147

Biochimica et Biophysics Acta, 1044 (1990) 147-157 Elsevier

BBALIP

53394

Conversion of stereoisomers of leukotriene B4 to dihydro and tetrahydro metabolites by porcine leukocytes William S. Powell 1 and Francine

Gravelle

*

’ Endocrine Laboratory, Royal Victoria Hospital and 2 Deparrment of Medicine, McGill University, Montreal (Canada)

(Revised

Key words:

Leukotriene

(Received 13 September 1989) manuscript received 16 January

1990)

B4 metabolism; 6-rrans-Leukotriene B4; Dihydroleukotriene; Polymorphonuclear leukocyte; (Porcine)

Tetrahydroleukotriene;

We have previously shown that porcine leukocytes convert leukotriene B4 (LTB,) to two major products, lO,ll-dihydroLTB, and lO,ll-dihydro-12-oxo-LTB,. Although we did not detect these products after incubation of LTB4 with human polymorphonuclear leukocytes, these cells converted It-epi-6-trans-LTB4 to the corresponding 6,11-dihydro metabolite (i.e., there appeared to be a shift in the positions of the remaining double bonds). The objective of the present investigation was to determine whether 6-truns isomers of LTB4 are metabolized by porcine leukocytes by a pathway similar to LTB,, or whether they are metabolized by a pathway analogous to that in human leukocytes. We found that 6-trans-LTB, and 12-egi-6-trans-LTB, are metabolized more much extensively than LTB, by porcine leukocytes. 6-iruns-LTB, appears to be converted by two different reductase pathways to two dihydro products differing in the positions of the two remaining double bonds between carbons 5 and 12. Dihydro-1Zoxo and dihydro-Soxo metabolites are also formed from this substrate. Porcine leukocytes also convert 6-trans-LTB,, presumably by a combination of the above two pathways, to tetrahydro, tetrahydro-12-oxo and tetrahydrod-oxo metabolites, none of which possesses any conjugated double bonds. 12-epi-Btrans-LTB~ is also converted to tetrahydro metabolites by these cells. Experiments with deuterium-labeled 6-truns-LTB4 indicated that the deuterium in the 5position was almost completely lost during the formation of tetrahydro-6-trans-LTB,, whereas about 80445% of the deuterium in the 1Zposition was lost, suggesting a requirement for a 5-0~0 intermediate. As with LTB,, 12-epi-8-cis-6-truns-LTB,, the product of the combined actions of 54ipoxygenase and lZlipoxygenase, was converted principally to dihydro and dihydro-12-oxo metabolites. Only a relatively small amount of the tetrahydro metabolite was detected.

Introduction The major metabolites of arachidonic acid formed by polymorphonuclear leukocytes (PMNL) are Shydroxy6,8,11,14-eicosatetraenoic acid and leukotriene B4 (LTB,), along with its two 6-truns stereoisomers, 6truns-LTB, and 12-epi-6-truns-LTB, [l]. LTB, is rapidly metabolized to 20-hydroxy-LTB, by human PMNL by a

Abbreviations: LT, leukotriene; HETE, hydroxyeicosatetraenoic acid; dh-LTB,, lO,ll-dihydro-LTB,,; dho-LTB,, lO,ll-dihydro-12-oxoLTB,; PMNL, polymorphonuclear leukocytes; ODS, octadecylsilyl; RP-HPLC, reversed-phase high pressure liquid chromatography; GCMS, gas chromatography-mass spectrometry; ~a, retention time; MSTFA, N-methyl-N-trimethylsilyltrifluoroacetamide; ETYA, 5,8, 11,14-eicosatetraynoic acid. Correspondence: W.S. Powell, Endocrine Laboratory, Royal Victoria Hospital, 687 Pine Avenue West, Montreal, Quebec, Canada H3A 1Al.

0005-2760/90/$03.50

0 1990 Elsevier Science Publishers

B.V. (Biomedical

specific 20-hydroxylase enzyme [2-51. This is followed by further conversion to the corresponding 20-0~0 [6] and w-carboxy metabolites [7-91. Human PMNL also convert 6-trans isomers of LTB, to their 20-hydroxy metabolites, but much more slowly than LTB, [3]. We recently identified a second pathway for the metabolism of these 6-tram isomers by human PMNL, culminating in the formation of dihydro compounds in which the characteristic conjugated triene chromophore had been reduced to a conjugated diene [lO,ll]. It appeared that there was a shift in the positions of the remaining double bonds, resulting in the formation of ‘6,11-dihydro’ metabolites [ll]. We did not detect dihydro metabolites of LTB, analogous to those described above in incubations with human PMNL, possibly because LTB, is so rapidly metabolized by the w-oxidation pathway in these cells. However, we found that LTB, is converted to a dihydro product by rat PMNL [12], which possess considerably Division)

148 less w-oxidation activity than human PMNL [3]. Other rat cells [13] as well as human alveolar macrophages [14], have also been shown to convert LTB, to a similar product. Porcine PMNL, which contain very little woxidation activity, metabolize LTB, to two major products, lO,ll-dihydro-LTB, and lO,ll-dihydro-12-oxoLTB, [15]. However, there are a number of differences between the reduction of 6-tram isomers of LTB, by human PMNL and the reduction of LTB, by porcine PMNL. In the first place, the positions of the conjugated double bonds in the resulting dihydro products appear to be different from one another. Secondly, the formation of dihydro metabolites from the 6-tram isomers by human PMNL appears to require prior oxidation of the 5-hydroxyl group of the substrate [ll], whereas lO,ll-dihydro-LTB, can be formed directly from the substrate [15]. There is no evidence for the formation of 5-0x0 metabolites of LTB, by porcine PMNL [15]. The differences in the dihydro products formed from LTB, and its 6-truns isomers could be due to differences in the reductase enzymes present in human and porcine leukocytes. Alternatively, the same enzymes could exist in both species, but the presence of a 6-truns double bond in the substrate could alter the course of the reaction. To determine which of these possibilities is correct, we examined the metabolism of 6-truns isomers of LTB, by porcine leukocytes. We found that these substrates are metabolized to dihydro and dihydro-120x0 products analogous to those formed from LTB, by porcine leukocytes. However, unlike LTB,, the 6-truns isomers are also converted to products similar to those formed from 12-epi-6-truns-LTB, by human PMNL. The combination of these two reductase pathways resulted in the formation of tetrahydro metabolites. Materials and Methods Preparation of substrates Unlabeled 6-truns-LTB, [ll], 6-truns-[5,6,8,9,11,12, 14,15-*H]LTB, [ll], 12-epi-6-truns-LTB, [ll] and 12epi-8-cis-6-truns-LTB, [3,16] were prepared as previously described using porcine leukocytes. The corresponding (1-i4C)-labeled products were synthesized by incubation of [1-i4C]arachidonic acid (New England Nuclear), with human PMNL [3]. Preparation of porcine leukocytes Leukocytes were prepared as described previously [3] by treatment of blood with Dextran T-500 (Pharmacia Fine Chemicals), followed by 0.135 M NH,Cl to remove red blood cells. The crude leukocyte preparation was resuspended in Dulbecco’s phosphate-buffered saline, containing 137 mM NaCl, 2.7 mM KCl, 1.5 mM KH,PO,, 8.1 mM Na,HPO,, 0.5 mM MgCl, and 0.9 mM CaCl,.

Incubation of 6-trans isomers of LTB, with porcine leukocytes and purification of metabolites Porcine leukocytes (10s cells/ml) were incubated with 6-truns-LTB,, 12-epi-6-truns-LTB,, or 12-epi-8-cis-6truns-LTB, at concentrations of 2 FM for 60 min at 37 o C. The incubations were terminated by centrifugation at 400 X g for 10 min at 4” C. The pellet was then washed by recentrifugation in an amount of ethanol sufficient to give a final concentration of 15% when combined with the first supematant. The combined supematants were extracted using cartridges (Waters C,, Sep-Paks) of octadecylsilyl (ODS) silica [17]. After loading the sample onto the cartridge, it was washed with 15% ethanol in water (20 ml), water (20 ml) and petroleum ether (20 ml). Metabolites were then eluted with redistilled methyl formate (Aldrich) (10 ml). The material in the methyl formate fraction was purified by RP-HPLC using a Waters solvent delivery system, a Raytest Ramona 5-LS radioactivity monitor and a Waters model 490 UV detector. The stationary phase was a Phenomenex Ultracarb ODS column (5 pm particle size; 4.6 X 250 mm), which was eluted under isocratic conditions with water/ acetonitrile/ acetic acid (63 : 37 : 0.05) at a flow rate of 2 ml/min. Gus chromatography-mass spectrometry (GC-MS) Electron impact GC-MS was performed on a VG ZAB instrument located in the Biomedical Mass Spectrometry Unit of McGill University. The stationary phase was a column (20 m X 0.32 mm) of DB-1 (J and W Scientific). Prior to analysis, products were methylated with diazomethane and converted to their trimethylsilyl ether derivatives by treatment with Nmethyl-N-trimethylsilyltrifluoroacetamide (MSTFA) (30 min, 23OC). In some cases, products were converted to methoxylamine derivatives prior to methylation by treatment with methoxylamine hydrochloride (1 mg) in pyridine (0.1 ml) overnight at room temperature. Hydrogenations were performed in the presence of PtO, at room temperature for 90 s, followed by purification of the products by RP-HPLC. Results Metabolism of 6-trans-LTB, by porcine leukocytes The chromatographic profile of metabolites formed from 6-trans-LTB, (Fig. 1A) was much more complex than that for LTB, (Fig. 1B). As with LTB,, there were two less polar peaks exhibiting UV Lbsorbance at 235 nm but not at 280 nm. However, in both cases these were double peaks with retention times at 42.4 min (T-la), 43.4 min (T-lb), 67.5 rnin (T-3a) and 69.3 min (T-3b). Unlike the situation with LTB,, the major metabolites of 6-truns-LTB, (T-2 (t,, 53.2 min) and T-4 (tR, 78.9 mm)) absorbed at neither 235 nm nor 280 nm, but could be detected at 200 nm (not shown). There was

LTB, 1280 nm

A

T-la _ b/T-lb

-

dh-LT@

~235 nm

dho-LTB,, A-

6- trans-LTB,

n T-Z RADIOACTIVITY

4b

TIME

8b (mid

0

4o TIME

LTB, \ 80

(mir$

Fig. 1. Reversed-phase high pressure liquid chromatogram of metabolites of 6-trans-LTB4 (A) and LTB, (B). 6-trans-[l-‘4C]LTB4 (2 pM, 0.1 pCi) (A) or LTB, (2 PM, 0.1 PCi) (B) were incubated with porcine leukocytes (10’ cells/ml) for 60 mm at 37OC. The chromatogram shown in (A) was obtained using isocratic conditions on a Phenomenex Ultracarb ODS column with a mobile phase consisting of water/acetonitrile/acetic acid (63 : 37 : 0.05). The chromatogram shown in (B) was obtained on a Waters-Millipore Novapak C,s column with a mobile phase consisting of a gradient over 52 mm between 20% and 35% acetonitrile in water containing acetic acid (final concentration, 0.05%). The concentration of acetonitrile was then kept constant at 35% for the duration of the chromatography. The flow rate was 2 ml/mm for both A and B. It should be noted that the scales showing UV absorbance at 235 nm (middle traces) and 280 nm (top traces) are not the same.

also an additional minor radioactive product (T-5; t,, 83.8 min), which could be detected in the UV only at 200 nm. This suggests that the reductase reaction proceeds further with 6-truns-LTB,, resulting in the formation of products which have no conjugated double bonds. Identification of metabolites T-la and T-lb. Metabolites T-la and T-lb were further purified by HPLC and then analyzed by GC-MS. Both products absorbed at 235 nm, but not at 280 nm, suggesting that they possess two conjugated double bonds. The mass spectrum (Fig. 2A) of the methyl ester, trimethylsilyl ether derivative of metabolite T-la (C value, 23.7) was quite similar to that of lO,ll-dihydro-LTB,, the major metabolite of LTB, by porcine PMNL [15]. Intense ions were observed at m/z 496 (M), 481 (M - 15), 465 (M - 31), 406 (M 90), 395 (C-5 to C-20), 385 (C-l to C-12), 305 (395 - 90), 295 (385 - 90), 269 (C-l to C-lo), 268, 237, 229 (C-l to C-7), 217, 213 (C-12 to C-20), 205 (385 - 2 x 90), 203 (C-l to C-5), 181, 159, 129, 119 and 105 (base peak). This mass spectrum would suggest that metabolite T-la is identical to lO,ll-dihydro-6-truns-LTB,. Although the mass spectrum of the trimethylsilyl ether, methyl ester derivative of metabolite T-lb (C value, 23.6) has many of the same ions as the corresponding derivative of metabolite T-la, the relative intensities are quite different. However, the mass spectrum of T-lb (Fig. 2B) is very similar to that which we previously reported for the dihydro metabolite of 12epi-6-trans-LTB, (5,12-dihydroxy-7,9,14-eicosatrienoic acid) formed by human PMNL [ll]. Intense ions were observed at m/z 481 (M - 15), 465 (M - 31), 406

(M - 90), 395 (C-5 to C-20), 385 (C-l to C-12), 305 (395 - 90), 295 (base peak; 385 - 90), 279 (C-7 to C-20), 269 (C-l to C-lo), 217, 203 (C-l to C-5), and 129. As was the case with dihydro-12-epi-6-trans-LTB,, the mass spectrum of this compound was dominated by ions formed cleavage of the bond between carbons 12 and 13 (m/z 385 and 295). The mass spectrum of the corresponding derivative of metabolite T-la shows much more extensive fragmentation (Fig. 2A). The mass spectra of T-la and T-lb indicate that both substances have hydroxyl groups in the 5- and 1Zpositions. The difference must therefore be in the positions of the double bonds between these two hydroxyl groups. The ion at m/z 279 in the mass spectra of T-lb and dihydro-1Z epid-truns-LTB, [ll], which shifts to m/z 285 when substrates labeled with deuterium in positions 5, 6, 8, 9, 11, 12, 14 and 15 are used, suggests that the double bonds of these metabohtes are in the 7- and 9-positions (see Ref. 11). If this is the case, metabolite T-lb would be identical to 5,12-dihydroxy-7,9,14-eicosatrienoic acid (i.e., 6,11-dihydro-6-truns-LTB,). Identification of metabolite T-2. MetaboIite T-2 absorbed UV light at 200 nm, but not at either at 235 or 280 nm, indicating that it did not contain any conjugated double bonds. The mass spectrum of the methyl ester, trimethylsilyl ether derivative of unlabeled T-2 (Fig. 3), which has a C value of 23.5, has intense ions at m/z 483 (M - 15), 467 (M - 31), 393 (M - 15 - 90), 387 (C-l to C-12), 318 (M - 2 x 90), 307 (397 - 90; loss of Me,SiOH from C-5 to C-20), 297 (387 - 90), 213 (C-12 to C-20), 207 (387 - 2 x 90), 203 (C-l to C-5), 169, 149, 129 (base peak) and 107. These results indi-

150

RELATIVE

INTENSITY

_ 385

295

1oojy5

A~

395

?TMS

269

60

213

/

I

I

269

217

I

385 I ’ 268

100

‘L

200

~

M-15 481

300

400

500

m/z

RELATIVE

INTENSITY 333

29

100 r

279

395

B

OTMS

02Me 269

*

&MS

x5 203

601 385

129

20

M-31

217

-I

I

203

~305

1

269

d IJL 200

300

400

500

of metabolites

T-la (A) and T-lb (B).

m/z Fig. 2. Mass spectra of the methyl ester, trimethylsilyl

cate that T-2 is a tetrahydro metabolite of 6-trans-LTB, containing one double bond between the 5- and 12carbons, the location of which cannot be deduced from the mass spectrum.

ether derivatives

The corresponding deuterium-labeled analog of metabolite T-2 was prepared by incubation of 6-trans[5,6,8,9,11,12,14,15-‘H]LTB, with porcine PMNL. Mass spectral analysis (not shown) revealed that about 80% of

151

RELATIVE

INTENSITY 397

387 297

r

OTMS

x3

02Me

OTMS

213

60-

203

203

x10

207 213 I I ,J

387

r M-15 483 M-31 1

200 Fig. 3. Mass spectrum

300

m/z

400

500

of the methyl ester, trimethylsilyl ether derivative of metabolite T-2. Although the double bond between groups is shown to be in the 7-position, we have no direct evidence to support this.

deuterium-labeled tetrahydro-6-tram-LTB, had lost two of the deuterium atoms originally present in the substrate, with the remainder having lost one deuterium atom (on the basis of the intensities of the ions corresponding to M - 15 and C-l to C-12; see above). Almost all the deuterium was lost from position 5 (ion corresponding to C-l to C-5), whereas over 80% was lost from position 12 (ion corresponding to C-12 to C-20). Identification of metabolites T-3a and T-3b. Metabolites T-3a and T-3b were further purified by HPLC and analyzed by GC-MS after conversion to the trimethylsilyl ether derivatives of their methyl esters. The mass spectrum (not shown) of this derivative of metabolite T-3a is virtually identical to that of the corresponding derivative of the enolic form of lO,ll-dihydro-12-oxoLTB, [15]. Major ions were observed at m/z 494 (M), 479 (M - 15), 463 (A4 - 31), 404 (M - 90) 393 (C-5 to C-20), 303 (393 - 90), 291 (C-6 to C-20), 277, 268, 225 (base peak, CH,-(CH,),-CH = CH-CH,-C(OSiMe,)CH), 203 (C-l to C-5), 167, 129, 119, and 105. Other 12-oxo-eicosanoids, such as the 12-0~0 derivative of 1ZHETE [18], enolize quite readily in the presence of trimethylsilylating reagents. After hydrogenation and treatment with diazomethane and MSTFA, metabolite T-3a formed 0x0 and enol derivatives which had mass spectra very similar to those of the corresponding de-

the two hydroxyl

rivatives of lO,ll-dihydro-12-oxo-LTB, [15]. Thus metabolite T-3a is identical to lO,ll-dihydro-12-oxo-6truns-LTB,. When the methyl ester, trimethylsilyl ether derivative of metabolite T-3b was analyzed by GC-MS, two peaks were observed, with C values of 23.5 and 23.7 and similar mass spectra, which were both quite distinct from that of the corresponding derivative of T-3a. The mass spectrum of the product with a C value of 23.7 (Fig. 4) exhibited intense ions at m/z 479 (A4 - 15), 463 (M31), 404 (M90), 383 (C-l to C-12), 293 (383 - 90), 281 (C-l to C-11) 267 (C-l to C-lo), 241 (C-l to C-8), 217 (base peak), 207, 167, 132 and 129. The base peak at m/z 217 could possibly be due to Me,SiO+= CH-CH = CH-OSiMe,, which could have been formed by a rearrangement. Intense ions at m/z 217 are also observed in the mass spectra of 6-tram isomers of LTB, [ll] and their dihydro metabolites (see Figs. 2A and B). The intense ion at m/z 241 could have been formed by cleavage between carbons 8 and 9. Since this bond occurs within a conjugated triene system, it is possible that an intermediate ion is first formed. A similar cleavage, apparently within a conjugated system, is observed at m/z 229 in the mass spectra of product T-la (Fig. 2A) and lO,ll-dihydroLTB, [15]. The mass spectrum of the isomer with a C-value of 23.5 was similar to that described above,

152

RELATIVE INTENSITY

r

383

2 17

?TMS

0,Me 241

OTMS 60

267

281

241

132 293

383

X5

20

267 i

M-90 : 404

r 281

1dlL

II//

Ill II

300

100

M-31 463

I

400 m/z

Fig. 4. Mass spectrum

of the methyl ester, trimethylsilyl

except that the ion at m/z 217 was much less intense and the base peak was at m/z 241. Unfortunately, the small amounts of product T-3b obtained in these reactions precluded a more conclusive determination of the origin of the above ions. The mass spectra of the two isomers of product T-3b were completely different from that of product T-3a. No ions were observed at m/z 225, which is the base peak in the mass spectrum of metabolite T-3a. On the other hand, the ion at m/z 241 in the mass spectra of the isomers of metabolite T-3b is virtually absent from the mass spectrum of metabolite T-3a. The ion at m/z 203, which is observed in the mass spectra of nearly all 5-hydroxy metabolites related to LTB, is quite weak in the mass spectrum shown in Fig. 4 and may have arisen from a different fragmentation or be due to the presence of a small amount of an interfering substance. This ion is absent from the mass spectrum of the second isomer of T-3b, which has a C value of 23.5. The ions at m/z 281 (C-l to C-11) 267 (C-l to C-10) and 241 (C-l to C-8) in the mass spectrum of product T-3b (Fig. 4) would be consistent with the presence of double bonds in the 5, 7- and 9-positions of the methyl ester, trimethylsilyl ether derivative of the enol isomer of dihydro-5-oxo-6-trans-LTB,. The two isomers of this compound could possibly have different configurations about the enolic double bond. Since it is possible that

ether derivative

of metabolite

T-3b.

the positions of the double bonds could have shifted during the enolization reaction, it is not possible to draw conclusions as to their original positions in the underivatized compounds on the basis of their mass spectra after derivatization. The mass spectrum of the trimethylsilyl ether, methyl ester derivative of hydrogenated T-3b (in the enof form) provides further support for the above structure, since very intense ions are present at m/z 215 and 297 (387 - 90) due to cleavage on either side of the trimethylsilyloxy group at carbon-12, confirming that the nonenolic hydroxyl group of T-3b is in the 12-position rather than the 5-position. The presence of only very weak ions at m/z 203 (C-l to C-5) and 309 (399 - 90, C-5 to C-20), possibly due to an impurity, is in agreement with this. Identification of metabolite T-4. Like metabolite T-2, metabolite T-4 did not absorb at either 235 or 280 nm, but could be detected at 200 nm, suggesting the absence of conjugated double bonds. It formed syn and anti O-methyloxime derivatives, which could be separated by HPLC, indicating the presence of an 0x0 group. The mass spectra (Fig. 5) of the methyl ester, trimethylsilyl ether derivatives of the methyloximes (C values 23.4 and 23.6) were similar and displayed intense ions at m/z 422 (M - 31), 406 (M - 47), 396 (loss of C-17 to C-20), 348 (M - 15 - 90) 342 (C-l to C-12) 332 (M -

153

RELATIVE

INTENSITY 210

100

i

vO,Me 3g6

129

OTMS

342

203

\

60

I

100

183 I 203

M-31-90

200

300

/tM-15-90

400

m/z Fig. 5. Mass spectrum

of the 0-methyloxime, methyl ester, trimethylsilyl ether derivative of metabolite T-4. Although the double bond b&tween the two hydroxyl groups is shown to be in the 7-position, we have no direct evidence to support this.

31 - 90), 310, 280, 264, 222, 210 (base peak; 342 - 101 - 31; loss of 31 from C-5 to C-12), 203 (C-l to C-5), 183 and 129. Hydrogenation of metabolite T-4, followed by treatment with diazomethane and MSTFA, resulted in the formation of a mixture of the 5-trirnethylsilyloxy-l2-oxo derivative and the corresponding trimethylsilylated enol, with the latter predominating. The mass spectrum of the enol derivative of the hydrogenated compound exhibited intense ions at m/z 485 (M - 15) 469 (M - 31), 415 (M - 85; loss of c,, to C,, from the 12,13-enol isomer), 410 (M - 90), 399 (C-5 to C-20), 325 (415 - 90), 309 (399 - 90), 267, 241 (C-10 to C-20, cleavage between C-9 and C-10 in the 11,12-enol isomer), 203 (C, to C,), 171, 143, 130 (base peak) and 129. This mass spectrum is almost identical to the mass spectrum of the enol isomer of the corresponding derivative of hydrogenated lO,ll-dihydro-lZoxo-LTB, and clearly indicates that metabolite T-4 contains an 0x0 group in the 12-position and a hydroxyl group in the 5-position. Product T-4 is therefore a lZoxo-tetrahydro metabdlite of 6-trans-LTB,, with only one double bond between carbons 5 and 12 (i.e., 5-hydroxy-lZoxo-eicosadienoic acid). Identification of metabolite T-5. Metabolite T-5 was a relatively minor product and only small amounts were obtained for structural analysis. It absorbed in the UV region at 200 nm, but not at 235 or 280 nm, indicating

the absence of conjugated double bonds. Treatment with diazomethane and MSTFA resulted in the formation of an enol derivative. The mass spectrum of this product (not shown) exhibited intense ions at m/z 496 (M), 481 (M - 15), 465 (M - 31), 423 (loss of CH,CO,Me from the 4,5-enol isomer), 406 (M - 90) 385 (base peak; C-l to C-12) 333 (423 - 90), 295 (385 - 90) 281 (C-7 to C-20) 269, 268, 255 (C-l to C-9), 241 (C-l to C-8), 239, 216, 213 (C-12 to C-20), 169, 159, 143 and 129. The mass spectrum of the enol of the trimethylsilyl ether, methyl ester derivative of hydrogenated T-5 was similar to that of the corresponding derivative of metabolite T-3b, with intense ions at m/z 387, 297 (387 - 90) and 215 due to cleavages on either side of the trimethylsilyloxy group in the 12-position. These mass spectra suggest that T-5 contains a hydroxyl group at carbon-12 and an enolizable 0x0 group at carbon-5. The ion at m/z 241 would suggest that double bonds are present in the enol isomer at positions 5 and 7, but the position of the C-7 double bond could have shifted during the enolization reaction. Metabolite T-5 would therefore appear to be the 5-oxo-tetrahydro metabolite of 6-trans-LTB, (i.e., 12-hydroxy-5-oxo-eicosadienoic acid). Metabolism of 12-epi-6-trans-LTB, by porcine leukocytes The second 6-trans isomer of LTB, formed from LTA, by PMNL (i.e., IZepid-tram-LTB,) was metabo-

154 lized much more slowly than 6-trans-LTB, by porcine leukocytes. The radioactive substrate was converted to two radioactive products, both of which had retention times longer than that of the substrate (data not shown). Neither of these products absorbed in the UV region at 235 nm or 280 nm, suggesting that they were both tetrahydro metabolites. The product with the shorter retention time had a mass spectrum identical to that of metabolite T-2 (tetrahydro-6-trans-LTB,; methyl ester, trimethylsilyl ether derivative). The second product had a mass spectrum identical to that of metabolite T-4 (12-oxo-tetrahydro-6-trans-LTB,; O-methyloxime, methyl ester, trimethylsilyl ether derivative). 12-epi-6trans-LTB, is therefore metabolized to tetrahydro-12epi-6-trans-LTB, and tetrahydro-12-oxo-6-trans-LTB, by porcine leukocytes.

Metabolism

of

I2-epi-8-cis-6-trans-LTB,

by

porcine

leukocytes

Porcine leukocytes metabolized 12-epi-8-cis-6-transLTB, (the product formed by metabolism of arachidonic acid by the combined actions of 5lipoxygenase and 12-lipoxygenase [16]) in a manner similar to LTB,. Two major products (E-l and E-2) with retention times (55.5 and 75.0 min, respectively) longer than that of the substrate were formed (Fig. 6). Both products absorbed at 235 nm, but not at 280 nm, suggesting that one of the three conjugated double bonds of the substrate had been reduced. The mass spectrum of the trimethylsilyl ether, methyl ester derivative of product E-l is very similar to those of the corresponding derivatives of lO,ll-dihydro-LTB, and lO,ll-dihydro-6-trans-LTB, (T-la; Fig. 2A). Intense ions were observed at m/z 496 (M), 481 (M - 15), 465 (M - 31), 406 (M - 90), 395 (C-5 to C-20), 385 (C-l to C-12) 375 (M - 31 - 90), 305 (395 - 90), 295 (385 - 90), 269 (C-l to C-lo), 268, 237, 229 (C-l to C-7), 213 (C-12 to C-20), 205 (385 - 2 x 90), 203 (C-l to C-5), 181, 159, 129 (base peak), 119 and 105. These results strongly suggest that metabolite E-l is lO,ll-dihydro-12-epi-8-cis-6-trans-LTB,. Similarly, the mass spectrum of the trimethylsilyl ether, methyl ester metabolite of E-2 was virtually identical to that of the corresponding derivatives of lO,lldihydro-12-oxo-LTB, and lO,ll-dihydro-12-oxo-6trans-LTB, (T-3a; see above), suggesting that E-2 is the lO,ll-dihydro-12-oxometabolite of 12-epi-8-cis-6trans-LTB, (i.e., 12-oxo-l0,ll,dihydro-8-cis-6-transLTB,). Gas chromatography of metabolite E-2 revealed that a second minor product was also present. This product had a mass spectrum similar to that of the tetrahydro metabolite of 6-trans-LTB, (T-2). However, the amount of the tetrahydro metabolite formed from 12-epi-8-cis-6-trans-LTB4 was much less than the amount of the corresponding metabolite formed from 6-trans-

1%epi-Gt-Bc-LTB,

40

80

TIME

(min)

Fig. 6. Reversed-phase high pressure liquid chromatogram of metabolites of 12-epi-8-cis-6-trans-LTB,. 12-epi-8-cis-6-rr~ns-[l-‘~C]LTB~ (2 PM, 0.1 Ci) was incubated with porcine leukocytes (10’ cells/ml) for 60 min at 37 o C. The chromatography was performed on a Phenomenex Ultracarb ODS column under isocratic conditions with a mobile phase consisting of water/acetonitrile/acetic acid (63 : 37 : 0.05) and a flow rate of 2 ml/min.

LTB,, and further not undertaken.

studies

on its characterization

were

Discussion

The metabolism of 6-trans-LTB, by porcine PMNL is much more extensive than that of LTB, due to the reduction of an additional double bond and oxidation of the 5-hydroxyl group. The initial reduction appears to proceed by two pathways in parallel (Fig. 7) resulting in the formation of two isomeric dihydro metabolites with hydroxyl groups in the 5 and 12-positions, one of which (T-la) appears to be analogous to lO,ll-dihydro-LTB,, the major metabolite of LTB, formed by porcine PMNL [15]. The second product (T-lb) is sirnilar to a dihydro product (5,12-dihydroxy-7,9,14-eicosatrienoic acid), which we previously identified after incubation of 12-epi-6-trans-LTB, with human PMNL [ll]. Unfortunately, only small amounts of these products were formed in the present study and it was not possible to rigorously determine the positions of the double bonds. Studies using 6-trans-LTB, labeled with deuterium in the 5- and 12-positions indicated that the deuterium in either one of these positions is largely lost in the above two dihydro metabolites. This suggests that 5-0~0 and 12-0~0 metabolites are also formed in these reactions. In agreement with this, different dihydro-oxo metabolites of 6-trans-LTB, were identified, one with an 0x0 group in the 12-position (T-3a), and another with an 0x0 group in the 5-position (T-3b). At least one of

155

I

dh-2

I

I

I R,hb

s

R,&&

I

I

I

dh-2

dh-2

dh-2

OH

red-2

red-2

red-2

R,&yq,%

z

R,&qRz

R,+WAfRZ

T_16

Fig. 7. Scheme showing The products identified bonds in the metabolites the reversible formation keto reductases are also

OH

T_2

b-i

T-4

OH

the possible mechanism for the formation of dihydro and tetrahydro metabolites of 6-trans-LTB4 by porcine leukocytes. in this study are underlined (T-la, T-lb, etc), whereas other products are hypothetical. The localizations of the double of 6-truns-LTB, have not been demonstrated unambiguously. Although single enzymes are shown in the figure to catalyze of 5-0x0 and 12-0~0 products from the corresponding hydroxy compounds, it is possible that additional enzymes, such as involved. Abbreviations: red-l, lO,ll-reductase; red-2, reductase-2, possibly a 6,11-reductase; dh-1, 12-hydroxy dehydrogenase or an epimerase; dh-2, 5-hydroxy dehydrogenase.

the above dihydro products is metabolized further to a series of tetrahydro products. The reactions discussed above would appear to require at least two enzymes, a reductase and a dehydrogenase. In the simplest case, the reductase could require the presence of 2 conjugated double bonds associated with a hydroxyl group in either the 5- or 1Zpositions. If there were a requirement for the double bond closest to the hydroxyl group to have the tram configuration, this could explain the fact that LTB, is not metabolized to tetrahydro products. The dehydrogenase could have similar specificities, explaining the lack of 5-0~0 metabolites of LTB,. However, this would not be consistent with the fact that only relatively small amounts of tetrahydro products were formed after incubation of 12-epi-8-cis-6-trans-LTB, with porcine leukocytes, in spite of the rapid conversion of this substrate to its lO,ll-dihydro and lO,ll-dihydro-12-oxo metabolites. Moreover, although human PMNL convert 6-tram isomers of LTB, to 6,11-dihydro metabolites, none of the corresponding tetrahydro metabolites was detected [ll]. On the other hand, preliminary experiments with rat PMNL suggest that 6-trans-LTB, is rapidly metabolized to lO,ll-dihydro and lO,ll-dihydro-1Zoxo metabolites, but not to tetrahydro products (unpublished results). The above considerations would suggest that 6-transLTB, is metabolized by two distinct enzymatic pathways (Fig. 7), one responsible for the formation of lO,ll-dihydro products and the other for the formation of putative 6,11-dihydro products. It should be made clear, however, that the positions of the conjugated double bonds in the ‘6,11-dihydro’ products have not

been firmly established. The first pathway, involving a lO,ll-reductase (‘red-l’ in Fig. 7), appears to be associated with another enzyme(s) ‘dh-1’ in Fig. 7) which catalyzes the reversible formation of 12-0~0 products. We have recently shown that porcine PMNL reversibly convert lO,ll-dihydro-LTB, to its 12-epi isomer, along with lO,ll-dihydro-12-oxo-LTB, [19], raising the possibility that an epimerase-like enzyme may be involved in this process. Alternatively, the reaction could be catalyzed by a 12-hydroxy dehydrogenase alone or in combination with a keto reductase. For example, one or two keto reductases may be involved in the reduction of 12-oxo-5,8,10,14-eicosatetraenoic acid to the corresponding 12(S) and 12(R)-hydroxy compounds by rat liver microsomes [20]. A second pair of enzymes (‘red-2’ and ‘dh-2’ in Fig. 7) could be responsible for the formation of the other dihydro (T-lb) and dihydro-5-oxo (T-3b) metabolites. The reductase and 5-hydroxy dehydrogenase required for the latter pathway could be similar to the enzymes in human PMNL which convert 12-epi-6-trans-LTB, to 6,11-dihydro-12-epi-6-trans-LTB,, apparently via a 50x0 intermediate [ll]. The mechanism for the reduction of 6-tram isomers of LTB, by the second reductase is quite different from that for the lO,ll-reductase, which simply catalyzes the 1,Zaddition of hydrogen to the lO,ll-double bond of 12-hydroxy-eicosanoids. The former enzyme would appear to catalyze the 1,6-reduction of 5-oxo-6-trans-LTB, to give a ‘6,11-dihydro’ product (T-3b), presumably by the addition of a hydride anion to carbon-11, followed by ketonization of the resulting enolate intermediate as shown in Fig. 8.

156 OH

OH

H+

OH

H-

1.6~Reduction

H-

1.4~Reduction Fig. 8. Possible mechanism

for the reduction

of 6-rruns isomers of LTB, by the second reductase

The same enzyme could also catalyze the 1,4-reduction of lO,ll-dihydro metabolites of 6-truns-LTB, to tetrahydro products by addition of hydride at carbon-9 instead of carbon-11 (Fig. 8). Assuming that the Shydroxy dehydrogenase and the 12-hydroxy dehydrogenase (or epimerase) catalyze reversible reactions, a combination of these two enzymes with the two reductases could be responsible for the formation of the tetrahydro metabolites of 6-truns-LTB,, as shown in Fig. 7. However, the metabolism of 6-truns-LTB, may well be more complex than indicated in Fig. 7 and other enzymes, such as keto reductases, could also be involved. We have previously shown that when deuteriumlabeled LTB, was incubated with porcine PMNL, nearly all the deuterium in the Sposition and about two-thirds of the deuterium in the 1Zposition of lO,ll-dihydroLTB, was retained [15]. However, when deuterium-

labeled 6-truns-LTB, was converted to tetrahydro-6truns-LTB, by porcine leukocytes, nearly all of the deuterium in the 5-position and over 80% of the deuterium in the l%-position was lost. This suggests that oxidation of the 5-hydroxyl group may be required prior to reduction by the second reductase enzyme (i.e., reductase-2, Fig. 7) as appears to be the case for the reduction of 12-epi-6-truns-LTB, by human PMNL [ll]. The retention of 15 to 20% of the deuterium in the 1Lposition in the tetrahydro metabolite of 6-truns-LTB, would suggest that, as with LTB,, prior oxidation of the 1Zhydroxyl group is not essential for reduction of the lO,ll-double bond. However, although not shown in Fig. 7, it is also possible that a 12-hydroxy dehydrogenase could oxidize 6-truns-LTB, directly to 12-0x0-6truns-LTB,, which could then be reduced by the lO,llreductase.

STRUCTURAL REQUIREMENTS

PATH WAY

enzyme.

SUBSTRATES

SPECIES

6-trans-LTE4

lO,ll-reductase

02H

12-dehydrogenase

f2-epi-6-trans-LTB4 12-epi-8-cis-6-trans-LTB4

PIG RAT

12-HETE

HO 6-trans-LTB4 reductase-2 5-dehydrogenase

Fig. 9. Characteristics

of the two reductase

‘zH

pathways

which metabolize the two pathways

12-epi-6-trans-LTB, 12-epi-8-cis-6-trans-LTE4

eicosanoids in PMNL. are enclosed in boxes.

The structural

PIG HUMAN

regions

important

for metabolism

by

157 In conclusion, our results suggest that there are two reductase pathways which convert dihydroxy-triene eicosanoids to dihydro metabolites (Fig. 9). The lO,llreductase pathway results in the reduction of LTB,, 6-trans-LTB,, 12-epi-8-cis-6-trans-LTB,, 12-HETE [21] and, to a lesser extent, 12-e@-6-truns-LTB,. It appears to require a 1Zhydroxyl group preceded by lO,ll-trans and 8,9-cis or 8,9-tram double bonds. We have identified this pathway in porcine [15] and rat [12,22] PMNL and there is evidence that it also exists in rat mesangial cells and fibroblasts [13], mouse macrophages and Tlymphocytes [13] and human alveolar macrophages [14]. The second pathway appears to involve another distinct reductase enzyme, along with a 5-hydroxy dehydrogenase. 6-truns-LTB,, 12-epi-6-trans-LTB,, and, to a lesser extent, 12-epi-8-cis-6-tram-LTB,, but not LTB,, are metabolized by this pathway. It is present in both human and porcine PMNL. The lO,ll-reductase pathway may be important in the biological inactivation of LTB,. A dihydro metabolite of LTB,, probably lO,ll-dihydro-LTB,, was found to be significantly less active than LTB, in its effects on human PMNL [23]. Alternatively, this pathway could be responsible for the formation of biologically active products such as 12( R)-hydroxy-5,8,14-eicosatrienoic acid (i.e., the lO,ll-dihydro metabolite of 12(S)HETE *), which has been reported to have potent proinflammatory effects [24]. We have recently identified this product as the major metabolite of 12( S)-HETE in porcine PMNL (unpublished work). The associated 12-hydroxy dehydrogenase or epimerase could catalyze the inversion of the stereochemistry at carbon-12 of 12-hydroxy-eicosanoids and could therefore possibly convert 12(S)-HETE to 12(R)-HETE, which is a more potent proinflammatory agent than the former compound [25]. The significance of the ‘6,11-reductase’ pathway is less obvious, since only 6-tram isomers of LTB, and apparently not LTB, itself are substrates. It is possible that this pathway could be involved in the metabolism of other related substances which, unlike 6-trans-LTB,, are biologically active.

* Due to the priority rules in assigning R and S configurations, the 12-lipoxygenase product, 12(S)-HETE has the 12s configuration, whereas its lO,ll-dihydro metabolite has the 12R configuration, even though the stereochemistry of the 1Zhydroxyl group has not been inverted.

Acknowledgements The authors are grateful to Dr. 0. Mamer of the Department of Medicine, McGill University, for assistance with the gas chromatography-mass spectrometry. This work was supported by grants from the Medical Research Council of Canada and Quebec Heart Foundation. References 1 Borgeat, P. and Samuelsson, B. (1979) Proc. Natl. Acad. Sci. USA 76, 3213-3217. 2 Lindgren, J.A., Hansson, G. and Samuelsson, B. (1981) FEBS Lett. 128, 329-335. 3 Powell, W.S. (1984) J. Biol. Chem. 259, 3082-3089. 4 Shak, S. and Goldstein, I.M. (1985) J. Clin. Invest. 76, 1218-1228. 5 Soberman, R.J., Okita, R.T., Fitzsimmons, B., Rokach, J., Spur, B. and Austen, K.F. (1987) J. Biol. Chem. 262, 12421-21427. 6 Soberman, R.J., Sutyak, J.P., Okita, R.T., Wendelbom, D.F., Roberts, L.J. and Austen, K.F. (1988) J. Biol. Chem. 263, 79968002. 7 Hansson, G., Lindgren, J.A., DahlCn, S.-E., Hedqvist, P. and Samuelsson, B. (1981) FEBS Lett. 130, 107-112. 8 Jubiz, W., Radmark, O., Malmsten, C., Hansson, G., Lindgren, J.A., Palmblad, J., UdCn, A.-M. and Samuelsson, B. (1982) J. Biol. Chem. 257, 6106-6110. 9 Sumimoto, H., Takeshigi, K. and Minakami, S. (1985) B&hem. Biophys. Res. Commun. 132, 864-870. 10 Powell, W.S. (1986) Biochem. Biophys. Res. Commun. 136, 707712. 11 Powell, W.S. and Gravelle, F. (1988) J. Biol. Chem. 263,2170-2177. 12 Powell, W.S. (1987) Biochem. Biophys. Res. Commun. 145, 991998. 13 Kaever, V., Martin, M., Fauler, J., Marx, K.-H. and Resch, K. (1987) Biocbim. Biophys. Acta 922, 337-344. 14 Schonfeld, W., Schliiter, B., Hilger, R. and Konig, W. (1988) Immunology 65, 529-536. 15 Powell, W.S. and Gravelle, F. (1989) J. Biol. Chem. 264,5364-5369. 16 Borgeat, P., Picard, S., Vallerand, P. and Sirois, P. (1981) Prostaglandins Med. 6, 557-570. 17 Powell, W.S. (1980) Prostaglandins 20, 947-957. 18 Fruteau de Laclos, B., Maclouf, J., Poubelle, P. and Borgeat, P. (1987) Prostaglandins 33, 315-337. 19 Wainwright, S., Falck, C., Yadagiri, P. and Powell, W.S. (1990) Biochemistry, 29, 1180-1185. 20 Falgueyret, J.P., Leblanc, Y., Rokach, J. and Riendeau, D. (1988) Biochem. Biophys. Res. Commun. 156,1083-1089. 21 Wainwright, S. and Powell, W.S. (1989) New Trends in Lipid Mediators Research, Vol. 3 (Zor, U., Naor, Z. and Danon, A., eds.), pp. 200-203, Karger, Basel. 22 Powell, W.S. and Gravelle, F. (1990) J. Biol. Chem., in press. 23 Kaever, V., Damerau, B., Wessel, K. and Resch, K. (1988) FEBS Lett. 231, 385-388. 24 Murphy, R.C., Falck, J.R., Lumin, S., Yadagiri, P., Zirrolli, J.A., Balazy, M., Masferrer, J.L., Abraham, N.G. and Schwartzman, M.L. (1988) J. Biol. Chem. 32, 17197-17202. 25 Cunningham, F.M. and Woollard, P.M. (1986) Prostaglandins 32, 387-399.