Zygophore-stimulating Precursors (Pheromones) of Trisporic Acids Active in (-)-Phycomyces blakesleeanus ACID-CATALYZED ANHYDRO DERIVATIVES OF METHYL 4-DIHYDROTRISPORATE-C AND 4-DIHYDROTRISPORATE-C*

The same pheromones apparently initiate sexual de- velopment in all mucoraceous fungi. We isolated methyl trisporates and methyl 4-dihydrotrisporates as pheromones from (+) cultures of Blakeslea trispom and used Sephadex LH-20 chromatography with ethyl ace- tate for the purification. The pheromones stimulated the development of zygophores (sex cells) in bioassays with (-) cultures of Phycomyces blakesleeanus and Mucor rnucedo. Labeled methyl 4-dihydrotrisporate-C and methyl trisporate-C, prepared chemically from labeled trisporic acid-C, were incorporated into trisporic acids in (-), but not (+), cultures of P. blakesleeanus. Two other labeled compounds, Cpd85 and Cpd76, also were isolated from the incubation with methyl 4-dihydrotri- sporate-C. Cpd85 and Cpd76 apparently were not precursors of trisporic acids, zygophore-stimulating pheromones, or zygotropic pheromones, but were proce- dural artifacts. Methyl 4-dihydrotrisporate-C dehydrated to Cpd85 and 4-dihydrotrisporate-C dehydrated to Cpd76 in slightly acidic solutions. Cpd85 was formed when Cpd76 was treated pheromone Pheromones-The relative zygophore-stimulating activities of the pheromones methyl 4-dihydroTA-B, methyl 4-dihydroTA-C, methyl TA-B, and methyl TA-C and their metabolites TA-B and TA-C upon cultures

crosses with wild type cultures. These observations demonstrate that each mating type of P. blakesleeanus makes a pheromone, a metabolite of /?-carotene, which initiates zygophore formation only in cultures of the opposite mating type.
Apparently these pheromones are the same in all species of mucoraceous fungi because many make zygophores in interspecies crosses (5,6). Furthermore, extracts isolated from the medium of a single mating type culture of BZakeslea trispora stimulate zygophore formation only in the opposite mating type when tested in bioassays with Mucor mucedo (7,8). Bu'Lock et al. (9,10) isolated and identified the sex pheromones from B. trispora; Nieuwenhuis and van den Ende (1 1) and van den Ende (12) found the same pheromones in M . mucedo. They found 2 pheromones made by (+) cultures and 4 pheromones made by (-) cultures. Both groups hypothesized that the pheromones are mating type-specific precursors of TAs' and that each mating type contributes a unique step to TA biosynthesis resulting in the formation of the sexspecific pheromones. For example, the (+) mating type supposedly cannot oxidize the hydroxyl group on carbon atom 4 of TA precursors, resulting in the formation of methyl 4-dihydroTA-B and methyl 4-dihydroTA-C.

RESULTS
Zygophore-stimulating Pheromones in (+) Culture Medium-The existence of pheromones made by both (+) and (-) cultures of P. blakesleeanus has been demonstrated in ' The abbreviations used are: TAs, trisporates (trisporic acids); TA, trisporate; TA-B, trisporate-B; T A X , trisporate-C; VJV,, the elution volume of the fraction with the maximum sample/total bed volume of the column.
Portions of this paper (including "Materials and Methods," "Results," and Tables 1V and VI) are presented in miniprint a t the end of the paper. Miniprint  crosses of mutants and wild types (4). However, we were unable to detect pheromones at the 0.1 n g / d level in the medium of 5-day-old (+) cultures of P. blakesleeanus, suggesting it does not make pheromones while undergoing hyphal elongation (nuclear division) or while committed to an alternate developmental pathway such as sporangiophore formation.
Extracts isolated from (+) B. trispora grown in undefined medium stimulated both zygophore formation and the accumulation of TAs in (-), but not ,(+), cultures of P . blakesleeanus. Zygophore formation was stimulated by 0.6 A 2 8 5 unit of extract in (-) cultures, whereas 80 AzH:, units had no effect on (+) cultures. Twenty A32R units of purified acid fraction (TAs) were recovered from (-) cultures incubated overnight with 294 AZ8:, units of extract. Thus, extracts isolated from 7day-old cultures of (+) B. trispora have the same effect upon young cultures of (+) and (-) P . blakesleeanus as upon M. rnucedo ( 7 , 8 ) and B. trispora (7,13 Components A, B, and C were resolved by thin layer chromatography into 2,4, and 3 UV-absorbing spots, respectively. Only component spots A-1, B-2, and C-2 stimulated zygophore formation (in both bioassays). Some properties of these component spots are given in Table I. C-2 was identified as methyl 4-dihydroTA-C from the following observations. C-2 and authentic methyl 4-dihydroTA-C exhibited identical UV spectra, V,/V,, and zygophore-stimulating activities upon dilution in the Mucor bioassay. A mixture of C-2 and methyl 4-dihydroTA-C migrated as a single spot upon thin layer chromatography with solvent systems I, 11, and 111. B-2 was identified as 85% methyl 4-dihydroTA-B and 15% methyl TA-C. One AZH8 unit of B-2 migrated as a single discrete spot upon thin layer chromatography with solvent systems I, 11, and 111. B-2 migrated further than C-2 in all 3 solvent systems demonstrating that B-2 is less polar than C-2. The absorbance at 276, 283, 296 nm indicated the presence of a methyl 4-dihydroTA-like compound and the absorbance a t 328 nm indicated the presence of a methyl TA-like compound. Cochromatography of B-2 and methyl TA-C using a Sephadex LH-20 column with ethyl acetate resulted in an elution profile with a constant A2X5/A320 ratio. The sodium borohydride reduction product of B-2 and authentic methyl 4-dihydroTA-C had identical UV spectra and VJV,; a mixture of the 2 Sephadex LH-20 chromatography. Sephadex LH-20 chromatography with ethyl acetate of ( A ) neutral fractions isolated from the medium of (+) cultures of B. trispora grown in a potato extract-glucose medium (adjusted to pH 2 prior to extraction) (solid line) and in a defined medium (pH unadjusted prior to extraction) (dotted line) and a composite profile ( B ) from analyses of known compounds: 1, methyl trisporate-B; 2, methyl trisporate-C and methyl 4-dihydrotrisporate-B; 3, methyl 4-dihydrotrisporate-C; 4, trisporic acid-B; 5, trisporic acid-C. Absorbance measurements a t h.,,, were made after fractions were diluted 50.. 60-, 60-, and 100-fold for neutral fractions, methyl trisporates, methyl 4-dihydrotrisporate-C, and trisporic acids, respectively. A separate experiment revealed that methyl 4-dihydrotrisporate-B co-chromatographed with methyl trisporate-C. VJV,, in this figure only, represents the elution volume of a fraction divided by the total bed volume of the column.

Recovery ofpheromones from the medium of (+) B. trispora
Pheromones were isolated from neutral fraction extract by Sephadex LH-20 chromatography with ethanol, Sephadex LH-20 chromatography with ethyl acetate, and silica gel thin layer chromatography with the indicated solvent systems. The losses of AZH5 units in each type of chromatography were 10.25, and 258, respectively. A285 units recovered are the amounts of component spot recovered from IO00 AZX, units of extract isolated from cultures grown on undefined medium. Criteria for identification of compounds are described in the text. Micromoles were calculated from absorbance data; 3.2 was used for the A328/AZX.r of methyl trisporate-C. The abbreviations used are mTA-B, methyl trisporate-B; mTA-C, methyl trisporate-C; m4dTA-B, methyl 4-dihydrotrisporate-B; mBdTA-C, methyl 4-dihydrotrisporate-C. giving a distinct positive response equals 108 pmol), and the recovery of Cpd85 after treating the sodium borohydride reduction product of A-1 with dilute acid. We conclude that the same pheromones stimulate zygophore formation in (-) cultures of both M. mucedo and P . blakesleeanus.

Component
A partially purified extract isolated from cultures grown on defined medium was resolved into 7 components designated A, B, B*, C, D, E, and F by Sephadex LH-20 chromatography with ethyl acetate (Fig. l A , dotted line) Table 11. The activity of each compound is expressed as the lowest quantity of that compound required to elicit a distinctly positive response in that species relative to the compound which was active in the lowest absolute picomole amount. Two-fold differences in activity are within experimental error and not significant. In bioassays with both species, methyl 4-dihydroTAs were active only on (-) cultures, methyl TAs were about 100 times more active on (-) than (+) cultures, and TAs were equally active on both (+) and (-) cultures. The B form (oxygen at C-13) of each class of compounds was more active than the C form

2336
Pheromones and Artifacts in Phycomyces (hydroxyl at C-13). This difference was a t least 4-fold in P. blakesleeanus and 10-fold in M. mucedo. There was one major difference between the bioassays with P . blakesleeanus and M. mucedo. The pheromone methyl 4-dihydroTA-B was the most active compound in bioassays with P. blakesleeanus. It was 3 times more active than methyl TA-B and 90 times more active than TA-B. In contrast, TA-B was the most active compound in bioassays with M. mucedo, being 4 times more active than the pheromones methyl 4-dihydroTA-B and methyl TA-B. The labeled compounds isolated from (+) and (-) cultures of P. blakesleeanus incubated with the labeled pheromones methyl 4-dihydroTA-C and methyl TA-C are presented in Table 111. TA-C, formed only by (-) cultures, was the major metabolite of both pheromones. It represented 45% and 64% of the label originally added to cultures as methyl 4-dihydroTA-C and methyl TA-C, respectively. (In a separate experiment (data not shown), 360 pg of TA-B and 140 pg of TA-C were recovered from (-) cultures incubated with 1.1 mg of unlabeled methyl 4-dihydroTA-B). Two unexpected compounds, Cpd76 and Cpd85, were found in (-) cultures incubated with labeled methyl 4-dihydroTA-C. Cpd76 was present in both the acid and neutral fractions. Cpd85 was found only in the neutral fraction. Unlabeled compounds absorbing UV radiation were not detected among the labeled compounds isolated from the cultures. '' Compounds were isolated from cultures of B. trispora; compounds with an * were prepared by chemical modification of trisporic acids. Abbreviations are described in legend to Table I.   TABLE I11 Metabolites of the pheromones After cultures were incubated for 14 h with 42,600 dpm of methyl 4-dihydrotrisporate-C (m4dTA-C) and 19,100 dpm of methyl trisporak-C (mTA-C) (see Table IV for specific activities), the acid and neutral fractions were isolated and the compounds were resolved as described under "Materials and Methods." Compounds were identified by their RF, A, , , , and specific activities. The percentages in brackets represent dpm for the compound indicated by the RF value but the A, , , and/or specific activity values differed from those expected for that compound.
ND, not detected UIM, uninoculated medium. Identification and Characterization of Cpd76 and Cpd85-Ninety per cent of the 4-dihydroTA-C in acidified 6% KH2P0, dehydrated to Cpd76 upon extraction with CHCL and evaporation to dryness (Fig. 2). Over 21 pmol of Cpd85 were formed when 24 pmol of methyl 4-dihydroTA-C in 5 ml of CHC13 were exposed to 2.5 mg ofp-toluene sulfonic acid for 30 min a t 22°C. (Less than 0.1 pmol of Cpd85, if any, was formed when 24 pmol of methyl TA-C were treated identically.) This Cpd85 was identical (as judged by UV spectral data and co-chromatography with solvent systems 111, V, and VI) to the: 1) Cpd85 isolated in the tracer experiments, 2) product formed upon treating Cpd76 with diazomethane, and 3) major compound (90%) of component spot A-2 isolated from (+) culture medium acidified prior to extraction with UV spectra of Cpd85 and Cpd76 exhibited X, , , 276,286,298 nm, and molar absorptivities 1.7 times greater than TAs (Table IV). IR spectra of Cpd85 and Cpd76 exhibited absorbances at 1715 cm-I and 1703 cm", respectively, but none in the 3450 cm" region, indicating that neither compound has a hydroxyl group. The mass spectrum of Cpd85 is shown in Fig.   3. The parent peak indicates a M , = 304. The parent peak plus one is 20.3% as abundant as the parent peak, which is consistent with a molecular formula of CI9H&3 for Cpd85. A 60 MHz NMR spectrum of Cpd85 indicated one OCH, group sample with a 250 MHz NMR spectrometer indicated the presence of 4 diastereoisomers as judged from the 0CH:l peaks: 3.687 pprn (52.4%), 3.672 ppm (40.5%), 3.644 ppm (3.3%), and 3.636 pprn (3.8%). (Cpd85 had been prepared from an unresolved mixture of TA-C (-80%) and TA-B.) The NMR spectra for the 2 diastereoisomers of Cpd85 prepared from 13R-TA-C are given in Table V. We conclude (based upon chromatographic, synthetic, and spectral data) that Cpd85 is 1,3-dimethyl-2-[3-(tetrahydro-5-methyl-2-furanyl)-2-butenylidene]-3-cyclohexene-l-carboxylic acid methyl ester. We have assumed in Fig. 2 that Cpd85 has the same stereochemistry at carbon atoms 1 and 13 as natural TA-C (14,15).
Cpd85 and Cpd76 did not have any detectable effects upon (+) and (-) cultures. For example, a micromole of either compound did not stimulate the formation of zygophores in either P . blakesleeanus or M . mucedo. A micromole of compound did not attract any zygophores of M . mucedo when placed upon a 4-mm2 fiiter paper between (+) and (-) cultures which were 0.7 mm apart, even though zygophores of opposite mating types attracted each other. In tracer experiments, 80% of the label was recovered after (+) and (-) cultures and uninoculated medium were incubated with 14,400 dpm of Cpd76 for 14 h. All of the label recovered was in Cpd76. Only 30% of the label was recovered after (+) and (-) cultures were incubated with 44,400 dpm of Cpd85, whereas 60% was recovered from uninoculated medium. All of the label recovered  Table V for composition of the mixture) was obtained with a unit resolution mass spectrometer. The sample (an oil) was applied to a probe which was inserted into the spectrometer and heated to 50°C. The spectrum was recorded 2 min after inserting the probe. Scan time was 5 s (0-450 rn/e), the ion source was 250°C. and the ionization potential was 70 eV. The spectrum was corrected for background and the accuracy of the mass/ charge (rn/e) values was corroborated in a separate analysis (not shown) with sample containing perfluorokerosene.

TABLE V NMR spectra for diastereoisomers of Cpd8S
The NMR spectrum of a mixture of diastereoisomers in CDCL was recorded after 20 scans by a 250 MHz proton correlation NMH spectrometer. The chemical shifts are expressed in parts per million (6) relative to tetramethylsilane, an internal standard. The coupling constants (J) are expressed in Hertz. Major and minor diastereoisomers represent 558 and 45% of the sample derived from 13R-trisporic acid-C and 93% of the entire sample; the remaining 7% of the sample is distributed between 2 diastereoisomers derived from 13s-trisporic acid-C.The signals for the four pair of methylene hydrogens on carbons 2,3, 11, and 12 are between 1.40 and 2.36 ppm.

Pheromones and Artifacts in Phycomyces
from uninoculated medium was Cpd85. Over 98% of the label recovered from cultures was in the neutral fraction, primarily as Cpd85, but traces of 2 and 4 unidentified compounds were found in (-) and (+) cultures, respectively. The purified acid fraction contained the remaining label. In summary, Cpd85 and Cpd76 apparently are not zygophore-stimulating pheromones, zygotropic pheromones, or precursors of trisporic acids.

DISCUSSION
Zygophore-stimulating Pheromones in (+) Culture Medium-Four pheromones were isolated from the medium of (+) cultures of B. trispora: methyl TA-B, methyl TA-C, methyl 4-dihydroTA-B, and methyl 4-dihydroTA-C ( Table I). This is the fist report of the isolation of methyl TAs from (+) cultures. The amount of pheromones isolated from a given (+) culture is a function of both the composition of the medium used for growing the culture (3) and the extraction procedure. Over 90% of the TA-stimulating components (pheromones) was recovered when (+) culture medium was extracted directly, whereas only 44% was recovered when the culture medium was adjusted to pH 2 prior to extraction (16). If the appropriate corrections are made for losses of pheromone during the purification of the extract and if the difference in pheromone levels between neutral fraction extract and culture medium were due sol.ely to the destruction of methyl 4-dihydroTA-C by acid with the corresponding formation of Cpd85, then there are 0.6 pmol of methyl TA-B, 1.1 pmol of methyl TA-C, 6.4 pmol of methyl 4-dihydroTA-B, and 15.5 pmol of methyl 4-dihydroTA-C/liter of medium from (+) cultures grown on undefined medium. Trisporic acids were found previously in the medium of (+) cultures grown under these conditions (7). The ratio of methyl 4-dihydroTAs to methyl TAs to TAs in the medium was 100 to 10 to 1. These findings demonstrate that (+) cultures are able to oxidize the hydroxyl group at carbon atom 4 of trisporate precursors.
Sephadex LH-20 chromatography with ethyl acetate resolves trisporate-related compounds unaltered (except for Cpd76 and 4-dihydroTA-C) and in excellent yields ( Fig. 1 and Table VI). The 75% recovery reported in Table I was due to an 8% loss of sample upon injecting the column and to pooling only 81% of the material that eluted from the column. Separation apparently is based upon the number and kind of oxygen-containing functional groups in the compounds. The functional groups may interact through hydrogen bonding with the hydroxyl groups in the gel. The affinity of the functional groups for Sephadex differs from their affinity for silica gel (Table VI). This was a definite asset in resolving the pheromones from closely related metabolites. A closely related procedure was developed independently for the separation of vitamin A derivatives (17).
Zygophore-stimulating Pheromones and Sexual Development-The results obtained with the bioassays for zygophorestimulating pheromones demonstrate that (-) cultures of P. blakesleeanus develop zygophores in response to the same (+) pheromones, methyl 4-dihydroTAs and methyl TAs, as (-) cultures of M. mucedo (Table 11). Reschke (14) reported previously that methyl TAs, isolated from (+/-) cultures, acted in a sex-specific manner in Mucor bioassays. In contrast, Bu'Lock et al. (10,18) reported that only twice as much methyl TA was needed for zygophore formation in (+) as in (-) Mucor. We are unable to explain this discrepancy.
Previous studies in which unlabeled methyl 4-dihydroT'As and methyl TAs were incubated with intact cultures and crude mycelial extracts of M. mucedo suggested that these compounds are mating type-specific precursors of TAs (11,19). The tracer studies reported here demonstrate that (-) cultures of P. blakesleeanus are a t least 100-fold more effective than (+) cultures in converting the pheromones methyl 4-dihydroTA-C and methyl TA-C into TA-C (Table 111).
Bu'Lock et al. (10) had reported isolating 2 anhydro compounds, an acid and the corresponding ester, from (-) cultures of B. trispora incubated with unlabeled methyl 4-dihydroTA-C. The significance of the compounds was not determined although they reported that the anhydro ester was converted quantitatively to TAs upon incubation with (-) B. trispora.
They proposed a structural formula for the anhydro ester based upon its polarity cnd m/e 304. UV, IR, NMR, and mass spectral data of Cpd85 verify that it has the structure which Bu'Lock proposed for the anhydro ester (see Fig. 2). We were unable to find evidence that Cpd85 or Cpd76 possessed either zygophore-stimulating or zygotropic pheromone activity, or that they were metabolized significantly by separate mating type cultures of P. 62akesleeanus. The artifactual origin of Cpd85 and Cpd76 is supported by the finding that, in the absence of cultures, traces of acid catalyzed their formation from methyl 4-dihydroTA-C and 4-dihydroTA-C, respectively. The reactions appear irreversible. We propose that any metabolite of TAs with hydroxyl groups at carbon atoms 4 and 13, such as 4-dihydrotrisporin-C, will form the corresponding anhydro derivative when exposed to acid.
The identification of Cpd76 as an anhydro derivative of 4-dihydroTA-C indicates that the esterase in (-) cultures is not absolutely specific for methyl TAs but can catalyze the hydrolysis of methyl 4-dihydroTAs too. The presence of large amounts of Cpd76 in the medium of (-) cultures incubated with methyl 4-dihydroTA-C implies either that the cultures are ineffective in oxidizing the hydroxyl group on carbon atom 4 of 4-dihydroTA-C or that 4-dihydroTA-C dehydrates to Cpd76 in the presence of cultures. In either case, methyl TAs would be the predominant biosynthetic intermediates of TAs made from methyl 4-dihydroTAs.
We conclude that under physiological conditions both methyl 4-dihydroTAs and methyl TAs are mating type-specific zygophore-stimulating compounds in P. blakesleeanus because TAs, which are more active in bioassays with (+) cultures than these compounds, are inactive in crosses of (+) and (-) P. blakesleeanus (3,4). We do not know which compounds initiate and maintain (regulate) the formation of zygophores. Are they the pheromones themselves or their metabolites, intrahyphal TAs? The regulatory molecules could be inactivated either by being converted to TAs or by being released into the medium. The prevalent opinion that TAs are the regulatory compounds is based upon circumstantial evidence. First, TAs were the initial compounds to be identified as zygophore-stimulating agents in M . mucedo (20). This was 3 years before it was realized that TAs might be synthesized by way of mating type-specifc precursors (21) and 7 years before mutant studies verified the existence of mating type-specific pheromones (4). Second, TAs were about 4 times more active than methyl 4-dihydroTAs in stimulating zygophore-formation in (-) M . mucedo based upon a dilution bioassay (10). However, 1) the relative stability of the two compounds under bioassay conditions is unknown, and 2) other workers have reported that TAs and methyl 4-dihydroTAs have equal zygophore-stimulating activities in (-) M. mucedo based upon the amount of compound needed for a half-maximal response (22). Third, (-) mutants of M . mucedo such as N301 are defective in their response to TAs under certain conditions, and to methyl 4-dihydroTAs under all conditions tested (22). Although these mutants demonstrate that TAs enter the cell, they do not indicate which compounds are the regulatory molecules. M. mucedo could have enzymes which recycle small amounts of TAs to regulatory molecules (an idea which has not been tested). It seems that the dilemma of what serves as the regulatory compound can be solved only after mutants unable to metabolize the pheromones are isolated.
Overview of Pheromone Metabolism in Mucorales- Fig. 4 summarizes what is known and what is believed about the interaction of (+) and (-) cultures in TA biosynthesis. Separate (+), but apparently not (-), cultures biosynthesize trace amounts of TAs. The methyl 4-dihydroTAs and methyl TAs are (+) pheromones, mating type-specific precursors of TAs; trisporins and trisporols are (-) pheromones. The distinguishing feature of (+) and (-) pheromones is the degree of modification of the pro-S methyl group on carbon atom 1 of 4-dihydrotrisporin, a putative intermediate made by both mating types. The methyl group is either unaltered (e.g. trisporins) or oxidized to an alcohol (e.g. trisporols) in (-) pheromones and oxidized to a carboxyl group which is esterified in (+) pheromones. The finding that (+) cultures do oxidize the hydroxyl group on carbon atom 4, possibly by the same enzyme protein which catalyzes this reaction in (-) cultures (23), implies that (+) cultures possess a mechanism, unrecognized to date, for inactivating TA precursors which would self-stimulate the formation of zygophores.   ----------

Polarities. Cpd85 and Cpd76
were recoqnlzed lnltlally as possible dehydration products by thelr behavior upon chromatography (Table  VI)