Lignin Peroxidase-Catalyzed Oxidation of Nonphenolic Trimeric Lignin Model Compounds: Fragmentation Reactions in the Intermediate Radical Cations

Article abstract of DOI:10.1021/jo035052w

The H₂O₂-promoted oxidations of the two nonphenolic β-O-aryl lignin model trimers 1 and 2, catalyzed by lignin peroxidase (LiP) at pH = 3.5, have been studied. The results have been compared with those obtained in the oxidation of 1 and 2 with the genuine one-electron oxidant potassium 12-tungstocobalt(III)ate. These models present a different substitution pattern of the three aromatic rings, and by one-electron oxidation, they form radical cations with the positive charge, which is localized in the dialkoxylated ring as also evidenced by a pulse radiolysis study. Both the oxidations with the enzymatic and with the chemical systems lead to the formation of products deriving from the cleavage of C-C and C-H bonds in a β position with respect to the radical cation with the charge residing in the dialkoxylated ring (3,4-dimethoxybenzaldehyde (5) and a trimeric ketone 6 in the oxidation of 1 and a dimeric aldehyde 8 and a trimeric ketone 9 in the oxidation of 2). These products are accompanied by a dimeric aldehyde 7 in the oxidation of 1 and 4-methoxybenzaldehyde (10) in the oxidation of 2. The unexpected formation of these two products has been explained by suggesting that 1^.+ and 2^.+ can also undergo an intramolecular electron transfer leading to the radical cations 1a^.- and 2a ^.+ with the charge residing in a monoalkoxylated ring. The fast cleavage of a C-C bond β to this ring, leading to 7 from 1^.+ and to 10 from 2^.+, is the driving force of the endoergonic electron transfer. A kinetic steady-state investigation of the LiP-catalyzed oxidation of the trimer 2, the dimeric model 1-(3,4-dimethoxyphenyl)-2-phenoxy-1-ethanol (4), and 3,4-dimethoxybenzyl alcohol (3) has indicated that the turnover number (k_cat) and the affinity for the enzyme decrease significantly by increasing the size of the model compound. In contrast, the three substrates exhibited a very similar reactivity toward a chemical oxidant [Co ^IIIW]. This suggests a size-dependent interaction of the enzyme with the substrate which may influence the efficiency of the electron transfer.

Full text of DOI:10.1021/jo035052w

Lign in P er oxid a se-Ca ta lyzed Oxid a tion of Non p h en olic Tr im er ic
Lign in Mod el Com p ou n d s: F r a gm en ta tion Rea ction s in th e
In ter m ed ia te Ra d ica l Ca tion s
Enrico Baciocchi,* Claudia Fabbri, and Osvaldo Lanzalunga*
Dipartimento di Chimica, Universita` La Sapienza, P.le A. Moro, 5 I-00185 Rome, Italy
enrico.baciocchi@uniroma1.it; osvaldo.lanzalunga@uniroma1.it
Received J uly 22, 2003
The H2O2-promoted oxidations of the two nonphenolic â-O-aryl lignin model trimers 1 and 2,
catalyzed by lignin peroxidase (LiP) at pH ) 3.5, have been studied. The results have been compared
with those obtained in the oxidation of 1 and 2 with the genuine one-electron oxidant potassium
12-tungstocobalt(III)ate. These models present a different substitution pattern of the three aromatic
rings, and by one-electron oxidation, they form radical cations with the positive charge, which is
localized in the dialkoxylated ring as also evidenced by a pulse radiolysis study. Both the oxidations
with the enzymatic and with the chemical systems lead to the formation of products deriving from
the cleavage of C-C and C-H bonds in a â position with respect to the radical cation with the
charge residing in the dialkoxylated ring (3,4-dimethoxybenzaldehyde (5) and a trimeric ketone 6
in the oxidation of 1 and a dimeric aldehyde 8 and a trimeric ketone 9 in the oxidation of 2). These
products are accompanied by a dimeric aldehyde 7 in the oxidation of 1 and 4-methoxybenzaldehyde
(10) in the oxidation of 2. The unexpected formation of these two products has been explained by
suggesting that 1•+ and 2•+ can also undergo an intramolecular electron transfer leading to the
radical cations 1a •+ and 2a •+ with the charge residing in a monoalkoxylated ring. The fast cleavage
of a C-C bond â to this ring, leading to 7 from 1•+ and to 10 from 2•+, is the driving force of the
endoergonic electron transfer. A kinetic steady-state investigation of the LiP-catalyzed oxidation
of the trimer 2, the dimeric model 1-(3,4-dimethoxyphenyl)-2-phenoxy-1-ethanol (4), and 3,4-
dimethoxybenzyl alcohol (3) has indicated that the turnover number (kcat) and the affinity for the
enzyme decrease significantly by increasing the size of the model compound. In contrast, the three
substrates exhibited a very similar reactivity toward a chemical oxidant [CoIIIW]. This suggests a
size-dependent interaction of the enzyme with the substrate which may influence the efficiency of
the electron transfer.
In tr od u ction
Lignin is a highly irregular three-dimensional biopoly-
mer composed of oxygenated phenylpropane units linked
together by different bonds, among which the most
common is the so-called â-O-4 bond shown in Figure
1, where two possible phenylpropane units are repre-
sented.1,2 Lignin is synthesized by plants mainly to
provide strength, rigidity, and protection from oxidative
processes and from attacks by microorganisms.1,3
F IGURE 1. Two phenylpropane units of lignin linked by the
â-O-4 bond.
The oxidative degradation of lignin is a process of
fundamental importance not only because it can convert
lignin into low molecular weight aromatic compounds,
thus making this polymer a renewable source for the
industrial preparation of a number of chemicals,4 but also
because the selective degradation of lignin and its
removal from the carbohydrate component of wood is a
key step in the pulp and paper industry.5
To reduce the environmental impact and energy con-
sumption of the lignin degradation processes, particular
attention has recently been given to the natural degrada-
tion promoted by fungi, mainly of the basidiomycetous
group.6-9 Among the various fungi, the white rot basidi-
(1) Sarkanen, K. V. Lignins: Occurrence, Formation, Structure and
Reactions; Sarkanen, K. V., Ludwig, C. H., Eds.; Wiley-Interscience:
New York, 1971; pp 95-195.
(2) â refers to the Câ carbon atom of the phenylpropane unit, -O-
indicates the oxygen atom of the Câ, and -4 refers to the carbon atom
in the aromatic ring with C-1 carrying the propyl side chain.
(3) Argyropoulos, D. S.; Menachem, S. B. Biotechnology in the Pulp
and Paper Industry; Eriksson, K. E. L., Ed.; Springer-Verlag: Berlin
Heidelberg, 1997; pp 127-158.
(5) Roberts, J . C. The Chemistry of Paper; The Royal Society of
Chemistry: Cambridge, 1996.
(6) J anshekar, H.; Fiechter, A. Advances in Biochemical Engineer-
ing/ Biotechnology; Fiechter, A., Ed.; Springer-Verlag: Berlin, 1983;
Vol. 27.
(4) Biørsvik, H.-R.; Minisci, F. Org. Process Res. Dev. 1999, 3, 330.
10.1021/jo035052w CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/23/2003
J . Org. Chem. 2003, 68, 9061-9069
9061

Baciocchi et al.
SCHEME 1
bearing the SOMO) are the most important since they
should play a key role in lignin oxidative degradation.
In the past years, many studies have been carried out
aimed at investigating the mechanism involved in the
oxidative degradation of lignin. Because of the enormous
practical problems posed by the lignin macromolecule,
these studies have generally dealt with simple mono-
meric, dimeric, or trimeric models21 containing structural
motifs of the polymer.22-24 Even though interesting
information has been collected, the structural factors,
which may play a role in determining the particular type
of bond that undergoes the cleavage, have not been fully
understood. One reason is that generally these models
contained aromatic rings similarly activated toward
electron transfer, making it practically impossible to
know the localization of the positive charge in the radical
cation and hence its structure.
To overcome this problem, we have directed our at-
tention to the nonphenolic trimeric lignin model com-
pounds 1 and 2, both containing two â-O-4 linkages,
which have been synthesized as a mixture of the anti-
anti diastereomers (see Experimental Section).25 1 and
2 present a substitution pattern such that upon one-
electron oxidation, they should form the radical cations
1•+ and 2•+, in which the positive charge is reasonably
expected to be predominantly localized in the dialkoxy-
lated ring A.26
omycetous Phanerochaete chrysosporium is one of the
most studied;10 its ligninolitic activity is due to the
production of manganese peroxidase (MnP) and lignin
peroxidase (LiP). LiP is a glycoprotein containing an
iron(III)-protoporphyrin IX as prosthetic group, with a
histidine residue (His 173) coordinated to the iron
atom,11,12 which differs from other heme peroxidases for
its ability to oxidize aromatic nonphenolic substrates with
relatively high redox potential (up to 1.4 V vs NHE).13-20
It is widely accepted that the catalytic cycle of LiP
involves first the reaction of the native enzyme [Por-
(Fe(III))] with H2O2 (Scheme 1, path a), leading to the
formation of LiP compound I (LiP I), an iron(IV)-oxo
porphyrin radical cation [Por•+FeIVdO]. LiP I is the
enzyme-active oxidant able to abstract one electron from
the aromatic ring of the substrate, leading to [Por-
FeIVdO], LiP compound II (LiP II), and the substrate
radical cation (Scheme 1, path b).12 One-electron oxida-
tion of the substrate by [PorFeIVdO] (Scheme 1, path c)
regenerates the native enzyme.
We felt that by knowing the structure of the radical
cation, it would be possible to obtain unambiguous
information on the bonds which are broken and on how
the structure of the radical cation may influence the
relative rate of cleavage of these bonds. Thus we under-
took a products investigation of the LiP-catalyzed oxida-
tion of 1 and 2, and the results of the enzymatic study
were compared with those obtained when 1 and 2 were
reacted with the genuine one-electron oxidant K5-
[CoIIIW12O40], from now on simply indicated as CoIIIW.29,30
Moreover, to establish to what degree the size of the
model compounds may influence the rate of the enzy-
matic oxidation, the kinetic enzymatic constants KM and
kcat for the trimer 2, the monomeric model 3,4-dimeth-
oxybenzyl alcohol (3), and the dimeric model 1-(3,4-
dimethoxyphenyl)-2-phenoxy-1-ethanol (4) were deter-
mined. The results were compared with kinetic data
obtained in the oxidation of the same substrates with the
chemical oxidant CoIIIW. Some pulse radiolysis experi-
ments were also carried out, aimed at obtaining ad-
ditional information on the structure of 1•+ and 2•+. The
results of this investigation are reported herewith.
Once formed, the radical cation may undergo a number
of reactions,13,14 among which those involving the cleav-
age of C-C and C-H â bonds (â with respect to the ring
(7) Eriksson, K. E. L.; Blanchette, R. A.; Ander, P. Microbial and
Enzymatic Degradation of Wood and Wood Components; Springer-
Verlag: Berlin, 1990.
(8) Levy, J . F. Cellulose Sources and Exploitation; Kennedy, J . F.,
Phillips, G. O., Williams, P. A., Eds.; Ellis Horwood: New York, 1990;
pp 397-408.
(9) Pedlar, S. L.; Betts, W. B. Cellulose Sources and Exploitation;
Kennedy, J . F., Phillips, G. O., Williams, P. A., Eds.; Ellis Horwood:
New York, 1990; pp 435-442.
(10) Kersten, P. J .; Stephens, S. K.; Kirk, T. K. Biotechnology in
the Pulp and Paper Manufacture; Kirk, T. K., Chang, H. M., Eds.;
Butterworth-Heinemann: Stoneham, MA, 1990; pp 457-463.
(11) Tien, M.; Kirk, T. K. Science 1983, 221, 661.
(12) Glenn, J . K.; Morgan, M. A.; Mayfield, M. B.; Kuwahara, M.;
Gold, M. H. Biochem. Biophys. Res. Commun. 1983, 114, 1077.
(13) Schoemaker, H. E. Recl. Trav. Chim. Pays-Bas, 1990, 109, 255.
(14) Labat, G.; Meunier, B. Bull. Soc. Chim. Fr. 1990, 127, 553.
(15) Hammel, K. E.; Tien, M.; Kalyanaraman, B.; Kirk, T. K. J . Biol.
Chem. 1985, 260, 8348.
(16) Schoemaker, H. E.; Harvey, P. J .; Bowen, R. M.; Palmer, J . M.
FEBS Lett. 1985, 183, 7.
(17) Kersten, P. J .; Tien, M.; Kalyanaraman, B.; Kirk, T. K. J . Biol.
Chem. 1985, 260, 2609.
(18) Miki, K.; Renganathan, V.; Gold, M. H. Biochemistry 1986, 25,
4790.
(19) Renganathan, V.; Miki, K.; Gold, M. H. Arch. Biochem. Biophys.
1986, 246, 155.
(20) Harvey, P. J .; Floris, R.; Lundell, T.; Palmer, J .; Schoemaker,
H. E.; Wever, R. Biochem. Soc. Trans. 1992, 20, 5.
9062 J . Org. Chem., Vol. 68, No. 23, 2003

Lignin Peroxidase-Catalyzed Oxidation
Resu lts a n d Discu ssion
P u lse Ra d iolysis Stu d ies. The radical cations 1•+ and
2•+ were generated from 1 and 2, respectively, by pulse
radiolysis using Tl2+ as the oxidant.31 Tl2+ was produced
by irradiating N2O-saturated aqueous solutions (pH 3.5)
of Tl2SO4 (5 mM) (eqs 1-4):
H2O ' H+, OH•, e-
(1)
(2)
(3)
(4)
aq
e- + N2O + H2O f N2 + OH• + OH-
aq
OH• + Tl+ + H+ f Tl2+ + H2O
Tl2+ + ArCH(OH)R f Tl+ + Ar•+CH(OH)R
The function of N2O is to scavenge eaq-, leading to the
formation of an additional hydroxyl radical (eq 2), with
F IGURE 2. Time-resolved absorption spectrum observed on
reaction of Tl2+ with 1 (0.1 mM) at T ) 25 °C, recorded after
pulse radiolysis of an N2O-saturated aqueous solution (pH )
3.5), containing 0.5 mM Tl2SO4, 200 µs after the 500-ns, 10-
MeV electron pulse.
k ) 9.1 × 109 M-1 s-1 34
.
Tl2+ is then produced by oxidation
of Tl+ by OH• (eq 3) with k ) 1.2 × 1010 M-1 s-1 33
.
Tl2+
reacts with alkoxyaromatics by one-electron transfer to
give the corresponding radical cations (eq 4) with k ≈ 5
× 108 M-1 s-1 35
.
The absorption spectra obtained from 1
and 2 recorded 200 µs after the pulse are displayed in
Figures 2 and 3, respectively.
Both these spectra present absorption bands at 310
and 420 nm, that is in a region typical for aromatic
radical cations and can be safely assigned to 1•+ (Figure
2) and 2•+ (Figure 3). More importantly, these spectra
closely resemble the spectrum of 3,4-dimethoxybenzyl
alcohol radical cation (bands at 310 and 420 nm)28 while
they are significantly different from the spectrum of
4-methoxybenzyl alcohol radical cation which displays
two main absorption bands at 290 and 450 nm.36 Thus,
in line with predictions, this result indicates that in both
1•+ and 2•+ the positive charge is localized on the
dialkoxylated ring. This conclusion is also supported by
the observation of a very slow decay of the absorption
bands of 1•+ and 2•+ (too slow to be followed in the 1-ms
time scale of the experiment) as expected for a dimethoxy-
substituted benzene radical cation. For example, the half-
life time is 29 ms for the 3,4(MeO)2C6H3CH(OH)tBu
radical cation28 but only 4.6 µs (almost 4 orders of
F IGURE 3. Time-resolved absorption spectrum observed on
reaction of Tl2+ with 2 (0.1 mM) at T ) 25 °C, recorded after
pulse radiolysis of an N2O-saturated aqueous solution (pH )
3.5), containing 0.5 mM Tl2SO4, 200 µs after the 500-ns, 10-
MeV electron pulse.
magnitude smaller) for the corresponding 4-methoxy
derivative.36 Another interesting observation is that the
spectrum of 2•+ exhibits a broad band in the near-infrared
(NIR) region of the spectrum (λ > 700 nm), which can be
assigned to an intramolecular charge resonance (CR)
interaction between a neutral donor ring and the charged
acceptor ring in analogy with what was previously
observed for the diarylmethanol radical cations.37 Since
the NIR band is absent in 1•+, it would appear that only
in 2•+, where the positive charge is located in the central
aromatic ring, the geometry of the system fulfills the
requirement of a cofacial arrangement between donor and
acceptor rings necessary for efficient charge transfer and
optimum electronic coupling.38,39
(21) With these terms, it is generally indicated the number of
phenylpropane units present in the model.
(22) Ten Have, R.; Teunissen, P. J . M. Chem. Rev. 2001, 101, 3397.
(23) Kirk, T. K.; Tien, M.; Kersten, P. J .; Mozuch, M. D.; Kalya-
naraman, B. Biochem. J . 1986, 236, 279.
(24) Mester, T.; Ambert-Balay, K.; Ciofi-Baffoni, S.; Banci, L.; J ones,
A. D.; Tien, M. J . Biol. Chem. 2001, 25, 22985.
(25) Ciofi-Baffoni, S.; Banci. L.; Brandi, A. J . Chem. Soc., Perkin
Trans. 1 1998, 3207.
(26) The oxidation potential of 4-methoxybenzyl alcohol27 is 0.54 V
higher than that of 3,4-dimethoxybenzyl alcohol.28
(27) Brown, O. R.; Chandra, S.; Harrison, J . A. J . Electroanal. Chem.
1972, 38, 185.
(28) Bietti, M.; Baciocchi, E.; Steenken, S. J . Phys Chem. 1998, 102,
7337.
(29) Eberson, L. J . Am. Chem. Soc. 1983, 105, 3192.
(30) Baciocchi, E.; Crescenzi, M.; Fasella, E.; Mattioli, M. J . Org.
Chem. 1992, 57, 4684.
(31) Similar results have been obtained using Br2•-, which is a more
selective oxidant (1.63 V vs NHE)32 than Tl2+ (2.2 V vs NHE).33
(32) Mohan, H.; Mittal, J . P. J . Phys. Chem. A 1997, 101, 10012.
(33) Schwarz, H. A.; Dodson, R. W. J . Phys. Chem. 1984, 88, 3643.
(34) J anata, E.; Schuler, R. H. J . Phys. Chem. 1982, 86, 2078.
(35) O’Neill, P.; Steenken S.; Shulte-Frohlinde, D. J . Phys. Chem.
1975, 79, 2773.
En zym a tic a n d Ch em ica l Oxid a tion of 1 a n d 2.
Oxidations of 1 or 2 with H2O2 promoted by LiP were
(37) Bietti, M.; Lanzalunga, O. J . Org. Chem. 2002, 67, 2632.
(38) Rathore, R.; Lindeman, S. V.; Kochi, J . K. J . Am. Chem. Soc.
1997, 119, 9393.
(36) Baciocchi, E.; Bietti, M.; Stenkeen, S. Chem.-Eur. J . 1999, 5,
1785.
(39) Badger, B.; Brocklehurst, B. Trans. Faraday Soc. 1969, 65,
2583.
J . Org. Chem, Vol. 68, No. 23, 2003 9063

Baciocchi et al.
TABLE 1. P r od u cts a n d Yield s in th e LiP /H2O2- a n d CoIIIW-P r om oted Oxid a tion of th e Lign in Mod el 1
a
Recovered starting material and yields of products are referred to the initial amount of the substrate. The average error is (10%.
carried out in 50 mM aqueous sodium tartrate buffer,
pH 3.5, with 6% of acetonitrile added as cosolvent at 25
°C, and under an argon atmosphere. H2O2, equimolar
with the substrate, was added gradually over 90 min
using a syringe pump. Oxidations promoted by CoIIIW
(2:1 molar ratio with the substrate) were also carried out
under identical conditions for 48 h at 60 °C. After
addition of an internal standard, the reaction products
98% in the enzymatic reaction and 87% in the oxidation
by CoIIIW.
The formation of the same products in the LiP/H2O2-
and CoIIIW-induced reactions of both 1 and 2 indicates
that the enzymatic and chemical oxidation proceed by
the same mechanism. Since it is well ascertained that
the oxidations promoted by CoIIIW take place by an
electron-transfer mechanism, this observation fully con-
firms that also the oxidation of 1 and 2 by LiP are one-
electron-transfer processes (Scheme 1), the radical cation
being the key intermediate en route to the observed
products.
1
were analyzed by HPLC, HPLC-MS, and H NMR. The
results are in Tables 1 and 2.
The enzymatic oxidation of 1 promoted by the LiP/H2O2
system leads mainly to the formation of 3,4-dimethoxy-
benzaldehyde (5), the ketone 6, and the dimeric aldehyde
7 (Table 1). The structure of these products was deter-
mined by comparison with authentic specimens (5 and
7) or by NMR and MS analysis (6) (see Experimental
Section). The same products, in similar proportions, were
obtained in the oxidation of 1 induced by CoIIIW at pH )
3.5. Products and yields, referred to the initial amount
of the substrate, are reported in Table 1, where the
amount of the unreacted substrate is also indicated. The
mass balance (with respect to the products containing
the dimethoxylated ring) is satisfactory accounting for
more than 90% of the starting material.
Thus, in the oxidation of 1, 3,4-dimethoxybenzaldehyde
(5) should be formed, together with the carbon-centered
radical 11, by CR-Câ bond cleavage in 1•+ (path a in
Scheme 2), whereas CR-H bond cleavage in the same
radical cation leads to the trimeric ketone 6 (paths b and
d in Scheme 2).
The formation of the dimeric aldehyde 7, the major
reaction product, is more difficult to rationalize since a
direct pathway leading from 1•+ to 7 would involve the
breaking of a bond which is very far apart from the
positively charged ring. Thus, we suggest that an ad-
ditional pathway for 1•+ is an intramolecular electron
transfer to give the radical cation 1a •+ where the charge
is now localized in the central ring of the model (path e
in Scheme 2). Cleavage of the C-C bond â to this ring
(the CR′-Câ′ bond) in 1a •+ can lead to 7. Certainly, 1a •+
is much less stable than 1•+, but the thermodynamically
uphill electron transfer is a unimolecular process and
could be driven by the C-C bond cleavage reaction in
1a •+, which is expected to be much faster in the
monoalkoxylated aromatic radical cation than in 1•+, as
also clearly shown by the half-life time data reported
The oxidation of 2 promoted by the LiP/H2O2 system
leads mainly to the formation of the dimeric aldehyde 8,
the trimeric ketone 9, and 4-methoxybenzaldehyde (10)
(Table 2, comparison with authentic specimens). The
same products, but in different relative amounts, were
also observed in the oxidation of 2 induced by CoIIIW at
pH ) 3.5. Products and yields, referred to the initial
amount of the substrate, are reported in Table 2. Also,
in this case, the mass balance (with respect to the
products containing the monomethoxylated ring) is good,
9064 J . Org. Chem., Vol. 68, No. 23, 2003

Lignin Peroxidase-Catalyzed Oxidation
TABLE 2. P r od u cts a n d Yield s in th e LiP /H2O2- a n d CoIIIW-P r om oted Oxid a tion of th e Lign in Mod el 2
a
Recovered starting material and yields of products are referred to the initial amount of the substrate. The average error is (10%.
above.28,36 In addition, since products coming from both
1•+ and 1a •+ are observed, and the UV-vis spectrum
(Figure 2) presents only absorptions due to 1•+, we can
estimate for the 1•+ f 1a •+ intramolecular electron
transfer a rate comparable with the decay rate of 1•+ (20-
30 s-1, pulse radiolysis experiments). Now, on the basis
of the Marcus equation, such a rate is consistent with
an electron-transfer process endoergonic by 13 kcal mol-1
(the estimated free-energy difference between the two
radical cations) by assuming a quite reasonable value of
25 kcal mol-1 for the intrinsic barrier λ.
The alternative possibility of an intermolecular elec-
tron transfer is very unlikely. In the first place, intramo-
lecular reactions are entropically favored with respect to
the intermolecular ones. In the second place, and more
specifically, it has been clearly shown that no inter-
molecular electron transfer takes place between the 3,4-
dimethoxybenzyl alcohol radical cation and 4-methoxy-
mandelic acid (a system very close to those investigated
by us) even though 4-methoxymandelic acid radical
cation can undergo an extremely fast C-C bond cleav-
age.40
If the hypothesis of the intramolecular electron trans-
fer is correct, it would result that C-C bond cleavage is
the exclusive fragmentation process in 1a •+, while in 1•+
there is competition between C-C and C-Η bond cleav-
ages. This is not surprising since it is well known that
competition between C-C and C-Η â-bond cleavage is
observed in the oxidation of dimethoxylated lignin model
dimers23,41 whereas monomethoxylated models undergo
only C-C bond cleavage.36
It is also important to note that the C-C bond cleavage
either from 1•+ or 1a •+ is also expected to lead to the
carbon-centered radicals 11 and 11a , respectively, and
to the products derived therefrom. No such products were
however observed. A possible explanation is that oxida-
tion and hydrolysis of 11 and 11a may lead to phenolic
compounds (paths c and g in Scheme 2), which are then
overoxidized to water-soluble and/or polymeric species
under the reaction conditions.42 Fortunately, the lack of
recovery of the products deriving from 11 and 11a does
not impair the mechanistic conclusions concerning the
rupture modes in 1•+. In view of the excellent mass
balances noted before, there is no doubt that compounds
5, 6, and 7 account for the main fragmentation routes
available to 1•+
.
Coming to the products from 2, the aldehyde 8 and the
ketone 9 derive from 2•+ by CR-Câ and CR-H bond
cleavage processes, respectively, as shown in Scheme 3
(paths a and b), analogously to what proposed for the
corresponding reactions in 1•+. In this case too, however,
there is a product, namely, 4-methoxybenzaldehyde (10),
which is by far the major product, whose formation from
2•+ is more difficult to rationalize. Accordingly, the
possibility that 10 derives from the cleavage of the CR′-
Câ′ bond in 2•+ is unlikely since this is a γ bond with
respect to the positively charged ring and, at the best of
our knowledge, there is no report concerning this kind
of cleavage in an aromatic radical cation. Thus, we
suggest that 2•+ can also undergo an intramolecular
electron transfer to give the radical cation 2a •+, a
monoalkoxylated radical cation (path e in Scheme 3),
which then undergoes the fast cleavage of the CR′-Câ′
bond, a â bond with respect to the positively charged ring.
(40) Candeias, L. P.; Harvey, P. J . J . Biol. Chem. 1995, 270, 16745.
(41) Baciocchi, E.; Bietti, M.; Gerini, M. F.; Lanzalunga, O.; Man-
cinelli, S. J . Chem. Soc., Perkin Trans. 2 2001, 1506.
(42) Koduri, R. S.; Tien, M. J . Biol. Chem. 1995, 270, 22254.
J . Org. Chem, Vol. 68, No. 23, 2003 9065

Baciocchi et al.
SCHEME 2
The feasibility of the 2•+ f 2a •+ intramolecular electron
transfer can be supported by the same reasoning pre-
sented before for the 1•+ f 1a •+ corresponding reaction.43
As already observed for 1, also in the oxidation of 2,
the products expected from the carbon radicals 12 and
known that LiP is unable to catalyze the oxidation of
monomethoxylated substrates.13 For example, LiP cannot
oxidize 4-methoxymandelic acid unless a mediator is
present.44,45 Thus, our observation clearly indicates that
the initial interaction of the enzyme with 1 or 2 concerns
the more electron-rich dialkoxylated ring from which the
hole can be transferred to the less activated monoalkoxy-
lated ring. In other words, the former ring acts as an
intramolecular redox mediator.
Finally, to get some information about the interaction
of the lignin model trimers with the enzyme, kinetic
experiments were carried out and the Michaelis-Menten
constants, kcat and KM, were determined for the LiP-
catalyzed oxidation of the trimeric model 2. For compari-
son, we also determined kcat and KM for the oxidation of
3,4-dimethoxybenzyl alcohol (3), a monomeric model, and
the dimeric model 1-(3,4-dimethoxyphenyl)-2-phenoxy-
12a formed by C-C bond cleavage in 2•+ and 2a •+
,
respectively, were not observed. The same explanation
presented before (formation of easily oxidizable phenolic
products, paths c and g in Scheme 3) may also hold in
this case. Anyway, in this case too, the excellent mass
balance noted above makes us confident that the products
8, 9, and 10 describe the main reaction pathways of 2.
The observation that the LiP-catalyzed oxidation of 1
and 2 efficiently promotes the cleavage of a bond â to a
monoalkoxylated ring is of great interest as it is well
(43) In this case, another mechanism leading to 10 might be an
intramolecular nucleophilic attack of the alcoholic OH group to the
dialkoxylated aromatic ring bearing the positive charge in 2•+ followed
by ring opening of the cyclohexadienyl radical41 leading to an alkoxy
radical, which undergoes â scission. However, this mechanism should
also form an isomer of 2 which has not been observed.
(44) Harvey, P. J .; Schoemaker, H. E.; Palmer, J . M. FEBS Lett.
1986, 195, 242.
(45) Baciocchi, E., Gerini, M. F.; Lanzalunga, O.; Mancinelli S.
Tetrahedron 2002, 58, 8087.
9066 J . Org. Chem., Vol. 68, No. 23, 2003

Lignin Peroxidase-Catalyzed Oxidation
SCHEME 3
against a chemical oxidant, namely, CoIIIW, the three
models exhibited practically the same reactivity as shown
by the rate constants in Table 3. It is therefore probable
that the reactivity trend observed with LiP is due to
variations in the enzyme-substrate interactions related
to the substrate structure. Even though 3, 4, and 2
exhibit a similar affinity for the enzyme (KM values), the
way the substrate is oriented with respect to the enzyme
might somewhat change with the size of the substrate
itself, which might influence the efficiency of the long-
range electron transfer.
TABLE 3. Kin etic P a r a m eter s of th e Oxid a tion of 3, 4,
a n d 2 P r om oted by LiP /H2O2 a n d by CoIII
W
LiP/H2O2
CoIII
W
kcat
KM
(mM)
kcat/KM
substrate
k (M-1 s-1
)
(s-1
)
(s-1 mM-1
)
3
4
2
0.80
0.76
0.99
1.85
0.38
0.13
0.27
0.28
0.44
6.85
1.35
0.29
1-ethanol (4). The results are reported in Table 3 together
with kinetic data concerning the oxidation of the same
models with CoIIIW.
It can be noted that the efficiency of the enzymatic
catalysis, as expressed by kcat, decreases by increasing
the size of the substrate.46 Thus, 3 is ca. 5 times more
reactive than the dimer 4, and 4 is ca. 3 times more
reactive than the trimer 2. In contrast, when tested
Apart from the possible interpretations of the kinetic
parameters in Table 3, it is remarkable that the trimeric
model still exhibits a substantial reactivity. This is an
(46) Similar conclusions were established in the oxidation of phenolic
lignin oligomers.47
(47) Banci, L.; Ciofi-Baffoni, S.; Tien, M. Biochemistry 1999, 38,
3205.
J . Org. Chem, Vol. 68, No. 23, 2003 9067

Baciocchi et al.
SCHEME 4 a
a
Reaction conditions: (a) LDA, THF, -78°; (b) NaBH4, THF, H2O; (c) PPTS, acetone, H2O; (d) TsOH, acetone; (e) LDA, THF, -78°;
(f) NaBH4, THF, H2O; (g) PPTS, acetone, H2O.
interesting result since LiP is reported to be the peroxi-
dase with the most restricted access to the active site,48
and it seems therefore highly unlikely that a molecule
as large as 2 may approach the heme edge. Thus, the
significant reactivity of 2 (and probably also of the dimer
4) might support the recent hypothesis of another, much
more accessible, interaction site in LiP, located on the
external surface of the enzyme and, namely, in the
proximity of the residue Trp171.49,50 However, direct
mutagenesis and/or crystallographic studies are required
to confirm this hypothesis.
where the positive charge is initially localized, can
efficiently be broken.
A steady-state kinetic study has showed that the size
of the model substrate influences the efficiency of the
oxidation catalyzed by LiP, the efficiency decreasing as
the model becomes larger. In contrast, the three models
exhibited a very similar reactivity toward CoIIIW, a
chemical oxidant, suggesting a size-dependent interaction
of the enzyme with the substrate, which may influence
the efficiency of the electron transfer. An additional
observation is that with the trimeric model a substantial
activity of the enzyme is still observed. Since the enzyme
active site and the access channel appear almost inac-
cessible to the trimeric model, the above observation
would support the recent suggestion of a substrate
oxidation mediated by the LiP surface residue Trp171.
Con clu sion s
The results reported in this work provide significant
indications about the fragmentation routes available to
the intermediate radical cation in the LiP-induced oxida-
tive degradation of trimeric model compounds 1 and 2.
The most important result is that with both models, not
only the cleavage of the C-C bond â with respect to the
dialkoxylated ring (where the positive charge is initially
localized) was observed but also the cleavage of the C-C
bond â to a monoalkoxylated ring. In fact, the major
products were those coming from the latter type of
cleavage. This observation was explained by suggesting
the occurrence, both in 1•+ and in 2•+, of an intramolecu-
lar electron transfer thereby the hole is transferred from
the dialkoxylated ring (which would act as an electron
transfer internal mediator) to a monoalkoxylated ring.
The driving force for this energetically uphill process may
be provided by the very fast cleavage of the C-C bond â
to the positively charged monoalkoxylated ring. Apart
from the mechanistic interpretation, our results indicate
that in the oxidative degradation of lignin, C-C bonds,
which are not in a â position with respect to the ring
Exp er im en ta l Section
Su bstr a tes. The substrates 1 and 2 have been synthesized
by condensing the three aromatic units with two aldol-type
reactions as described in Scheme 4.25
The detailed description of the synthetic sequences is
available in Supporting Information. The relative anti config-
uration of the stereocenters has been assigned on the basis of
the well-documented25 preferential formation of the anti
product in the condensation of the lithium enolate of 14a or
14b with veratraldehyde or 4-methoxybenzaldehyde (step a
in Scheme 4) and in the condensation of the lithium enolate
of compound 20 with the dimeric aldehydes 17a or 17b (step
e in Scheme 4). The assignment of the relative anti configu-
ration for the two stereocenters formed in step a of Scheme 4
has been also confirmed by 1H and 13C NMR analysis of the
1,3-diol acetonides 17a and 17b (see Supporting Informa-
tion).25 1-(3,4-Dimethoxyphenyl)-2-phenoxy-1-ethanol (4) was
synthesized according to a literature procedure.41
P r od u ct s. 4-Methoxybenzaldehyde and 3,4-dimethoxy-
benzaldehyde are commercially available. The dimeric alde-
hydes
7 and 8 have been synthesized as precursors of
(48) Choinowski, T.; Blodig, W.; Winterhalter, K. H.; Piontek, K. J .
Mol. Biol. 1999, 286, 809.
(49) Bloding, W.; Doyle, W. A.; Smith A. T.; Winterhalter, K.;
Choinowski, T.; Piontek, K. Biochemistry 1998, 37, 8832.
(50) Doyle, W. A.; Bloding, W.; Veitch, N. C.; Piontek, K.; Smith, A.
T. Biochemistry 1998, 37, 15097.
compounds 1 and 2. The ketone 9 was synthesized starting
from the precursor 19b of compound 2 as described in the
Supporting Information.
En zym a tic Oxid a tion . The substrate 1 or 2 (125 µL of a
solution 0.038 M in acetonitrile) and LiP (0.4 units, 0.9 nmol)
9068 J . Org. Chem., Vol. 68, No. 23, 2003

Lignin Peroxidase-Catalyzed Oxidation
were added to 2 mL of sodium tartrate buffer solution (50 mM
at pH 3.5, previously degassed with argon for 15 min, under
magnetic stirring in a Schlenk tube at 25 °C). H2O2 (4.7 µmol)
in 0.8 mL of H2O was added gradually over a period of 90 min
by an infusion pump. At the end, 10 µL of a 0.1 M solution of
the internal standard, 4-methoxyacetophenone, in acetonitrile
was added. The reaction was quenched with methanol and
analyzed by HPLC and LC-MS or extracted with dichloro-
methane, dried over anhydrous Na2SO4, and, after evaporation
of the solvent, analyzed by 1H NMR analysis. No products were
observed in experiments carried out without the presence of
enzyme. The reaction of 1 led to 3,4-dimethoxybenzaldehyde
(5) and the dimeric aldehyde (7), which were recognized by
comparison with authentic specimens. Moreover, the LC-MS
spectrum of the mixture showed the presence of a compound
with M ) 482 (M + Na+ ) 505), which was assigned to the
ketone 6 also on the basis of the presence in the 1H NMR
spectrum of signals at 7.4-7.5 δ, attributable to the protons
ortho to the carbonyl group in the dimethoxylated ring. For
the quantitative determination (HPLC), the response factor
of 6 was taken equal to that of the ketone 9. The reaction of 2
led to the dimeric aldehyde 8, the ketone 9, and 4-methoxy-
benzaldehyde (10), which were recognized by comparison with
authentic specimens.
Ch em ica l Oxid a tion . The substrate 1 or 2 (125 µL of a
solution 0.038 M in acetonitrile) and CoIIIW (31 mg, 9.4 µmol)
were added to 2 mL of sodium tartrate buffer solution (50 mM
at pH 3.5, previously degassed with argon for 15 min under
magnetic stirring in a Schlenk tube). The mixture was heated
at 60 °C in an argon atmosphere for 48 h. At the end, 10 µL of
a 0.1 M solution of the internal standard, 4-methoxyacetophe-
none, in acetonitrile was added. The solution was extracted
with dichloromethane and dried over anhydrous Na2SO4, and
the solvent was evaporated. The products were analyzed by
1H NMR, HPLC, and LC-MS.
of products containing a carbonyl group in the R position with
respect to the dialkoxylated aromatic ring, that is, 3,4-
dimethoxybenzaldehyde (oxidation of 3), 3,4-dimethoxy-
benzaldehyde and 1-(3,4-dimethoxyphenyl)-2-phenoxy-1-etha-
none (oxidation of 4),41 and aldehyde 8 and ketone 9 (oxidation
of 2). Since two products are formed in the oxidation of 2 and
4, the rate of the process was calculated according to the
following equation v ) (dA/dt)(1/(ꢀA + ꢀBCB/CA)), where dA/dt
is the change of absorbance measured at 310 nm, ꢀA and ꢀB
are the molar extinction coefficient of reaction products A and
B at 310 nm, and CA and CB are the molar concentrations of
products A and B. Kinetic parameters were not determined
for the enzymatic oxidation of the trimer 1 because it was not
possible to isolate the trimeric ketone 6 from the reaction
mixture in a state of purity sufficient for the determination of
its extinction coefficient.
Kin etics w ith CoIIIW. A solution of 2, 3, or 4 (3 µmol) and
AcOK (0.6 mmol) in AcOH/H2O (70/30) was thoroughly purged
with argon in a 3-mL cuvette and placed in a thermostated
compartment (25 °C) of the UV-vis spectrophotometer. The
reaction was started by rapid addition of 0.1 mL of a CoIII
W
solution in AcOH/H2O (70/30) (6 mM). The rate of disappear-
ance of cobalt(III) was followed spectrophotometrically by
measuring the absorbance of CoIIIW at 390 nm (ꢀ ) 1207).30
First-order rate constants were obtained from the plot of log-
(At - Ainf)/(A0 - Ainf) against time: (9.9 ( 0.5) × 10-4 s-1 for 2,
(8.0 ( 0.4) × 10-4 s-1 for 3, and (7.6 ( 0.6) × 10-4 s-1 for 4.
Ack n ow led gm en t. Thanks are due to the Ministero
dell’Istruzione, dell’Universita` e della Ricerca and the
Consiglio Nazionale delle Ricerche for financial support.
Pulse radiolysis experiments were performed at the Free
Radical Research Facility of Daresbury, Cheshire, U.K.
with the support of the European Commission through
the Access to Large-Scale Facilities activity of the TMR
Program. Thanks also to Prof. Suppiah Navaratnam for
technical assistance.
En zym a tic Kin etics. The kinetics were performed in a
1-mL cuvette containing 800 mL of tartrate buffer (50 mM,
pH 3.5 at 25 °C). For each measure, 50 µL of a solution of LiP
4.6 mM in tartrate buffer (50 mM at pH 4) was added to the
buffer followed by the solution of substrates in acetonitrile
(final concentration 10%). The concentration of the substrates
varied from 0.18 to 1.4 mM for 3, from 0.11 to 0.56 mM for 4,
and from 0.10 to 0.52 mM for 2. The reaction started by
addition of 50 mL of a 0.012 M solution of H2O2. The kcat and
Su p p or tin g In for m a tion Ava ila ble: Pulse radiolysis
experiments, materials, syntheses of the trimeric lignin model
compounds 1, 2, and the ketone 9, and 13C NMR spectra of
compounds 1, 2, 17a , and 17b. This material is available free
of charge via the Internet at http://pubs.acs.org.
K
M values were determined at 310 nm following the formation
J O035052W
J . Org. Chem, Vol. 68, No. 23, 2003 9069