APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Feb. 1978, p. 435-4380099-2240/78/0035-0435$02.00/0Copyright © 1978 American Society for Microbiology

Vol. 35, No. 2

Printed in U.S.A.

NOTES

Production of Penicillic Acid and Patulin by an AtypicalPenicillium roqueforti Isolatet

F. J. OLIVIGNI AND L. B. BULLERMAN*

Department ofFood Science and Technology, University of Nebraska-Lincoln, Lincoln, Nebraska 68583

Received for publication 28 June 1977

Simultaneous production of penicillic acid and patulin by an atypical strain ofPenicillium roqueforti isolated from cheddar cheese is reported. Mycotoxinproduction was confirmed by thin-layer and gas-liquid chromatography and byultraviolet, infrared, and mass spectral analyses. Culture extracts were toxic toBacillus megaterium and chicken embryos. Commercial strains of P. roquefortiused in production of blue-veined cheeses were shown not to produce penicillicacid and patulin.

In a recent study of the mold flora of cheddarcheese, we reported that a number of mold iso-lates were capable of producing patulin and pen-icillic acid in yeast-extract sucrose broth (1). Insubsequent work with one of the isolates desig-nated M-247, it was observed that this organismappeared to be capable of producing penicillicacid and patulin simultaneously. Preliminary ex-amination of the culture indicated that the or-ganism appeared to be an atypical Penicilliumroqueforti. Since we were unaware of previousreports of the simultaneous production of peni-cillic acid and patulin by P. roqueforti, the pur-pose of the work reported here was to verify theidentity of the organism, confirm the presenceof penicillic acid and patulin in culture extracts,and to determine whether commercial culturesof P. roqueforti might also possess this ability.Upon close examination, it was found that the

culture clearly belonged in the Asymmetricasection of the genus Penicillium. The correctplacement of the species, however, was some-what in doubt, even though the classificationsystems of Thom (17) and Raper and Thom (12)were consulted. Initial attempts to identify theisolate to species level, using the key of Raperand Thom (12), led to uncertainty as to whetherit was P. cyclopium or P. roqueforti. Culture M-247 differed from P. cyclopium in that the colonycolor was dark green rather than the typicalblue-green, the mycelial mat was very thin, andthe culture lacked the typical strong Actinomy-cetes-like odor. Reverse color of the isolate was

t Published as paper no. 5322, Journal Series, NebraskaAgricultural Experiment Station, Lincoln.

a pale green, not the dark greenish black nor-mally associated with P. roqueforti, and thetypical arachnoid margins of P. roqueforti werenot produced. The conidia were smooth andglobose, never deviating toward subglobose orelliptical, as is common in P. cyclopium. Conidialchains were produced in long columnar massesthat clung together, often even in a wet mount.Conidiophores, branches, and metulae were con-spicuously encrusted when grown in Czapek ormalt agar (Fig. 1). These echinulations were verycoarse and knobby, much more similar to P.roqueforti than to P. cyclopium. When we re-ferred to earlier classification keys, this organismshowed a striking similarity to Thom's descrip-tion of P. asperulum Bainier, which he listedimmediately after the species he believed to benearly inseparable from P. roqueforti (17). Intheir 1949 manual, Raper and Thom assigned P.asperulum as "probably some member of the P.roqueforti series" (12). From these observations,the organism was identified as an atypical P.roqueforti.To study toxin production the mold was grown

on 900 ml of 2% yeast extract-15% sucrose brothin Fernbach flasks at 12°C for 14 days. Thecultures were macerated in a blender, filtered toremove particulate material, and extracted withtwo equal 900-ml volumes of chloroform. Theextracts were then concentrated in vacuo, trans-ferred to vials, and dried in a stream of nitrogen.Penicillic acid and patulin fractions of the ex-tract were separated, identified by comparisonto standards, and collected by using preparativethin-layer chromatography plates (20 by 20 cm;

435

on July 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

APPL. ENVIRON. MICROBIOL.

FIG. 1. Photomicrograph ofM-247 isolate showing two views of the roughened conidiophores and metulae(oil immersion, xl,OOO).

Silica Gel G-HR, Brinkmann Instruments Inc.)developed in chloroform-ethyl acetate-90%formic acid (60:30:10) (CEF). The bands corre-sponding to the Rf of penicillic acid and patulinwere previously determined by derivatization

with ammonia and phenylhydrazine (14). Thesebands were collected, and the respective myco-toxin was eluted from the silica gel by refluxingwith chloroform for 1 to 2 h in a Mini-Soxhletapparatus. The fractions were checked by thin-

436 NOTES

on July 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

NOTES 437

layer chromatography to confirm the presence

of only one of the desired metabolites in eachfraction; if the fraction contained both com-

pounds, preparative thin-layer chromatographywas repeated until a clean fraction was obtained.The resulting fractions, designated as PA (pen-icillic acid) and PAT (patulin), were used toconfirm the identity of the compounds by deter-mining the various spectral and chromato-graphic properties of the fractions.On thin-layer chromatography analysis, the

PA fraction had an Rf of 0.73, and the PATfraction had an Rf of 0.64 (Silica Gel G-HRplates, 0.25 mm thick, developed in CEF). Whenderivatized with ammonia, phenylhydrazine,and p-anisaldehyde (14), the Rf values and colorreactions were identical to those of standard PAand PAT. Ultraviolet absorption spectra (Beck-man model 25) gave results for the PA (241 nm)and PAT (276 nm) fractions identical to thosefor the PA and PAT standards. The infraredspectra (Beckman model IR-5A) of the two frac-tions corresponded to those of the two myco-

toxin standard compounds, with absorptionbands for the PA fraction at 5.7 and 6.1 cm-'and for the PAT fraction at 5.7, 6.0, and 6.1cm-'. Gas-liquid chromatography (model 1740Varian Aerograph with model 30 Varian Aero-graph recorder; column conditions: stainlesssteel, 1.7 m by 3.2 mm OD, 3.0% OV-7 and 1.5%OV-22 on 80- to 100-mesh Chromosorb G, VarianAerograph; conditioned at 2500C for 24 h; N2carrier gas flow of 25 ml/min; injector block anddetector temperatures of 2500C; temperatureprogram of 160 to 2500C in increments of6°C/min; electometer range, 1010 A/mV) was

done, comparing both the nonderivatized andthe trimethylsilyl-derivatized {Tri-Sil-BSA-[N,O-Bis-(trimethylsilyl)-acetamide]; PierceChemical Co., Rockford, Ill.) forms of the twofractions to the same forms of the mycotoxinstandards (11). Retention times for the PA stan-dard were 6.3 and 5.4 min, and those for thePAT standard were 8.7 and 7.3 nir. (nonderiva-tized and derivatized, respectively). The PA andPAT fractions had similar retention times. Fi-nally, the mass spectral patterns of the PA andPAT fractions were compared with authenticPA and PAT standards by using an AE1 MS 50high-resolution mass spectrometer (source tem-perature, 200°C; ionizing voltage, 70 eV). It wasfound that the mass spectral peaks of the frag-mentation patterns of the fractions matchedthose of the authentic compounds. The valuespresented are in agreement with those reportedin the literature (6, 10, 11, 14-16, 18, 20) andconfirm that both penicillic acid and patulinwere produced by the culture. The culture ex-

tract was inhibitory to Bacillus megaterium and

toxic to chicken embryos. Both penicillic acidand patulin inhibit the growth of B. megaterium(7), and both are toxic to chicken embryos, inwhich the mean lethal dose of penicillic acid is0.85 mg/egg via air sac (3) and that of patulin is68.7,g/egg via air sac (4).

P. roqueforti, the essential fungus used in themanufacture of Roquefort and other varieties ofblue cheeses, has been shown to produce othertoxic compounds. These include PR toxin (19)and roquefortine, a neurotoxic alkaloid, whichhas been detected in blue cheese samples fromseven different countries (13). Orth (9) reportedfinding PR-toxin-producing P. roqueforti iso-lates in nut ice cream, a dessert cream, and on ared wine bottle cork. However, the significanceof these toxins in terms of human health isunclear, particularly because blue-veinedcheeses have been consumed for centuries with-out apparent ill effects.The potential carcinogenic nature of penicillic

acid and patulin (5) possibly makes them a causefor concern. However, work that we have re-ported elsewhere with M-247 has shown thatthese toxins may not be produced in proteina-ceous substrates, such as cheese (Olivigni andBullerman, J. Food Sci., in press). Also, it ispossible that commercial cultures of P. roque-forti used in blue cheese production do not pro-duce penicillic acid and patulin. This latter pointwas studied in our laboratory, using culturesobtained from commercial suppliers of P. roque-forti cultures (Dairyland Food Laboratories Inc.,Waukesha, Wis., and Laboratorium Visby, T0n-der, Denmark) and cultures isolated directlyfrom blue-veined cheeses. Three commercialstrains of P. roqueforti and seven isolates ob-tained from domestic and imported blue, Roque-fort, and Gorgonzola cheeses were tested fortheir ability to produce penicillic acid and pa-tulin. Each culture was grown on 50 ml each of2% yeast extract-15% sucrose and potato-dex-trose (8) broths in 250-ml Erlenmeyer flasks at120C for 14 days. In addition, culture M-247 anda patulin-producing strain of P. patulum, alsoisolated from cheese, were grown under the sameconditions and used as controls. All cultureswere separated into mold mat and broth frac-tions, and each was extracted separately. Theyeast extract-sucrose broth and correspondingmold mats were each extracted with two 50-mlvolumes of chloroform. The potato-dextrosebroth and corresponding mold mats were eachextracted with two 50-ml volumes of ethyl ace-tate. Samples were concentrated and examinedby thin-layer chromatography, as previously de-scribed. All commercial cultures and isolates ofP. roqueforti obtained from the blue-veinedcheeses were found negative for both penicillic

VOL. 35, 1978

on July 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

438 NOTES APPL. ENVIRON. MICROBIOL.

acid and patulin. However, control culture M-247 produced detectable amounts of patulin (75yg/ml) on potato-dextrose broth and both peni-cillic acid and patulin (100 and 35 ,tg/ml, respec-tively) on yeast extract-sucrose broth. P. patu-lum likewise produced detectable amounts ofpatulin in both potato-dextrose (2.8 mg/ml) andyeast extract-sucrose (1.4 mg/ml) broths. Thus,whereas the known toxin-producing controlsproduced easily detectable amounts of penicillicacid and patulin under the conditions of thistest, the commercial strains of P. roqueforti didnot.

Several species of Penicillium and Aspergil-lus have been reported to produce either patulinor penicillic acid (2). Ciegler (2) previously listedP. roqueforti as capable of producing penicillicacid, but no reference has been found associatingit with patulin production. Also, to our knowl-edge this is the first report of any isolate pro-ducing both mycotoxins.

We are grateful to A. Ciegler for the penicillic acid andpatulin standards. We thank D. I. Fennell and P. B. Mislivecfor their assistance in confirmation of the identification of themold isolate. We also thank Pat Holzemer for her technicalassistance with portions of this study.

This research was conducted under project no. 16-022 andwas supported by Public Health Service research grantCA14260 from the National Cancer Institute.

LITERATURE CITED

1. Bullerman, L. B., and F. J. Olivigni. 1974 Mycotoxinproducing potential of molds isolated from Cheddarcheese. J. Food Sci. 39:1166-1168.

2. Ciegler, A. 1973. Mycotoxins, p. 526-576. In A. I. Laskinand H. A. Lechevalier (ed.), CRC handbook of micro-biology, vol. 3, Microbial products. CRC Press, Cleve-land.

3. Ciegler, A., A. C. Beckwith, and L. K. Jackson. 1976.Teratogenicity of patulin and patulin adducts formedwith cysteine. Appl. Environ. Microbiol. 31:664-667.

4. Ciegler, A., H.-J. Mintzlaff, D. Weisleder, and L.Leistner. 1972. Potential production and detoxificationof penicillic acid in mold-fermented sausage (salami).

AppI. Microbiol. 24:114-119.5. Dickens, F., and H. E. H. Jones. 1961. Carcinogenic

activity of a series of reactive lactones and relatedsubstances. Br. J. Cancer 15:85-88.

6. Ford, J. H., A. R. Johnson, and J. W. Hinman. 1950.The structure of penicillic acid. J. Am. Chem. Soc.72:4529-4531.

7. Geiger, W. B., and J. E. Conn. 1945. The mechanism ofthe antibiotic activity of clavacin and penicillic acid. J.Am. Chem. Soc. 67:112-116.

8. Norstadt, F. A., and T. M. McCalla. 1969. Patulinproduction by Penicillium urticae Bainier in batchculture. Appl. Microbiol. 17:193-196.

9. Orth, R. 1976. PR-toxin production by Penicillium ro-queforti strains. Z. Lebensm. Unters. Forsch.160:131-136.

10. Pero, R. W., and D. Harvan. 1973. Simultaneous detec-tion of metabolites from several toxigenic fungi. J. Chro-matogr. 80:225-258.

11. Pero, R. W., D. Harvan, R. G. Owens, and J. P. Snow.1972. A gas chromatographic method for the mycotoxinpenicillic acid. J. Chromatogr. 65:501-506.

12. Raper, K. B., and C. Thom. 1949. A manual of thepenicillia. The Williams & Wilkins Co., Baltimore.

13. Scott, P. M., and B. P. C. Kennedy. 1976. Analysis ofblue cheese for roquefortine and other alkaloids fromPenicillium roqueforti J. Agric. Food Chem.24:865-868.

14. Scott, P. M., and E. Somers. 1968. Stability of patulinand penicillic acid in fruit juices and flour. J. Agric.Food Chem. 16:483-485.

15. Stott, W. T., and L. B. Bullerman. 1975. Influence ofcarbohydrate and nitrogen source on patulin productionby Penicillium patulum. Appl. Microbiol. 30:850-854.

16. Suzuki, T., M. Takeda, and H. Tanabe. 1971. Studieson chemical analysis of mycotoxins. II. Gas chromatog-raphy of penicillic acid and its derivatives and gaschromatography analysis of penicillic acid in rice. J.Food Hyg. Soc. Jpn. 12:495-500.

17. Thom, C. 1930. The penicillia. The Wiiliams & WilkinsCo., Baltimore.

18. van Eijk, G. W. 1969. Isolation and identification oforsellinic acid and penicillic acid produced by Penicil-lium fennelliae Stolk. Antonie van Leeuwenhoek J.Microbiol. Serol. 35:497-504.

19. Wei, R., P. E. Still, E. B. Smalley, H. K. Schnoes, andF. M. Strong. 1973. Isolation and partial characteriza-tion of a mycotoxin from Penicillium roqueforti. Appl.Microbiol. 25:111-114.

20. Woodward, R. B., and G. Singh. 1949. The structure ofpatulin. J. Am. Chem. Soc. 71:758-759.

on July 7, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from