Mold spoilage of dairy products is a major concern, which not only causes substantial food waste, economic losses and even brand image deterioration, but also becomes a public health concern due to mycotoxin production in some cases. Due to the growing concern with food safety issues and the demand for “clean label” food products from consumers, using lactic acid bacteria (LAB) bioprotective cultures and their metabolites as potential alternatives to chemical preservatives in dairy products has attracted much attention in recent years. Some compounds produced by LAB have been reported as having antifungal activities against certain fungi, and some competition for nutrients may also play a role in the antifungal activity of LAB. In the present PhD project, we aimed to investigate the aspects of the susceptibility of a panel of spoilage molds (Chapter 6) isolated from freshly fermented dairy products towards twelve potential bioprotective LAB cultures (Chapter 7) and the selected metabolites (Chapter 9).
The growth potential of a panel of spoilage molds (four Mucor and nine Penicillium strains) isolated from freshly fermented dairy products was investigated, Mucor strains grew faster than Penicillium strains (Paper Ⅰ). Fungal cell growth and morphology in malt extract broth (MEB) were monitored using oCelloScope at 25 ℃ for 24 h. Mucor plumbeus (M. plumbeus) 01180036 was the fastest-growing isolate, while Penicillium roqueforti (P. roqueforti) ISI4 was the slowest. On yoghurt-agar plates, all molds were able to grow at 5, 16, and 25 °C, and the growth potential was in a temperaturedependent manner. Although P. roqueforti ISI4 grew slowly initially, results in both tests revealed a later fast increase in the growth of this mold compared to the other Penicillium strains.
The inhibitory effects of the twelve LAB cultures on these molds varied based on a high-throughput overlay method. All the tested molds except P. roqueforti ISI4 was strongly inhibited by the 12 LAB cultures (Paper I). In particular, the minimal inhibitory inoculum levels and the inhibition zone of the 12 cultures against the Penicillium strains were determined to compare the inhibitory effects of these cultures. Lacticaseibacillus rhamnosus (L. rhamnosus) LRH01, Lactiplantibacillus plantarum (L. plantarum) LP01, L. plantarum LP37 and Lentilactobacillus parabuchneri (L. parabuchneri) LPB04 exhibited more potent antifungal activity, whereas Lacticaseibacillus paracasei (L. paracasei) LPC46 was less active (Manuscript I). The sensitivity of the 13 molds to LAB cell-containing fermentates varied although inhibition was observed. However, if the cells were removed, the inhibitory effects of LAB cell-free fermentates markedly decreased. The antifungal activity of volatiles produced by LAB cultures was very modest (Paper I).
As one of the most potent cultures, Lactiplantibacillus plantarum (L. plantarum) LP37 was chosen as the main culture to combine with other eleven LAB cultures for investigation of the antifungal activity of binary combinations in yoghurt serum, which was used to simulate yoghurt. The combination of L. plantarum LP37 and L. plantarum LP48 significantly improved the antifungal activity against the growth of several molds beyond that of single cultures (Paper I).
L. rhamnosus LRH01, L. plantarum LP01, and L. parabuchneri LPB04 were employed to screen for synergistic effects (Manuscript I). In yoghurt, the two binary combinations composed of L. plantarum LP01 and either L. rhamnosus LRH01 (combination A) or L. parabuchneri LPB04 (combination B) interacted synergistically with regards to antifungal activity. Synergistic interactions between the cell-containing or cell-free fermentates in combination A and B were observed. The volatiles produced by combination A exhibited a more potent antifungal effect than combination B (Manuscript I). A challenge test was performed using the selected binary combinations (the combination of L. plantarum LP37 and L. plantarum LP48, and combination A) as the bioprotective adjunct cultures in yoghurt production. Unlike the results in laboratory media and commercial yoghurt, very weak inhibition of mold growth was observed in the yoghurt prepared with the selected cultures at 107 CFU/mL (Manuscript I).
Competitive exclusion through manganese depletion seems to play a key role in mold inhibition. The addition of manganese with increasing concentrations of up to 0.1 mM resulted in partly or fully restored growth of six Penicillium and two Mucor strains in yoghurt. Interestingly, P. roqueforti ISI4 was almost non-affected by either L. plantarum LP37 or manganese (Paper I). Similar results were observed with other LAB species, including L. rhamnosus LRH01, L. paracasei LPC44 and L. parabuchneri LPB04 (Manuscript I).
The susceptibility of these dairy-associated molds towards some selected compounds (Chapter 9) produced by LAB was evaluated (Paper II). Diacetyl exhibited the most potent inhibitory effect. Penicillium commune (P. commune) was more sensitive to these compounds, whereas the two Mucor strains generally exhibited higher tolerance. The growth of P. commune 01180002 and P. crustosum 01180001 in yoghurt serum (pH 4.6) was inhibited by diacetyl in a dose- or temperature-dependent manner. In order to control the growth of tolerant molds, P. roqueforti ISI4 and M. circinelloides 01180023, the inhibitory effects of binary combinations of the selected compounds were tested. Two binary combinations composed of octanoic acid and either diacetyl (COD) or 3-Phenylpropanoic acid (3-PPA, COP) displayed a synergistic inhibitory effect (FICI=0.5) on the two tolerant molds (Manuscript II).
The underlying mechanism behind the antifungal activity of the compounds seems to be associated with plasma membrane damage and reactive oxygen species (ROS) accumulation. The observed morphological changes, cellular materials leakage, and propidium iodide (PI) stained-spores demonstrated that the plasma membrane was damaged by diacetyl. The accumulation of ROS indicated that diacetyl induced oxidative stress (Paper II). As expected, the two combinations (C_OD or C_OP) potentiated the membrane disruption, ROS generation and lipid peroxidation, and thereby led to severe cellular dysfunction and eventually cell death (Manuscript II).
These findings clarify the growth potential and variations in susceptibility of a panel of spoilage molds. They highlight the possibility of using selected LAB cultures and their microbial metabolites as an effective treatment strategy to control spoilage isolates and provide a preliminary insight in some of the mechanisms behind the antifungal activity of cultures and metabolites. In terms of practical application, further research is needed employing the actual processed products on a larger scale in order to understand all the factors of importance for an efficient inhibitory action. Regarding the antifungal mechanisms behind the antifungal LAB cultures, more investigations such as transcriptomic analysis of bioprotective cultures as well as target organisms could be carried out in the future.