JDS
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Journal of Dairy Science Vol. 85 No. 11 2786-2794
© 2002 by American Dairy Science Association ®
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Suhr, K. I.
Right arrow Articles by Larsen, T. O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Suhr, K. I.
Right arrow Articles by Larsen, T. O.

Factors Affecting Growth and Pigmentation of Penicillium caseifulvum

K. I. Suhr, I. Haasum, L. D. Steenstrup and T. O. Larsen

Mycology Group, Biocentrum-DTU, Søltofts Plads, Building 221, Technical University of Denmark, DK-2800 Lyngby, Denmark

Corresponding author:
Karin Isabel Suhr; e-mail:
kasu{at}biocentrum.dtu.dk.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 ACKNOWLEDGEMENTS
 REFERENCES
 
Color formation, metabolite production and growth of Penicillium caseifulvum were studied in order to elucidate factors contributing to yellow discoloration of Blue Cheese caused by the mold. A screening experiment was set up to study the effect of pH, concentration of salt (NaCl), P, K, N, S, Mg and the trace metals Fe, Cu, Zn, Mn on yellow color formation, metabolite production and mold growth. Multivariate statistical analysis showed that the most important factor affecting yellow color formation was pH. The most pronounced formation of yellow color, supported by highest amount of colored metabolites, appeared at low pH (pH 4). Mold growth was not correlated to the yellow color formation. Salt concentration was the most important factor affecting mold growth and length of lag phase. Production of secondary metabolites was strongly influenced by both pH and salt concentration. The screening results were used to divide the metabolites into the following three groups: 1) correlated to growth, 2) correlated to color formation, and 3) formed at high pH. Subsequently, a full factorial experiment with factors P, Mg and Cu, showed that low P concentrations (2000 mg/kg) induced yellow color formation. Among the factors contributing to yellow color formation, pH and salt concentration are easy to control for the cheesemaker, while the third factor, P-concentration, is not. Naturally occurring variations in the P-concentration in milk delivered to Blue Cheese plants, could be responsible for the yellow discoloration phenomenon observed in the dairy industry.

Key Words: blue cheese • Penicillium caseifulvum • pigmentation

Abbreviation key: aw = water activity, HPLC = High performance liquid chromatography, PLS = partial least square regression


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 ACKNOWLEDGEMENTS
 REFERENCES
 
Penicillium caseifulvum is a recently described species isolated from Roqueforti-fermented cheese types, i.e. blue cheese, and their production environment (Lund, 1998). Experiments have shown that metabolites produced by P. caseifulvum are responsible for yellowish spots appearing on the surface of these cheeses (Lund, 1995). Although the mold is present and grows in the environment and on the cheese all year round, the discoloration only appears in certain periods, and is most pronounced in the summer. This suggests that unknown factors in the cheese control the production of the colored metabolites. Although discoloration of cheese by contaminating molds has been previously described (Chapman and Sharpe, 1990; Estrup, 1994; Piantanida et al., 1995), knowledge of the biochemical mechanisms and essential factors for discoloration is limited.

Cheese contains all the essential nutrients for growth of fungi (Northolt et al., 1995). However, the proportions of nutrients in the cheese will depend upon the production process, such as coagulation, whey removal, and salting (Martin-Hernández et al., 1992; Coni et al., 1995). As process parameters are held relatively constant when making a specific cheese type, variability in the mineral content of the cheese is mainly due to changes in the milk. In modern dairying, feeding practice is the most important factor for variability, and this is normally linked with season. The greatest seasonal effect occurs in the spring-summer, when cows are fed fresh grass (Grandison et al., 1984; Mariani et al., 1993). Grazing correlates with higher pH in the milk, due to a changed balance of salts (Hermansen et al., 1994). Variability of soluble inorganic phosphorus is the main contributor to changes in the salt balances (Donnelly and Horne, 1986). The concentration of Cu in the milk is lowest in the summer, while Zn and Fe are reported to have the highest level during this time of the year (von Kreuzer and Kirchgessner, 1990; Larsen and Werner, 1985). Generally, the lowest levels of minerals in the milk occur from April to September (in Denmark), with changes between 5 and 10% (Larsen and Werner, 1984), coinciding with grazing of cows. The amount of minerals in vat milk is not standardized prior to cheese making, and as a result mineral levels in blue cheese fluctuate according to the natural variation in the raw milk.

Growth of fungi depends on the composition of the growth media, water activity (aw), pH, temperature, light, and the surrounding atmospheric gas mixture. The effect of environmental factors on growth of fungi is generally less specific and restricted than the effect on secondary metabolite production. For example, the ranges of aW and pH within which formation of certain secondary metabolites occur, is more narrow, than the range for conidial growth (Northolt and Bullerman, 1982).

The effect trace elements such as Zn, Cu, Fe, and Mn have on the secondary metabolism of fungi has been investigated (Smith, 1949; Francis, 1985; Griffin, 1994). For example, production of the yellow pigment citrinin, by Penicillium implicatum, was induced by addition of Cu and Zn to the growth media (Smith, 1949), and other minerals have been reported to be involved in the formation of other pigments (Francis, 1985).

The purpose of this study was to determine the effect of pH, salt, and selected minerals on the growth and pigmentation of P. caseifulvum.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 ACKNOWLEDGEMENTS
 REFERENCES
 
Organisms and Preparation of Inoculum
Two isolates of Penicillium caseifulvum (IBT 15151 ex. Danablu, DK and IBT 15157 ex. Danablu, DK), obtained from the culture collection at Biocentrum-DTU, Technical University of Denmark, originating from two different dairy plants, were used for the experiments. Conidiated cultures were prepared by spreading silica-dried spores on Czapeck Yeast autolysate Agar (Pitt, 1979) with modifications (Samson et al., 1995). Plates were incubated for 7 d at 25°C in the dark, after which colonies were transferred to fresh media and reincubated for 7 d under the same conditions. Suspensions of spores for inoculation were made as 106 spores /ml in double distilled water with 0.5% agar and 0.5% Tween-80.

Media Preparation
Media for the experiments were modified from inorganic Raulin-Thom (RT) media, which was modified from Clutterbuck et al. (1932). Salt (NaCl) content and pH of the media was varied from 3 to 10% (w/w) and from 4 to 6, respectively. Adjustment of pH was done prior to autoclaving by adding HCl or NaOH. The minerals varied in the media were P, K, Mg, Fe, Cu, Zn and Mn, and their concentrations ranged from 2000 to 6000 mg/kg (ppm) for P ((NH4)2HPO4), 500 to 3000 ppm for K (KCl), 150 to 300 ppm for Mg (MgSO4 • 7H2O), 0.5 to 3 ppm for Fe (FeSO4 • 7H2O), 0.3 to 1.5 ppm for Cu (CuSO4 • 5H2O), 30 to 250 ppm for Zn (ZnSO4 • H2O), and 0.3 to 10.6 ppm for Mn (MnSO4 • H2O). All chemicals were from Merck (Darmstadt, Germany), except Zn, which was from Sigma Chemicals (St. Louis, MO). The mineral ranges used correspond to the typical range of these minerals in Danablu cheese 50+ (National Food Agency of Denmark, 1989). All minerals were dissolved in water and added as water solutions to the media. Besides the minerals, all media contained 0.26% ammonium(+)tartrate (BDH Chemicals, Poole, England), 0.26% L(+)tartaric acid (Merck), 5% D(+)glucose (BDH) and 2% agar (B&B, Rødovre, Denmark). Agar, glucose and water were autoclaved in a separate mixture to avoid acid hydrolysis of the agar and Maillard reactions.

Inoculation and Incubation
The prepared media were inoculated with spore suspension in three points with a needle. Two Petri dishes of each medium were covered with adhesive autoclavable cellophane for later evaluation of reverse colony color. Cellophane-covered plates were inoculated with a droplet of 10 µl spore suspension. Fungi were cultured at 25°C in the dark for a 3-wk period.

Experimental Design
A 2-level fractional factorial design (1/25 x 29) with three center points was constructed for testing the nine factors: pH, salt, P, K, Mg, Fe, Cu, Zn and Mn (Table 1Go). The design yielded 17 different media. Subsequently, a full factorial experiment (23) with three center points was designed to investigate further the factors: P, Cu and Mg. The ranges of P, Cu and Mg in the media were identical to those in the screening experiment while the other factors were held constant at the center point values from the screening experiment, except salt, which was added in the amount of 3% (w/w). The design yielded eight different media. All experiments were designed using the software program Modde version 3.0 (Umetri AB, Umeå, Sweden).


View this table:
[in this window]
[in a new window]

  		Table 1. Screening experiment for Penicillium caseifulvum growth and color formation on media with different pH levels and different concentrations of salt and minerals.
 
Colony Measurements
The colony diameter was recorded (in millimeters) after 3, 7, 10, and 14 d of incubation. The mean of six colony measurements was used for data analysis. The lag phase was defined as the number of days until the colony diameter exceeded 2 mm.

The color was measured on both the front (obverse) and reverse side of the mold colony. Color measurements of the reverse were done using the cellophane-covered plates. The cellophane was removed, thereby separating the mycelium from the agar surface, uncovering the reverse side of the colony. Color was measured with a colorimeter (Croma Meter, CCR-200, Minolta, Japan) as L*, a* and b* values, which, going from low to high numbers, measures the color spectra as follows: black to white (L*), green to red (a*) and blue to yellow (b*). In order to use values that better described colony colors, L* and b* values were referred to as lightness and yellow color, respectively. Measures of a* were reversed, so the spectra, going from high to low numbers, described red to green, and thus a* was referred to as green color. Measurements were performed after 14 d of incubation. The mean of triplicate measurements was used for data analysis.

Extraction and Analysis of Metabolites
Fungal biomass from 21-d-old colonies (from the screening experiment only) was extracted with a mixture of 50 ml methanol, dichloromethane and ethylacetate in the proportion 3:2:1 and added 0.5 % formic acid. The organic extract was filtered through a Whatman 1 PS phase separator filter and evaporated on a rotary evaporator to dryness. The residue was dissolved in 1.5 ml methanol and transferred to a vial through a Minisart RC 15 filter (0.45 µm) before analysis. The extracts were analyzed by reverse-phase high-performance liquid chromatography (HPLC) on a Hewlett Packard 1090 series II liquid chromatograph (Hewlett Packard, Germany) with a built-in diode array detector measuring full UV-VIS spectra (200 to 600 nm) for each peak detected at 210 and 340 nm. The gradient ran from 15.0% acetonitrile (Merck 707) in water to 100% acetonitrile, and contained 0.005% trifluoracetic acid (Merck 633). It was held at a flow rate of 1 ml/min. The column was a 4 mm id x 100 mm HP Hypersil BDS C-18 (3 µm particles).

Metabolites were quantified by peak area (milliabsorption units (mAu) x s) at {lambda}max, or maximum absorption (mAu) if interfering substances were found within the area. The peak area of ergosterol in the samples, which is a suitable measure of mold growth (Seitz et al., 1979), was measured and used for standardizing the metabolite content according to biomass units. Retention indices (RI) of fungal metabolites were calculated according to Frisvad and Thrane (1987).

Response Variables and Data Analysis
The response variables in the screening experiment, 17 in all, were the following: mold diameter recorded after 7 and 14 d, the calculated lag phase, color coordinates (i.e. L*, a*, and b* values) for obverse and reverse sides of the colony, and finally eight metabolites were quantified and used for analysis. In the subsequent full-factor experiment, the metabolite extraction was omitted, and only colony diameter after 14 d and reverse side color coordinates were used for analysis, thus four responses were used for analysis.

The data were analyzed by partial least square (PLS) regression using SIMCA-P for Windows (version 2.1, 1995; Umetri AB, Umeå, Sweden). We previously described the advantages of using this method for physiological characterization of fungi (Haasum and Nielsen, 1998). The statistical principle of the method was described by Wold et al. (1984). The X-matrix for the screening experiment consisted of the design variables given in Table 1Go. All minerals except P were added to the media as sulfate compounds, and in order to determine the effect of sulfate, it was added to the matrix as an uncontrolled variable (P was added as an ammonia compound to the media, and since this was the only mineral added as such, the effect of ammonia was completely confounded with P).

Prior to statistical analysis, all variables were standardized to mean zero and variance one. In the screening experiment the two P. caseifulvum isolates responded alike, thus only one representative is shown in the figures, whereas for the full factor experiments, results for both isolates are shown.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 ACKNOWLEDGEMENTS
 REFERENCES
 
Secondary Metabolites
Seven chromophores represented the majority of the metabolites in all mold extracts. As the examples of chromatograms for media N10 and N13 show, the seven metabolites were not all present in all extracts (Figure 1Go). For example, N10 supported low metabolite F production and high metabolite I, whereas the opposite was true for media N13. Metabolite D had two analog forms, which were called D1 and D2. Using results from Frisvad and Thrane (1987) the metabolites B and C were identified as rugulovasine and cyclopeptine, respectively. The chemical structures of the other metabolites (i.e. A, D1, D2, E, F and I) were not identified.


Figure 1
View larger version (18K):
[in this window]
[in a new window]

  		Figure 1. Chromatograms of Penicillium caseifulvum (IBT 15151) on media N10 and N13 obtained at 254 nm (full line) and 340 nm (dashed line). Metabolites A, B (rugulovasine), C (cyclopeptine), D1, D2, E, F, and I were quantified for statistical analysis (metabolite E was not produced on the media shown here). Norm: Normalized absorption in milliabsorption units.

 
Strong absorption of light was detected for metabolites E and F in the visible wavelength area (380 to 780 nm), indicating that these metabolites were colored products (Figure 2Go). The spectrum of metabolite F showed especially strong absorption at 350 to 500 nm indicating a yellow to orange-yellow color, while metabolite E absorbed at 350 to 420 nm, indicating a more greenish-yellow color (Figure 2Go). Metabolites D1 and D2 had slight absorption above 380 nm, indicating that these might have some coloring properties (Figure 2Go). The remaining metabolites (A, rugulovasine, cyclopeptine and I) did not absorb light in the visible wavelength area.


Figure 2
View larger version (22K):
[in this window]
[in a new window]

  		Figure 2. UV-spectra and retention index (RI) of metabolites from Penicillium caseifulvum. Metabolites E and F showed absorption of light in the visible area (380 to 780 nm). mAU: milliabsorption units.

 
All metabolites were quantified at the wavelength where peak absorbance was observed. In this fashion, the metabolites A, rugulovasine, cyclopeptine, and ergosterol (used as a biomass indicator) were quantified in chromatograms obtained at 210 nm; metabolites D1, D2, and I were quantified at 335 nm, while E was quantified at 400 nm. Detachment of metabolite F from the column was not optimal, causing a broad peak and a span in the retention index. In order to reduce interference from other substances, it was quantified at 432 nm.

Effect of Salt and pH on Growth and Color of P. caseifulvum
Growth of P. caseifulvum was strongly correlated with salt content of the media. On media with 10% salt, the average mold colony diameter was 8.4 ± 3.5 mm after 2 wk, whereas on media containing 3% salt, it averaged 38.4 ± 8.8 mm (Figure 3aGo). Measurements of reverse yellow color (reverse b) correlated well with pH values of the media (Figure 3bGo). At pH 4 the b average was 44.0 ± 7.7, while at pH 6 it was 14.0 ± 8.2. Thus at low pH the reverse colony color of P. caseifulvum was more yellow compared with high pH. Production of metabolite F also followed this pattern, so that the mean production of metabolite F was 5.6 ± 2.0 and 1.2 ± 1.0 mAu x s on media with pH 4 and 6, respectively. Thus, metabolite F could be used as a measure of yellow color development of P. caseifulvum colonies. It should be noted, that all extractions were performed at identical pH values (in 0.5% formic acid), thus the observed correlation between pH and yellow color was not caused by any extra-cellular pH effect.


Figure 3
View larger version (18K):
[in this window]
[in a new window]

  		Figure 3. Salt content in the media correlated to measurements of growth (a), and pH correlated to measurements of yellow color (coordinate b on the Chroma Meter) and production of metabolite F (b).

 
The PLS model obtained by statistical analysis consisted of four significant components, explaining 70.4% of the variation in the data. Loading component one showed that salt was positively correlated with lag phase, and negatively correlated with growth (colony diameter), and production of rugulovasine, cyclopeptine, and metabolite A (Figure 4aGo). Thus, high salt content in the media produced longer lag phase, less mold growth, and inhibited production of some metabolites. Extension of lag phase caused by high salt percent is well known for P. roqueforti (Godinho and Fox, 1981). Furthermore, high salt-media produced less green and more white colony obverse colors, which agreed with the fact that mold growth was reduced in high salt media, resulting in fewer conidia and consequently less green color.


Figure 4
View larger version (24K):
[in this window]
[in a new window]

  		Figure 4. Loading (a) and scores (b) plot of the screening experiment of Penicillium caseifulvum (IBT 15157). The explained variance for each component is given in parenthesis, and the independent variables are shown in bold writing (S was added to the analysis as an uncontrolled factor). The loading plot (a) shows that salt content of the media (Salt) was positively correlated with obverse lightness (Lightness-o) and lagphase (Lagphase), and negatively correlated with mold growth (Growth), obverse green color (Green-o), and production of metabolites A (met-A), rugulovasine (Rugulovasine), and cyclopeptine (Cyclopeptine). pH was positively correlated with production of metabolite I (met-I), and negatively correlated with reverse and obverse yellow color (Yellow-r, Yellow-o), reverse lightness (Lightness-r) and production of metabolites D1, D2, E, and F (met-D1, met-D2, met-E, met-F). The scores plot (b) shows that some observations divided into groups according to visually observed color of the reverse colony side. The numbers identify sample number, which is identical to media number in the screening experiment (table 1Go).

 
Loading component two showed the strong negative correlation between reverse yellow color and pH (Figure 4aGo), which was described earlier. Obverse yellow and reverse lightness were also negatively correlated with pH and thus positively correlated with reverse yellow. This suggests that measuring the colony color of the obverse side of P. caseifulvum gives a good indication of the yellowness on the reverse. Visually observed color of the reverse also showed the pH effect, as groupings of yellow and orange-yellow colonies appeared in media with low pH (4), as indicated in the score-plot (Figure 4bGo). This observation was notable after 7 d, and very clear after 21 d.

Media pH was also negatively correlated with production of metabolites D1, D2, E, and F (Figure 4aGo). Thus low pH (4) induced production of these metabolites. Component two also described production of metabolite I, however, contrary to the other metabolites, production of I was stimulated at high pH.

Overall, the model divided the metabolites into three groups: 1) rugulovasine, cyclopeptine and metabolite A, which were associated with mold growth; 2) metabolites D1, D2, E and F, which were associated with low pH and yellow color; and 3) metabolite I, which was associated with high pH values. It is likely that there is a biosynthetic relationship between metabolites D1 and D2 and metabolite F because of the similarities in their UV-spectra (Figure 2Go) and their loading values in the PLS model (Figure 4aGo).

Effect of Minerals on Growth and Color of P. caseifulvum
All media except N3 supported mold growth. This medium had a pH of 4, 10% salt, low content of Fe (0.5 mg/kg), Cu (0.3 mg/kg) and Mn (150 mg/kg); this combination of ingredients may have caused the lack of mold growth. However, the PLS loading plot showed that the effects of minerals on mold growth, color and metabolite production were small or insignificant (all loading values of minerals were smaller than 0.5) compared with those of pH and salt.

In order to focus on the effect of minerals on color formation, an additional PLS analysis was performed using the colored metabolites E and F and the three reverse color values as response variables. The model, which explained 73% of the variation with three loading components, exposed again the impact of pH on color formation, and showed that green color was not correlated to production of the colored metabolites E and F. Of the minerals the most important ones correlated with yellow color and colored metabolites were P, Mg and Cu. Consequently, these three minerals were used in a full factorial experiment to further analyze the mineral effect on yellow color development in P. caseifulvum colonies.

Growth of P. caseifulvum on media with different levels of P, Mg and Cu showed only small differences in growth and morphology. After 14 d the average colony diameter for isolate 15151 and 15157 was 40.3 ± 3 mm and 42.9 ± 3 mm, respectively. Yellow color measurements averaged 34.6 ± 6.5 and 30.7 ± 7.2 for isolate 15151 and 15157, respectively.

Media containing high levels of P (6000 mg/kg) had, on average, a 1% higher pH value, equal to 0.05 pH units, after autoclaving compared to media with low P content (2000 mg/kg). The difference in pH was significant by paired t-test (P < 0.05), and therefore pH was included as an additional factor in the data analysis. Minor variations of pH in all media were also recorded. The PLS model, explaining 72.9% of the variation with four components, showed that although the difference in pH was small (i.e. 5.10 ± 0.01 for low-P media and 5.15 ± 0.03 for high P media). The pH was significantly and negatively correlated with reverse yellow color, again confirming the importance of pH for yellow color development (Figure 5Go).


Figure 5
View larger version (14K):
[in this window]
[in a new window]

  		Figure 5. Loadings plot of the full factorial experiment. The explained variance for each component is given in parenthesis and the independent variables are shown in bold writing. Media pH (pH) (added to the analysis as an uncontrolled factor) and concentration of P (P) were the most significant factors for yellow color (Yellow-isolate 1, Yellow-isolate 2) and growth (Growth-isolate 1, Growth-isolate 2). High levels of pH or P (6000 mg/kg) gave less yellow color, and increased growth.

 
Of the minerals, P was the most important factor for yellow color and mold growth. High levels of P (6000 mg/kg) gave less yellow color and increased growth. It has previously been shown that P can be crucial for pathways in the secondary metabolism (Turner, 1971; Garraway and Evans, 1984). The results of this experiment suggest that P is important for metabolism that support mold growth but it inhibits metabolism causing yellow color formation. Thus, high levels of P may inhibit or reduce production of the colored metabolites E and F.

The P-level in milk is reported to drop during the summer season (Larsen and Werner, 1985), coinciding with the period where pigmentation problems mostly occur. In conclusion, addition of phosphorous salts to the cheese-milk may prevent or reduce formation of pigments by contaminating molds. However, it may also increase the basis for mold growth.

In this experiment P was added as (NH4)2HPO4, and since it was the only mineral added as an ammonia salt the effect of P and N could not be statistically separated. Therefore, future research on this subject should investigate the effect of P and N separately.

Similar to P, high levels of Mg (300 mg/kg) were positively correlated with mold growth and negatively correlated with yellow color formation. In contrast, high levels of Cu (1.5 mg/kg) were positively correlated with yellow color formation. Smith (1949) also showed that Penicillium implicatum production of the yellow pigment citrinin was induced when Cu was added to the growth media. However, the effect on yellow color formation was very small within the levels of Cu found in Blue Cheese (0.3 to 1.5 mg/kg). This agrees with the fact that levels of Cu in milk used for Blue Cheese production are highest in the winter, but the yellow color formation is mostly a problem in the summer. Therefore, if high concentrations of minerals initiate yellow color formation in P. caseifulvum colonies, minerals other than Mg, Cu, or P are the cause. Zn, for instance, may trigger yellow color formation, because it was positively correlated with yellow color formation (Figure 4aGo), and it is present in milk in highest levels during the summer (Larsen and Werner, 1985; von Kreuzer and Kirchgessner, 1990). Additionally, high levels of S (520 to 842 mg/kg) were also positively correlated with yellow color (Figure 4aGo). However, S was not a controlled factor in this experiment.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
ACKNOWLEDGEMENTS
 REFERENCES
 
Of the chromophores representing the majority of metabolites produced by P. caseifulvum in the different media, E and F were identified as the only colored products, i.e. greenish-yellow and orange-yellow, respectively. The production of these metabolites was also correlated with measurements of yellow color on the reverse side of mold colonies as well as visual observations.

Yellow pigmentation of P. caseifulvum colonies was induced at low pH (4). Even a very small decrease of pH (0.1 pH units) in the media caused a significant increase in yellow color development. Thus, these findings stress the importance of controlling the pH in the Blue Cheese process.

Production of the colored metabolites was not affected by salt content or mold growth. Instead, increased salt content caused retardation of mold growth and increased the length of the lagphase.

Of the minerals, P was the only one that affected yellow color formation of the colonies. High levels of P (6000 mg/kg) caused less yellow color development. Thus, the P-level may have a regulatory effect on production of the colored metabolites E and F, and therefore addition of phosphorous salts to the cheese-milk may help prevent or reduce formation of yellow metabolites by P. caseifulvum on Blue Cheese. However, high levels of P also increased mold growth, which could limit the effectiveness of P added to the cheese-milk. This study could not identify any significant effects from any of the other minerals studied, i.e. Mg, Cu, Mn, Fe, and S (which was an uncontrolled factor).

Two of the seven chromophores that were analyzed were identified as rugulovasine and cyclopeptine, while the remaining metabolites were not identified. The seven chromophores were divided into three groups. One group consisted of metabolites related to mold growth, these were metabolite A, rugulovasine, and cyclopeptine. A second group consisted of colored metabolites produced at low pH (4 to 5); these were metabolites E, F, D1, and D2 (D1 and D2 were only slightly colored). Finally, a third group, consisting of just one metabolite, was produced at high pH (7); this was metabolite I.


   ACKNOWLEDGEMENTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 ACKNOWLEDGEMENTS
REFERENCES
 
The support of the Danish Dairy Foundation (Danish Dairy Board) and the Danish Research and Development Programme for Food Technology is acknowledged.

Received for publication March 15, 2002. Accepted for publication June 25, 2002.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 ACKNOWLEDGEMENTS
 REFERENCES
 


Chapman, H. R., and E. M. Sharpe. 1990. Microbiology of Cheese. Pages 157–243 in Dairy Microbiology. Vol. 2. R. K. Robinson, Elsevier Science Publishers, London.

Clutterbuck, P. W., A. E. Oxford, H. Raistrick, and G. Smith. 1932. Studies in the biochemistry of microorganisms. The metabolic products of the Penicillium brevi-compactum series. Biochem. J. 26:1441–1458.[Medline]

Coni, E., A. Bocca, D. Ianni, and S. Caroli. 1995. Preliminary evaluation of the factors influencing the trace element content of milk and dairy products. Food Chem. 52:123–130.

Donnelly, W. J., and D. S. Horne. 1986. Relationship between ethanol stability of bovine milk and natural variations in milk composition. J. Dairy Res. 1:23–33.

Estrup, S. 1994. Microorganisms responsible for spoilage in ripened cheese. A review. Dairy Science and Technology, Hjørring, Denmark.

Francis, F. J. 1985. Pigments and other colorants. Pages 545–584 in Food Chemistry. O. R. Fennema, Marcel Dekker, Inc., New York.

Frisvad, J. C., and U. Thrane. 1987. Standardized high-performance liquid chromatography of 182 mycotoxins and other fungal metabolites based on alkylphenone retention indices and UV-VIS spectra (diode array detection), J. Chromatogr. 404:195–214.

Garraway, M. O., and R. C. Evans. 1984. Fungal nutrition and physiology. John Wiley and Sons, Inc., New York, U.S.A.

Godinho, M., and P. H. Fox. 1981. Effect of NaCL on the germination and growth of Penicillium roqueforti. Milchwissenschaft. 36:205–208.

Grandison, A. S., G. D. Ford, A. J. Owen, and D. Millard. 1984. Chemical composition and coagulating properties of renneted Friesian milk during transistion from winter rations to spring grazing. J. Dairy Res. 51:69–78.

Griffin, D. H. 1994. Fungal Physiology. 2nd ed. Wiley-Liss, New York, U.S.A.

Haasum, I., and P. V. Nielsen. 1998. Ecophysiological characterization of common food-borne fungi in relation to pH and water activity under various atmospheric composition. J. Appl. Microbiol. 84:451–460.[Medline]

Hermansen, J. E., S. Ostersen, and O. Aaes. 1994. Effect of the level of N fertilizer, grass and supplementary feeds on nitrogen composition and renneting properties of milk from cows at pasture. J. Dairy Res. 61:179–189.[Medline]

von Kreuzer, M., and M. Kirchgessner. 1990. Zn, Fe, Cu und Mn in der Milch von Kuhen bei unterschiedlicher Rohproteinzufuhr. J. Anim. Physiol. Anim. Nutr. 63:204–209.

Larsen, J., and H. Werner. 1984. Calcium, magnesium, sodium, potassium, phosphorus, chloride and iron in liquid milk products. 257th report. National Institute of Dairy Science, Hillerød, Denmark.

Larsen, J., and H. Werner. 1985. Heavy metals in market milk products. 262nd report. National Institute of Dairy Science, Hillerød, Denmark.

Lund, F. 1995. Penicillium solitum and a new Penicillium species causing discoloration of P. roqueforti fermented cheese. Pages 41–49 in Mycological methods for detecting the associated funga of cheese and ryebread in relation to hygiene. Ph.D. Diss., IBT, Technical University of Denmark, Lyngby, Denmark.

Lund, F., O. Filtenborg, and J. C. Frisvad. 1998. Penicillium caseifulvum, a new species found on fermented blue cheese. J. Food Mycol. 2:95–100.

Mariani, P., G. Zanzucchi, P. Blanco, and M. Masoni. 1993. Seasonal variations in the phosphorus content of bulk milk from individual dairy herds. Industria-del-Latte. 1:39–53.

Martin-Hernández, C., L. Amigo, P. J. Martin-Alvarez, and M. Juárez. 1992. Differentiation of milks and cheeses according to species based on the mineral content. Z. Lebensm. Unters. Forsch. 194:541–544.

National Food Agency of Denmark. 1989. Nutrient Composition of Danish Foods. Page 462 in Food Composition Tables. A. MØller, ed. Stougaard Jensen, Copenhagen, Denmark.

Northolt, M. D., and L. B. Bullerman. 1982. Prevention of mold growth and toxin production through control of environmental conditions. J. Food Prot. 6:519–526.

Northolt, M. D., J. C. Frisvad, and R. A. Samson. 1995. Occurrence of food-borne fungi and factors for growth. Pages 243–250 in Introduction to Food-Borne Fungi. 4th ed. R. A. Samson, E. S. Hoekstra, J. C. Frisvad, and O. Filtenborg, Centraalbureau voor Schimmelcultures, Baarn, Holland.

Piantanida, L., C. Cantoni, I. Dragoni, and A. Modolo. 1995. Alterazione gialla superficiale del Taleggio. Industrie Alimentari. 34:734–735.

Pitt, J. I. 1979. The genus Penicillium and its teleomorphic states Eupenicillium and Talaromyces. Acad. Press, London.

Samson, R. A., E. S. Hoekstra, J. C. Frisvad, and O. Filtenborg. 1995. Introduction to Food-Borne Fungi. 4th ed. Centraalbureau voor Schimmelcultures, Baarn, Holland.

Seitz, L. M., D. B. Sauer, R. Burroughs, H. E. Mohr, and J. D. Hubburd. 1979. Ergosterol as a measure of fungal growth. Phytopathology. 11:1202–1203.

Smith, G. 1949. The effect of adding trace elements to Czapek-Dox Medium. Trans. Br. Mycol. Soc. 32:280–283.

Turner, W. B. 1971. Fungal Metabolites. Acad. Press, London.

Wold, S., C. Albano, and W. J. Dunn III. 1984. Multivariate data analysis in chemistry. Pages 17–95 in Chemometrics: Mathematics and Statistics in Chemistry. B. R. Kowalski, ed., D. Reidel Publishing Company, Dordrecht, Holland.



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Suhr, K. I.
Right arrow Articles by Larsen, T. O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Suhr, K. I.
Right arrow Articles by Larsen, T. O.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS