Báo cáo khoa học: The hepoxilin connection in the epidermis docx

MINIREVIEW
The hepoxilin connection in the epidermis
Alan R. Brash, Zheyong Yu*, William E. Boeglin and Claus Schneider
Department of Pharmacology, Vanderbilt University Medical Center, Nashville, TN, USA
Introduction
A relatively quiet area of eicosanoid research, the oxi-
dative metabolism of polyunsaturated fatty acids in
skin, recently jumped to prominence with the discovery
of a genetic connection between two lipoxygenase
(LOX) genes and a rare form of inherited ichthyosis
[1]. The findings gave life to a LOX enzyme thought to
be inactive and linked its function to the second LOX,
both of which appear to be essential in creating the
normal permeability barrier of the skin. Our recent
analysis of the catalytic activities of these proteins,
12R-LOX and eLOX3, suggested that the formation of
epoxy-hydroxy fatty acid derivatives (hepoxilins) could
be an integral part of the biological pathway disrupted
in the LOX-related form of ichthyosis and, by infer-
ence, responsible for the wellbeing of the epidermis in
all normal subjects [2].
From the discovery of essential fatty acids (EFAs)
around 1930, it has been clear that there is some con-
nection between these particular lipids and the proper
functioning of the water-impermeable barrier of the
epidermis [3,4]. One of the hallmark symptoms of
EFA deficiency is the development of a scaly skin phe-
notype, and this is cured by topical application of lino-
leic, arachidonic and other EFAs [4–8]. Although LOX
enzymes were not identified in animal tissues until the
mid-1970s, thereafter several lines of evidence sugges-

ted that the ameliorative effects of EFAs in the EFA-
deficient animal involve the LOX-catalyzed conversion
Keywords
arachidonic acid; epidermis; epoxyalcohol;
essential fatty acid; hepoxilin; ichthyosis;
linoleic acid; lipoxygenase; psoriasis; trioxilin
Correspondence
A. R. Brash, Department of Pharmacology
RRB Room 510, Vanderbilt University
Medical Center, 23rd Ave at Pierce,
Nashville, TN 37232-6602, USA
Fax: +1 615 3224707
Tel: +1 615 343 4495
E-mail:
*Present address
Howard Hughes Medical Institute, Washing-
ton University, St Louis, MO, USA
(Received 13 October 2006, accepted
12 March 2007)
doi:10.1111/j.1742-4658.2007.05909.x
The recent convergence of genetic and biochemical evidence on the activit-
ies of lipoxygenase (LOX) enzymes has implicated the production of hep-
oxilin derivatives (fatty acid epoxyalcohols) in the pathways leading to
formation of the water-impermeable barrier of the outer epidermis. The
enzymes 12R-LOX and eLOX3 are mutated in a rare form of congenital
ichthyosis, and, in vitro, the two enzymes function together to convert
arachidonic acid to a specific hepoxilin. Taken together, these lines of evi-
dence suggest an involvement of these enzymes and their products in skin
barrier function in all normal subjects. The natural occurrence of the speci-
fic hepoxilin products, and their biological role, whether structural or sign-

aling, remain to be defined.
Abbreviations
AA, arachidonic acid; EFA, essential fatty acid; HETE, hydroxyeicosatetraenoic acid; HPETE, hydroperoxyeicosatetraenoic acid; KETE,
ketoeicosatetraenoic acid; LOX, lipoxygenase; NCIE, nonbullous congenital ichthyosiform erythroderma.
3494 FEBS Journal 274 (2007) 3494–3502 ª 2007 The Authors Journal compilation ª 2007 FEBS
of the fatty acid substrate to oxygenated products
[9,10]. This is now considered to include further trans-
formation of the primary LOX products to epoxy-
alcohols (hepoxilins), triols and possibly also
x-hydroxylated derivatives. There may or may not be
a direct mechanistic connection between the effects of
EFAs and the recent findings on ichthyosis, but
together they reinforce the fact that there is an obliga-
tory involvement of EFAs and their products in the
normal process of forming the water-impermeable
barrier in skin. Herein is reviewed the connection of
EFAs, LOXs, hepoxilins and the structure–function of
the human epidermis.
The six human LOX genes
Of the six different LOXs in the human genome, the
best known by far is the 5-LOX of leukocytes, the
enzyme that gives rise to the leukotrienes, the inflam-
matory mediators implicated in asthma [11–13]. The
role of the other LOXs is less certain, although they
are known to be specific to certain cell types and to
produce distinct fatty acid hydroperoxide products
[12,14–16]. The prototypical cells of expression of one
group of LOX enzymes are distinct blood cell types,
namely the 5-LOX in leukocytes, the 12S-LOX in
platelets and the 15-LOX-1 in reticulocytes. Three

other LOX enzymes are epithelial cell-specific – and
two of these three (15-LOX-2 and 12R-LOX) were dis-
covered in human skin by our group [17,18]. The third,
cloned initially from mouse skin, is called eLOX3 [19].
In contrast to other family members, it is incapable of
forming fatty acid hydroperoxides [2,19] and therefore
has no ‘5-’, ‘12-’, or ‘15-’ designation to its name. It is
simply named from the fact that it was the third epi-
thelial LOX to be discovered – eLOX3. It is one of the
two LOX enzymes that are central to this review, the
other being 12R-LOX.
Ichthyosis has multiple genetic origins
The name ichthyosis comes from the Greek word for
fish. Taken together, the ichthyoses are a group of der-
matological conditions caused by genetic abnormalities
and characterized by a scaly skin phenotype. The dif-
ferent mutations typically cause problems associated
with construction of the water-impermeable barrier in
the outermost cornified layer of the epidermis. Ichthy-
osis vulgaris is the most common type (with an inci-
dence worldwide of 1 in 250), as implied by its name,
and is associated with mutations in the filaggrin gene
[20]; this leads to defects in formation of the cornified
cell envelope, an important component of the barrier
layer. Other types of ichthyosis typically have inci-
dences of 1 in 100 000 or less, and mutations in a
number of genes involved in barrier function have
been implicated [21]. Among the more recent additions
to the expanding list are the genes encoding 12R-LOX
and eLOX3.

12R-LOX and eLOX3 mutations: the
connection to ichthyosis
A major breakthrough in understanding the role of
LOX enzymes in skin was provided by the genetic
study of Fischer and colleagues in 2002 [1]. The
authors pinpointed a previously reported locus of
inherited ichthyosis on chromosome 17p [22] with
mutations in the coding regions of either the 12R-LOX
(ALOX12B) or eLOX3 (ALOXE3) genes [1]. The phe-
notype is classified as nonbullous congenital ich-
thyosiform erythroderma (NCIE, in layman’s terms
translating as a nonblistering, inherited, scaly red
skin). An independent study later extended these find-
ings following the identification of 17 additional famil-
ies with mutations in the same two LOX genes [23];
for classification of the disease, these authors used the
more general designation of autosomal-recessive con-
genital ichthyosis, of which NCIE could be considered
a subdivision [21]. Because one or the other LOX
genes was mutated in the affected families, producing
a similar phenotype, Fischer and colleagues speculated
that the two enzymes operate in the same metabolic
pathway [1]. This served to revamp our thinking on
the potential catalytic activities of the apparently non-
functional oxygenase, eLOX3.
LOX enzyme expression in epidermis
All the LOX genes are expressed in skin, as detected
by activity, immunohistochemistry and⁄ or PCR of the
mRNA. The activity of 12R -LOX, detected as 12R-
hydroxyeicosatetraenoic acid (12R-HETE) formation,

is quite low in normal human epidermis [24], in which
the dominant LOX activities are 12S-LOX and
15-LOX [24–27]. The synthesis of 12R-HETE is
strongly elevated in the inflammatory and proliferative
skin disease of psoriasis [28,29]. (The pro-inflammatory
bioactivity of 12R-HETE is quite weak [30,31] and we
speculate that its elevated synthesis is a result of the
keratinocyte hyperproliferation of psoriasis.) For many
years, the enzyme making 12R-HETE was unknown.
Then, in 1998, we reported the discovery of a 12R-
LOX in human skin and showed that it can account
for the selective formation of 12R-HETE [18]. Others
found that the mouse 12R-LOX is first expressed on
A. R. Brash et al. Hepoxilins in the epidermis
FEBS Journal 274 (2007) 3494–3502 ª 2007 The Authors Journal compilation ª 2007 FEBS 3495
embryonic day 15.5 at the location and time that the
epidermis is being formed [32].
The second LOX gene implicated in NCIE, eLOX3,
is strongly expressed in the epidermis, as indicated by
RT-PCR [33,34]. In addition, Krieg and coworkers
detected expression of the human mRNA in additional
tissues such as placenta, pancreas, ovary, testis, brain
and some secretory epithelia. In general, the expression
pattern of human and mouse eLOX3 was paralleled
by the expression of 12 R -LOX and was highest in the
skin [34].
eLOX3 as a hydroperoxide isomerase
(hepoxilin synthase)
The genetic findings concerning 12R-LOX and eLOX3
mutations in ichthyosis are intriguing from the bio-

chemical point of view, partly because the eLOX3 pro-
tein has been expressed and studied in at least two
laboratories, including our own, and no oxygenase
activity was detectable with any of a selection of
potential fatty acid substrates [2,19]. This conundrum,
the association of eLOX3 mutations with an ichthyosis
phenotype in the apparent absence of any LOX activ-
ity in the expressed eLOX3 protein, led us to examine
the possibility that the primary products of other LOX
enzymes were substrates for eLOX3. We found that,
indeed, eLOX3 will metabolize fatty acid hydroperox-
ides, although not through oxygenation as is typical of
LOX enzymes. eLOX3 reacts with the hydroperoxide
moiety and induces an isomerization of the hydro-
peroxide to specific epoxyalcohol (hepoxilin-type)
products and a ketoeicosatetraenoic acid (KETE) [2].
Among the three arachidonate-derived hydroperoxides
that are most likely to be found in human epidermis
[12S-hydroperoxyeicosatetraenoic acid (HPETE), 15S-
HPETE and 12R-HPETE – Fig. 1 shows the structures
of the main products] the best substrate was 12R-
HPETE. As this is formed by 12R-LOX, the other
gene implicated in the LOX-related form of NCIE, this
sets up a potential biochemical rationalization of the
genetic findings in NCIE [2].
An unusual aspect of eLOX3 catalysis is that its
activity is stimulated by typical LOX-reducing inhibi-
tors such as nordihydroguaiaretic acid. The explan-
ation is that, in contrast to LOX enzymes acting as
dioxygenases, the active form of the enzyme utilizes

the reduced, ferrous, form of the iron (Fig. 2). The
hepoxilin product contains both the original oxygen
atoms of the hydroperoxide substrate (Fig. 2) and thus
eLOX3 functions as a hydroperoxide isomerase [2].
It was implicit in the report on the link between
mutations in 12R-LOX and eLOX3 in NCIE that the
enzyme activities would be compromised [1]. It has
now been demonstrated experimentally that these and
more recently identified mutations in both 12R-LOX
and eLOX3 inactivate the enzymes [23,35]. Several
O
HO
O
HO
2
C
HO
O
HO
HO
2
C
HO
2
C
O
HO
HO
2
C

From 12
R
-HPETE :-
+ 12-KETE
+ 12-KETE
+ 15-KETE
From 12
S
-HPETE :-
From 15
S
-HPETE :-
Fig. 1. Structures of the epoxyalcohol products of eLOX3. The
structural analysis is given in a previous publication [2].
NDGA
12R-HPETE
OOH
128
12-KETE
O
3
1
4
2
O
O
OH
product
Epoxyalcohol
Fe

2+
Fe
3+
Fe
3+
-OH
Fe
3+
-OH
+ H
2
O
O
Fig. 2. Proposed mechanism for eLOX3 catalysis. The Fe
2+
enzyme
initiates homolytic cleavage of the O–O bond of the fatty acid
hydroperoxide, forming an Fe
3+
–OH complex and a substrate alk-
oxyl radical (RO
•
) (step 1). The alkoxyl radical instantly reacts with
the adjacent double bond, forming an epoxyallylic carbon radical
(step 2); this is hydroxylated by oxygen rebound from the Fe
3+
–OH
complex, thus completing the catalytic cycle (step 3). KETE is
formed as a minor by-product (step 4).
Hepoxilins in the epidermis A. R. Brash et al.

3496 FEBS Journal 274 (2007) 3494–3502 ª 2007 The Authors Journal compilation ª 2007 FEBS
mutations are remote from the LOX active site, and
we surmise that they may destabilize the protein, as
the mutants failed to accumulate during heterologous
expression in Escherichia coli [35].
A primer on hepoxilins and their
formation from HPETEs
As for the prostaglandins, leukotrienes and lipoxins, it
is useful to have a group name for the fatty acid
epoxyalcohols. The name ‘hepoxilin’, coined by Pace-
Asciak, serves a useful function in this regard, the first
three letters of the name (hep) standing for hydroxy-
epoxy; trioxilins are the corresponding trihydroxy
hydrolysis products [36–38]. Although applied strictly
only to derivatives of 12-HPETE, the hepoxilin termin-
ology is useful for describing two general classes of
epoxyalcohol, hepoxilin A-type and B-type (Fig. 3).
The terms hepoxilin A
3
or hepoxilin B
3
(‘3¢ for the
three double bonds in hepoxilins derived from
arachidonic acid) each refer to any of a mixture of
diastereomers and⁄ or enantiomers, and therefore the
hepoxilin nomenclature runs into difficulties when pre-
cise definition is required. In this review, individual
molecules are named with the hydroxyl and epoxide
configurations specified, as in the 12R-LOX–eLOX3
product, 8R-OH,11R,12R-epoxy-hepoxilin A

3
.
The transformation of fatty acid hydroperoxides to
epoxyalcohols is a facile nonenzymatic reaction, the
chemistry of which has been studied extensively and
found to be complex (reviewed in ref. 39). Free heme
or transition metals will initiate the reaction. For any
one fatty acid hydroperoxide, for example 12-HPETE,
there are three separate routes for conversion to
epoxyalcohols: the new hydroxy group can be formed
by rearrangement of the two hydroperoxide oxygens,
by the reaction of intermediates with O
2
, or by reac-
tion with water [39–42]. Nonenzymic reactions give a
mixture of hepoxilin A-type and B-type products with
a cis or trans epoxide and with R or S in stereochemis-
try of the hydroxyl group. From this knowledge of the
extensive nonenzymic possibilities for reaction, it is
easy to see why the appearance of a single distinct
epoxyalcohol isomer is taken as one hallmark denoting
the potential involvement of an enzyme.
The hepoxilin A-type epoxides are much more sensi-
tive to acid-catalyzed hydrolysis than the B-type
(Fig. 3), and they may also be more readily hydrolyzed
enzymatically. Accordingly, in biological extracts the
hepoxilin B-type epoxides are often detectable, whereas
the A-type are recovered as their trihydroxy hydrolysis
products.
Detection of hepoxilins and triols in

the epidermis
Nugteren and coworkers were the first to provide evi-
dence that LOX-derived epoxyalcohol and triol fatty
acids are important to the structure–function of the
epidermal water barrier [9]. They applied different
14
C-labeled unsaturated fatty acids onto the skin of
live fatty acid-deficient rats and followed the metabolic
fate over the course of 1–4 days. The radiolabeled sub-
strates [linoleic acid, or its trans ⁄ cis, cis ⁄ trans and
trans ⁄ trans isomers, or arachidonic acid (AA)] were
transformed through multiple pathways [9], including
incorporation into complex acylceramide lipids that
are a characteristic of the epidermal barrier layer
[43,44]. Formed specifically from arachidonic and lino-
leic acids (and not the trans isomers) were polyhydrox-
ylated fatty acid derivatives (epoxyalcohols and triols).
Their synthesis in vivo was blocked by co-application
of the LOX inhibitor eicosatetraynoic acid, and this
paralleled its inhibition of the ameliorative effects of
the applied EFA. The authors speculated that these
OOH
hepoxilin B-type
(more stable)
hepoxilin A-type
(easily hydrolyzed)
O
HO
8
12

8
R
,11
R
,12
R
-
hepoxilin A
3
O
HO
O
HO
O
HO
O
HO
O
HO
O
HO
O
HO
O
HO
O
HO
Fatty acid
hydroperoxide
Fig. 3. General structures of hepoxilin A-type and B-type. Nonenzy-

mic transformation from racemic hydroperoxide could lead to for-
mation of all the individual isomers shown (plus the corresponding
cis-epoxides, not shown). 12R-lipoxygenase (12R-LOX) and eLOX3
form exclusively the 8R-OH,11R,12R-epoxy-hepoxilin A
3
(boxed).
A. R. Brash et al. Hepoxilins in the epidermis
FEBS Journal 274 (2007) 3494–3502 ª 2007 The Authors Journal compilation ª 2007 FEBS 3497
LOX-derived products contributed to the formation of
the lamellar lipid phase that helps constitute the water-
impermeable barrier [44], or that they serve as a signal
to promote differentiation [9]. Notably, these LOX-
derived products could not be detected in normal epi-
dermis using a similar approach [45]. It remains an
open debate of whether LOX-derived products have
any structural role and ⁄ or act as specific signaling mol-
ecules in contributing to the epidermal water barrier.
Also awaiting clarification is linoleic acid metabolism
by 12R-LOX, the primary LOX enzyme strongly impli-
cated through genetic evidence as being involved in the
skin barrier function.
The biosynthesis of hepoxilin-type products and
their triol derivatives from
14
C-AA has been reported
in isolated human epidermal fragments [46]. Vila and
colleagues showed a predominant 12-LOX pathway of
metabolism leading to both hepoxilin A
3
-derived triols

and hepoxilin B
3
products. The results are of special
interest, not only for the characterization of specific
products, but also for the finding of predominant
12-LOX metabolism. By contrast, cultured human ker-
atinocytes are found typically to convert AA mainly
via a 15-LOX pathway [47,48]. In fact ‘12-LOX’ would
encompass 12S-LOX and 12R-LOX and ‘15-LOX’
could reflect 15-LOX-1 and ⁄ or 15-LOX-2. Each is rep-
resented in human epidermis [18,27,49] and the relative
proportions probably reflect differences in the site of
epidermis collection and the stage(s) of differentiation
of the keratinocytes.
Another significant finding was the well-documented
synthesis of a single predominant hepoxilin B
3
product
in epidermal fragments and in the microsomal fraction
[46,50]. This hepoxilin B
3
product had the same
GC-MS characteristics as the synthetic standard of
10R-hydroxy-11S,12S-hepoxilin B
3
[50], which is a
product we identified as specifically formed from 12S-
HPETE by eLOX3 [2]. (Note, however, that Anto
´
n&

Vila’s method could not distinguish between this hep-
oxilin B
3
and its enantiomer, 10S-hydroxy-11R,12R-
hepoxilin B
3
.) The product was formed from AA in
epidermal microsomes and at a much lower yield using
recombinant platelet-type 12S-LOX. Formation from
12S-HPETE could not be demonstrated. In the
absence of other candidate enzymes (the activity of
eLOX3 being unknown at the time), the authors con-
cluded that 12S-LOX is probably the hepoxilin B
3
syn-
thase [50]. This work ranks as one of the very few in
which a single hepoxilin diastereomer has been shown
to be produced in mammalian cells or tissue. Produc-
tion of a single diastereomer (as opposed to an equi-
molar mixture of a diastereomeric pair) probably
denotes its enzymatic synthesis. Vila and colleagues
went on to use GC-MS to demonstrate the presence of
CO
2
H
CO
2
H
CO
2

H
CO
2
H
OOH
CO
2
H
OH
CO
2
H
O
HO
O
HO HO OH
+
Arachidonic acid
12R-HPETE
12R-HETE
8R,11R,12R-epoxyalcohol
12-KETE
8,11,12-triol
12R-LOX
eLOX3
epoxide
hydrolase
peroxidase
mutations
mutations

CH
2
OH
omega-oxidation
products
CYP 4F22
?
?
CYP 4F22
Fig. 4. Proposed metabolism of arachidonic
acid (AA) in human epidermis through the
12R-lipoxygenase (12R-LOX)–eLOX3 path-
way. The putative x-hydroxylase, CYP4F22,
was able to react at other points on
the pathway [e.g. with AA or 12R-
hydro(pero)xyeicosatetraenoic acid (12R-
HPETE)], or with unrelated fatty acids (see
the text under Recent developments).
Hepoxilins in the epidermis A. R. Brash et al.
3498 FEBS Journal 274 (2007) 3494–3502 ª 2007 The Authors Journal compilation ª 2007 FEBS
elevated levels of endogenous hepoxilins and trioxilins
in human psoriatic scales [50,51]. One aspect of these
observations that should be revisited is the occurrence
of a novel hepoxilin isomer in psoriatic scales. This
putative hepoxilin B
3
isomer was separated from the
authentic standards on GC-MS [51]. The finding of the
product specifically in psoriasis is intriguing and
worthy of further study.

eLOX3 converts fatty acid hydroperoxides to
a KETE byproduct in addition to the hepoxilins
(Figs 1, 2 and 4) [2]. This type of unsaturated ketone
is highly reactive with cellular nucleophiles, such as
glutathione [52], and consequently it may not appear
as a peak of KETE upon chromatography of cell
extracts. Although reaction with glutathione is usually
considered a pathway of inactivation, some derivatives
are bioactive (e.g. leukotrienes); the fate and potential
bioactivity of the eLOX3-derived ketones remains to
be evaluated.
Recent developments: further aspects
of the LOX pathway
Two independent reports describe a dramatic pheno-
type associated with deletion of the 12R-LOX gene in
mice [53,54]. The homozygous– ⁄ – neonates die within
hours of birth because of excessive transepidermal
water loss. As is typical for gene defects that disrupt
barrier function, the phenotype is more severe in the
mouse compared with the human as a result of the
much larger surface to volume ratio, but generally is
consistent with the findings in the ichthyosis patients.
Although the gross morphology of the epidermis was
not affected, the upper granular layer of the skin in
the knockout animals showed evidence of disruption
of the normal processing of the lipid-rich lamellar bod-
ies that play a crucial role in formation of the water-
impermeable barrier. These studies add a convincing
new line of evidence for a key role of 12R-LOX in the
normal functioning of the epidermis and, furthermore,

provide models in which the mechanism of action can
be investigated. The findings do, however, present a
conundrum for the proposed role of hepoxilins in
maintaining the epidermal water barrier. Whereas
mouse eLOX3 has the required activity with fatty acid
hydroperoxides, mouse 12R-LOX is very unusual in
apparently lacking oxygenase activity with AA or any
other polyunsaturated lipid tested to date [55–57].
Arachidonate methyl ester is metabolized to the corres-
ponding 12R-hydroperoxide, but the methyl ester is
not naturally occurring. Thus, there remains an open
question of whether mouse 12R -LOX could generate
a fatty acid hydroperoxide with a suitable natural
substrate, or whether the protein functions in some
other way to promote the correct differentiation of the
epidermis.
Meanwhile, the geneticists continue to provide pro-
vocative new insights. The association of mutations in
a putative membrane protein giving a similar pheno-
type as in the LOX-related form of NCIE led Fischer
and colleagues to speculate that it constitutes a hepoxi-
lin or trioxilin receptor [58]. They named this new gene
ichthyin. Another new gene implicated in ichthyosis by
the same group encodes a cytochrome P450 [59]; it is
classified as CYP4F22 [60], an uncharacterized member
of the CYP4 family, which are generally fatty acid
x-hydroxylases. Again, based on the similarity to the
ichthyosis phenotype, the genetics group questioned if
this potential x-hydroxylase is involved in producing
the biologically active end-product of the LOX path-

way. There is biochemical precedent for the P450
x-hydroxylases having substrate specificity for an
oxygenated fatty acid: thus, CYP4F8 efficiently
x-hydroxylates the prostaglandin endoperoxide PGH
2
,
whereas AA and PGE
2
are comparatively feeble sub-
strates [61]. Similarly, several CYP4A isoforms more
efficiently x-hydroxylate epoxyeicosatrienoic acids than
AA [62]. Perhaps the active principal of the epidermal
LOX pathway is x-hydroxylated hepoxilin or trioxilin
(Fig. 4). Incidentally, the linoleate-containing acylcera-
mide of the epidermal barrier layer is composed of
sphingosine in amide linkage to the carboxyl group of
a very long x-hydroxy acid (mainly C
30
,C
32
,orC
34
chain length), which in turn is esterified to the carb-
oxyl group of linoleate; the enzyme responsible for the
x-hydroxylation of the long-chain acid is uncharacter-
ized and could be the P450 enzyme that Lefevre et al.
have implicated in ichthyosis [59].
Interestingly, the hepoxilin derived via 12R-LOX
and eLOX3 is hydrolyzed specifically in keratinocytes
to a single triol, tentatively identified as 8R,11S,12R-

trihydroxyeicosa-5Z,9E,14Z-trienoic acid formed by
S
N
2 hydrolysis of the epoxide at C-11 [35]. In human
keratinocytes, this hydrolase activity may be a down-
stream enzyme in the pathway consisting of 12R-LOX
and eLOX3 to form an active mediator in the regula-
tion of keratinocyte differentiation (Fig. 4). A bioac-
tive trioxilin is precedented: in vascular endothelial
cells a specific triol is implicated as one of the endo-
thelium-derived hyperpolarizing factors [63].
Concluding remarks ) unresolved
issues
Linoleate is usually considered to be a structural
component of the ceramides in the stratum corneum
A. R. Brash et al. Hepoxilins in the epidermis
FEBS Journal 274 (2007) 3494–3502 ª 2007 The Authors Journal compilation ª 2007 FEBS 3499
[6,8], whereas arachidonate is viewed currently as the
initial substrate of the hepoxilin signaling pathway
[1,2,23]. So, it is questionable as to whether the EFA-
deficiency phenotype is attributable to lack of the
LOX products that are missing in the LOX-depend-
ent ichthyosis. This highlights the two poles in cur-
rent views on the activities of EFA in the epidermis:
structural and ⁄ or signaling. Part of this debate con-
cerns whether hepoxilin-type derivatives of linoleate
itself might function in a critical role in the epidermal
barrier function.
The recent 12R-LOX knockout studies in mice leave
little doubt about the crucial involvement of this LOX

gene, but raise issues of its proposed metabolic coup-
ling with eLOX3 (which itself is known to be critical
in human genetic studies [1,23]). Identification of a
natural substrate for the mouse 12R-LOX or condi-
tions under which it exhibits oxygenase activity will be
necessary to substantiate the hepoxilin connection to
epidermal differentiation.
With regard to the LOX-dependent phenotype, the
nature, occurrence and bioactivity of the active prod-
ucts remain to be defined. Currently, the lack of
authentic hepoxilins is a significant holdup. The prime
candidates for assay and for pharmacological testing
include the hepoxilin derivatives of eLOX3, the corres-
ponding trihydroxy hydrolysis products and their
x-hydroxylated derivatives. Regarding mechanism of
action, this is also currently an open issue. Over
20 years ago the main proponents of the LOX connec-
tion in barrier function noted either a structural [44] or
a signaling function in differentiation [9], and we are
not much further advanced in defining the mechanism
today. Genetic analyses of mutant skin phenotypes
have paved the way for unraveling the LOX ⁄ hepoxilin
pathway in the epidermis, and they continue to pro-
vide fresh impetus with identification of the putative
receptor protein, ichthyin. Defined candidates are on
the table and the search is on to determine their
involvement.
Acknowledgements
This work was supported by NIH grant AR51968 to
ARB.

References
1 Jobard F, Lefe
`
vre C, Karaduman A, Blanchet-Bardon
C, Emre S, Weissenbach J, O
¨
zgu
¨
c M, Lathrop M,
Prud’homme JF & Fischer J (2002) Lipoxygenase-3
(ALOXE3) and 12(R)-lipoxygenase (ALOX12B) are
mutated in non-bullous congenital ichthyosiform eryth-
roderma (NCIE) linked to chromosome 17p13.1. Hum
Mol Genet 11, 107–113.
2 Yu Z, Schneider C, Boeglin WE, Marnett LJ & Brash
AR (2003) The lipoxygenase gene ALOXE3 implicated
in skin differentiation encodes a hydroperoxide isom-
erase. Proc Natl Acad Sci USA 100, 9162–9167.
3 Burr GO & Burr MM (1929) A new deficiency disease
produced by the rigid exclusion of fat from the diet.
J Biol Chem 82, 345–367.
4 Burr GO & Burr MM (1930) On the nature and roˆ le of
the fatty acids essential in nutrition. J Biol Chem 86,
587–621.
5 Burr GO, Burr MM & Miller ES (1932) On the fatty
acids essential in nutrition. III. J Biol Chem 97, 1–9.
6 Elias PM, Brown BE & Ziboh VA (1980) The permeab-
ility barrier in essential fatty acid deficiency: evidence
for a direct role for linoleic acid in barrier function.
J Invest Dermatol 74, 230–233.

7 Houtsmuller UMT & van der Beek A (1981) Effects of
topical application of fatty acids. Prog Lipid Res 20,
219–224.
8 Wertz PW, Swartzendruber DC, Abraham W, Madison
KC & Downing DT (1987) Essential fatty acids and epi-
dermal integrity. Arch Dermatol 123, 1381–1384.
9 Nugteren DH, Christ-Hazelhof E, van der Beek A &
Houtsmuller UM (1985) Metabolism of linoleic acid
and other essential fatty acids in the epidermis of the
rat. Biochim Biophys Acta 834, 429–436.
10 Nugteren DH & Kivits GA (1987) Conversion of lino-
leic acid and arachidonic acid by skin epidermal lipoxy-
genases. Biochim Biophys Acta 921, 135–141.
11 Samuelsson B, Dahlen SE, Lindgren JA, Rouzer CA &
Serhan CN (1987) Leukotrienes and lipoxins: structures,
biosynthesis, and biological effects. Science 237, 1171–
1176.
12 Funk CD (2001) Prostaglandins and leukotrienes:
advances in eicosanoid biology. Science 294, 1871–1875.
13 Drazen JM, Israel E & O’Byrne PM (1999) Treatment
of asthma with drugs modifying the leukotriene path-
way. N Engl J Med 340, 197–206.
14 Funk CD (1996) The molecular biology of mammalian
lipoxygenases and the quest for eicosanoid functions
using lipoxygenase-deficient mice. Biochim Biophys Acta
1304, 65–84.
15 Brash AR (1999) Lipoxygenases: occurrence, functions,
catalysis, and acquisition of substrate. J Biol Chem 274,
23679–23682.
16 Ziboh VA, Miller CC & Cho Y (2000) Significance of

lipoxygenase-derived monohydroxy fatty acids in cutane-
ous biology. Prostaglandins Other Lipid Med 63, 3–13.
17 Brash AR, Boeglin WE & Chang MS (1997) Discovery
of a second 15S-lipoxygenase in humans. Proc Natl
Acad Sci USA 94, 6148–6152.
18 Boeglin WE, Kim RB & Brash AR (1998) A 12R-lip-
oxygenase in human skin: Mechanistic evidence,
Hepoxilins in the epidermis A. R. Brash et al.
3500 FEBS Journal 274 (2007) 3494–3502 ª 2007 The Authors Journal compilation ª 2007 FEBS
molecular cloning and expression. Proc Natl Acad Sci
USA 95, 6744–6749.
19 Kinzig A, Heidt M, Fu
¨
rstenberger G, Marks F & Krieg
P (1999) cDNA cloning, genomic structure, and chro-
mosomal localization of a novel murine epidermis-type
lipoxygenase. Genomics 58, 158–164.
20 Smith FJ, Irvine AD, Terron-Kwiatkowski A, Sandi-
lands A, Campbell LE, Zhao Y, Liao H, Evans AT,
Goudie DR, Lewis-Jones S et al. (2006) Loss-of-func-
tion mutations in the gene encoding filaggrin cause ich-
thyosis vulgaris. Nat Genet 38, 337–342.
21 Akiyama M (2006) Harlequin ichthyosis and other auto-
somal recessive congenital ichthyoses: The underlying
genetic defects and pathomechanisms. J Dermatol Sci
42, 83–89.
22 Krebsova A, Kuster W, Lestringant GG, Schulze B,
Hinz B, Frossard PM, Reis A & Hennies HC (2001)
Identification, by homozygosity mapping, of a novel
locus for autosomal recessive congenital ichthyosis on

chromosome 17p, and evidence for further genetic het-
erogeneity. Am J Hum Genet 69, 216–222.
23 Eckl KM, Krieg P, Kuster W, Traupe H, Andre F,
Wittstruck N, Furstenberger G & Hennies HC (2005)
Mutation spectrum and functional analysis of
epidermis-type lipoxygenases in patients with auto-
somal recessive congenital ichthyosis. Hum Mutat 26,
351–361.
24 Baer AN & Green FA (1993) Fatty acid oxygenase
activity of human hair roots. J Lipid Res 34, 1505–1514.
25 Baer AN, Costello PB & Green FA (1991) In vivo acti-
vation of an w-6 oxygenase in human skin. Biochem
Biophys Res Commun 180, 98–104.
26 Takahashi Y, Reddy GR, Ueda N, Yamamoto S &
Arase S (1993) Arachidonate 12-lipoxygenase of plate-
let-type in human epidermal cells. J Biol Chem 268,
16443–16448.
27 Hussain H, Shornick LP, Shannon VR, Wilson JD,
Funk CD, Pentland AP & Holtzman MJ (1994) Epider-
mis contains platelet-type 12-lipoxygenase that is over-
expressed in germinal layer keratinocytes in psoriasis.
Am J Physiol 266, C243–C253.
28 Hammarstro
¨
m S, Hamberg M, Samuelsson B, Duell
EA, Stawiski M & Voorhees JJ (1975) Increased con-
centrations of nonesterified arachidonic acid, 12L-hyd-
roxy-5,8,10,14-eicosatetraenoic acid, prostaglandin E
2
,

and prostaglandin F
2a
in epidermis of psoriasis. Proc
Natl Acad Sci USA 72, 5130–5134.
29 Woollard PM (1986) Stereochemical difference between
12-hydroxyeicosa-tetraenoic acid in platelets and psori-
atic lesions. Biochem Biophys Res Commun 136, 169–
176.
30 Cunningham FM & Woollard PM (1987) 12(R)-Hyd-
roxy-5,8,10,14-eicosatetraenoic acid is a chemoattractant
for human polymorphonuclear leucocytes in vitro.
Prostaglandins 34, 71–78.
31 Fretland DJ, Widomski DL, Zemaitis JM, Tsai BS,
Djuric SW, Penning TD, Miyashiro JM & Bauer RF
(1989) 12(R)-hydroxyeicosatetraenoic acid is a neutro-
phil chemoattractant in the cavine, lapine, murine and
canine dermis. Prostaglandins Leukot Essent Fatty Acids
37, 79–81.
32 Sun D, McDonnell M, Chen X-S, Lakkis MM, Li H,
Isaacs SN, Elsea SH, Patel PI & Funk CD (1998)
Human 12(R)-lipoxygenase and the mouse ortholog.
Molecular cloning, expression, and gene chromosomal
assignment. J Biol Chem 273, 33540–33547.
33 Heidt M, Fu
¨
rstenberger G, Vogel S, Marks F & Krieg
P (2000) Diversity of mouse lipoxygenases: identification
of a subfamily of epidermal isozymes exhibiting a differ-
entiation-dependent mRNA expression pattern. Lipids
35, 701–707.

34 Krieg P, Marks F & Fu
¨
rstenberger G (2001) A gene
cluster encoding human epidermis-type lipoxygenases at
chromosome 17p13.1: cloning, physical mapping, and
expression. Genomics 73, 323–300.
35 Yu Z, Schneider C, Boeglin WE & Brash AR (2005)
Mutations associated with a congenital form of ichthy-
osis (NCIE) inactivate the epidermal lipoxygenases
12R-LOX and eLOX3. Biochim Biophys Acta 1686,
238–247.
36 Pace-Asciak CR & Asotra S (1989) Biosynthesis, cata-
bolism, and biological properties of HPETEs, hydroper-
oxide derivatives of arachidonic acid. Free Rad Biol
Med 7, 409–433.
37 Pace-Asciak CR, Reynaud D, Demin P & Nigam S
(1999) The hepoxilins. A review. Adv Exp Med Biol 447,
123–132.
38 Pace-Asciak CR, Reynaud D & Demin PM (1995) Hep-
oxilins: a review on their enzymatic formation, metabo-
lism and chemical synthesis. Lipids 30, 107–114.
39 Gardner HW (1989) Oxygen radical chemistry of poly-
unsaturated fatty acids. Free Rad Biol Med 7, 65–86.
40 Pace-Asciak CR (1984) Arachidonic acid epoxides.
Demonstration through [18O]oxygen studies of an intra-
molecular transfer of the terminal hydroxyl group of
(12S)-hydroperoxyeicosa-5,8,10,14-tetraenoic acid to
form hydroxyepoxides. J Biol Chem 259, 8332–8337.
41 Dix TA & Marnett LJ (1985) Conversion of linoleic
acid hydroperoxide to hydroxy, keto, epoxyhydroxy,

and trihydroxy fatty acids by hematin. J Biol Chem 260,
5351–5357.
42 Chang MS, Boeglin WE, Guengerich FP & Brash AR
(1996) Cytochrome P450-dependent transformations of
15R- and 15S-hydroperoxyeicosatetraenoic acids: Stereo-
selective formation of epoxy alcohol products. Bio-
chemistry 35, 464–471.
43 Wertz PW & Downing DT (1983) Ceramides of pig epi-
dermis: structure determination. J Lipid Res 24, 759–765.
44 Bowser PA, Nugteren DH, White RJ, Houtsmuller
UMT & Prottey C (1985) Identification, isolation and
A. R. Brash et al. Hepoxilins in the epidermis
FEBS Journal 274 (2007) 3494–3502 ª 2007 The Authors Journal compilation ª 2007 FEBS 3501
characterization of epidermal lipids containing linoleic
acid. Biochim Biophys Acta 834, 419–428.
45 Wertz PW & Downing DT (1990) Metabolism of lino-
leic acid in porcine epidermis. J Lipid Res 31, 1839–
1844.
46 Anto
´
n R, Abian J & Vila L (1995) Characterization of
arachidonic acid metabolites through the 12-lipoxyge-
nase pathway in human epidermis by high-performance
liquid chromatography and gas chromatography ⁄ mass
spectrometry. J Mass Spectrom Rapid Commun Mass
Spectrom (suppl.) S169–182.
47 Burrall BA, Wintroub BU & Goetzl EJ (1985) Selective
expression of 15-lipoxygenase activity by cultured
human keratinocytes. Biochem Biophys Res Commun
133, 208–213.

48 Green FA (1989) Generation and metabolism of lip-
oxygenase products in normal and membrane-damaged
cultured human keratinocytes. J Invest Dermatol 93,
486–491.
49 Brash AR, Jisaka M, Boeglin WE & Chang MS (1997)
Molecular cloning of a second human 15S-lipoxygenase
and its murine homologue, an 8S-lipoxygenase: their
relationship to other mammalian lipoxygenases. In
Lipoxygenases and Their Products: Biological Functions
(Pace-Asciak CR & Nigam S, eds), pp. 29–36. Plenum
Press, New York, NY.
50 Anto
´
n R & Vila L (2000) Stereoselective biosynthesis of
hepoxilin B3 in human epidermis. J Invest Dermatol
114, 554–559.
51 Anto
´
n R, Puig L, Esgleyes T, de Moragas JM & Vila L
(1998) Occurrence of hepoxilins and trioxilins in psoriatic
lesions. J Invest Dermatol 110, 303–310.
52 Murphy RC & Zarini S (2002) Glutathione adducts of
oxyeicosanoids. Prostaglandins Other Lipid Med 68–69,
471–482.
53 Moran JL, Qiu H, Turbe-Doan A, Yun Y, Boeglin WE,
Brash AR & Beier DR (2007) A mouse mutation in the
12(R)-lipoxygenase, Alox12b, disrupts formation of
the epidermal permeability barrier. J Invest Dermatol,
doi: 10.1038/sj.jid.5700825.
54 Epp N, Fu

¨
rstenberger G, Mu
¨
ller K, de Juanes S,
Leitges M, Hausser I, Thieme F, Liebisch G, Schmitz G
& Krieg P (2007) 12R-lipoxygenase deficiency disrupts
epidermal barrier function. J Cell Biol 177, 173–182.
55 Krieg P, Siebert M, Kinzig A, Bettenhausen R, Marks
F&Fu
¨
rstenberger G (1999) Murine 12(R)-lipoxygenase:
functional expression, genomic structure and chromoso-
mal localization. FEBS Lett 446, 142–148.
56 Siebert M, Krieg P, Lehmann WD, Marks F & Fu
¨
rsten-
berger G (2001) Enzymic characterization of epidermis-
derived 12-lipoxygenase isoenzymes. Biochem J 355,
97–104.
57 Meruvu S, Walther M, Ivanov I, Hammarstrom S, Fur-
stenberger G, Krieg P, Reddanna P & Kuhn H (2005)
Sequence determinants for the reaction specificity of
murine (12R)-lipoxygenase: Targeted substrate modifica-
tion and site-directed mutagenesis. J Biol Chem 280,
36633–36641.
58 Lefe
`
vre C, Bouadjar B, Karaduman A, Jobard F,
Saker S, Ozguc M, Lathrop M, Prud’homme JF &
Fischer J (2004) Mutations in ichthyin a new gene on

chromosome 5q33 in a new form of autosomal reces-
sive congenital ichthyosis. Hum Mol Genet 13 , 2473–
2482.
59 Lefe
`
vre C, Bouadjar B, Ferrand V, Tadini G,
Me
´
garbane
´
A, Lathrop M, Prud’homme JF &
Fischer J (2006) Mutations in a new cytochrome P450
gene in lamellar ichthyosis type 3. Hum Mol Genet 15,
767–776.
60 Nelson DR, Zeldin DC, Hoffman SM, Maltais LJ,
Wain HM & Nebert DW (2004) Comparison of cyto-
chrome P450 (CYP) genes from the mouse and human
genomes, including nomenclature recommendations
for genes, pseudogenes and alternative-splice variants.
Pharmacogenetics 14, 1–18.
61 Bylund J, Hidestrand M, Ingelman-Sundberg M & Oliw
EH (2000) Identification of CYP4F8 in human seminal
vesicles as a prominent 19-hydroxylase of prostaglandin
endoperoxides. J Biol Chem 275, 21844–21849.
62 Cowart LA, Wei S, Hsu MH, Johnson EF, Krishna
MU, Falck JR & Capdevila JH (2002) The CYP4A iso-
forms hydroxylate epoxyeicosatrienoic acids to form
high affinity peroxisome proliferator-activated receptor
ligands. J Biol Chem 277, 35105–35112.
63 Pfister SL, Spitzbarth N, Zeldin DC, Lafite P, Mansuy D

& Campbell WB (2003) Rabbit aorta converts 15-HPETE
to trihydroxyeicosatrienoic acids: potential role of cyto-
chrome P450. Arch Biochem Biophys 420, 142–152.
Hepoxilins in the epidermis A. R. Brash et al.
3502 FEBS Journal 274 (2007) 3494–3502 ª 2007 The Authors Journal compilation ª 2007 FEBS