Increased access to high-calorie and fast foods has significantly contributed to higher daily caloric intake and a growing prevalence of metabolic syndrome. According to Bruce and Hanson (2010), metabolic syndrome comprises a constellation of symptoms arising from various cardiometabolic risk factors, including obesity, insulin resistance, dyslipidemia, and hypertension. Epidemiological data show that metabolic syndrome affects 20%–25% of the global population. The Framingham Offspring Study reported a prevalence of 29.4% in men and 23.1% in women aged 26–82 years (Ingelsson et al., 2007). In Indonesia, 23.34% of the population suffers from metabolic syndrome, with a prevalence of 26.2% among men and 21.4% among women (Hadaegh et al., 2013). An important consequence of metabolic syndrome is impairment of the immune system, which renders individuals more vulnerable to disease.

Recent advancements in food production extend beyond merely satisfying nutritional requirements; they also prioritize the health benefits of food for humans, enabling these foods to function as functional foods. Functional foods are foods that have been fortified or enriched to improve their nutritional content, thereby fulfilling the nutritional requirements of the body and conferring positive health effects (Hasler, 2002). Examples of functional foods that offer health benefits include those that contain probiotics.

Probiotics are live microorganisms that are incorporated into food products, either individually or in combination, to enhance digestive health. Probiotics are acknowledged for their health benefits to the host when they are administered in adequate amounts (10⁶–10⁷ CFU/mL) (Food and Agriculture Organization of the United Nations and World Health Organization, 2006). The global probiotic market expanded from USD 79.6 billion in 2024 to USD 86.8 billion in 2025, with projections reaching USD 132.8 billion by 2029 (The Business Research Company, 2024). In Indonesia, the probiotic market is predominantly supplied by imported products, primarily from Europe, Japan, and the United States. This reliance not only elevates retail prices for consumers, but it also poses challenges related to the availability of probiotics and their adaptability to local human gastrointestinal conditions. In contrast, Indonesia possesses a rich microbial biodiversity derived from traditional fermented food products and other local sources. This presents substantial opportunities to develop indigenous probiotics that are not only scientifically competitive, but that also offer adaptive benefits for local consumers.

Notable local probiotics that have been developed to date include Lacticaseibacillus casei AP and Pediococcus acidilactici BE (Widodo and Taufiq, 2017; Widodo et al., 2019; Widodo et al., 2021; Widodo et al., 2023). Lacticaseibacillus is a newly established genus that originated from the reclassification of Lactobacillus (Zheng et al., 2020). Lacticaseibacillus casei AP was isolated and identified from the feces of naturally born, breastfed Indonesian infants aged less than 1 month. This strain was able to acidify milk resulted in a high viscosity (Widodo Tono et al., 2012; Widodo and Taufiq, 2017). Milk fermented with Lacticaseibacillus casei AP has also been reported to be an effective antihypercholesterolemic and antihyperglycemic agent (Widodo et al., 2021). Widodo Septiana et al. (2012) successfully isolated the P. acidilactici strain BE from the same feces samples. Milk fermented with P. acidilactici BE has been shown to reduce blood glucose levels in diabetic rats, consistent with increased insulin production and an increase in both the number and percentage of immunoreactive pancreatic beta cells (Widodo et al., 2023).

Probiotics are typically incorporated into food products such as fermented dairy products, which require storage under constant refrigeration along with the implementation of cold chain technology during distribution to consumers. Consequently, the incorporation of probiotics into liquid products is expensive owing to the high costs associated with refrigeration. In addition, the viability of bacterial cells decreases due to excessive acidification and prolonged storage times (Terpou et al., 2019). Therefore, innovations in the development of probiotic powders are important alternatives to liquid formulations (Ferdousi et al., 2013).

Dehydration represents a preservation technique that reduces the water content and activity (Aw). A low Aw value (i.e., <0.6) is essential to avoid the growth of spoilage microorganisms and to extend the shelf life of food products. For example, spray drying is a dehydration technology that removes moisture from food products and decreases their Aw values. However, the heat treatment employed during spray-drying can decrease the viability of probiotic cells and adversely affect their metabolic activities (Soukoulis et al., 2014). Spray drying involves various parameters, including the inlet air temperature, outlet air temperature, air flow rate, product feed rate, and atomized droplet size (Santivarangkna et al., 2008). Among these parameters, the outlet air temperature affects the viability of spray-dried probiotic cultures the most (Santivarangkna et al., 2008). Low outlet temperatures have been reported to increase productivity (Felfoul et al., 2022) and maintain bacterial viability during spray drying (Ananta et al., 2005; Desmond et al., 2002). Habtegebriel et al. (2018) reported that spray drying at an outlet temperature range of 60 °C–80 °C caused only limited denaturation of milk. In our previous study, we optimized the spray-drying process by setting the inlet air temperature to 160 °C and conditioning the feed pump to maintain an outlet air temperature of 67 °C (unpublished data). This parameter was identified as the most suitable for scaling up production because of its high productivity, short processing time, and ability to maintain both physicochemical quality and microbial viability. The response of probiotics to heat stress during spray drying with an outlet air temperature at 67 °C is therefore important to investigate. In this study, bacterial viability and metabolic profiles under heat treatment at 67 °C were compared with those under normal growth conditions at 37 °C, as well as with those under heat treatment at 55 °C, which was intended to result in 50% bacterial cell injury according to a previous study (Irie et al., 2014).

The metabolic profiles of probiotics reflect their metabolic activity and physiological status under specific conditions. Variations in these profiles can also serve as significant indicators of how probiotics respond and adapt to heat stress. For example, the levels of metabolites, such as glutamate, mannitol, and ethanol, may increase in response to oxidative stress and disruptions in energy metabolism. Using metabolomic techniques based on gas chromatography–mass spectrometry (GC-MS) and liquid chromatography–mass spectrometry (LC-MS), changes in the concentrations of these metabolites have been precisely detected and compared before and after heat treatment (Zhang et al., 2022; Li et al., 2025). Overall, metabolic profiling can serve as a diagnostic tool for evaluating cellular adaptation to heat stress and indicating physiological resilience. The resulting findings can be used to select probiotic cultures with enhanced metabolic stability and to design fermentation media or process conditions that are optimal for maintaining cell viability and the bioactive functions of probiotics (Timilsena et al., 2020). However, the resistance of Lacticaseibacillus casei AP and P. acidilactici BE to heat stress and their resulting metabolic adaptations remain unknown. Consequently, the aim of the current study is to evaluate the effects of heat stress on the viability and metabolic profiles of Lacticaseibacillus casei AP and P. acidilactici BE.

Heat stress is a critical environmental challenge for bacterial cells during spray drying, since temperatures exceeding 40 °C can disrupt the integrity of cellular macromolecules, including proteins and lipids, within the cell membrane. Table 1 shows the viability of the Lacticaseibacillus casei AP and P. acidilactici BE cells after heat treatment under different conditions. As detailed in the table, both Lacticaseibacillus casei AP and P. acidilactici BE cultures experienced a decrease in cell viability after growth at elevated temperatures. Specifically, Lacticaseibacillus casei AP experienced 7.5% and 11.9% decreases at 55 °C and 67 °C, respectively. In contrast, P. acidilactici BE experienced 11.1% and 16.1% decreases at the same temperatures. Both cultures exhibited similar bacterial viabilities prior to heat treatment when both strains were incubated at 37 °C (Table 1).

The metabolic profiles of Lacticaseibacillus casei AP and P. acidilactici BE were analyzed using a heatmap and are presented in Figure 1. Specifically, the metabolic profiles are shown for these cultures after heat treatment at 37, 55, and 67 °C. In this analysis, dark red indicates a high concentration of metabolites, whereas dark blue indicates a low concentration. The metabolic profiles obtained at 37 °C indicate that Lacticaseibacillus casei AP and P. acidilactici BE produce primary metabolites such as isoleucine, hypoxanthine, 4-oxoproline, cyclohexanol, campechic acid, gingerol, galaxolidone, and malevamide D (Figure 1). These observations suggest that these metabolites are closely associated with essential metabolic processes, particularly amino acid and purine metabolism (Kim et al., 2024; Aoki et al., 2025).

Compared with the profile obtained at 37 °C, the metabolic profile recorded at 55 °C was substantially altered in both strains (Figure 1). Metabolites that were present at high concentrations at 37 °C underwent a marked reduction, with decreases in the concentrations of cyclohexanol, campechic acid, and gingerol being particularly pronounced, as evidenced by the predominance of blue coloration. These reductions suggest that the bacterial cells encountered thermal stress, which resulted in modifications to their metabolic pathways, reduced enzymatic activity, or adjustments in metabolism as an adaptive response to elevated temperature conditions (Castaldo et al., 2006; Liu et al., 2021).

The metabolic profile observed at 67 °C exhibited significant alterations, notably in the levels of nucleotide metabolites such as cytidine 5′-monophosphate and uridine monophosphate, as well as lipid compounds such as erucamide, certonardosterol, N,N-bis(2-hydroxyethyl), and additional compounds such as 28-epimutalomycin and phenol dodecyl (Figure 1). This increase in nucleotide metabolites is intricately linked with bacterial responses to heat-induced DNA stress, including mechanisms like DNA repair and the synthesis of new RNA as a protective response (Qian et al., 2014).

Multivariate analysis was subsequently performed using PLS-DA to compare the metabolic profiles of Lacticaseibacillus casei AP and P. acidilactici BE after heat treatment. PLS-DA is a statistical method designed to differentiate sample groups based on their metabolic patterns and identify metabolites that significantly contribute to these distinctions. The outcomes of the PLS-DA analysis are shown in Figure 2, wherein the effects of the different heat treatment temperatures on the metabolic patterns are clearly illustrated. The PLS-DA results can be divided into two components, namely component 1 and component 2. In the principal component analysis (PCA) biplot depicted in Figure 2A, a distinct separation between strains is evident along PC1, which accounts for 27.9% of the variance, with an additional contribution from PC2 at 45.1%. Replicates of P. acidilactici BE (represented by green dots) are clustered predominantly on the right side of the plot, mainly within the upper right quadrant, and exhibit a positive correlation with several features that load strongly in the same direction, including citric acid, tri(2-ethylhexyl) ester, N-(3-aminopropyl) hexadecanamide, and valine. In contrast, replicates of Lacticaseibacillus casei AP (indicated by red dots) are clustered on the lower left side and are associated with features exhibiting negative loadings on PC1/PC2, such as guanosine-5′-monophosphate (GMP), rhabdopeptide-1, and several other minor components. This pattern suggests that the variation in metabolites projected onto PC1 is the primary driver of discrimination between strains, with P. acidilactici BE demonstrating relatively high levels of certain amino/amide compounds and esters, whereas Lacticaseibacillus casei AP is more closely associated with specific nucleotides and peptides. The broader distribution of P. acidilactici BE points along the PC2 axis indicates slightly greater metabolic heterogeneity in this strain than in Lacticaseibacillus casei, yet both groups remain consistently separated, which confirms the presence of distinct metabolic signatures for each strain.

Additionally, PLS-DA clearly revealed differences in the metabolic profiles at the three treatment temperatures. As shown in Figure 2B, area_AP37 and area_AP67 are closely positioned in the upper-left quadrant of the plot, whereas area_AP55 is notably displaced downward, albeit still within the AP-class ellipse. The distinct positioning of area_AP55 suggests that at 55 °C, the Lacticaseibacillus casei AP strain experiences a marked alteration in metabolite expression, differing from the profiles observed at 37 °C and 67 °C. This implies that a temperature of 55 °C represents a critical physiological threshold, potentially triggering stress responses, such as the synthesis of protective metabolites or metabolic adjustments. The metabolic profiles of P. acidilactici BE cultures were tightly clustered in the upper right quadrant of the plot, indicating clear homogeneity, although area_BE55 was positioned slightly lower and remained within the same cluster, indicating resilience to heat stress. Compared with the Lacticaseibacillus casei AP cultures, the P. acidilactici BE cultures exhibit greater metabolic stability across all temperatures, including at 55 °C. These results imply that P. acidilactici BE potentially sustains a consistent core metabolic function when it is subjected to heat stress. Nevertheless, distinct protective strategies or alterations in metabolite biosynthesis under these conditions seem to be lacking.

Multivariate analysis was performed to evaluate the metabolic profiles after subjecting the cultures to various heat-stress temperatures. The results were evaluated using PCA, as presented in Figure 3. Notably, PCA reduces the dimensionality of a large dataset into several principal components (PCs) that encapsulate the most significant variations within the data. The first principal component (PC1) and second principal component (PC2) explained 61.4% and 14.1% of the total data variance, respectively. The colored points on the plot correspond to the different temperature treatments, wherein red represents 37 °C, green represents 55 °C, and blue represents 67 °C. The observed shifts in the positions of these points within the plot indicates notable differences in the metabolic profiles across the temperature spectrum, wherein greater separation between points represents more pronounced differences. PCA modeling indicated separation driven by temperature on PC1 and strain-specific modulation on PC2 (variance proportions are displayed on the axes of Figure 3). In essence, the score plot is depicted in Figure 4, while Figure 3 clarifies the link between score separation and the direction and magnitude of metabolite contributions (loading), with the 95% confidence ellipses for each group assisting in visualizing the consistency of replicates and the proximity of clusters.

Figure 4 provides a visual representation of the sample distribution based on the scores or projection values for PC1 (61.4%) and PC2 (14.1%). Unlike a biplot that includes metabolite vectors, a score plot focuses solely on the positioning of each sample, thereby accentuating separation and grouping according to the treatment conditions. Each point on the plot corresponds to an individual sample; area_AP37B corresponds to Lacticaseibacillus casei AP at 37 °C, area_BE37B denotes P. acidilactici BE at 37 °C, area_AP55B represents Lacticaseibacillus casei AP at 55 °C, area_BE55B indicates P. acidilactici BE at 55 °C, area_AP67B refers to Lacticaseibacillus casei AP at 67 °C, and area_BE67B pertains to P. acidilactici BE at 67 °C. Primary separation is observed along PC1 (61.4%), which is predominantly influenced by temperature. Both strains exhibit a marked shift to the right (positive PC1) at 55 °C, with Lacticaseibacillus casei AP and P. acidilactici BE demonstrating the most significant alterations in metabolite profiles compared with the conditions at 37 °C and 67 °C. In contrast, the conditions at 37 °C and 67 °C tend to cluster on the left (negative PC1), with Lacticaseibacillus casei AP at both temperatures grouped in the left quadrant and P. acidilactici BE at 37 °C and 67 °C positioned within the lower PC1 range. The secondary separation along PC2 (14.1%) highlights differences in the strain responses at specific temperatures. At 37 °C, the abundance of Lacticaseibacillus casei AP is greater (positive PC2) than that of P. acidilactici BE (negative PC2), indicating that distinct metabolite components differentiate the two strains under basal conditions. Conversely, at 55 °C, compared with Lacticaseibacillus casei AP, P. acidilactici BE showed a slightly higher PC2 score, suggesting a strain–temperature interaction in the metabolite response patterns. At 67 °C, both strains are positioned near the PC2 axis (approximately zero to slightly negative), indicating reduced interstrain variation compared with that at 37 °C and 55 °C. Overall, this pattern confirms that temperature, particularly 55 °C, is the primary driver of metabolomic variation (PC1), while strain-specific differences (PC2) modulate responses at certain temperatures. These findings align with the results of the present study, which revealed distinct metabolite signatures for each strain under heat stress.

Lacticaseibacillus casei AP and P. acidilactici BE exhibited similar bacterial viabilities prior to heat treatment when both strains were incubated at 37 °C (Table 1). This observation confirmed the equivalent growth rates of both cultures under standard temperature conditions, ensuring that any subsequent variations could be clearly attributed to heat treatment (Tripathi and Giri, 2014). At 55 °C, the Lacticaseibacillus casei AP cells exhibited significantly greater viability than the P. acidilactici BE cells. Increasing the temperature to 67 °C revealed a clearer difference in viability between the two strains, with the P. acidilactici BE culture experiencing a substantial growth decline relative to Lacticaseibacillus casei AP. These findings suggest that Lacticaseibacillus casei AP cells are more tolerant than P. acidilactici BE cells when exposed to temperatures of 55 °C and 67 °C. The differences in heat stress tolerance among the bacterial cultures can be partially attributed to the proportion of saturated fatty acids present in the cell membrane. Previous studies by Adu et al. (2018) and Zhang et al. (2021) indicated that Lacticaseibacillus casei cultures possess a relatively high concentration of saturated fatty acids, which contributes to membrane stability at moderate temperatures. In contrast, P. acidilactici requires external protection to preserve its membrane fluidity (Adu et al., 2018; Zhang et al., 2021).

Heatmap analysis represents a comprehensive method for assessing variations in metabolic expression patterns across different samples or treatments. Specifically, a heatmap visually represents the relative intensities of various metabolites identified in research samples, with red indicating a high level of metabolite expression and blue indicating a low level of expression (Goodacre et al., 2004). Such analysis offers a clear visual depiction of sample clustering based on similarities in the metabolic profiles, thereby facilitating the interpretation of the physiological changes resulting from heat treatment (Liu et al., 2025). Comparative analysis of the metabolic profiles of the Lacticaseibacillus casei AP and P. acidilactici BE strains revealed distinct metabolic responses to heat stress in each strain (Figure 1). The heatmap shows that compared with P. acidilactici BE, Lacticaseibacillus casei AP exhibits a more robust response, particularly at 67 °C. This enhanced adaptation in the cells of Lacticaseibacillus casei AP was evidenced by significant alterations in the metabolite intensity, including those related to nucleotides and membrane lipids, whereas P. acidilactici BE demonstrated comparatively moderate changes in the metabolite intensity. According to a previous study by Adu et al. (2018), Lacticaseibacillus casei possesses an effective protective mechanism against heat stress, such as the elevated expression of heat shock proteins. In contrast, P. acidilactici is less heat tolerant (Jonathan et al., 2023).

Heatmap analysis of the metabolic profiles revealed that exposure to high-temperature stress induced substantial alterations in the metabolic profiles of both probiotic strains. The observed responses included increases in the levels of nucleotide metabolites, membrane lipids, and other protective compounds that function as mechanisms for cellular adaptation. Nucleotides such as cytidine 5′-monophosphate and uridine monophosphate generally serve to safeguard bacterial cells and facilitate adaptation to heat stress by regulating secondary metabolism and synthesizing the nucleotides necessary for the repair and protection of critical molecular structures within the cell (Zhen et al., 2020). Moreover, an increased lipid intensity (e.g., for erucamide and cerotanordesterol) suggests that the bacteria actively modify their cell membrane structure to maintain membrane fluidity, which is crucial for preserving cellular integrity under high-temperature conditions (Fonseca et al., 2019; Siroli et al., 2020). Consequently, augmentation of these metabolites is part of the intricate adaptive mechanisms employed by bacteria in response to severe heat stress. Compared with P. acidilactici BE, Lacticaseibacillus casei AP demonstrated a more pronounced metabolic response, suggesting significant differences in the metabolic strategies and adaptation mechanisms employed by each strain. As reported by Liao et al. (2010), heat stress compels probiotic bacteria to modify their metabolism by enhancing the synthesis of secondary metabolites, such as nucleotides and lipid compounds, which ultimately preserves membrane integrity and protects the cellular structure against heat stress-induced damage.

Using PLS-DA analysis, different dominant metabolites were observed between the two probiotic strains. Specifically, citric acid, tri (2-ethylhexyl)ester, N-(3-aminopropyl)hexadecanamide, and valine were detected as metabolites in the P. acidilactici BE culture, whereas pentaoxahexacosan-1-ol, rhabdopeptide 1, guanosine-5′-monophosphate (5ʹ-GMP), and molybdenite were detected in the P. acidilactici BE culture. Citric acid and valine represent critical metabolites in the tricarboxylic acid cycle and protein biosynthesis pathways, suggesting that P. acidilactici BE produces more energy through primary metabolism in response to osmotic stress conditions under increased growth temperatures (Cronan, and Laporte, 2005; Ye et al., 2012; Akram, 2014). Moreover, 5ʹ-GMP and peptides (e.g., rhabdopeptides), which are linked to metabolic regulation and antimicrobial activity, are likely involved in the adaptive responses of Lacticaseibacillus casei AP to elevated temperatures (Rossi et al., 2016). It is also known that monogalactosyl diacylglycerols, including galactolipid derivatives such as 1,2-diacyl-3-alpha-glucopyranosyl-sn-glycerol, play a significant role in preserving membrane integrity during heat stress. Additionally, elevated levels of sterol compounds in Lacticaseibacillus casei have been reported to fortify membranes and mitigate ion loss at extreme temperatures (Papadimitriou et al., 2016). Sterol compounds increase membrane stability and protect lipids from maintaining cellular homeostasis. Moreover, in Lactobacillus casei, the ratio of saturated to unsaturated fatty acids increases at extreme temperatures, which helps maintain membrane fluidity and stability (Machado et al., 2004). The metabolism of citric acid in lactic acid bacteria is altered under stress conditions. Citric acid plays a role in mitigating oxidative damage commonly associated with heat stress and contributes to the stabilization of proteins at elevated temperatures (Książek, 2024). N-(3-aminopropyl) hexadecanamide, a fatty acid derivative, is recognized for its ability to modulate membrane properties in response to stress. Long-chain amides can physically influence the membrane, thereby reducing permeability at high temperatures (Bandi et al., 2025). Rhabdopeptide-1 functions as a protease inhibitor, limiting excessive proteolysis and modulating the microbial community when heat stress induces protein lysis (Reimer et al., 2013). MGDG undergoes dynamic changes under heat stress through lipid remodeling mechanisms to preserve membrane fluidity and integrity as temperatures increase. Lactic acid bacteria adapt their membrane lipids in response to heat stress (Walczak-Skierska et al., 2020). Nucleotides, such as 5ʹ-GMP, serve as precursors for (p)ppGpp alarmones involved in the stringent response. The levels of (p)ppGpp increase rapidly during heat shock, aiding bacteria in adapting to the proteotoxicity induced by heat stress (Schäfer et al., 2020). Valine acts as a precursor for branched-chain fatty acids (BCFAs), and an increased proportion of BCFAs stabilizes the membrane and reduces permeability at high temperatures, a process known as homeoviscous adaptation, thereby increasing heat tolerance. Alterations in membrane fatty acid composition represent a response of lactic acid bacteria to heat stress (Walczak-Skierska et al., 2020).

A previous study reported that Lacticaseibacillus casei employs multiple strategies to counteract heat stress, such as increasing the expression of heat shock proteins, modifying membrane lipids, and activating purine metabolic pathways (Desmond et al., 2004; Adu et al., 2018). Modifications in the membrane lipid composition contribute to the maintenance of membrane stability at elevated temperatures. Specifically, heat stress induces changes in the lipid composition, particularly in the synthesis of fatty acids, which promote membrane stability and protect against thermal damage. Moreover, Jonathan et al. (2023) reported that P. acidilactici maintains substantial cell viability under heat stress conditions up to 60 °C, with marked reductions being observed at 75 °C and 90 °C. However, the expression of heat shock protein genes, such as groEL, did not change significantly under heat stress conditions.

The vectors depicted by arrows in the PCA-based biplot shown in Figure 3 illustrate the direction and magnitude of the influence of each specific metabolite on the bacterial metabolic profile. The length of each vector corresponds to the strength of the contribution of the specific metabolite (PC), with longer vectors indicating a more substantial contributions. The predominant metabolites in the negative PC1 included certonardosterol M, N-(3-aminopropyl)hexadecanamide, citric acid tri(2-ethylhexyl) ester acetate, and monogalactosyl diacylglycerol (MGDG). These metabolites were more frequently observed at temperatures corresponding to the negative side of PC1. Conversely, the dominant metabolites in the positive PC2 included methoxy-2-propanediyl bis(octylcarbamate).1, cyclo-(L-ala-trans-4-hydroxy-L-pro), cyclooctene, 1,4-bis[(2-ethylhexyl)oxy]-1,4-dioxobutane-2-sulfonic acid, and 3,6,9,12,15-pentaoxahexacosan-1-ol. These findings suggest that the above compounds play a significant role in the bacterial response to specific temperatures and that their levels tend to increase at increasing temperatures. The metabolite ornithine, which aligns with the negative PC2, also significantly contributes to the variation in the data and is associated with metabolic responses to elevated temperatures. The findings of this analysis revealed substantial differences in the metabolic profiles produced by bacteria at 37, 55, and 67 °C. Specifically, each temperature elicited a distinct set of metabolites in response to heat stress adaptation, indicating that heat stress induced specific alterations in their metabolic pathways, as evidenced by the changes in the metabolic profiles. As reported previously, the metabolic profiles of various probiotics are significantly affected by temperature (Papadimitriou et al., 2016). For example, MGDG serves as a crucial membrane modulator that facilitates the maintenance of lipid fluidity at elevated temperatures, consistent with the adaptation mechanisms observed in thermotolerant organisms (Sun et al., 2022).

Figure 4 shows a PCA-based score plot that captures the changes in metabolite levels across a range of temperatures. The samples subjected to heat treatment at 55 °C (area_AP55B and area_BE55B) formed a distinct cluster in the upper right quadrant of Figure 4. These observations suggest that compared with exposure to temperatures of 37 °C and 67 °C, exposure at 55 °C elicits markedly different metabolite responses in both Lacticaseibacillus casei AP and P. acidilactici BE. This temperature can therefore be classified as the active metabolite stress zone. Conversely, the samples exposed to a temperature of 67 °C (area_AP67B and area_BE67B) were positioned in the lower left quadrant, indicating a unique metabolite pattern associated with high heat stress, distinct from that observed at 37 °C. This finding implies the existence of an emergency metabolic response, potentially involving increases in the levels of protective lipids or adaptive molecules, such as MGDG, sterols, and lipid amines. At 37 °C, the samples (area_AP37B and area_BE37B) were somewhat separated but remained within the basal or non-stress metabolite zone, characterized by a metabolic profile that was indicative of basic cellular metabolism. The results of this score plot analysis reveal that metabolite responses vary with temperature and culture conditions; heat stress at 55 °C can be considered a transitional temperature that induces the active expression of protective metabolites, whereas heat stress at 67 °C activates specific metabolic pathways distinct from those at 37 °C and 55 °C, leading to the activation of protective metabolites.

Under heat stress conditions, there is a notable increase in the levels of ether compounds, which serve as protectors of cell membranes by reducing the surface tension between the lipid molecules within the membrane (Gianni de Carvalho et al., 2019; Yang et al., 2021). Similarly, the levels of sterol compounds increased significantly at elevated temperatures, suggesting that enhanced activation of the sterol biosynthesis pathway plays a role in fortifying the membrane structure (Lee et al., 2018; Abedin et al., 2023). Additionally, the secondary metabolites eucaramide and campechic acid contribute to oxidative homeostasis and function as antioxidants, thereby providing essential biochemical protective responses that enable cell survival under high-temperature conditions (Ou et al., 2012; Fadlillah et al., 2021). Collectively, the interplay between lipid metabolites, sterols, and heat shock protein expression can serve as a metabolic marker to assess the quality and resilience of probiotics during high-temperature processing or in response to heat stress.

In conclusion, this study advances our understanding of how two probiotic strains, namely Lacticaseibacillus casei AP and P. acidilactici BE, respond metabolically to heat stress. This work highlights not only the distinct adaptive capabilities of these strains at relatively high temperatures, but also the essential compounds that play a role in cellular defense mechanisms. The insights gained from this investigation lay crucial groundwork for developing probiotic applications that can withstand extreme environmental conditions. Several key conclusions could be drawn from the obtained results. Specifically, compared with P. acidilactici BE, Lacticaseibacillus casei AP demonstrates superior viability at elevated temperatures (i.e., 55 °C and 67 °C), implying that a more robust thermal stress protection mechanism exists in Lacticaseibacillus casei AP. Compounds such as citric acid, tri(2-ethylhexyl) ester, N-(3-aminopropyl) hexadecanamide, and valine in P. acidilactici BE, as well as 5ʹ-GMP, rhabdopeptide1, and 1,2-diacyl-3-α-glucopyranosyl-sn-glycerol (a monogalactosyl diacylglycerol derivative) in Lacticaseibacillus casei AP, may act as metabolic markers that are indicative of responses to increased temperatures. Moreover, the presence of sterol compounds under heat stress conditions suggests that thermal stress triggers changes in the lipid composition of the cell membrane as a primary adaptive response in P. acidilactici BE cells, whereas the synthesis of heat shock proteins in Lacticaseibacillus casei AP supports cellular functions during stress. In the industrial application of heat marker metabolites, compounds such as 5′-GMP, rhabdopeptide-1, and MGDG derivatives for Lacticaseibacillus casei AP, as well as valine and N-(3-aminopropyl) hexadecanamide for P. acidilactici BE, serve as distinctive markers for the efficient identification of robust strains via LC-HRMS. Additionally, the systematic observation of these metabolites during production enables the fine-tuning of fermentation and drying parameters, including the regulation of inlet and outlet temperatures, air flow rates, and the proportion of protectants from proteins or polysaccharides, with the objective of optimizing the viability and stability of the end product.