- Email: [email protected]
16α-Dehydration of corticoids by bacteria isolated from rat fecal flora
(w)??-47~1~H2/020231-07SO3.00/0 Copyrtght 0 IVX2 Pcr~mon Press Lid
16a-DEHYDRATION OF CORTICOIDS BY BACTERIA ISOLATED FROM RAT FECAL FLORA JEAN-
WINTER*~,SHERYLO'ROURKE~, VICTOR D. BOKKENHEUSER?, FQILIPB.HYLEMON$ ~~~THOMAS L.GLASS$
tDepartment of Pathology, St. Luke’s_Roosevelt Hospital Center-New York, NY 10025 and IFrom the Department of Microbiology, Medical College of Virginia, Richmond VA 23298, U.S.A. (Receioed 22 April 1981) SUMMARY Two
strains (No. 144 and No. 146) of rat intestinal anaerobic bacteria, phenotypically similar to Euboccorticoids. The initial step in the 16a-dehydration of 16a-hydroxyprogesterone was dehydration at the C-16 and C-17 position with the accumulation of 16-dehydroprogesterone. This step required the side chain at C-17. In bacterial cultures the 16-dehydroprogesterone was then slowly reduced to &-progesterone. 16a-Hydroxypregnanolone was also converted to iso-pregnanolone by these bacteria. 16u-Dehydratase was easily demonstrated in cell fractions of strain No. 144 incubated either aerobically or anaerobically. The same extracts did not convert 16-dehydroprogesterone to iso-progesterone under similar assay conditions. 16a-Dehydration occurred at all substrate concentrations tested (20 to 200 pgjml) provided the pH of the growth medium was between 6 and 8 and the Eh below - 130mV. Strain No. 146 had both 16a-dehydration and 21-dehydroxylation activities. The two enzymes functioned independently. A role for intestinal bacteria in the biotransformation of biliary 16a-hydroxylated steroids and subsequent excretion in the urine is proposed. terium lentum, were isolated and found capable of 16a-dehydrating
INTRODUmION
In 1962 Calvin and Liebertnanrl] demonstrated that 16a-hydroxyprogesterone is excreted in the urine as
iso-pregnanolone. In both humans and rats, 16a-hydroxyprogesterone is the precursor of biliary 16ahydroxypregnanolone [2-81. This steroid is exposed to the intestinal flora during the enterohepatic circulation. Eriksson and Gustafsson[9] demonstrated that gnotobiotic rats excreted 16a-hydroxypregnanolone unchanged while normal rats l&-dehydrated the compound and converted the side chain from B to a. Recently we isolated two bacteria from rat feces, designated strains No. 144 and No. 146, which carried * Please address reprint requests to Dr J. Winter. This investigation was supported by grant 25763, awarded by the National Cancer Institute, DHEW; by grant 25324, awarded by the National Institute of Arthritis, Metabolism and Digestive Diseases, DHEW; and by a grant from the Fannie E. Rippel Foundation. Abbreviations used: 16a-hydroxyprogesterone 16a-hydroxy+pregnene-3,20-dione; 16-dehydroprogesterone4,16pregnandiene-3,20_dione; iso-progesterone l’la-pregn4 ene-3,2O-dione;progesterone4-pregnene-3,2~dione; 16a-hydroxypregnanolone 3~, 16a-dihydroxy-S/J-pregnan-20-one; iso-pregnanolone 3a-hydroxy-58, 17a-pregnan-M-one; deoxycorticosterone (DOC) 21-hydroxy4pregnene-3,20dione; pregnanolone 3a-hydroxy-5p-pregnan-3.2O-dione; tetrahydrodeoxycorticosterone 3a,21-dihydroxy-5/?-pregnan-20-one; 16a-hydroxy-estrone 3,16a-dihydroxy-1,3,5 (IO)-estratrien-17-one; 16a-hydroxy-androsterone 3a,16adihydroxy-5a-androstan-17-one. LB. lb,‘2--ca
out the l&-dehydration of corticoids. These bacteria are obligate anaerobes and were identified as organisms phenotypically similar to Eubacterium lenturn’ [lo]. In this paper $e describe the kinetics of 16adehydration by these bacteria. MATERIALS AND METHODS
Culture media Bacterial 16a-dehydration was examined in prereduced Brain Heart Infusion Broth (PR) supplemented with 0.05’4 cysteine-HCl (Scott Laboratories Inc.) or in BHIC medium with the following composition per liter: dehydrated brain heart infusion broth (Baltimore Biological Lab) 37 g; cysteine-HCl, 0.5 g; NaHCO,, 1 g; 4 ml of 0.025% aqueous resazurin (J. T. Baker Chemical Co). The medium was distributed in 50ml amounts and sterilized at 121°C for 20 min. For certain experiments, media were biologically reduced [l l] by addition of 0.1 ml of a 24-h old culture of Escherichia coli (BHIC-EC). Microorganisms
(a) Culture No. 144 and No. 146 were isolated from the fecal flora of rats [lo].
(b) E. lenrum neotypc strain (V.P.I. 0225) and other strains of E. lenturn were kindly supplied by Drs L. V. Holdeman and W. E. C. Moore, Anaerobe Laboratory, Virginia Polytechnic Institute (V.P.I.), Blacksburg, Virginia.
231
232
JEANETTE WINTERet al.
of 5 fig/ml and l”/;, CH30H (v/v). Incubation was continued another 5 h. The culture was then harvested by TLC* HPLQ GLC$ centrifugation at 8000 g for 15 min at WC under Steroid tR tR R, argon. The cell pellets were combined and suspended in l.Oml of boiled and cooled (under argon) 50mM 16u-Hydroxyprogesterone 0.06 1.70 1.83 16-Dehydroprogesterone 0.59 2.50 1.06 potassium phosphate buffer, pH 6.8 containing 1 mM Iso-progesterone 0.53 1.90 0.85 dithiothreitol (anaerobic buffer). The cell suspension Progesterone 1.15 0.59 2.50 was centrifuged for 15 min at 12,000 g at -4°C. The Deoxycorticosterone 0.37 2.00 0.90 cell pellet was stored overnight at -4°C under argon. The frozen pellet was thawed in 5 ml of anaerobic * Solvent system, chloroform-acetone (95 : 5, v/v) t Solvent system, acetonitrile-water (70: 30, v/v) buffer under argon. The cells were broken by two $Oxime derivatives on OVlOl column, 2R at 230°C. passages through a chilled French Pressure Cell at Relative rR/5atholestane 1034-1266 kg cm-‘. The broken cell suspension was then centrifuged at 12,000g for 20min at 04°C. The supematant fluid was removed and centrifuged at (c) Clostridium paraputr@um and E. coli were stock 105,ooOg for 2 h at (14°C to yield the soluble superstrains recovered from human fecal flora [12,13]. The natant fluid and membrane pellet. The supernatant fluid was removed and stored under argon at 0°C. cultures were maintained in a lyophilized form and, prior to use, passed two or three times in PR medium The membrane pellet was washed twice with 4ml of at 37°C. anaerobic buffer and stored at 0-4”C under argon. The assay conditions for 16z-hydroxyprogesterone Steroids and solvents dehydratase and 16dehydroprogesterone reductase Unlabeled steroids were purchased from Steraloids are described in the legend to Table 2. Briefly, assays were carried out at 37°C for 10 min (16adehydratase) Inc. and Makor Chemicals Ltd. Israel; isoprogesterone was synthesized in our laboratory as previously and 30 min (16dehydroprogesterone reductase). The described [lo]. Solvents were reagent grade except for reactions were terminated by extracting 0.15 ml aliquots twice with 2ml of ether. The ether phase was methanol which was technical grade; acetonitriie and blown off and the residue dissolved in 0.25 pll water were HPLC grade. CHsOH. Steroids were separated and quantitated by Conversion experiments. Prior to incubation, HPLC using a Beckman model 332 HPLC equipped steroids were dissolved in 0.25 ml methanol and and a added to the medium to give a concentration of 20 ,ug with a Hitachi u.v.-VIS Spectrophotometer steroids/ml, if not otherwise stated. The media were Shimadzu C-RlA integrator. The column used was a Beckman 4.6mm x 15cm C-18 ultrasphere-ODS then seeded with 0.25ml of a 48-h old culture and reverse phase and the system was operated isocraticincubated at 37°C for 7 days. At the end of the incubation, pH and Eh were measured (Beckman Zeroally using CHsOH/H1O (78:22 v/v) at a flow rate of matic II). A platinum electrode was used for the Eh 1.25 ml/min. The steroids were detected by monitormeasurement. Steroid extraction and identification ing at 240 nm and quantitated by external standardizwere carried out as previously described [lo, 143. ation. Briefly, the steroids were extracted with organic Protein was determined by the method of Lowry et a/.[151 after precipitation with 10% TCA, dissolving solvents, concentrated and purified by thin-layer the precipitate in 0.1 N NaOH and heating at 100°C chromatography (t.1.c.). a&Unsaturated carbonyl groups were detected from their absorbance at for 15min. Bovine serum albumin was used as the protein standard. 254 nm. The metabolites were identified by their relative R, values on t.l.c, their tR in the high performance liquid chromatograph (HPLC) and by gas-liquid chromatography. Ultimately the metabolites were RESULTS characterized by their mass spectrum as described J6a-Dehydration of !6a-hydroxyprogesterone previously [ 143. Chromatographic data of reference The results described below in experiments with steroids are given in Table 1. strain No. 144 apply equally well to strain No. 146 Preparation of cell extracts and assay for f6a-dehydraunless otherwise stated. tase and ltiehydroprogesterone reductase activities PR medium with initial pH ranging from 5.0 to 9.0 with 16a-hydroxyprogesterone Eubacterium sp. No. 144 was grown in 3 liters of was supplemented BHIC medium. The medium, after autoclaving was (2O&ml), seeded with strain No. 144, and incubated occurred rapidly when cooled on ice to room temperature while gassing with for 7 days. &-Dehydration the initial pH of the conversion medium was between C02. Prior to inoculation with 10 ml of strain No. 144, the gas phase was shiRed to 100% argon. The 6.0 and 8.0 (Fig. 1). Small amounts of acid developed culture was incubated for 17 h at 37°C. At this time during incubation, but the pH decrease was ~0.2 units. Growth was reduced at pH 5.0 and completely the substrate, 16a-hydroxyprogesterone or S-dehyinhibited ~~...._... _ at DH , 9.0. droprogesterone, was added to a final concentration Table 1. Chromatographic
data of reference steroids
16uDehydration Table 2. 16a-Hydroxyprogesterone
dehydratase and 16dehydroprogesterone Eubucrerium sp. No. 144
105,000 g Preparation
233
of corticoids
reductase activities in cell fractions of
16x-Hydroxyprogesterone dchydrata.se* (nmol 16_dehydroprogesterone/min/mg protein)
16Dehydroprogesterone reductaset (nmol i-progesterone/min/mg protein)
36 35
Supernatant fluid Membrane enriched fraction
* The assay mixture contained 50 mM potassium phosphate buffer (pH 6.8); 75 pg 16a-hydroxyprogesterone; CH,OH,
25% (v/v) and 0.5 mg protein (supernatant) or 0.5 mg protein (membrane) in a final volume of 0.25 ml.
t The assay mixture for 16dehydroprogesterone reductase was similar to “*“, except that 25 c(g lbdehydroprogesterone were added instead of 16x-hydroxyprogesterone and the reactions were carried out under an argon gas atmosphere. The following cofactors or combinations were individually tested as electron donors: NADH, NADH + FAD. NADH + FMN, NADPH + FAD, and NADPH + FMN. The reduced pyridine nucleotides and flavin nucleotides were added to a final concentration of 1 mM and 0.05 mM. respectively. In no case was iso-progesterone detected under these conditions nor was this product formed when the supernatant and pellet were mixed and assayed as above. In each case, a small fraction of the 16dehydroprogesterone added was hydrated to 16a-hydroxyprogesterone.
Eflect of co-culturing strain No.144 anaerobic bacteria
with facultatively
Aerobic BHIC medium does not support growth of obligate anaerobes. However, simultaneous seeding of the medium with strain No. 144 and either E. co/i, Proteus mirabilis or a Klebsiella species, permitted
of strain No. 144 and subsequent Madehydration. None of the facultative anaerobes transformed 16a-hydroxyprogesterone. In other experiments, BHIC medium containing 16a-hydroxyprogesterone (20 c(BIml) was inoculated with 0.1 ml broth culture of E. coli and incubated at 37°C. Regardless of whether strain No. 144 was added early or after 1 week of incubation with E. coli, 16ahydroxyprogesterone was quantitatively converted to iso-progesterone within 7 days after inoculation. Alternatively, BHIC medium inoculated simultaneously with E. coli and strain No. 144, was incubated at 37°C. Whether the substrate was added immediately or after 1 week incubation of the mixed culture, complete conversion to iso-progesterone was observed. Identical results were obtained when PR medium was inoculated with strain No. 144 and the substrate was added at different times. growth
Effect of substrate concernration on 16adehydration
16aDehydration by dilute fecal flora depends in part on the concentration of substrate [14]. To deter-
mine if this were true for pure cultures, strain No. 144 was inoculated into PR and BHIC-EC media at concentrations of substrate ranging from 20 @ml to 200 &ml. 16a-Hydroxyprogesterone was quantitatively converted to iso-progesterone at all levels of substrate tested. Because of the limited solubility of 16clhydroxyprogesterone in methanol and the desire to keep the concentration of the bactericidal solvent in the growth medium below lo/ it was not possible to achieve higher substrate concentrations. Kinetics of 16adehydration
Eight vials of PR medium supplemented with Ma-hydroxyprogesterone (20 pg/ml) were seeded with strain No. 144 and incubated at 37°C. Culture samples were taken at different times and the contents of each vial was submitted to steroid analysis. The results in Fig 2 shows that, within 12 h, the substrate was quantitatively converted to Mdehydroprogesterone. The second step of the reaction, the saturation of the double bond required about 36 h for complete conversion. Identical results were obtained with BHIC-Ec medium. Substrate specificity of 16a-dehydratase
To determine the substrate specificity of 16c+dehydration, deoxycorticosterone (DOC), 16a-hydroxyA-b
ISd-hydroxy
0-0
A’!pro~rrtrron. Ir0-pros.*t.ronr
x---i
~rog.rc.rons
0 5
6
7
6
9
PH
Fig 1.
lb-Dehydration
by strain No. 144 as a function of PH.
Time,
hrs.
Fig. 2. Kinetics of l&x-dehydration by strain No. 144.
JEANEITEWINTERet al.
234
too/
,/__/--*
/
f
/”
:! so-
9
0
0
,P/lP
/
z
0
ti
x--x
BCIlC.EC
o---D
pR
I’
I
I2
-
24
36
40
72
60
Tim.. hrr
0
12
24
36
42
Fig 3. Kinetics of i~hydros~~d No. 144.
Fig 4. 2LDehydroxylation capacity in two media at various Eh. Substrate: DOC; end product: progesterone.
Ttme. hrr
reductase by strain
have found that dialysis of the supernatant fluid had no effect on the 1fSadehydratase activity indicating no estrone or Ma-hydroxy-androsterone fall at 20 Irg/ ml) requirement for low molecular weight dialyxable were incubated with strain No. 144 in PR medium. cofactors. Furthermore, anaerobic precautions and None of these substrates were detectably dehydrated assay conditions are not necessary for measuring 16aby this bacterium The same negative results were dehydratase activity. Thus, the enzyme appears to be obtained in BHIC-ECmedium. However, strain No. 146 relatively oxygen stable. In contrast, we were unable converted DOC to progesterone and tetrahydro- to demonstrate lddehydrosteroid reductase activity deoxycorticosterone to pregnanolont indicating the in these extracts using the assay conditions described. presence of 21-dehydroxylation activity in this strain. l&z-Dehy&ationand 2l~ehy~roxy~~ion ofsteroids by To determine if related steroids inhibit the strain No. 146 l~d~ydration reaction, vials of PR and BHIC-EC Kinetic studies of 2ldehy~oxy~tion of DOC by media were supp~t~ with 1~“h~ox~ro~terstrain No. 146 were carried out in PR and BHIC-EC one and inoculated with strain No. 144.l&Hydroxymedium. Figure 4 shows that the 21dehydroxyiation estrone (100 pg/ml) was added to one set of vials, 16a-hydroxy-androsterone (100 pgjml) to another set, was much faster in BHIC-EC than in PR medium. In additional studies, varying concentrations (3 JAM and DOC (100&ml) to a third set. Inhibition of to 55 PM) DOC were added to conversion media. The 16adehydration was not detected but traces of unresults presented in Fig. 5 shows the E. coli enhanced identified steroid metabolites of lower polarity were present in the vials containing C-18 and C-19 steroids. the capacity of strain No. 146 to Zldehydroxylate DOC. The maximum amounts of DOC converted in 50ml of PR medium and BHIC-EC medium were 24 Comersion of f6-dehydr5pr5gestefone to i.w-progesterand 39 rmol, respectively. one The kinetic experiments &mo~~at~ that the l~dehy~ation is a two step reaction, suggesting that the conversion involves at least two enzymes; one that converts the substrate to a M-dehydrosteroid, and another that reducea the double bond, leaving the side chain in the a-position @o-progesterone).
To confirm this hypothesis, M-dehydroprogesterone was added to media and incubated for 48 h. In PR medium, reduction of the 16dehydroprogesterone was slow for the first 12 h and then increased in velocity until all substrate was converted (Fig. 3). It was completed within 48 h. Interestingly, the rate of the reaction was not enhanced by the presence of E. co/i in the medium. 16a-Dehydratase
activity in cell fiactioas
The data in Table 2 shows that 16a-hydroxyprogesterone dehydratase activity was easily demonstrated in both the soluble and membrane fractions of E&Xterium sp. No. 144. The membrane enriched fraction was washed twice before assaying; hence, the membrane associated activity is not 1ikeIy to be due to gross contamination by soluble protein. Moreover, we
Si~~t~~s ~t5~~~rn of I ~-hy~oxypr~est~one and DOC by strain No. 146 Because strain No. 146 has both l&dehydration and 2ldehydroxylation activities, competition between these enzymes was investigated in growing cultures. PR and BHIC-EC media were supplemented with Ma-hydroxyprogesterone (200&ml) and DOC (20 @ml, 160&ml, or 260 pg/mlA respectively. l&x-
100kq
-x\ 3
t e” 0
x
\
._s u)
1 I l
so-
3
0
\
I I I I
-
I 25
x--x o--
BHIC-EC QR
X \
I 3s pM
t\. 45
55
DOC
Fig. 5. 21-Dehydroxylation capacity at various concentrations of substrates. Substrate: deoxycorticosterone; end product: progesterone.
lb-Dehydration
235
of corticoids
Stroin +I46
cn I
@
I 3 :=o
, m
Fig. 6. Bacterially mediated metabolic pathway of 16x-hydroxyprogesterone in vitro. I = 16x-hydroxyprogesterone; II = 16dehydroprogesterone; III = iso-progesterone; IV = Ha-hydroxypregnanolone; V = lddehydropregnanolone; VI = iso-pregnanolone; + demonstrated pathway ---a hypothetical pathway.
Hydroxyprogesterone w’as quantitatively converted to iso-progesterone in all cultures regardless of the concentration of DOC. Moreover, 21-dehydroxylation of DOC was complete in the presence of low concentrations of DOC but did not go to completion in the presence of high concentrations.
their similarity to E. lentum, we examined 10 strains of the latter species for Wx-dehydratase activity using 16a-hydroxyprogesterone as substrate [18]. However, no 16adehydratase activity was detected in these bacteria.
Metabolism of 16a-hydroxyprogesterone in mixed cultures of C. paraputrificum and strain No. 144
DISCUSSION
Experiments with mixed fecal flora suggest that the conversion of 16a-hydroxyprogesterone to iso-pregnanolone requires the combined enzyme activities of bacteria having the ability to reduce the 4-ene-fketo structures [16,17] and organisms having 16c+dehydration activity. This was confirmed in a series of experiments in which 16a-hydroxyprogesterone or 16 dehydroprogesterone were quantitatively converted to iso-pregnanolone by the combined action of C. paraputrificum and strain No. 144. Moreover, DOC was also quantitatively reduced to pregnanolone by C. paraputrijicum and strain No. 146. Distribution of bacteria having 16adehydration actioity
Isolates No. 144 and No. 146 are the first organisms known to synthesize 16x-dehydratasc. Because of
Mechanism of l&x-dehydration
16aDehydration was carried out rapidly and efficiently by strains No. 144 and No. 146 in both PR and BHIC-Ec media at all substrate concentrations tested (2O-2OOIrglml), provided the Eh was below - 130 mV and the pH between 6 and 8. Kinetic experiments confirmed Calvin and Lieberman’s hypothesis [l] that &dehydration of 16a-hydroxyprogesterone proceeded via an intermediary, 16dehydroprogcsterone. This compound was formed rapidly from the substrate by dehydration at C-16, C-17 positions. 16a-Hydroxy-estrone and 16a-hydroxy-androsteone, proved resistant to 16adehydration. Moreover, l&dehydration did not occur with C-18 or C-19 compounds with an OH group at C-17 (unpublished observations). This suggests that the 16adehydratasc requires the steroid side chain. Substrate specificity
JEANEITEWINTER et al.
236 pOti
$“3 c=o
p0
0
&
/ A&
-OH
0
0
16u- hydroxy- proQostuonr
Dqaaycorticoakronr
\
yz-O”
uvr
y3 c=o
c=o
-OR
Ro’PX
RO’
H THOOC
&
”
16d-hydroay-prrqnonolom
M
R*gnonolon*
ho- prognonolono
Fig. 7. Proposed metabolic pathways in viva.
and mechanism of reaction must await studies with purified enxymes. Because strains No. 144 and No. 146 dehydrated Ma-hydroxypregnanolone and Ma-hydroxyprogesteronc equally well in growing cultures, it follows that enzyme specifrcity is independent of the Gene-3-keto structure in ring A. However, studies with purified enxyme will he neassary to determine which might be the prekrmd steroid substrate. The second step of the biotnu&rmatios the reduction of the Sdehydrosteroid has been reported to involve the trans addition of the H to the Ma and 178 position [193. Because the sequential removal of the 16a-hydroxy group and the subsequent reduction of the Mdehydrosteroid could be separated in time and were of different velocity, we believe that more than one enzyme is required for the l&-dehydration and subsequent conversion to iso-progesterone.
16sDehydration strain No. 146
and 2Idehydroxylation
activities in
In previous studies we showed that the presence of E. coli in mixed cultures enhanced 21dehydroxyL ation activity of strain No. 116 apparently by reducing the Eh to - 230 mV [ 1l] or by supplying reducing equivalents in the form of molecular hydrogen [20]. The 2ldehydroxylation activity of strain No. 146 behaved in a similar manner. In contrast, l&x-dehydration activity in strain No. 146 functioned efhciently up to -13OmV and was not inhibited by 21-hydroxylated steroids. Steroid biotran@uwtarion patterns in vitro and in vivo In vitro, DOC is converted to pregnanolone by the joint action of C. puruputrificum and strain No. 146 by the mechanism previously proposed for this biotrans-
16xDehydration formation
112). Briefly, because of the rapid growth of
and the high velocity of 4-ene-3keto reduction, THDOC is observed as an intermediary metabolite. However, the reaction may begin with a tl-dehydroxylation of DOC to progesterone, which is ring-A reduced so rapidly that it never is observed in a culture containing C. p~~upuf~~~c~~ [12]. In uitro formation of iso-pregnanolone from 16a-hydroxyprogesterone is a multistep biotransformation performed by two anaerobic bacteria each containing separate enzyme activities (Fig. 6). C. parapurrificum reduces the 4-ene-3-keto structure and strain No. 144 removes the C-16OH group, reduces the 16dehydrosteroid and converts the side chain. fn GUO 16ff-hydroxyprogesterone and 21-hydroxycorticoids are excreted in the bile as ring-A reduced products conjugated with glucuronic acid [II J. The biliary steroids are deconjugated in the gut, 21-dehydroxylated by E. fe~r~m or phenotypi~iiy similar organisms [18] and 16a-dehydrated by bacteria similar to strains No. 144 and No. 146 (Fig. 7). The dehydrated products are absorbed, further reduced in the liver, conjugated with glucuronic acid and excreted in the urine. It is noteworthy that the position of the side chain of urinary metabolites reveals which steroids have been 16x-dehydrated. C. paraputrijcum
Taxonomic position of strains No. 144 and No. 146 Strains No. 144 and No. 146 are the first bacteria isolated that have 16a-dehydrative activity. These organisms appear to be phenotypi~liy similar to Eubacterium Ientum. It is interesting that bacteria observed this far capable of dehydrating ring-D and its side chain all are gram-positive, obligate anaerobes, non-spore forming rods that do not ferment carbohydrates [2l]. Evidence is accumulating that members of this group of bacteria are rich in enzymes involved in the biotransformation of steroids [ 18,223.
REFERENCES 1. Calvin H. L. and Lieberman S.: Studies on the metabolism of 16a-hydroxyprogesterone in humans. Conversion to urinary 1%isopregnanolone. Bjoc~ern~rry 1 (1962) 639-645. 2. Laatikainen f.: Identification of ClpOt and C2r02 steroids in the glucuronide fraction of human bile. Eur. J. Biochem. 10 11969) 165-171. Laatikainen T.t rd~ntification of C19C2 and C2i02 steroids in the mono- and disuiphate fractions of human feces. Steroids IS (1970) 139-150. Laatikainen T. and Vihko R.: Identification of C,.O, and CztOz steroids in the glucuronide fraction 01 human bile. Steroids 13 (1970) 534-538. Taylor W.: The excretion of steroid hormone metabolites in bite and feces. Warn. Horm. 29 (1970) 201-285.
237
of corticoids
Ashmore J.. Elliott W. H., Doisy E. A. Jr and Doisy E. A.: Excretion of metabolites of testosterone 4X14 in rats. J. &of. Chem. 200 (1953) 661668. Bocklage 8. C.. Daisy E. A. Jr, Elliott W. H. and Doisy E. A.: Absorption and metabolism of cortisone~C14. J. biol. Chek
212 (1955) 935-939.
Shen H.. Elliott W. H.. Doisv E. A. Jr and Doisv E. A.: Excretidn of rne~boi~t~ o~pro~esterone-21-~~4 after intragastric administration to rats. J. biol. Clrem. 20s (1954) 133-133. 9. Eriksson H., Gustafsson J.-A. and Sjovall J.: Steroids in germ-free and conventional rats. 4. Identification and bacterial formation of I Xx-pregnane derivatives. Eur. J. Biochem. 6 (1968) 219-226. V. D., Winter J., O’Rourke S. and Rit10. ~kkenheu~r chic A. E.: Isolation and characterization of fecal baa teria capable of I&-dehydroxylating corticoids. Appi. Environ. ~ierobiof. 40 (1980) 803-808. 11. Winter J. and Bokkenheuser V. D.: 21Dehydroxylation of corticoids by anaerobic bacteria isolated from human fecal flora. .I. steroid Biochem. 9 (1978) 379-384. 1.2. Bokkenheuser V. D., Winter J., Dehazya P. De Leon 0. and Kelly W. G.: Formation and metabolism of tetrahydrodeoxycorticosterone by human fecal flora. 1. steroid &o&em. 7 (1976) 837-843. 13. Bokkenheuser V. D., Winter J., Dehazya P. and Kelly, W. G.: Isolation and characterization of human fecal bacteria capable of 21-dehydroxylating corticoids. Appl. Environ. Microbial. 34 (197’7)571-575. 14. Bokkenheuser V. D., Winter J., Hyfemon P. B, Ayenear N. K. N. and Mosbach E. H.: Dehvdroxvlation of &x-hydroxyprogesterone by fecal flora of man and rat. J. Lipid Res. 22 (1981) 95-102. 15. Lowery 0. H., Rosebrough N. J., Farr A. L. and Randall R. 3.: Protein me~ure~nt with the folin phenol reagent. J. bioi. Chem. 193 (1951) 265-275. 16. Bokkenheuser V. D., Suzuki J. B., Polovsky S. B, Winter J. and Kelly W. G.: Metabolism of deoxy-corticosteronc by human fecal flora. Appf. Microhioi. 30 (197.5)82-90. 17. Glass T. L., Wheeler L. A., Sutter V. L. and Finegold S. M.: Transformation of 4-androsten-3,l%dione by growing cultures and cell extracts of ~10s~~~ parapurr&urn. Biochim. Biophys Acta 573 (1979) 332-342. 18. Bokkenheuser V. D., Winter J., Finegold S. M., Sutter V. L.. Ritchie A. E.. Moore W. E. C. and Holdeman L. V.: New markers for Eubacterium lenrum. Appf. Environ. M~crobjof. 37 (1979) 1001-1006. 19. Bjorkhem J., Eriksson H. and Gustaffson J. A.: Microbial formation of 17a-Cz, steroids. Sterochemistry of saturation of the Atb-double bond. Eur. J. Biochem 20 (1971) 340-343.
20. Feighner S. D. and Hylemon P. B.: Chara~er~tion
of corticosteroid 2l-dehydroxylase from the intestinal anaerobic bacterium, E~acterium fen&m. J. Lipid Res. 21 (1980) 583-593. 21. Holdeman L. V., Cato E. P. and Moore W. E. C. Eds.: Anaerobe Laboratory Manuaf, 4th edn,. Virginia Polytechnic Institute, Anaerobe Laboratory, Blacksburg, Virginia pp. 46-55. 22. Macdonald I. A., Jellett J. F, Nahony D. E. and Holdeman L. V.: Bile salt 3a and 12a-hydroxysteroid dehydrogenases from E~ucrerium fenrum and related organisms. Appl. Environ. Microbiof 37 (1979) 992-looo.
16a-DEHYDRATION OF CORTICOIDS BY BACTERIA ISOLATED FROM RAT FECAL FLORA JEAN-
WINTER*~,SHERYLO'ROURKE~, VICTOR D. BOKKENHEUSER?, FQILIPB.HYLEMON$ ~~~THOMAS L.GLASS$
tDepartment of Pathology, St. Luke’s_Roosevelt Hospital Center-New York, NY 10025 and IFrom the Department of Microbiology, Medical College of Virginia, Richmond VA 23298, U.S.A. (Receioed 22 April 1981) SUMMARY Two
strains (No. 144 and No. 146) of rat intestinal anaerobic bacteria, phenotypically similar to Euboccorticoids. The initial step in the 16a-dehydration of 16a-hydroxyprogesterone was dehydration at the C-16 and C-17 position with the accumulation of 16-dehydroprogesterone. This step required the side chain at C-17. In bacterial cultures the 16-dehydroprogesterone was then slowly reduced to &-progesterone. 16a-Hydroxypregnanolone was also converted to iso-pregnanolone by these bacteria. 16u-Dehydratase was easily demonstrated in cell fractions of strain No. 144 incubated either aerobically or anaerobically. The same extracts did not convert 16-dehydroprogesterone to iso-progesterone under similar assay conditions. 16a-Dehydration occurred at all substrate concentrations tested (20 to 200 pgjml) provided the pH of the growth medium was between 6 and 8 and the Eh below - 130mV. Strain No. 146 had both 16a-dehydration and 21-dehydroxylation activities. The two enzymes functioned independently. A role for intestinal bacteria in the biotransformation of biliary 16a-hydroxylated steroids and subsequent excretion in the urine is proposed. terium lentum, were isolated and found capable of 16a-dehydrating
INTRODUmION
In 1962 Calvin and Liebertnanrl] demonstrated that 16a-hydroxyprogesterone is excreted in the urine as
iso-pregnanolone. In both humans and rats, 16a-hydroxyprogesterone is the precursor of biliary 16ahydroxypregnanolone [2-81. This steroid is exposed to the intestinal flora during the enterohepatic circulation. Eriksson and Gustafsson[9] demonstrated that gnotobiotic rats excreted 16a-hydroxypregnanolone unchanged while normal rats l&-dehydrated the compound and converted the side chain from B to a. Recently we isolated two bacteria from rat feces, designated strains No. 144 and No. 146, which carried * Please address reprint requests to Dr J. Winter. This investigation was supported by grant 25763, awarded by the National Cancer Institute, DHEW; by grant 25324, awarded by the National Institute of Arthritis, Metabolism and Digestive Diseases, DHEW; and by a grant from the Fannie E. Rippel Foundation. Abbreviations used: 16a-hydroxyprogesterone 16a-hydroxy+pregnene-3,20-dione; 16-dehydroprogesterone4,16pregnandiene-3,20_dione; iso-progesterone l’la-pregn4 ene-3,2O-dione;progesterone4-pregnene-3,2~dione; 16a-hydroxypregnanolone 3~, 16a-dihydroxy-S/J-pregnan-20-one; iso-pregnanolone 3a-hydroxy-58, 17a-pregnan-M-one; deoxycorticosterone (DOC) 21-hydroxy4pregnene-3,20dione; pregnanolone 3a-hydroxy-5p-pregnan-3.2O-dione; tetrahydrodeoxycorticosterone 3a,21-dihydroxy-5/?-pregnan-20-one; 16a-hydroxy-estrone 3,16a-dihydroxy-1,3,5 (IO)-estratrien-17-one; 16a-hydroxy-androsterone 3a,16adihydroxy-5a-androstan-17-one. LB. lb,‘2--ca
out the l&-dehydration of corticoids. These bacteria are obligate anaerobes and were identified as organisms phenotypically similar to Eubacterium lenturn’ [lo]. In this paper $e describe the kinetics of 16adehydration by these bacteria. MATERIALS AND METHODS
Culture media Bacterial 16a-dehydration was examined in prereduced Brain Heart Infusion Broth (PR) supplemented with 0.05’4 cysteine-HCl (Scott Laboratories Inc.) or in BHIC medium with the following composition per liter: dehydrated brain heart infusion broth (Baltimore Biological Lab) 37 g; cysteine-HCl, 0.5 g; NaHCO,, 1 g; 4 ml of 0.025% aqueous resazurin (J. T. Baker Chemical Co). The medium was distributed in 50ml amounts and sterilized at 121°C for 20 min. For certain experiments, media were biologically reduced [l l] by addition of 0.1 ml of a 24-h old culture of Escherichia coli (BHIC-EC). Microorganisms
(a) Culture No. 144 and No. 146 were isolated from the fecal flora of rats [lo].
(b) E. lenrum neotypc strain (V.P.I. 0225) and other strains of E. lenturn were kindly supplied by Drs L. V. Holdeman and W. E. C. Moore, Anaerobe Laboratory, Virginia Polytechnic Institute (V.P.I.), Blacksburg, Virginia.
231
232
JEANETTE WINTERet al.
of 5 fig/ml and l”/;, CH30H (v/v). Incubation was continued another 5 h. The culture was then harvested by TLC* HPLQ GLC$ centrifugation at 8000 g for 15 min at WC under Steroid tR tR R, argon. The cell pellets were combined and suspended in l.Oml of boiled and cooled (under argon) 50mM 16u-Hydroxyprogesterone 0.06 1.70 1.83 16-Dehydroprogesterone 0.59 2.50 1.06 potassium phosphate buffer, pH 6.8 containing 1 mM Iso-progesterone 0.53 1.90 0.85 dithiothreitol (anaerobic buffer). The cell suspension Progesterone 1.15 0.59 2.50 was centrifuged for 15 min at 12,000 g at -4°C. The Deoxycorticosterone 0.37 2.00 0.90 cell pellet was stored overnight at -4°C under argon. The frozen pellet was thawed in 5 ml of anaerobic * Solvent system, chloroform-acetone (95 : 5, v/v) t Solvent system, acetonitrile-water (70: 30, v/v) buffer under argon. The cells were broken by two $Oxime derivatives on OVlOl column, 2R at 230°C. passages through a chilled French Pressure Cell at Relative rR/5atholestane 1034-1266 kg cm-‘. The broken cell suspension was then centrifuged at 12,000g for 20min at 04°C. The supematant fluid was removed and centrifuged at (c) Clostridium paraputr@um and E. coli were stock 105,ooOg for 2 h at (14°C to yield the soluble superstrains recovered from human fecal flora [12,13]. The natant fluid and membrane pellet. The supernatant fluid was removed and stored under argon at 0°C. cultures were maintained in a lyophilized form and, prior to use, passed two or three times in PR medium The membrane pellet was washed twice with 4ml of at 37°C. anaerobic buffer and stored at 0-4”C under argon. The assay conditions for 16z-hydroxyprogesterone Steroids and solvents dehydratase and 16dehydroprogesterone reductase Unlabeled steroids were purchased from Steraloids are described in the legend to Table 2. Briefly, assays were carried out at 37°C for 10 min (16adehydratase) Inc. and Makor Chemicals Ltd. Israel; isoprogesterone was synthesized in our laboratory as previously and 30 min (16dehydroprogesterone reductase). The described [lo]. Solvents were reagent grade except for reactions were terminated by extracting 0.15 ml aliquots twice with 2ml of ether. The ether phase was methanol which was technical grade; acetonitriie and blown off and the residue dissolved in 0.25 pll water were HPLC grade. CHsOH. Steroids were separated and quantitated by Conversion experiments. Prior to incubation, HPLC using a Beckman model 332 HPLC equipped steroids were dissolved in 0.25 ml methanol and and a added to the medium to give a concentration of 20 ,ug with a Hitachi u.v.-VIS Spectrophotometer steroids/ml, if not otherwise stated. The media were Shimadzu C-RlA integrator. The column used was a Beckman 4.6mm x 15cm C-18 ultrasphere-ODS then seeded with 0.25ml of a 48-h old culture and reverse phase and the system was operated isocraticincubated at 37°C for 7 days. At the end of the incubation, pH and Eh were measured (Beckman Zeroally using CHsOH/H1O (78:22 v/v) at a flow rate of matic II). A platinum electrode was used for the Eh 1.25 ml/min. The steroids were detected by monitormeasurement. Steroid extraction and identification ing at 240 nm and quantitated by external standardizwere carried out as previously described [lo, 143. ation. Briefly, the steroids were extracted with organic Protein was determined by the method of Lowry et a/.[151 after precipitation with 10% TCA, dissolving solvents, concentrated and purified by thin-layer the precipitate in 0.1 N NaOH and heating at 100°C chromatography (t.1.c.). a&Unsaturated carbonyl groups were detected from their absorbance at for 15min. Bovine serum albumin was used as the protein standard. 254 nm. The metabolites were identified by their relative R, values on t.l.c, their tR in the high performance liquid chromatograph (HPLC) and by gas-liquid chromatography. Ultimately the metabolites were RESULTS characterized by their mass spectrum as described J6a-Dehydration of !6a-hydroxyprogesterone previously [ 143. Chromatographic data of reference The results described below in experiments with steroids are given in Table 1. strain No. 144 apply equally well to strain No. 146 Preparation of cell extracts and assay for f6a-dehydraunless otherwise stated. tase and ltiehydroprogesterone reductase activities PR medium with initial pH ranging from 5.0 to 9.0 with 16a-hydroxyprogesterone Eubacterium sp. No. 144 was grown in 3 liters of was supplemented BHIC medium. The medium, after autoclaving was (2O&ml), seeded with strain No. 144, and incubated occurred rapidly when cooled on ice to room temperature while gassing with for 7 days. &-Dehydration the initial pH of the conversion medium was between C02. Prior to inoculation with 10 ml of strain No. 144, the gas phase was shiRed to 100% argon. The 6.0 and 8.0 (Fig. 1). Small amounts of acid developed culture was incubated for 17 h at 37°C. At this time during incubation, but the pH decrease was ~0.2 units. Growth was reduced at pH 5.0 and completely the substrate, 16a-hydroxyprogesterone or S-dehyinhibited ~~...._... _ at DH , 9.0. droprogesterone, was added to a final concentration Table 1. Chromatographic
data of reference steroids
16uDehydration Table 2. 16a-Hydroxyprogesterone
dehydratase and 16dehydroprogesterone Eubucrerium sp. No. 144
105,000 g Preparation
233
of corticoids
reductase activities in cell fractions of
16x-Hydroxyprogesterone dchydrata.se* (nmol 16_dehydroprogesterone/min/mg protein)
16Dehydroprogesterone reductaset (nmol i-progesterone/min/mg protein)
36 35
Supernatant fluid Membrane enriched fraction
* The assay mixture contained 50 mM potassium phosphate buffer (pH 6.8); 75 pg 16a-hydroxyprogesterone; CH,OH,
25% (v/v) and 0.5 mg protein (supernatant) or 0.5 mg protein (membrane) in a final volume of 0.25 ml.
t The assay mixture for 16dehydroprogesterone reductase was similar to “*“, except that 25 c(g lbdehydroprogesterone were added instead of 16x-hydroxyprogesterone and the reactions were carried out under an argon gas atmosphere. The following cofactors or combinations were individually tested as electron donors: NADH, NADH + FAD. NADH + FMN, NADPH + FAD, and NADPH + FMN. The reduced pyridine nucleotides and flavin nucleotides were added to a final concentration of 1 mM and 0.05 mM. respectively. In no case was iso-progesterone detected under these conditions nor was this product formed when the supernatant and pellet were mixed and assayed as above. In each case, a small fraction of the 16dehydroprogesterone added was hydrated to 16a-hydroxyprogesterone.
Eflect of co-culturing strain No.144 anaerobic bacteria
with facultatively
Aerobic BHIC medium does not support growth of obligate anaerobes. However, simultaneous seeding of the medium with strain No. 144 and either E. co/i, Proteus mirabilis or a Klebsiella species, permitted
of strain No. 144 and subsequent Madehydration. None of the facultative anaerobes transformed 16a-hydroxyprogesterone. In other experiments, BHIC medium containing 16a-hydroxyprogesterone (20 c(BIml) was inoculated with 0.1 ml broth culture of E. coli and incubated at 37°C. Regardless of whether strain No. 144 was added early or after 1 week of incubation with E. coli, 16ahydroxyprogesterone was quantitatively converted to iso-progesterone within 7 days after inoculation. Alternatively, BHIC medium inoculated simultaneously with E. coli and strain No. 144, was incubated at 37°C. Whether the substrate was added immediately or after 1 week incubation of the mixed culture, complete conversion to iso-progesterone was observed. Identical results were obtained when PR medium was inoculated with strain No. 144 and the substrate was added at different times. growth
Effect of substrate concernration on 16adehydration
16aDehydration by dilute fecal flora depends in part on the concentration of substrate [14]. To deter-
mine if this were true for pure cultures, strain No. 144 was inoculated into PR and BHIC-EC media at concentrations of substrate ranging from 20 @ml to 200 &ml. 16a-Hydroxyprogesterone was quantitatively converted to iso-progesterone at all levels of substrate tested. Because of the limited solubility of 16clhydroxyprogesterone in methanol and the desire to keep the concentration of the bactericidal solvent in the growth medium below lo/ it was not possible to achieve higher substrate concentrations. Kinetics of 16adehydration
Eight vials of PR medium supplemented with Ma-hydroxyprogesterone (20 pg/ml) were seeded with strain No. 144 and incubated at 37°C. Culture samples were taken at different times and the contents of each vial was submitted to steroid analysis. The results in Fig 2 shows that, within 12 h, the substrate was quantitatively converted to Mdehydroprogesterone. The second step of the reaction, the saturation of the double bond required about 36 h for complete conversion. Identical results were obtained with BHIC-Ec medium. Substrate specificity of 16a-dehydratase
To determine the substrate specificity of 16c+dehydration, deoxycorticosterone (DOC), 16a-hydroxyA-b
ISd-hydroxy
0-0
A’!pro~rrtrron. Ir0-pros.*t.ronr
x---i
~rog.rc.rons
0 5
6
7
6
9
PH
Fig 1.
lb-Dehydration
by strain No. 144 as a function of PH.
Time,
hrs.
Fig. 2. Kinetics of l&x-dehydration by strain No. 144.
JEANEITEWINTERet al.
234
too/
,/__/--*
/
f
/”
:! so-
9
0
0
,P/lP
/
z
0
ti
x--x
BCIlC.EC
o---D
pR
I’
I
I2
-
24
36
40
72
60
Tim.. hrr
0
12
24
36
42
Fig 3. Kinetics of i~hydros~~d No. 144.
Fig 4. 2LDehydroxylation capacity in two media at various Eh. Substrate: DOC; end product: progesterone.
Ttme. hrr
reductase by strain
have found that dialysis of the supernatant fluid had no effect on the 1fSadehydratase activity indicating no estrone or Ma-hydroxy-androsterone fall at 20 Irg/ ml) requirement for low molecular weight dialyxable were incubated with strain No. 144 in PR medium. cofactors. Furthermore, anaerobic precautions and None of these substrates were detectably dehydrated assay conditions are not necessary for measuring 16aby this bacterium The same negative results were dehydratase activity. Thus, the enzyme appears to be obtained in BHIC-ECmedium. However, strain No. 146 relatively oxygen stable. In contrast, we were unable converted DOC to progesterone and tetrahydro- to demonstrate lddehydrosteroid reductase activity deoxycorticosterone to pregnanolont indicating the in these extracts using the assay conditions described. presence of 21-dehydroxylation activity in this strain. l&z-Dehy&ationand 2l~ehy~roxy~~ion ofsteroids by To determine if related steroids inhibit the strain No. 146 l~d~ydration reaction, vials of PR and BHIC-EC Kinetic studies of 2ldehy~oxy~tion of DOC by media were supp~t~ with 1~“h~ox~ro~terstrain No. 146 were carried out in PR and BHIC-EC one and inoculated with strain No. 144.l&Hydroxymedium. Figure 4 shows that the 21dehydroxyiation estrone (100 pg/ml) was added to one set of vials, 16a-hydroxy-androsterone (100 pgjml) to another set, was much faster in BHIC-EC than in PR medium. In additional studies, varying concentrations (3 JAM and DOC (100&ml) to a third set. Inhibition of to 55 PM) DOC were added to conversion media. The 16adehydration was not detected but traces of unresults presented in Fig. 5 shows the E. coli enhanced identified steroid metabolites of lower polarity were present in the vials containing C-18 and C-19 steroids. the capacity of strain No. 146 to Zldehydroxylate DOC. The maximum amounts of DOC converted in 50ml of PR medium and BHIC-EC medium were 24 Comersion of f6-dehydr5pr5gestefone to i.w-progesterand 39 rmol, respectively. one The kinetic experiments &mo~~at~ that the l~dehy~ation is a two step reaction, suggesting that the conversion involves at least two enzymes; one that converts the substrate to a M-dehydrosteroid, and another that reducea the double bond, leaving the side chain in the a-position @o-progesterone).
To confirm this hypothesis, M-dehydroprogesterone was added to media and incubated for 48 h. In PR medium, reduction of the 16dehydroprogesterone was slow for the first 12 h and then increased in velocity until all substrate was converted (Fig. 3). It was completed within 48 h. Interestingly, the rate of the reaction was not enhanced by the presence of E. co/i in the medium. 16a-Dehydratase
activity in cell fiactioas
The data in Table 2 shows that 16a-hydroxyprogesterone dehydratase activity was easily demonstrated in both the soluble and membrane fractions of E&Xterium sp. No. 144. The membrane enriched fraction was washed twice before assaying; hence, the membrane associated activity is not 1ikeIy to be due to gross contamination by soluble protein. Moreover, we
Si~~t~~s ~t5~~~rn of I ~-hy~oxypr~est~one and DOC by strain No. 146 Because strain No. 146 has both l&dehydration and 2ldehydroxylation activities, competition between these enzymes was investigated in growing cultures. PR and BHIC-EC media were supplemented with Ma-hydroxyprogesterone (200&ml) and DOC (20 @ml, 160&ml, or 260 pg/mlA respectively. l&x-
100kq
-x\ 3
t e” 0
x
\
._s u)
1 I l
so-
3
0
\
I I I I
-
I 25
x--x o--
BHIC-EC QR
X \
I 3s pM
t\. 45
55
DOC
Fig. 5. 21-Dehydroxylation capacity at various concentrations of substrates. Substrate: deoxycorticosterone; end product: progesterone.
lb-Dehydration
235
of corticoids
Stroin +I46
cn I
@
I 3 :=o
, m
Fig. 6. Bacterially mediated metabolic pathway of 16x-hydroxyprogesterone in vitro. I = 16x-hydroxyprogesterone; II = 16dehydroprogesterone; III = iso-progesterone; IV = Ha-hydroxypregnanolone; V = lddehydropregnanolone; VI = iso-pregnanolone; + demonstrated pathway ---a hypothetical pathway.
Hydroxyprogesterone w’as quantitatively converted to iso-progesterone in all cultures regardless of the concentration of DOC. Moreover, 21-dehydroxylation of DOC was complete in the presence of low concentrations of DOC but did not go to completion in the presence of high concentrations.
their similarity to E. lentum, we examined 10 strains of the latter species for Wx-dehydratase activity using 16a-hydroxyprogesterone as substrate [18]. However, no 16adehydratase activity was detected in these bacteria.
Metabolism of 16a-hydroxyprogesterone in mixed cultures of C. paraputrificum and strain No. 144
DISCUSSION
Experiments with mixed fecal flora suggest that the conversion of 16a-hydroxyprogesterone to iso-pregnanolone requires the combined enzyme activities of bacteria having the ability to reduce the 4-ene-fketo structures [16,17] and organisms having 16c+dehydration activity. This was confirmed in a series of experiments in which 16a-hydroxyprogesterone or 16 dehydroprogesterone were quantitatively converted to iso-pregnanolone by the combined action of C. paraputrificum and strain No. 144. Moreover, DOC was also quantitatively reduced to pregnanolone by C. paraputrijicum and strain No. 146. Distribution of bacteria having 16adehydration actioity
Isolates No. 144 and No. 146 are the first organisms known to synthesize 16x-dehydratasc. Because of
Mechanism of l&x-dehydration
16aDehydration was carried out rapidly and efficiently by strains No. 144 and No. 146 in both PR and BHIC-Ec media at all substrate concentrations tested (2O-2OOIrglml), provided the Eh was below - 130 mV and the pH between 6 and 8. Kinetic experiments confirmed Calvin and Lieberman’s hypothesis [l] that &dehydration of 16a-hydroxyprogesterone proceeded via an intermediary, 16dehydroprogcsterone. This compound was formed rapidly from the substrate by dehydration at C-16, C-17 positions. 16a-Hydroxy-estrone and 16a-hydroxy-androsteone, proved resistant to 16adehydration. Moreover, l&dehydration did not occur with C-18 or C-19 compounds with an OH group at C-17 (unpublished observations). This suggests that the 16adehydratasc requires the steroid side chain. Substrate specificity
JEANEITEWINTER et al.
236 pOti
$“3 c=o
p0
0
&
/ A&
-OH
0
0
16u- hydroxy- proQostuonr
Dqaaycorticoakronr
\
yz-O”
uvr
y3 c=o
c=o
-OR
Ro’PX
RO’
H THOOC
&
”
16d-hydroay-prrqnonolom
M
R*gnonolon*
ho- prognonolono
Fig. 7. Proposed metabolic pathways in viva.
and mechanism of reaction must await studies with purified enxymes. Because strains No. 144 and No. 146 dehydrated Ma-hydroxypregnanolone and Ma-hydroxyprogesteronc equally well in growing cultures, it follows that enzyme specifrcity is independent of the Gene-3-keto structure in ring A. However, studies with purified enxyme will he neassary to determine which might be the prekrmd steroid substrate. The second step of the biotnu&rmatios the reduction of the Sdehydrosteroid has been reported to involve the trans addition of the H to the Ma and 178 position [193. Because the sequential removal of the 16a-hydroxy group and the subsequent reduction of the Mdehydrosteroid could be separated in time and were of different velocity, we believe that more than one enzyme is required for the l&-dehydration and subsequent conversion to iso-progesterone.
16sDehydration strain No. 146
and 2Idehydroxylation
activities in
In previous studies we showed that the presence of E. coli in mixed cultures enhanced 21dehydroxyL ation activity of strain No. 116 apparently by reducing the Eh to - 230 mV [ 1l] or by supplying reducing equivalents in the form of molecular hydrogen [20]. The 2ldehydroxylation activity of strain No. 146 behaved in a similar manner. In contrast, l&x-dehydration activity in strain No. 146 functioned efhciently up to -13OmV and was not inhibited by 21-hydroxylated steroids. Steroid biotran@uwtarion patterns in vitro and in vivo In vitro, DOC is converted to pregnanolone by the joint action of C. puruputrificum and strain No. 146 by the mechanism previously proposed for this biotrans-
16xDehydration formation
112). Briefly, because of the rapid growth of
and the high velocity of 4-ene-3keto reduction, THDOC is observed as an intermediary metabolite. However, the reaction may begin with a tl-dehydroxylation of DOC to progesterone, which is ring-A reduced so rapidly that it never is observed in a culture containing C. p~~upuf~~~c~~ [12]. In uitro formation of iso-pregnanolone from 16a-hydroxyprogesterone is a multistep biotransformation performed by two anaerobic bacteria each containing separate enzyme activities (Fig. 6). C. parapurrificum reduces the 4-ene-3-keto structure and strain No. 144 removes the C-16OH group, reduces the 16dehydrosteroid and converts the side chain. fn GUO 16ff-hydroxyprogesterone and 21-hydroxycorticoids are excreted in the bile as ring-A reduced products conjugated with glucuronic acid [II J. The biliary steroids are deconjugated in the gut, 21-dehydroxylated by E. fe~r~m or phenotypi~iiy similar organisms [18] and 16a-dehydrated by bacteria similar to strains No. 144 and No. 146 (Fig. 7). The dehydrated products are absorbed, further reduced in the liver, conjugated with glucuronic acid and excreted in the urine. It is noteworthy that the position of the side chain of urinary metabolites reveals which steroids have been 16x-dehydrated. C. paraputrijcum
Taxonomic position of strains No. 144 and No. 146 Strains No. 144 and No. 146 are the first bacteria isolated that have 16a-dehydrative activity. These organisms appear to be phenotypi~liy similar to Eubacterium Ientum. It is interesting that bacteria observed this far capable of dehydrating ring-D and its side chain all are gram-positive, obligate anaerobes, non-spore forming rods that do not ferment carbohydrates [2l]. Evidence is accumulating that members of this group of bacteria are rich in enzymes involved in the biotransformation of steroids [ 18,223.
REFERENCES 1. Calvin H. L. and Lieberman S.: Studies on the metabolism of 16a-hydroxyprogesterone in humans. Conversion to urinary 1%isopregnanolone. Bjoc~ern~rry 1 (1962) 639-645. 2. Laatikainen f.: Identification of ClpOt and C2r02 steroids in the glucuronide fraction of human bile. Eur. J. Biochem. 10 11969) 165-171. Laatikainen T.t rd~ntification of C19C2 and C2i02 steroids in the mono- and disuiphate fractions of human feces. Steroids IS (1970) 139-150. Laatikainen T. and Vihko R.: Identification of C,.O, and CztOz steroids in the glucuronide fraction 01 human bile. Steroids 13 (1970) 534-538. Taylor W.: The excretion of steroid hormone metabolites in bite and feces. Warn. Horm. 29 (1970) 201-285.
237
of corticoids
Ashmore J.. Elliott W. H., Doisy E. A. Jr and Doisy E. A.: Excretion of metabolites of testosterone 4X14 in rats. J. &of. Chem. 200 (1953) 661668. Bocklage 8. C.. Daisy E. A. Jr, Elliott W. H. and Doisy E. A.: Absorption and metabolism of cortisone~C14. J. biol. Chek
212 (1955) 935-939.
Shen H.. Elliott W. H.. Doisv E. A. Jr and Doisv E. A.: Excretidn of rne~boi~t~ o~pro~esterone-21-~~4 after intragastric administration to rats. J. biol. Clrem. 20s (1954) 133-133. 9. Eriksson H., Gustafsson J.-A. and Sjovall J.: Steroids in germ-free and conventional rats. 4. Identification and bacterial formation of I Xx-pregnane derivatives. Eur. J. Biochem. 6 (1968) 219-226. V. D., Winter J., O’Rourke S. and Rit10. ~kkenheu~r chic A. E.: Isolation and characterization of fecal baa teria capable of I&-dehydroxylating corticoids. Appi. Environ. ~ierobiof. 40 (1980) 803-808. 11. Winter J. and Bokkenheuser V. D.: 21Dehydroxylation of corticoids by anaerobic bacteria isolated from human fecal flora. .I. steroid Biochem. 9 (1978) 379-384. 1.2. Bokkenheuser V. D., Winter J., Dehazya P. De Leon 0. and Kelly W. G.: Formation and metabolism of tetrahydrodeoxycorticosterone by human fecal flora. 1. steroid &o&em. 7 (1976) 837-843. 13. Bokkenheuser V. D., Winter J., Dehazya P. and Kelly, W. G.: Isolation and characterization of human fecal bacteria capable of 21-dehydroxylating corticoids. Appl. Environ. Microbial. 34 (197’7)571-575. 14. Bokkenheuser V. D., Winter J., Hyfemon P. B, Ayenear N. K. N. and Mosbach E. H.: Dehvdroxvlation of &x-hydroxyprogesterone by fecal flora of man and rat. J. Lipid Res. 22 (1981) 95-102. 15. Lowery 0. H., Rosebrough N. J., Farr A. L. and Randall R. 3.: Protein me~ure~nt with the folin phenol reagent. J. bioi. Chem. 193 (1951) 265-275. 16. Bokkenheuser V. D., Suzuki J. B., Polovsky S. B, Winter J. and Kelly W. G.: Metabolism of deoxy-corticosteronc by human fecal flora. Appf. Microhioi. 30 (197.5)82-90. 17. Glass T. L., Wheeler L. A., Sutter V. L. and Finegold S. M.: Transformation of 4-androsten-3,l%dione by growing cultures and cell extracts of ~10s~~~ parapurr&urn. Biochim. Biophys Acta 573 (1979) 332-342. 18. Bokkenheuser V. D., Winter J., Finegold S. M., Sutter V. L.. Ritchie A. E.. Moore W. E. C. and Holdeman L. V.: New markers for Eubacterium lenrum. Appf. Environ. M~crobjof. 37 (1979) 1001-1006. 19. Bjorkhem J., Eriksson H. and Gustaffson J. A.: Microbial formation of 17a-Cz, steroids. Sterochemistry of saturation of the Atb-double bond. Eur. J. Biochem 20 (1971) 340-343.
20. Feighner S. D. and Hylemon P. B.: Chara~er~tion
of corticosteroid 2l-dehydroxylase from the intestinal anaerobic bacterium, E~acterium fen&m. J. Lipid Res. 21 (1980) 583-593. 21. Holdeman L. V., Cato E. P. and Moore W. E. C. Eds.: Anaerobe Laboratory Manuaf, 4th edn,. Virginia Polytechnic Institute, Anaerobe Laboratory, Blacksburg, Virginia pp. 46-55. 22. Macdonald I. A., Jellett J. F, Nahony D. E. and Holdeman L. V.: Bile salt 3a and 12a-hydroxysteroid dehydrogenases from E~ucrerium fenrum and related organisms. Appl. Environ. Microbiof 37 (1979) 992-looo.