Volatile compound generation in dry fermented sausages by the surface inoculation of selected mould species
The effect of the inoculation of dry fermented sausage surface with an atoxigenic, proteolytic and lipolytic strain of Mucor racemosus, Penicillium aurantiogriseum and Penicillium camemberti on the volatile composition was studied. The analysis of the headspace volatile compounds using gas chromatography/mass spectrometry enabled the identification of 55 volatiles. The study showed that every mould species produced a different volatile profile which was also different from that of the control sausages. Compounds derived from amino acid catabolism, i.e. branched aldehydes and the corresponding alcohols, were produced in higher amounts in sausages inoculated with Penicillium spp. On the other hand, volatiles coming from the microbial esterification were related to sausages inoculated with M. racemosus. The development of the fungal mycelia on the sausage surface protected lipids from oxidation, thus giving rise to fewer lipid oxidation products in the inoculated sausages.
KeywordsDry fermented sausages Moulds Gas chromatography/mass spectrometry Volatile compounds Aroma
Moulds are used as starter cultures for some fermented meat products such as dry fermented sausages and dry cured hams. They provide a pleasant appearance and also participate in the taste and aroma development of these products [1, 2, 3].
The numerous compounds involved in dry fermented sausage aroma arise from enzymatic (endogenous and/or microbial) and chemical reactions during ripening, and also from seasoning and smoking . They include a wide variety of substances, such as hydrocarbons, aldehydes, alcohols, ketones, acids, esters, sulphur derivatives and furans [5, 6]. Proteolytic and lipolytic activity of moulds together with their ability to degrade both free amino acids and fatty acids leads to the formation of branched aldehydes and alcohols, methylketones and secondary alcohols. The former branched compounds have been associated with a “ripened aroma” [7, 8]. 2-Methylpropanal, 2-methylbutanal and 3-methylbutanal have been detected in higher amounts in dry fermented sausages superficially inoculated with moulds than in controls [9, 10] and they have also been identified as aroma compounds produced by different Staphylococcus strains used as starter cultures in these meat products [11, 12].
The starter culture used can be decisive for the aroma of the final product. The effect of bacterial starter cultures on the production of volatile compounds has been extensively studied in dry sausages and in model systems resembling them. In this way, studies by Berdagué et al. , Montel et al.  and Stahnke  showed that different species of Staphylococcus produced different volatile compounds in different amounts. In relation to mould starter cultures, there are some works dealing with the volatile metabolites of several mould strains grown on different culture media [15, 16, 17, 18]. However, there are no references which compare the volatile compounds generated by different mould strains when growing on dry fermented sausages.
Therefore, the objective of the present study was to identify and quantify the volatile compounds of dry fermented sausages inoculated with different atoxigenic mould species (Mucor racemosus, Penicillium aurantiogriseum and Penicillium camemberti) in an attempt to establish the volatile pattern according to the mould species inoculated.
Materials and methods
Preparation of the spore suspensions
Three mould species were used: two of them were atoxigenic, proteolytic and lipolytic strains of Mucor and Penicillium isolated from Spanish fermented sausages [2, 19] and the other one was an atoxigenic, proteolytic and lipolytic strain isolated from commercial Camembert cheese . All these mould strains were identified in the International Mycological Institute (Egham, UK) as M. racemosus, P. aurantiogriseum and P. camemberti, respectively. To obtain the spores, these strains were grown in Roux flasks on Sabouraud agar (Oxoid, Unipath, Basingstoke, Hampshire, UK) at 22 °C for 7 days. Spores were harvested by washing the cultures with sterile saline solution and glass beads, which were added to help dislodge the spores from the mycelium. The spore suspension was filtered through a sterile gauze to remove the mycelial debris and to clarify the turbid suspension. Then, it was centrifuged at 4,800g for 10 min and resuspended again in saline solution. Finally, the spore concentration was adjusted to 106 spores/ml using a Thoma chamber (0.1-mm depth×0.0025-mm2 surface).
Preparation of the fermented sausages
The mixture for “salchichón” (salami-like) dry fermented sausages was prepared using the following formula: (percentage weight per weight): pork (55), beef (13.45), pork fat (25), NaCl (2.5), dextrin (1.8), lactose (1.0), glucose (0.8), monosodium glutamate (0.25), sodium ascorbate (0.046), NaNO3 (0.0095), NaNO2 (0.0065), and equal amounts of whole grain and ground black pepper (0.14). The ingredients were processed in a mincer equipped with an adjustable plate set at a hole diameter of 5 mm and inoculated with a starter culture of Lactobacillus plantarum 4045, Staphylococcus carnosus and Staphylococcus xylosus. The mixtures were stuffed into synthetic sausage casings (40 mm in diameter) and left to ripen in an Ibercex ripening cabinet, model G-28 (A.S.L., San Fernando de Henares, Spain). Four separate batches of fermented sausages were manufactured: batch C (control) consisted of the initial ingredients, seasonings, curing salts and starter culture; and batches M, PA and PCM were like batch C but they were superficially inoculated with a spore suspension of M. racemosus, P. aurantiogriseum and P. camemberti, respectively, immediately after stuffing. Potassium sorbate (25%) (Carlo Erba, Rodano, Italy) was sprayed onto the surface of batch C to prevent the growth of environmental moulds. Sausages were fermented at 22 °C and 90% relative humidity (RH) for 12 h. Afterwards, the temperature and RH were slowly reduced to 18 °C and 80%, respectively, in 60 h. Finally, the sausages were dried at 12 °C and 80% RH until the end of the ripening process (a total of 22 days). At this time, the whole surface of the sausages was uniformly colonised by the inoculated moulds. The results reported here are the mean data obtained with samples from three different manufacturing processes carried out with different ingredients but the same formulation and technology.
Dry matter (DM) was determined by drying the sample at 110 °C to constant weight.
Volatile compounds were analysed by gas chromatography/mass spectrometry (GC/MS) as described by Elmore et al. . Twenty-five grams of each sample were introduced into a glass flask and equilibrated for 30 min at 30 °C. Volatiles were extracted at 30 °C by a nitrogen flow of 40 ml/min for 1 h and adsorbed on a steel trap (105 mm×3-mm inner diameter, i.d.) containing 85 mg Tenax TA (Scientific Glass Engineering, Milton Keynes, UK). A standard of 131 ng 1,2-dichlorobenzene (Sigma) in 1 μl hexane (Panreac) was added to the trap at the end of the collection and excess solvent and any water retained on the trap were removed by purging the trap with nitrogen at 40 ml/min for 5 min.
Analyses were performed with a Hewlett-Packard 5972 mass spectrometer fitted with a HP5890 Series II gas chromatograph and a G1034 Chemstation (Hewlett-Packard, Palo Alto, CA, USA). A CHIS injection port (Scientific Glass Engineering) was used to thermally desorb the volatiles from the Tenax trap onto the front of a CP-Sil 8 CB low bleed/MS fused silica capillary column (60 m×0.25-mm i.d., 0.25-μm film thickness, Chrompack, Middelburg, The Netherlands). During a desorption period of 5 min, volatile compounds were cryofocused by immersing 15 cm of the column adjacent to the heater in a solid CO2 bath while the oven temperature was held at 40 °C. The bath was then removed and chromatography achieved by holding at 40 °C for 2 min followed by a programmed rise to 280 °C at 4 °C/min and holding for 5 min. A series of n-alkanes (C6–C22) (Sigma) were analysed under the same conditions to obtain linear retention index (LRI) values for the aroma components.
The mass spectrometer was operated in electron impact mode with an electron energy of 70 eV and an emission current of 50 μA. Compounds were identified by first comparing their mass spectra with those contained in the HP Wiley 138 Mass Spectral Database and then comparing the LRI values either with those of authentic standards or with published values. Approximate quantities of the volatiles were estimated by comparing their peak areas with those of the 1,2-dichlorobenzene internal standard, obtained from the total ion chromatograms, using a response factor of 1. Analyses were performed by triplicate.
A principal component analysis (PCA) was carried out on the correlation matrix of the volatile compounds. Pearson correlation coefficients between the retained principal components and the volatile compounds were calculated. Only those compounds that showed a significant correlation (p<0.05) with the principal component were considered as being represented in that component.
Results and discussion
Values around 70–75% at the end of the ripening period arose from the dry matter of the different fermented sausages. At this time the values reached by the aw were less than 0.90 in all batches. The final pH of the control batch was 4.6, whereas the inoculated ones reached pH values of 5.1 (M. racemosus and P. aurantiogriseum) and 5.4 (P. camemberti) owing to the production of ammonia as a consequence of proteolytic activity .
Average and standard deviation of the volatile compounds identified in the headspace of four batches of dry fermented sausages after 22 days of ripening. Association between the volatile compounds and the batches according to the principal components analysis (PCA). The linear retention index (LRI) was measured on a CP-Sil 8 CB low bleed/mass spectrometry column
Observed quantities in the batch (ng/100 g)a
PCA results: associated with
Amino acid degradation
A decrease in the concentration of compounds derived from lipid oxidation was observed in sausages inoculated with M. racemosus, P. aurantiogriseum and P. camemberti, among them aliphatic saturated (pentanal, hexanal, heptanal, octanal, nonanal) and unsaturated (2-hexenal, 2-heptenal) aldehydes, 2-methylfuran and 2-pentanone; the latter was only detected in the control batch. These results confirm the antioxidative effect exhibited by moulds because of their catalase activity, oxygen consumption, reduced oxygen penetration through the mycelium and protection against light . This is an important feature in the ripening process as many ketones and aldehydes possess low odour thresholds and they are formed during lipid oxidation .
It is interesting to note that the batch inoculated with P. camemberti showed the highest amount of 2-heptanone. This compound was also the main methylketone detected in dry sausages covered with P. nalgiovense . Molimard and Spinnler  stated that some moulds, like P. camemberti and P. roqueforti, possess an enzymatic system that permits the detoxification of free fatty acids through an oxidation to yield a β-ketoacyl-CoA, the subsequent formation of a β-ketoacid and a final decarboxylation to give a methylketone. β-Ketooctanoyl-CoA is the preferred substrate for the deacylation reaction  which yields 2-heptanone, which may be further reduced to the corresponding secondary alcohol such as occurs in cheese .
Aliphatic eight-carbon alcohols and ketones are very common fungal metabolites which are formed in the degradation of lipids [15, 18]. Only 1-octen-3-ol was detected in all batches, and especially in those superficially inoculated. This alcohol has a low odour threshold (1 μg/l)  and imparts a mushroom quality, being one of the key compounds in the global aromatic quality of Camembert cheese . Batches inoculated with Penicillium spp. showed the highest levels of 1-octen-3-ol, with values approximately twofold and sevenfold higher than those of batches M and C, respectively. The production of 1-octen-3-ol by P. aurantiogriseum and P. camemberti in culture media has been reported by other authors [16, 17].
The superficial inoculation produced an important rise in the volatile compounds derived from amino acids. The branched aldehydes (2-methylpropanal, 2-methybutanal and 3-methybutanal) and their corresponding alcohols (2-methyl-1-propanol, 2-methyl-1-butanol and 3-methyl-1-butanol) showed the highest levels in sausages inoculated with both Penicillium strains. However, M. racemosus was the mould which produced the lowest increase of these compounds. The levels of these volatiles in batches M, PA and PCM were twofold, ninefold and tenfold higher, respectively, than those of batch C. All of them are degradation products of the amino acids valine, isoleucine and leucine. The increase detected in the inoculated batches could be due to the L-amino oxidase activity of the fungal strains, especially of both Penicillium [28, 29], and also to a rise in the levels of certain free amino acids as a result of the proteolytic activity developed on the sausage surface, since it has been reported that all these moulds are endowed with an intense proteolytic activity [2, 19, 30, 31]. As these volatiles have been associated with a “ripened aroma” , the latter explanation could also be in agreement with the results of Geisen , who found a positive correlation between the aroma intensity in mould-ripened sausages and the degree of proteolysis. The branched aldehydes 2-methylpropanal, 2-methylbutanal and 3-methylbutanal can be transformed into their corresponding alcohols, acids and even esters, all of them being compounds of great significance in the final flavour of dry fermented sausages.
The other group of compounds that increased in the superficially inoculated sausages were the esters. They arise from the esterification of carboxylic acids and alcohols. The most abundant esters were the ethyl esters, which are generated from the esterification of ethanol and organic acids by microbial esterases . Ethanol derives mainly from carbohydrate fermentation, by, for example, lactic acid bacteria, which may divert the homolactic pathway at the level of pyruvate, yielding ethanol by the system pyruvate formate lyase . It is remarkable that ethanol was found in higher levels in batches PCM, PA and M, with values 16-, 9- and 4-fold higher, respectively, than those observed in the control batch. Esters have been reported as important volatiles in fermented sausages [6, 35] and they are also present, although in lower levels, in dry cured ham [8, 36] owing to the lower microbial counts found in the latter . They have low odour threshold values, impart fruity qualities  and have been associated, together with branched aldehydes, twith ripened flavour in cured meat products [7, 14, 37].
Pearson correlation coefficients between the volatile compounds and the first three principal components (significant correlations are in bold, i.e. p<0.05)
From the five volatiles that were not significantly correlated with any of the principal components (octanal, methyl hexanoate, 2-methyl pentanoic acid, 2-methylfuran, 2,3-butanedione), only 2-methylfuran showed slightly different values between batches. As it can be observed in Table 1, a greater amount of this compound was detected in the control batch and in that inoculated with M. racemosus.
Table 1 gives the interpretation and validation of these results according to the scores and the correlations between the PCs and the volatile compounds of Table 2. The control batch was mainly associated with the compounds derived from lipid oxidation, mainly aliphatic saturated (pentanal, hexanal, heptanal, nonanal, decanal) and unsaturated (2-hexenal, 2-heptenal, 2-decenal) aldehydes. The batch inoculated with M. racemosus was related to the compounds coming from microbial esterification. The two batches inoculated with Penicillium spp. were related to products of amino acid degradation, although the batch inoculated with P. aurantiogriseum was also characterised by volatile compounds coming from lipid oxidation.
This study shows that it is possible to generate aroma profiles in mould-ripened sausages depending on the fungal strain used. Thus, the superficial inoculation of M. racemosus could be used when ester formation is desired, while P. camemberti and P. aurantiogriseum inoculation would give rise to amino acid degradation compounds. Both procedures would also control the level of lipid autooxidation.
This work was supported by the Comisión Interministerial de Ciencia y Tecnología with the ALI 96-0928 project. J.M.B was a recipient of a grant from Formación de Personal Investigador of the Universidad Complutense de Madrid. We thank Dra. Selgas and Dra. García from the Departmento de Nutrición, Bromatología y Tecnología de los Alimentos (Universidad Complutense de Madrid, Spain) for kindly providing the P. aurantiogriseum and M. racemosus strains used in this work.
- 1.Cook PE (1995) In: Campbell-Platt G, Cook PE (eds) Fermented meats. Blackie Academic & Professional, London, pp 110–129Google Scholar
- 3.Lücke FK (1998) In: Wood BJB (ed) Microbiology of fermented foods. Applied Science Publishers, London, pp 441–483Google Scholar
- 7.Careri M, Mangia A, Barbieri G, Bolzoni L, Virgili R, Parolari G (1993) J Food Sci 58:968–972Google Scholar
- 15.Larsen TO, Frisvad JC (1995) Mycol Res 99:1153–1166Google Scholar
- 18.Jelen H, Wasowicz E (1998) Food Rev Int 14:391–426Google Scholar
- 19.Trigueros G, García ML, Casas C, Ordóñez JA, Selgas MD (1995) Z Lebensm Unters Forsch 201:298–302Google Scholar
- 22.Jessen B (1995) In: Campbell-Platt G, Cook PE (eds) Fermented meats. Blackie Academic & Professional, London, pp 130–159Google Scholar
- 23.Shahidi F, Rubin LJ, D’Souza LA (1986) CRC Crit Rev Food Sci Nutr 24:141–243Google Scholar
- 25.Molimard P, Spinnler HE (1996) J Dairy Sci 79:169–184Google Scholar
- 26.Okumura J, Kinsella JE (1985) J Dairy Sci 68:11–15Google Scholar
- 27.Rychlik M, Schieberle P, Grosch W (1998) Garching: Deutsche Forschungsanstalt für Lebensmittelchemie and Institut für Lebensmittelchemie der Technischen Universität MünchenGoogle Scholar
- 31.Gripon JC (1997) In: Law BA (ed) Microbiology and biochemistry of cheese and fermented milk. Blackie Academic & Professional, London, pp 193–218Google Scholar
- 33.Stahnke LH (1995) Meat Sci 41:193–209Google Scholar
- 36.Buscailhon S, Berdagué JL, Monin G (1993) J Sci Food Agric 63:69–75Google Scholar
- 37.Barbieri G, Bolzoni L, Parolari G, Virgili R (1992) J Agric Food Chem 40:2389–2394Google Scholar