(Received for publication, April 8, 1997, and in revised form, May 6, 1997)
From the Departments of Pharmacology and Pathology and the Center for Neurobiology and Behavior, College of Physicians & Surgeons, Columbia University, New York, New York 10032
Arachidonic acid is converted to (8R)-hydroperoxyeicosa-5,9,11,14-tetraenoic acid (8-HPETE) during incubations with homogenates of the central nervous system of the marine mollusc, Aplysia californica. 8-HPETE can be reduced to the corresponding hydroxy acid or be enzymatically converted to a newly identified metabolite, 8-ketoeicosa-5,9,11,14-tetraenoic acid (8-KETE). These metabolites were identified by high performance liquid chromatography, UV absorbance, and gas chromatography/mass spectrometry. Stereochemical analysis of the products demonstrate that the neuronal enzyme is an (8R)-lipoxygenase. Previously we have shown that the neurotransmitters, histamine and Phe-Met-Arg-Phe-amide, activate 12-lipoxygenase metabolism in isolated identified Aplysia neurons. We now show that acetylcholine activates the (8R)-lipoxygenase pathway within intact nerve cells. Thus, both (12S)- and (8R)-lipoxygenase co-exist in intact Aplysia nervous tissue but are differentially activated by several neurotransmitters. The precise physiological role of the 8-lipoxygenase products is currently under investigation, but by analogy to the well-described 12-lipoxygenase pathway, we suggest that (8R)-HPETE and 8-KETE may serve as second messengers in Aplysia cholinoceptive neurons.
We previously found that the 12-lipoxygenase pathway is activated by histamine and FMRF-amide,1 two neurotransmitters in identified Aplysia neurons (1-3). Activation occurs through receptor-mediated release of arachidonic acid by a phospholipase and is blocked by pertussis toxin (4, 5), indicating that arachidonate metabolism is regulated by a G protein. In electrophysiological studies, application of (12S)-lipoxygenase metabolites, including (12S)-hydroperoxyeicosatetraenoic acid ((12S)-HPETE), mimics the action of histamine (2, 6) and modulates the action of FMRF-amide (7-9), suggesting that these lipids function as second messengers. There is evidence that lipoxygenase metabolites also act as second messengers in the mammalian nervous system as well as in Aplysia. Bliss et al. (10, 11) showed that inhibiting lipoxygenase with nordihydroguaiaretic acid, a nonspecific antioxidant, interferes with the development of long term potentiation (LTP) in rat hippocampal slices. More recently, we found that a specific lipoxygenase inhibitor with no antioxidant activity also inhibits LTP (12), and others have reported that LTP is modulated by applying 12-HPETE to hippocampal slices (13). Carlen et al. (14) demonstrated that hippocampal homogenates can convert 12-HPETE to the epoxy alcohol, hepoxilin A3. The hepoxilin hyperpolarizes the cell and increases the amplitude and duration of inhibitory postsynaptic potentials when applied to CA1 pyramidal neurons in rat hippocampal slices (14). Our earlier work in Aplysia nervous tissue suggested that the biologically active epoxy alcohols are formed enzymatically (6). Additional evidence supporting enzymatic conversion of 12-HPETE to hepoxilin was obtained from studies of the rat pineal gland (15).
Lipoxygenases are widely distributed and highly conserved enzymes found in animals and plants. These enzymes have been cloned and have four different positional specificities, 5, 8, 12, and 15 (see Ref. 16 for a review). The 12-lipoxygenase appears in two distinct forms called the platelet type and the leukocyte type. Our biochemical studies have established that Aplysia neurons contain a 12-lipoxygenase with characteristics most like the mammalian leukocyte-type enzyme which is also the form present in vertebrate brain (17). While similar in many respects, the Aplysia enzyme differs in being particulate, however (18).
We reported the isolation of several biologically active lipids from Aplysia nervous tissue derived from the 12-lipoxygenase pathway including hepoxilin A3 and 12-ketoeicosatetraenoic acid (12-KETE) (2, 6). In addition, several metabolites remained to be identified. In the course of purifying radiolabeled lipoxygenase products from Aplysia nervous tissue, we noticed that a substantial fraction of the radioactivity did not correspond to known products of the 12-lipoxygenase pathway. We therefore examined the possibility that Aplysia neurons contain some related metabolic pathway in addition to the 12-lipoxygenase. We have found that Aplysia neurons contain an 8-lipoxygenase that converts arachidonic acid to (8R)-HPETE and a second enzyme that metabolizes the hydroperoxy acid to 8-ketoeicosatetraenoic acid (8-KETE). 8-Lipoxygenases have been found in coral (19, 20), crustacea (21), and echinoderms (22, 23). This is the first description of this pathway in the nervous system. The 8-lipoxygenase is activated in neurons by acetylcholine but not by histamine indicating specificity in signal transduction mechanisms linked to lipoxygenases.
Aplysia californica, weighing between 60 and 100 g from Marinus, Long Beach, CA, were maintained in aquaria at 15 °C prior to use. Central ganglia were removed from animals anesthetized with MgCl2 as described by Schwartz and Swanson (24). Ganglia, transferred in a small volume of cold artificial sea water (ASW (25)), were desheathed by dissection on Petri dishes coated with a silicone plastic (Sylgard) under a dissecting microscope.
Nervous Tissue PreparationsAfter removing the connective tissue sheath, we collected neural components (cell bodies and neuropil) with a forceps for homogenization in a Duall size 20 ground glass tissue homogenizer (Kontes) containing supplemented low-salt buffer (50 mM Tris-HCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 1 mM benzamidine, 0.1 mM leupeptin, 0.1 mM phenylmethylsulfonyl fluoride, 50 µg/ml aprotinin, and 20% glycerol, pH 7.6; 200 µl/3 animals). The homogenate was transferred to polypropylene microcentrifuge tubes and centrifuged for 5 min at 80 × g at 4 °C. The resulting supernatant was diluted in ASW and incubated as described below or fractionated further by centrifugation.
In some experiments, the 80 × g supernatant was transferred to polyallomer ultracentrifuge tubes for centrifugation at either 50,000 × g or 100,000 × g for 40 min at 4 °C. Supernatant fractions were assayed either directly or after dilution in ASW. Similarly, pellets were resuspended in ASW containing the protease inhibitors using a tuberculin syringe and assayed in ASW. Typically, neural components prepared by this method will yield 1 mg of protein per animal, and about half of this will be found in the high speed pellet (24).
Incubation of Tissue FractionsTypically, samples of the
80 × g supernatant (50 µl) were added to glass tubes
containing [3H]arachidonic acid (2 µCi, 200-230
Ci/mmol) and 13-hydroperoxyoctadecadienoic acid (2 µg) and brought to
a final volume of 1 ml with ASW and incubated for various times at
15 °C. The reactions were stopped by adding cold acetone (2 ml) and
stored at 20 °C for at least 30 min before extraction. In
experiments which included 13-hydroperoxyoctadecadienoic acid, the
reduction product (13-hydroxyoctadecadienoic acid) was used as an
internal standard to normalize recovery. Otherwise, 8- or 12-HETE was
added before extraction. The recovery of internal standard was measured
by UV absorbance at 235 nm.
In other experiments, homogenates were incubated with a mixture of [1-14C]- or [U-14C](8R)-HPETE (100,000 cpm) and unlabeled (8R)-HPETE (885 ng). The contribution of mass by the radiolabeled (8R)-HPETE was negligible. In some experiments, [14C](8S)-HPETE was substituted for (8R)-HPETE. In others, the added Aplysia nervous tissue extract was boiled (10 min) prior to the incubation, or hematin (5 µM) was added instead of tissue. All incubations were for 10 min at 15 °C, stopped by adding cold acetone, extracted with diethyl ether, and analyzed by RP- and SP-HPLC.
Lipid ExtractionWe centrifuged reaction mixtures for 5 min
at 150 × g at 4 °C. The resulting supernatants were
decanted into fresh tubes and, as noted in some experiments, 2 drops of
trimethylphosphite (TMP) were added to reduce hydroperoxides to the
corresponding hydroxy acids. Lipids were extracted with diethyl ether
(26). Briefly, supernatants were acidified to pH 3.5 with HCl, mixed
with 6 volumes of diethyl ether and 4 volumes of water. The resulting
organic phase was washed with 2 volumes of water and then with 1 volume. The ether extract was dehydrated by washing with brine (1 volume) and then adding anhydrous magnesium sulfate. After it was
filtered (Whatman No. 4 filter paper), the lipid extract was collected and evaporated to dryness under reduced pressure. Samples were dissolved in ethanol and stored at 70 °C.
Lipid extracts were dried, dissolved in HPLC mobile phase, purified on a Novapak C18 column (Waters), and eluted isocratically with acetonitrile/water (50:50, v/v; pH adjusted to 4.5 with acetic acid) at a flow rate of 0.7 ml/min. HPLC analyses were conducted on a Hewlett-Packard model 1090M HPLC system with a photodiode array UV detector. Internal standards were monitored at 235 nm to correct for losses during extraction. Between injections the column was washed with acetonitrile to remove any residual nonpolar lipids.
Products were collected in 30-s fractions, and radioactivity was
measured in a Packard Tri-Carb -counter. Background counts were
determined and subtracted. In most experiments, the HPLC eluent was
directed to a flow-through radioactivity detector (
RAM, IN/US) and
mixed with Universol. Sample and scintillation fluid were mixed in a
ratio of 1:3 to achieve a residence time in the detector cell of
21 s. Counting efficiency was estimated to be 70%.
Lipid extracts were either purified by SP-HPLC directly or after initial fractionation by RP-HPLC. Lipids, dissolved in mobile phase, were fractionated on a Nucleosil 100 silica column (5 µm; 250 × 4.6 mm; Alltech) eluted isocratically with hexane/isopropyl alcohol/acetic acid (98:2:0.1, v/v/v) at a flow rate of 1.0 ml/min. UV absorbance and radioactivity were measured as described for RP-HPLC.
Chiral Phase (CP)-HPLCProducts were reduced with TMP, re-extracted (where appropriate), and converted to the corresponding methyl esters with diazomethane. This material was injected onto two dinitrobenzoylphenylglycine columns connected in series (5 µm; 500 × 4.6 mm, J. T. Baker Inc.) eluted under isocratic conditions with hexane/isopropyl alcohol (100:0.5, v/v) at a flow rate of 0.5 ml/min. Ultraviolet absorbance was monitored at 235 nm.
Preparation of HPETE(8R)-HPETE was prepared
enzymatically using the (8R)-lipoxygenase in the gorgonian
coral, Plexaura homomalla, according to the method of Brash
and co-workers (19, 27). The coral was obtained from the Harbor Branch
Oceanographic Research Institute (Fort Pierce, FL) and was stored at
70 °C until used. Lipids produced from arachidonic acid were
extracted as described earlier and fractionated by preparative SP-HPLC
(Polygosil silica) eluted with hexane/isopropyl alcohol/acetic acid
(97:3:0.1, v/v/v) at 3 ml/min.
We prepared racemic 8-HPETE by the vitamin E-assisted autooxidation method of Peers and Coxon (28) using 3H-labeled or unlabeled arachidonate methyl esters as substrate. Racemic 8-HPETE-ME was purified by preparative SP-HPLC and then rechromatographed by CP-HPLC using a Chiracel OB column (J. T. Baker Inc.) eluted with hexane/isopropyl alcohol (100:5, v/v) at a flow rate of 1 ml/min. 8-HPETE-ME was hydrolyzed with base using freshly distilled tetrahydrofuran and lithium hydroxide (1 M) allowed to react at room temperature overnight with slow stirring. The HPETE free acids were extracted with ether, purified by SP-HPLC, dried, redissolved in ethanol, and quantified by UV spectrophotometry.
Bulk Preparation of 8-HPETE (Peak II) and 8-KETE (Peak III)Aplysia ganglia from 10 animals were dissected and homogenized in ASW. To prepare 8-HPETE (peak II), the 80 × g supernatant was diluted into ASW (50 ml), transferred to a flask containing arachidonic acid (50 µM) and 13-hydroperoxyoctadecadienoic acid, incubated for 10 min at 15 °C, and stopped with cold acetone. After the addition of 12-HETE (5 µg), the samples were extracted as described above. The material corresponding to peak II on RP-HPLC was dried and dissolved in ethanol.
To prepare 8-KETE (peak III), the 100,000 × g pellet was resuspended in ASW (10 ml) and incubated with [14C](8R)-HPETE (50 µM, 5 µCi) for 10 min at 15 °C. The reactions were stopped, extracted, and purified by RP-HPLC as described above. 8-KETE was detected by its absorbance at 270 nm, collected, dried under reduced pressure, and resuspended in ethanol.
Ultraviolet SpectrophotometryStandards and samples purified by RP- or SP-HPLC were dried under reduced pressure and dissolved in ethanol. The UV absorbance of standards and samples were measured using a Perkin-Elmer Lambda 4B UV-visible spectrophotometer. We used an extinction coefficient of 23,000 to calculate the concentration of HETE and HPETE and 22,000 for 8-KETE.
Gas Chromatography/Mass Spectrometry (GC/MS)Samples previously extracted and purified by HPLC were converted to their corresponding catalytically reduced, methyl ester trimethylsilyl ether derivatives. Conversion was accomplished in several steps. First, samples were treated with excess ethereal diazomethane for 10 min. After evaporating the solvent, the residue was dissolved in methanol (1 ml) on ice. A small amount of rhodium on alumina was added, and hydrogen gas was bubbled through the solution for 20 min. The rhodium-alumina was removed by filtration through glass wool, and the methanol filtrate was dried under reduced pressure. The residue was incubated with bis(trimethylsilyl)trifluoroacetamide/pyridine (1:1, v/v; 50 µl) for 15 min at 60 °C. Samples were dried under a stream of nitrogen gas and reconstituted in hexane (5 µl) for analysis. GC/MS analysis of the hydroxy acid derivatives was performed on a Hewlett-Packard 5987A equipped with an HP-1 capillary column (12 m) and using helium as the carrier gas. For electron impact analysis, we maintained a source temperature of 200 °C and an injection port temperature of 220 °C. Carrier flow was regulated by a constant head pressure of 52 kPa. The electron voltage was kept at 70 eV.
Samples believed to contain 8-KETE purified by HPLC were converted to their corresponding catalytically reduced methyl ester derivatives. This conversion was accomplished by treatment with excess ethereal diazomethane followed by exposure to hydrogen gas in the presence of the rhodium catalyst. The keto group was left underivatized. Samples were dried under a stream of N2 and reconstituted in hexane (5 µl) for analysis. The electron voltage was reduced from previous experiments to 20 eV to preserve the integrity of the keto group.
Sodium Borohydride Reduction of 8-KETEUnlabeled (2 µg) and radiolabeled (50,000 cpm) samples of 8-KETE (peak III) were dissolved in ethanol (100 µl) and incubated with sodium borohydride (1-2 mg) for 15 min at 4 °C. Thereafter, the sample was acidified to pH 3.5 with HCl and extracted with ether. We monitored RP-HPLC retention characteristics and UV spectra.
To identify the lipids produced in Aplysia
neurons, we incubated homogenates of neural components with
[3H]arachidonic acid. After extraction, products were
fractionated by RP-HPLC. A typical profile of the labeled metabolites
produced is shown in Fig. 1A. Two major
components were observed, peak I eluting at 27 min and peak II at 35 min. In some experiments, we added the reducing agent,
trimethylphosphite (TMP), to convert any hydroperoxy acids to hydroxy
acids. Under these conditions, only peak I appeared (Fig.
1B). The retention time of peak I, the absence of peak II
after the addition of TMP, the UV absorbance observed at 235 nm (which
is consistent with the presence of a conjugated diene), and the
incorporation of radioactivity into the product suggested that peak I
is a hydroxy acid derived from arachidonate and that peak II might be
its precursor hydroperoxy acid. At first we presumed that peak II is
12-HPETE because the retention time of peak I is the same as that of
12-HETE, a lipoxygenase product previously identified in
Aplysia neurons (1, 7). However, the retention time of peak
II is not the same as that of 12-HPETE. Nevertheless, reduction of the
material in peak II by TMP converted peak II to peak I (data not
shown). This conversion is consistent with the idea that the material
in peak II is a precursor of peak I. Although peak I might contain some
12-HETE, it seems likely that it also contains a second hydroxy acid
that is not separated from 12-HETE under these RP-HPLC conditions.
SP-HPLC of Peak II and Its Reduction Product
We found that
peak I from the RP-HPLC contains an unidentified component along with a
small amount of 12-HETE. Two distinct components were resolved by
SP-HPLC of the material in peak I (Fig. 2). The minor
component that eluted first had the retention characteristics of
authentic 12-HETE. The unknown component eluted much later (retention
time approximately 17 min). In the experiment shown in Fig. 2, 10%
(520 cpm) of the radioactivity applied eluted at the retention time of
12-HETE and 72% (3585 cpm) eluted as this second, unidentified
component. This major unknown radioactive product has the retention
time of 8-HETE, suggesting that peak II is 8-HPETE. The material in
peak I would then be a mixture of 12-HETE and 8-HETE.
Structural Analyses of Peaks I and II
To confirm the
identification of peak I as 8-HETE, we conducted UV and mass spectral
analyses. The UV absorbance spectrum of peak I (in ethanol) was
indistinguishable from authentic 8-HETE (data not shown) with a maximum
at 235 nm, consistent with the presence of a cis,trans-conjugated
diene. GC/MS analysis of the purified product was carried out using the
catalytically hydrogenated, methyl ester, trimethylsilyl ether
derivative (ME/TMS), prepared as described under "Experimental
Procedures." This derivative eluted from the GC at the same retention
time as the comparable derivative of authentic 8-HETE, and the mass
spectra of the two derivatives were indistinguishable (data not shown).
We found three major ion fragments in the spectrum at 245 atomic
mass units (-cleavage between C-7 and C-8), 271 atomic mass
units (
-cleavage between C-8 and C-9), and 399 atomic mass
units (M+
15), further supporting the conclusion that
the parent material is 8-HETE.
Large amounts of peak II were prepared from Aplysia nervous tissue for comparison with authentic 8-HPETE. We compared the chromatographic characteristics of the product purified from the tissue with 8-HPETE first as the free acid and then as the methyl ester, both before and after reduction with TMP. These tests were conducted on RP- as well as SP-HPLC. Peak II and its methyl ester behaved identically as 8-HPETE or its methyl ester under each condition tested (data not shown). Upon reduction with TMP and further purification on SP-HPLC, the material in peak II showed UV and GC/MS characteristics identical to 8-HETE, further evidence that the material in peak II is 8-HPETE.
Stereochemistry of 8-H(P)ETE from AplysiaSince lipid
hydroperoxides synthesized enzymatically by lipoxygenases are expected
to be stereochemically pure, we determined the stereochemistry of the
8-H(P)ETE isolated from Aplysia. 8-HPETE (5 µg) was
reduced with TMP and converted to the methyl ester with diazomethane.
The derivatized material was fractionated on a CP-HPLC system. The
elution time of each stereoisomer was verified by injection or
co-injection of standards, including racemic 8-HPETE prepared by
autooxidation and commercially obtained (8S)-HETE. The
resolution of the racemic 8-HETE methyl ester standard is shown in Fig.
3 (A). The S and R
isomers of 8-HETE-ME were resolved into a doublet eluting at
approximately 120 min. In contrast, the 8-HETE-ME isolated from
Aplysia eluted as a single peak, indicating the presence of
a single stereoisomer (Fig. 3, B). The 8-HETE-ME isolated
from Aplysia was mixed with racemic material and analyzed by
CP-HPLC. The Aplysia product co-eluted with the second of
the two stereoisomers, indicating that (8R)-HETE is the only
stereoisomer generated in Aplysia nervous tissue. In many
organisms, hydroperoxy acids are precursors for enzymatic conversion to
biologically active metabolites. Therefore, we looked for possible
conversion of 8-HPETE to novel products.
Identification of a Novel 8-HPETE Metabolite as 8-Ketoeicosatetraenoic Acid (8-KETE)
[3H](8R)-HPETE (50 µM, 0.1 µCi) was incubated with the resuspended
100,000 × g pellet fraction from Aplysia
nervous tissue (see "Experimental Procedures"). After extracting
lipids and fractionating them by RP-HPLC, we detected three major
radioactive products (Fig. 4A) in addition to
the solvent front. The first peak which eluted between 9 and 12 min is
most likely a mixture of epoxy alcohols and has not yet been studied
further. The second peak (24 min) had the HPLC retention
characteristics of 8-HETE which is the expected product of reduction of
the substrate, 8-HPETE (no remaining 8-HPETE was detectable in this
experiment). The third peak (32 min; Fig. 4, peak III) did not
correspond to any previously described eicosanoid and was collected for
further analysis.
To establish the identity of the material in peak III, we used several
approaches, including UV spectroscopy, chemical modification, and GC/MS
analysis. The ultraviolet absorbance spectrum of peak III was
bell-shaped with a single maximum at 279 nm when measured in SP-HPLC
mobile phase (solvent composition hexane/isopropyl alcohol/acetic acid,
98:2:0.1; data not shown). This spectrum is indicative of a
cis,trans-conjugated dienone. When measured in a more polar RP-HPLC
solvent (acetonitrile/water, 50:50, pH 4.5), a bathochromic shift of
the absorbance maximum was observed (283 nm; Fig.
5A, inset). This shift in absorbance maximum
with a change in solvent polarity is characteristic of conjugated
dienones and dienals (29, 30). The observed shift in UV maxima (279 nm
in hexane to 283 nm in acetonitrile/water) is further evidence that
peak III is a molecule that contains a conjugated system, probably a
dienone. In accord with our previously published studies of 12-HPETE
metabolism, these UV absorption data suggested that the material in
peak III is 8-KETE.
If peak III contains a ketone as a functional group, treatment with sodium borohydride should reduce it to a hydroxyl; thus, 8-KETE would be converted to 8-HETE. Upon borohydride treatment (Fig. 5B), 3H-labeled peak III was converted to a product with an UV absorbance maximum of 237 nm (Fig. 5B, inset) that eluted at about 30 min. These characteristics are consistent with the presence of a conjugated diene. The UV and HPLC properties of the material in peak III reduced by sodium borohydride are indistinguishable from authentic 8-HETE (Fig. 5C).
Conclusive evidence that the material in peak III is 8-KETE was
obtained by GC/MS analysis. GC/MS analysis was carried out using the
catalytically hydrogenated, methyl ester derivative of the material in
peak III. Although there is no synthetic 8-KETE standard with which to
compare the purified substance obtained from Aplysia, the
fragmentation pattern of lipids with similar structures has been well
characterized (31). The results with peak III are consistent with the
structure shown in Fig. 6. Six major ion fragments were
found in the spectrum: 212 (McLafferty rearrangement:
CH3-(CH2)11-C(+OH) = CH2), 197 (-cleavage between C-7 and C-8), 186 (McLafferty rearrangement: CH2 = C(+OH)-(CH2)6-COOCH3),
171 (
-cleavage between C-8 and C-9), 154 (186-CH3OH),
129 (
-cleavage between C-6 and C-7). A fragment of lower abundance
was seen at 309 atomic mass units, presumably representing the loss of
a methoxy group from the molecular ion. The fragmentation pattern
obtained is consistent with the conclusion that the structure of the
material in peak III is 8-ketoeicosatetraenoic acid.
Finally, when the material in peak III was reduced with sodium borohydride and analyzed by GC/MS as the TMS ether, methyl ester derivative, its mass spectrum was indistinguishable from 8-HETE (data not shown) confirming the identification of peak III as 8-KETE.
Enzymatic Formation of 8-KETEIf the formation of 8-KETE from
(8R)-HPETE is catalyzed by an enzyme, there should be a
stereochemical preference for substrate. To test this preference, we
compared the ability of the 100,000 × g pellet of
Aplysia ganglia homogenates to metabolize
[3H](8R)-HPETE and
[3H](8S)-HPETE. As shown in Fig.
7A, 8-KETE was formed only when (8R)-HPETE was added as substrate; none was detected with
(8S)-HPETE (Fig. 7B). Both 8-HETE and 8-HPETE
(unreacted starting material) were present in all of the incubations.
From this experiment, we conclude that the Aplysia enzyme
that generates 8-KETE is specific for the naturally occurring
(8R)-HPETE. In other experiments, we showed that metal ions
or metalloproteins do not catalyze the formation of the ketone
nonenzymatically. No 8-KETE was formed when the Aplysia
enzyme preparation was heat-denatured prior to the incubation with
(8R)-HPETE (Fig. 7C). In addition, a nonenzymatic iron-catalyzed rearrangement is unlikely since no 8-KETE was formed when (8R)-HPETE was incubated with hematin (5 µM; data not shown). Heat denaturation of the
100,000 × g pellet fraction also decreased the
generation of 8-HETE (Fig. 7C). This decrease suggests that a heat-labile peroxidase activity is present in the Aplysia
homogenate that catalyzes the reduction of (8R)-HPETE to
(8R)-HETE.
Differential Activation of 8- and 12-Lipoxygenases
8-Lipoxygenase was activated in
Aplysia neural components (nerve cell bodies and neuropil)
after application of acetylcholine (ACh) (Fig. 8).
Radioactive metabolites were detected in four regions: unidentified
material in the solvent front, hepoxilin-like molecules (retention time
10-15 min), 8-HETE, and 8-KETE. This pattern is similar to that
reported for histamine activation of the 12-lipoxygenase in this neural
preparation (2). Activation of the lipoxygenases was measured after the
application of either ACh or histamine. 8- and 12-HETE were quantified
by HPLC. As previously reported, 12-HETE was the major product
generated in the neural components exposed to histamine (50 µM). Under these conditions, 8-HETE was barely detectable
(Fig. 9, top). In contrast, in the same
tissue preparation application of ACh (100 µM; Fig. 9,
bottom) resulted in the generation of 8-HETE.
Aplysia neural components contain an (8R)-lipoxygenase that converts arachidonic acid to (8R)-HPETE. This hydroperoxy acid is the precursor for the enzymatic synthesis of 8-HETE and 8-KETE. We previously showed that Aplysia nervous tissue also contains a (12S)-lipoxygenase that leads to the generation of similar products, 12-HETE and 12-KETE, as well as to an epoxy alcohol, hepoxilin A3 (1, 2, 6). An (8R)-lipoxygenase has recently been cloned from the coral, P. homomalla, that has a catalytic mechanism that is closely related to the (12S)-lipoxygenase (20). In earlier studies of intact neurons, we found that histamine, a neurotransmitter in Aplysia, causes the activation of the 12-lipoxygenase pathway (1, 2). Moreover, the neurons that synthesize the 12-lipoxygenase products also respond to them. For example, the open probability of K+s channels in isolated sensory cells is increased in response to 12-HPETE (3, 8) by a mechanism thought to require conversion of the hydroperoxide to hepoxilin A3 (9). These metabolites either mimic or modulate the action of the Aplysia neuropeptide, FMRF-amide (3, 7). In addition, pharmacologic application of 12-KETE induces depolarization in L32, an identified neuron in Aplysia (2) and hepoxilin A3 causes L32 to hyperpolarize (6). These observations suggest that 12-lipoxygenase is part of a second messenger pathway. Although we have not yet tested the action of specific 8-lipoxygenase metabolites on identified neurons, we report here that the 8-lipoxygenase pathway is activated by applying ACh to intact Aplysia neurons.
The similarity of the two lipoxygenase pathways prompts us to predict that one or more of the lipids generated by 8-lipoxygenase also function as second messengers. Since this report is the initial observation of an 8-lipoxygenase in the nervous system, its potential neurophysiological activity remains to be determined. Previous studies have identified (8R)-lipoxygenase metabolites in coral (20, 32), crustacea (21, 33), and echinoderms (23), and various physiological activities have been described. For example, nanomolar concentrations of 8-hydroxyeicosapentaenoic acid, an 8-lipoxygenase metabolite of eicosapentaenoate, activates larval hatching in several species of barnacle (33). A trihydroxy fatty acid, presumed to arise from the decomposition of an epoxy alcohol generated by 12-lipoxygenase, is also a potent hatching factor for these species. It is interesting that, even though these barnacles do not contain 12-lipoxygenase activity, they can generate hepoxilin-like molecules. The absence of 12-lipoxygenase suggests that related epoxy alcohols may be formed from 8-hydroperoxy fatty acids as well as the better characterized 12-HPETE pathway. Nanomolar concentrations of (8R)-HETE also induce the maturation of oocytes in starfish (23, 34). This lipid, but not its stereoisomer, stimulates the production of cAMP and results in the protein phosphorylation that initiates meiosis in these oocytes. Thus, (8R)-HETE mimics the cellular effects of 1-methyladenine, a natural activator of oocyte maturation. Although the oocytes can generate (8R)-HETE from exogenous arachidonic acid (23), 8-lipoxygenase metabolism does not appear to be induced by application of 1-methyladenine (35). More recently, products of the 8-lipoxygenase pathway similar to those isolated in coral have been identified in starfish oocytes suggesting that signaling through these lipids might involve additional molecules derived from this pathway. In coral, (8R)-HPETE is enzymatically converted through an allene oxide into a prostaglandin-like molecule (19, 27) of unknown function. In Aplysia we see no evidence of this pathway. Nevertheless, we show that (8R)-HPETE can be enzymatically converted to (8R)-HETE and 8-KETE.
Lipoxygenases typically have a restricted cell distribution. Nevertheless, we find that Aplysia nervous tissue contains both 8- and 12-lipoxygenases and that the enzymes are independently activated by different neurotransmitters (that is, ACh or histamine). Since lipoxygenases are thought to be constitutively active and controlled by the availability of substrate, this raises the question about how these pathways are regulated. Do both enzymes exist in the same neurons? If so, receptor-mediated activation of a phospholipase resulting in the release of free arachidonate ought to yield products of both pathways. 12-Lipoxygenase has been shown to exist in several identified Aplysia neurons including L14 and mechanosensory cells of the abdominal and pleural ganglia (3, 6). Other identified neurons contain the 8-lipoxygenase (36). The specificity of activation would be explained if each lipoxygenase is restricted to a subset of the neuronal population with the appropriate neurotransmitter receptor (12-lipoxygenase in histamine-sensitive neurons and 8-lipoxygenase in ACh-responsive cells).
Alternatively, a neuron might have both types of lipoxygenase. Selectivity in the production of metabolites would then have to be regulated by intracellular mechanisms. But differential activation is difficult to envision since both pathways would be expected to have the same initial step, release of arachidonic acid from membrane phospholipid. Although the best characterized lipoxygenases are cytosolic, there have been several reports of enzymes that appear in the membrane fraction. For example, human epidermal tissue contains a microsomal platelet-type 12-lipoxygenase (37, 38), and most of the 15-lipoxygenase in the fungus, Sparolegnia parasitica, is membrane bound (39). In addition, our earlier studies showed that the Aplysia 12-lipoxygenase tends to be concentrated in a membrane fraction after differential centrifugation of homogenates of nervous tissue (18). Thus, it is possible that the 12- and 8-lipoxygenase are located in different cellular compartments. In vitro studies of the soybean lipoxygenase have revealed a conformational change in the enzyme due to changes in the enzyme's microenvironment when water-soluble co-solvents are added (40). In association with the observed changes of the protein's conformation was a significant shift in the positional specificity of the lipoxygenase. The mechanistic similarities between (12S)- and (8R)-lipoxygenation, including antarafacial insertion of oxygen after abstraction of the same proton (20), might suggest that the two activities are catalyzed by the same enzyme that displays a different positional specificity depending upon its local environment. Alternatively, specific neurotransmitters might cause a differential translocation of the cytosolic lipoxygenases to a membrane where the enzyme would have access to its fatty acid substrate, since some lipoxygenases are known to be activated by translocation from the soluble compartment to intracellular membranes (41, 42). Moreover, several reports suggest that at least one isoform of 12-lipoxygenase is activated by a calcium-dependent translocation (43).
Most, if not all, Aplysia neurons are responsive to ACh (44). ACh responses are mediated by at least three distinct receptors (45). It is therefore reasonable to think that the 8-lipoxygenase pathway is activated by a specific subset of ACh receptors whose distribution is restricted to a sub-population of neurons. We are continuing to study the relationship of lipoxygenase activation to specific ACh receptors with the expectation that some indication of the physiological role of these lipids may be obtained. Identified neurons in Aplysia are convenient for these studies since specific cells can be dissected and cultured for biochemical and electrophysiological analysis.
We thank Dr. JacSue Kehoe for helpful discussions and Dr. Irving Kupfermann for critically reading the manuscript.