, e.g. 2 and 15
(Table 1). In the case of 21, a steric clash between the methyl group and K186 may be an
alternative explanation for its nonꢀreaction. However, in the case of 18, a similar situation is
unlikely since other substituents in the 8ꢀposition in the phenolic ring (e.g. 8ꢀchloro and 8ꢀ
methoxy in 17 and 19, respectively) did not preclude reaction. Based on the crystal structure
of AlnA, the reactivity of 19, which contains a bulky 8ꢀmethoxy substituent can be considered
surprising as there appears to be only a limited amount of space available in the area of the
substituent due to a nearby loop comprising residues T123, A124 and G125 (Figure 1).
In this paper we have demonstrated that AlnA and AlnB can be utilized to attach a ribose
moiety via a C–C bond to a variety of substrates. Such modifications are challenging to
accomplish through chemical means and since glycosylation reactions have been noted to
ACS Paragon Plus Environment

ACS Chemical Biology
Page 10 of 22
10
1
2
3
4
5
6
7
8
9
have an important role in the biological activity of natural products,4,7 this natural promiscuity
of the enzymes is encouraging. In particular, the reaction of 27 indicates that the 2Hꢀpyran
ring of the natural substrate 2 may be amenable for further exploration. In addition,
establishing the principle requirements for substrate reactivity for the native enzymes paves
the way for structureꢀbased protein engineering attempts to expand the substrate acceptance of
the enzymes for the generation of novel Cꢀribosylated metabolites. An attractive starting point
for these studies would be the T123ꢀA124ꢀG125 loop region to allow more variation in the
substitution pattern of the phenolic ring of 15.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The results described above are in line with a Michaelꢀtype 1,4ꢀaddition reaction mechanism
that has been proposed previously for AlnA,17 but a mechanism involving Schiff base
formation cannot be entirely excluded. A key difference between the two mechanisms
proposed for AlnA17 and YeiN19 is that they theoretically prefer different atomic charges at
the reacting carbon of the substrates prealnumycin and uracil, respectively (Figure S1). Of the
compounds tested, 21 should be a good substrate for the Schiffꢀbase mechanism, but a poor
one for the Michaelꢀaddition due to the high electron density at Cꢀ3 (Table 2). The compound
could not be converted into the corresponding Cꢀribosylate, which is consistent with the
Michaelꢀaddition mechanism. However, the binding studies suggest that 21 binds very weakly
to AlnA and might even be bound in the wrong orientation (Table 2), and therefore additional
experiments are required in the future to conclusively verify the reaction mechanism of AlnA.
METHODS
Chemical Syntheses of Substrates 16, 17, 19, 20 and 27. The condensation of maleic
anhydride separately with mꢀcresol, 4ꢀchloroꢀ3ꢀmethyl phenol and 1ꢀnaphthol under Friedl–
ACS Paragon Plus Environment

Page 11 of 22
ACS Chemical Biology
11
1
2
3
4
5
6
7
8
Crafts conditions afforded 16, 17 and 27, respectively.21 General reaction conditions consisted
of melting 25 g AlCl3 and 5 g NaCl at 463 K under vigorous stirring followed by addition of
the appropriate phenol (15 mmol) and maleic anhydride (15 mmol). After 2 minutes, the
9
reaction mixture was poured into a mixture of ice and 12
M HCl and kept for one hour. The
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
reaction product was extracted with toluene, washed with deionized water and isolated by
column chromatography (Silica Gel Mesh 60, Merck) followed by semiꢀpreperative HPLC.
The Oꢀmethylation of juglone (15, 49.8 mg) and naphthazarin (18, 61 mg) was accomplished
by their treatement with 0.56 mmol methyl iodide (MeI) and silver (I) oxide (15, 0.26 mmol;
18, 0.24 mmol) in 0.9 mL dichloromethane at room temperature.22 After 20 hours, additional
charges of MeI (0.56 mM) and Ag2O (15, 195 mmol; 18, 205 mmol) were added and the
reaction stirred for a further 3 hours. The methylated compounds 20 and 19, respectively,
were isolated by column chromatography (Silica Gel Mesh 60, Merck) and stored in CHCl3.
NMR spectroscopic analysis of the compounds is presented in the Supporting Information.
In vitro Reaction Assays. The production and purification of the recombinant AlnA and
AlnB was done according to the protocol reported previously.15 The enzymes were stored in
phosphate buffer (50 m
−20 °C. A typical reaction was done in a buffered solution (50 m
pH 6.85) with 50 ꢁ AlnA, 20 ꢁ AlnB, 10 mM Dꢀriboseꢀ5ꢀphosphate and 250 ꢁ
screened compound together with 5 or 10% DMSO. Oxygenꢀreduced conditions were
effected with 100 n glucoseoxidase and 70 mM Dꢀglucose in the presence of 5 ꢁ catalase.
M
Na2HPO4, 150 m
M
NaCl, 5 m
M
MgCl2, 50 % glycerol, pH 6.85) at
Na2HPO4, 100 m NaCl,
of the
M
M
M
M
M
M
M
The reactions were incubated at 288 K for 4 hours and quenched by lowering the pH either by
addition of formic acid (1% final concentration) or ammonium acetate buffer (pH 3.6).
Products were isolated either by solvent extraction using either CHCl3 or ethyl acetate, or,
ACS Paragon Plus Environment

ACS Chemical Biology
Page 12 of 22
12
1
2
3
4
5
alternatively, SPEꢀcolumn extraction (C18 or phenyl phase, Sulpelco) using MeOH or ACN
as eluent.
6
7
8
9
Analysis of Metabolites. The reactions were analyzed by HPLC (Shimadzu SCLꢀ10Avp)
with a Cꢀ18 column (Sulpelco Discovery C18, 4.6 × 50 mm, 5 ꢁm) using UVꢀvis detection
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(250–600 nm). A gradient from 15% methanol in 60 mM ammonium acetate buffer (pH 3.6)
to 100% MeOH was used with a flow rate of 0.5 mL min−1. The ribosylated compounds were
isolated from the reaction extract by semiꢀpreparative HPLC (Merck Hitachi Lꢀ6200A) with a
reverse phase column (SunFire Prep C18, 10 × 250 mm, 5 ꢁm) using a gradient run similar to
the analytical reactions. The fractions collected were extracted with CHCl3 or, alternatively,
by SPEꢀcolumn extraction (Cꢀ18, Sulpelco), and stored in MeOH.
LC-MS and NMR Measurements. HPLCꢀESIꢀMS were acquired in either negative or
positive mode using a MicroTOFꢀQ mass spectrometer (Bruker Daltonics) linked to a HPLCꢀ
system (Agilent Technologies 1200 series) using the Discovery column and similar conditions
as in the analytical HPLC runs. NMR spectra were acquired at 25 °C with a Bruker Avance
500 MHz NMR spectrometer in CDCl3 (D2O/H2O for the binding experiments). A full
description of the NMR methodology is presented in the Supporting Information.
Molecular Modeling. DFT quantum chemical calculations were performed using Gaussian09
(version A.01) and analyzed using GaussView (version 3.07). Structures were preliminarily
geometryꢀoptimized as per previously described.13 A full description of the molecular
modeling is presented in the Supporting Information.
ASSOCIATED CONTENT
ACS Paragon Plus Environment

Page 13 of 22
ACS Chemical Biology
13
1
2
Supporting Information
3
4
5
6
7
8
Details of NMR methodology, molecular modeling methodology and compound
characterizations. This material is available free of charge via the Internet at
http://pubs.acs.org.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Authors
* (M.M.ꢀK.) Eꢀmail: mikko.mk@gmail.com; (K.D.K.) Eꢀmail: klikakd@yahoo.co.uk.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGEMENTS
The financial support of the Academy of Finland (grant no.s 136060 and 251486) is
acknowledged and the CSC–IT Center for Science Ltd. is thanked for providing
computational resources.
REFERENCES
(1) Cipollone, A., Berettoni, M., Bigioni, M., Binaschi, M., Cermele, C., Monteagudo, E.,
Olivieri, L., Palomba, D., Animati, F., Goso, C., and Maggi, C. A. (2002) Novel anthracycline
oligosaccharides: influence of chemical modifications of the carbohydrate moiety on
biological activity. Bioorg. Med. Chem. 10, 1459–1470.
(2) Schlünzen, F., Zarivach, R., Harms, J., Bashan, A., Tocilj, A., Albrecht, R., Yonath, A.,
and Franceschi, F. (2001) Structural basis for the interaction of antibiotics with the peptidyl
transferase centre in eubacteria. Nature 413, 814–821
ACS Paragon Plus Environment

ACS Chemical Biology
Page 14 of 22
14
1
2
3
4
5
6
7
8
(3) Mrozik, H., Linn, B. O., Eskola, P., Lusi, A., Matzuk, A., Preiser, F. A., Ostlind, D. A.,
Schaeffer, J. M., and Fisher, M. H. (1989) Syntheses and biological activities of 13ꢀ
substituted avermectin aglycons. J. Med. Chem. 32, 375–381.
9
(4) WeymouthꢀWilson, A. C. (1997) The role of carbohydrates in biologically active natural
products. Nat. Prod. Rep. 14, 99–110.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(5) Luzhetskyy, A., Mendez, C., Salas, J. A., and Bechthold, A. (2008) Glycosyltransferases,
important tools for drug design. Curr. Med. Chem. 8, 680–709.
(6) Thibodeaux, C. J., Melancon, C. E., and Liu, H. W. (2007) Unusual sugar biosynthesis
and natural product glycodiversification. Nature 446, 1008–1016.
(7) Kren, V., and Martínková, L. (2001) Glycosides in medicine: The role of glycosidic
residue in biological activity. Curr. Med. Chem. 8, 1303–1328.
(8) Weatherman, R. V., Mortell, K. H., Chervenak, M., Kiessling, L. L., and Toone, E. J.
(1996) Specificity of Cꢀglycoside complexation by mannose/glucose specific lectins.
Biochemistry 35, 3619–3624.
(9) Bililign, T., Griffith, B. R., and Thorson, J. S. (2005) Structure, activity, synthesis and
biosynthesis of arylꢀCꢀglycosides. Nat. Prod. Rep. 22, 742–760.
(10) Hultin, P. G. (2005) Bioactive Cꢀglycosides from bacterial secondary metabolism. Curr.
Med. Chem. 5, 1299–1331.
(11) Bieber, B., Nuske, J., Ritzau, M., and Grafe, U. (1998) Alnumycin a new
naphthoquinone antibiotic produced by an endophytic Streptomyces sp. J. Antibiot. 51, 381–
382.
(12) Naruse, N., Goto, M., Watanabe, Y., Terasawa, T., and Dobashi, K. (1998) K1115 A, a
new anthraquinone that inhibits the binding of Activator Proteinꢀ1 (APꢀ1) to its recognition
sites. II. Taxonomy, fermentation, isolation, physicoꢀchemical properties and structure
determination. J. Antibiot. 51, 545–552.
ACS Paragon Plus Environment

Page 15 of 22
ACS Chemical Biology
15
1
2
3
4
5
6
7
8
(13) Oja, T., Tähtinen, P., Dreiack, N., Mäntsälä, P., Niemi, J., MetsäꢀKetelä, M., and Klika,
K. D. (2012) Alnumycins A2 and A3: new inverseꢀepimeric pairs stereoisomeric to
alnumycin A1. Tetrahedron: Asymmetry 23, 670–682.
9
(14) Oja, T., Palmu, K., Lehmussola, H., Leppäranta, O., Hännikäinen, K., Niemi, J.,
Mäntsälä, P., and MetsäꢀKetelä, M. (2008) Characterization of the alnumycin gene cluster
reveals unusual gene products for pyran ring formation and dioxan biosynthesis. Chem. Biol.
15, 1046–1057.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(15) Oja, T., Klika, K. D., Appassamy, L., Sinkkonen, J., Mäntsälä, P., Niemi, J., and Metsäꢀ
Ketelä, M. (2012) Biosynthetic pathway toward carbohydrateꢀlike moieties of alnumycins
contains unusual steps for CꢀC bond formation and cleavage. Proc. Natl. Acad. Sci. U. S. A.
109, 6024–6029.
(16) Preumont, A., Snoussi, K., Stroobant, V., Collet, J. F., and van Schaftingen, E. (2008)
Molecular identification of pseudouridineꢀmetabolizing enzymes. J. Biol. Chem. 283, 25238–
25246.
(17) Oja, T., Niiranen, L., Sandalova, T., Klika, K. D., Niemi, J., Mäntsälä, J., Schneider, G.,
and MetsäꢀKetelä, M. (2013) Structural basis for Cꢀribosylation in the alnumycin A
biosynthetic pathway. Proc. Natl. Acad. Sci. U. S. A. 110, 1291–1296.
(18) Allen, K. N., and DunawayꢀMariano, D. (2004) Phosphoryl group transfer: evolution of a
catalytic scaffold. Trends Biochem. Sci. 29, 495–503.
(19) Huang, S., Mahanta, N., Begley, T. P., and Ealick, S. E. (2012) Pseudouridine
monophosphate glycosidase: a new glycosidase mechanism. Biochemistry 51, 9245–9255.
(20) Fielding, L. (2007) NMR methods for the determination of protein–ligand dissociation
constants. Prog. Nucl. Magn. Res. Spectrosc. 51, 219–242.
(21) Mahapatra, A., Mativandlela, S. P., Binneman, B., Fourie, P. B., Hamilton, C. J., Meyer,
J. J., van der Kooy, F., Houghton, P., and Lall, N. (2007) Activity of 7ꢀmethyl juglone
ACS Paragon Plus Environment

ACS Chemical Biology
Page 16 of 22
16
1
2
3
4
5
derivatives against Mycobacterium tuberculosis and as subversive substrates for mycothiol
disulfide reductase. Bioorg. Med. Chem. 15, 7638–7646.
6
7
8
9
(22) Tietze, L. F., Stewart, S. G, Polomska, M. E., Modi, A., and Zeeck, A. (2004) Towards a
total synthesis of the new anticancer agent mensacarcin: synthesis of the carbocyclic core.
Chemistry 10, 5233–5242.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 17 of 22
ACS Chemical Biology
17
1
2
3
4
5
Table 1. Evaluation of the Most Important Substrates in Regards to Binding to AlnA by NMR
and Potential Reactivity by Molecular Modeling. The Results Are ColorꢀCoded Based on
Good (Green), Tolerated (Orange) and Unsuitable (Red) Characteristics
6
7
8
9
rel. binding
from
model. pred.
ꢀꢀG / kcal mol−
0.00
binding orientation by
waterLOGSY
model.
pred. QC3
rel. reactivity by
HPLC / %
compound
1
ν
½
2
n/aa
n/aa
weak, especially H-2
and H-3
81
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
−0.175
−0.186
13
medium to weak
3.85
no reaction
15
16
17
18
strong
weakb
n/aa or c
strong
signals even
signals even
n/aa or c
4.62
0.17
2.43
7.09
77
100
5.4
−0.176
−0.175
−0.172
−0.173
n/ad
no reaction
signals even except
for OMe (slightly low)
H-6 low, H-8 high,
OMe low
signals even except
for Me (low)
n/ac
19
20
strong
11
−0.174
−0.188
−0.74
4.56
7.23
medium to weak
no reaction
21
27
very weak
n/ac
no reaction
99
−0.249
−0.175
3.41
Legend: n/a, not applicable. aNot accessible for binding studies due to micelle formation. Difficult to assess as
the compound only sparingly soluble. cCompound solubility too low. dOnly one signal for the non-exchangeable
Hs due to symmetry and exchange, therefore waterLOGSY unable to provide binding orientation information.
b
ACS Paragon Plus Environment

ACS Chemical Biology
Page 18 of 22
18
1
2
Figures and Legends to the Figures
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Model for the biosynthesis of alnumycin A (1)15,17 and the structures of the AlnA
(red) and AlnB (green) proteins that are responsible for the overall Cꢀribosylation reaction.17
The active site of AlnA containing the reacting molecules is derived from a crystal structure
of AlnA in complex with Dꢀriboseꢀ5ꢀphosphate onto which prealnumycin (2) has been
docked.17 The active site of AlnB containing alnumycin P (3) is derived from a crystal
structure of AlnB onto which 3 has been docked.17
ACS Paragon Plus Environment

Page 19 of 22
ACS Chemical Biology
19
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Structures of the 23 substrates, 5–27, tested in this work and the novel Cꢀ
ribosylated reaction products 28–32. Tick marks designate accepted substrates.
ACS Paragon Plus Environment

ACS Chemical Biology
Page 20 of 22
20
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Design rules for the Cꢀribosylation of naphthoquinones.
ACS Paragon Plus Environment

Page 21 of 22
ACS Chemical Biology
21
1
2
For Table of Contents Use Only
3
4
5
6
7
8
9
Chemoenzymatic Synthesis of Novel CꢀRibosylated Naphthoquinones
Bastian Blauenburg, Terhi Oja, Karel D. Klika and Mikko MetsäꢀKetelä
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

ACS Chemical Biology
Page 22 of 22
1
2
3
4
ToC:
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment