Study of the thermo-rheological and biochemical parameters of Afuega'l Pitu PDO cheese

Resumo

The great cheese heritage in Spain is reflected in the production of more than 100 varieties of traditional cheeses, with different flavours, textures and shapes. Currently, there are 28 varieties, recognized by the European Union, with Protected Designation of Origin (PDO) and Protected Geographical Indication (PGI). In Asturias (northern Spain), which covers an area of 2% of the total area of Spain, the livestock tradition together with its orographic and climatic conditions is reflected in the production of more than 20 traditional cheeses. Cow’s milk is used in the manufacture of more than 65% of these cheeses, although there are varieties that are produced with sheeps’, goats’ milk and/or their mixture. Afuega’l Pitu cheese, which has been named PDO since 2004, is a predominantly acid coagulated traditional Asturian cheese. It is made by adding starter cultures and a small amount of rennet to cow’s milk (Frisona and Asturiana de los Valles breeds and their crosses), in several municipalities of Asturias. The name Afuega’l Pitu (“to choke the throat” in Asturian dialect) has an uncertain origin and numerous interpretations of this unusual name have been proposed. Among these, the name might refer to the occasional difficulties that can be encountered when swallowing the cheese, due to its peculiar compact and firm texture. Afuega’l Pitu cheese with PDO is made under four traditional varieties that differ in: (1) the use of moulds with a truncated cone shape (atroncau) or a gauze tied at its ends giving the shape of a courgette (trapu); and (2) the absence (blancu) or presence of paprika (roxu). Thus, the following mentions are obtained: atroncau blancu, atroncau roxu, trapu blancu and trapu roxu. In addition, there are technological differences such as the kneading of the curd, typical of the trapu variety and the cheeses with paprika, i.e., trapu blancu, trapu roxu and atroncau roxu. The weight of the cheeses varies between 200 and 600 grams, and it is consumed both fresh and ripened. Afuega’l Pitu cheese is highly appreciated for its sensory properties, which according to the product specifications has a slightly sour taste, not (or slightly) salty, creamy and quite dry. On the other hand, the addition of paprika provides a stronger and spicier flavour. Cheese has a mild aroma, which becomes more intense as it ripens. It is during the ripening process when numerous biochemical reactions take places, such as protein degradation and consequent rehydration of the cheese network, notably influencing the physical properties. Proteolysis involves the degradation of the casein matrix to a variety of peptides and amino acids that have a direct and indirect role in the formation of the cheese texture and flavour. In the xi case of acid curd cheeses, this would imply an increase in protein-water interactions, which affects the firmness and melting properties of the cheese. Ripening also involves other biochemical changes, such as lipolysis, the metabolism of residual lactose, lactate and citrate, and the formation of volatile compounds, which are of great importance for the development of aroma. Therefore, ripening is a complex process that involves microbiological, biochemical and viscoelastic changes, giving rise to the characteristic flavour and texture of each particular variety of cheese. In acid-curd cheeses, coagulation occurs when the pH drops to the isoelectric point (pI) of casein (pH 4.6), which destabilizes casein micelles and solubilized colloidal calcium phosphate (CCP). Therefore, casein particles are held together primarily by physical protein-protein interactions, forming a three-dimensional gel network. This physical gel represents a highly demineralized curd. Typically, these types of structures, when deformed, tend to fracture rather than flow when deformed. It is worth mentioning that some denatured whey proteins, due to previous heat treatments of the milk, could be associated with the casein micelle, contributing to the strengthening of the curd. Therefore, predominantly acid coagulated cheeses have a highly packed protein matrix (continuous phase) that mainly occludes fat and serum (dispersed phases), resulting in particulate compounds with viscoelastic properties. The type of interactions between caseins and the amount of associated Ca are decisive factors in the physical properties of the cheese, such as meltability/stretchability and body/texture. The characteristic low levels of pH and CCP reached in acid cheeses elaboration increase the interactions casein-casein, resulting in cheeses with limited functional properties of melting and stretching. The cheese microstructure is closely related to the final characteristics, flavour and/or texture, which results in complex rheological behaviour. The rheological properties of cheese are obtained after the application of stress (cutting, shearing and chewing) or deformations (compression, extension and shear). Furthermore, stress can also be the result of heat-induced physicochemical and microstructural changes (fat liquefaction, protein dehydration, fat coalescence and matrix collapse). Understanding the microstructure of cheese is not only important for creating a consumer-desirable texture, but also for creating an environment conducive to flavour perception. Rheology, therefore, has an essential role in the cheese industry, for example in processes with pumps, mixers, homogenizers, etc.; in the evaluation of the functionality of each ingredient, quality controls and food texture studies. Small amplitude oscillatory shear (SAOS) deformations have been used by observing the dependence of stress with frequency in the linear viscoelastic (LVE) range. These data provide structural information related to the viscoelastic properties of foods in the LVE region, through parameters such as elastic or storage modulus (G'), viscous or loss modulus (G''), complex modulus (G') and the loss factor (tanδ = G''/G'). However, as explained above, in most processing operations, ingredients or final products can undergo large and sudden deformations that can xii reversibly or irreversibly change the structure. For that reason, large amplitude oscillatory shear (LAOS) experiments have been relevant in the last decade, as they provide useful information on microstructural responses in complex gels such as acid-coagulated cheese, that cannot be obtained by conventional rheometry. In the LVE region, the material response to a small sinusoidal strain/stress is a linearly proportional sinusoidal stress/strain. Only one harmonic is studied in the LVE region: the viscoelastic modulus G', G'' and/or compliance J' (storage compliance) and J'' (loss compliance), which correspond to G'1, G''1 and J'1, J''1. However, when the strain is large enough (outside the LVE range), the stress/strain response is not a perfectly sinusoidal waveform, but rather contains higher harmonic contributions. In the linear region, both linear viscoelastic modulus and compliances are easily convertible. Nevertheless, in nonlinear rheology, this does not occur, and applying sinusoidal stress or strain as the input signal will produce different functions of the material. The nonlinear response of a material can be visually analyzed using Lissajous-Bowditch curves (also called Lissajous curves). If stress is applied, curves typically represent strain (γ) (or strain rate) against stress (σ) from the (raw) sample test data. The curves of γ vs σ are called elastic Lissajous curves and strain rate vs σ are viscous Lissajous curves. In an elastic Lissajous curve (γ vs σ), a strongly elastic material is represented as a straight line, while a viscous material as a circumference. Therefore, an ellipse-shaped curve corresponds to a linear viscoelastic material. Lissajous curves become distorted when the response changes from linear to nonlinear behaviour. Recently, new viscoelastic parameters have been defined from the physical interpretation of these curves, which allow quantifying the type of intra-cycle nonlinear response. From an elastic Lissajous curve, the minimum-stress elastic compliance or tangent compliance at zero stress (J'M ), and the large-stress elastic compliance or secant compliance at maximum stress (J'L) are obtained. In this way, their ratio (J'L =J'M ) describes the elastic (nonlinear) behaviour in a cycle, such as stress-stiffening or stress-softening, in a soft solid material. The general objective of this doctoral dissertation was to evaluate Afuega’l Pitu atroncau blancu and roxu cheese from all PDO cheese factories and also to study the effect of ripening time on biochemical parameters, microstructural and rheological properties, comparing blancu and roxu varieties. The following specific objectives were established: (1) Study the compositional and physicochemical parameters, as well as the intensity of protein and lipid degradation; (2) Analyze the microstructural properties; (3) Determine the viscoelastic properties at constant temperature (20 ºC) using SAOS, in terms of strain (γcrit) and stress (σcrit) amplitudes; (4) Study the thermoviscoelastic characteristics, within LVE range, based on the influence of temperature on the mechanical spectra at 20 ºC, 50 ºC and 75 ºC, and the thermal profiles from 20 ºC to 90 ºC at 1 ºC/min; (5) Study the nonlinear thermoreological properties using LAOS, at various temperatures (20 ºC, 50 ºC and 75 ºC) and angular frequencies (ω = 0,628 rad/s, 6,28 rad/s and 31,4 rad/s), and; (6) Correlate the biochemical magnitudes with microstructure and viscoelastic properties to explain changes in the physical properties of cheeses. To study the biochemical, microstructural and thermo-rheological characteristics of Afuega’l Pitu atroncau blancu (B) and roxu (R) cheeses from the different cheese factories (F1–F9) with PDO, three batches per factory and by type of cheese were analyzed, with a ripening time of 30 days. The ripening conditions were specific to each manufacturer. The compositional, physicochemical and lipolytic parameters were determined. F1B–F9B and F1R–F9R were acid cheeses (pH < pI) and with high-fat content (> 45% of total solids, TS). The moisture content (%) varied between F1–F9, especially between the lowest moisture content in F5 (B and R: ~30%) and the highest in F8 (B and R: ~49%); similarly for B and R. From cryo-SEM (Scanning Electron Microscopy) analysis, B and R exhibited a protein matrix (continuous phase) made up of fused casein particles and with a level of compaction that depends on the moisture content. The pores of the matrix contain water/aqueous phase and also fat (dispersed phases). The size and shape of fat were different for the F1B–F9B cheeses, which showed approximately spherical fat particles homogeneously dispersed in the matrix, compared to F1R–F9R. This variety (R), showed large non-globular fat particles located in specific areas in the protein matrix. Stress sweeps (20 ºC) showed that a higher moisture content produced lower stress amplitude (σcrit) and lower gel strength (G*), in both cheese varieties. Thus, it was observed that σcrit varied from values of ~170 Pa (F8) to ~940 Pa (F5) and G* between ~51 Pa (F8) and ~360 Pa (F5), similarly for B and R. These data reflected the moisture-to-protein ratio (MPR), lower in F5 (~1.44) compared to the highest in F8 (~2.77). In general, samples B exhibited higher values of γcrit and σcrit than R, but with similar solid character (tanδ) between both varieties. The influence of temperature on the mechanical spectra was analyzed at 20 ºC, 50 ºC and 75 ºC. The G' and G'' modules were lower in F8 and higher in F5, regardless of temperature and cheese variety. At 50 ºC, all cheeses showed an increase in tanδ values throughout the frequency range, whereas at 75 ºC a notable reinforcement of the solid character (lower tan δ) was observed at low frequencies, indicating some gelling induced by shear. The thermal profiles obtained by dynamic thermomechanical analysis indicated that the cheeses from all the factories maintained their gel property (G' > G'') during heating from 20 ºC to 90 ºC. Specifically, for T> 60 ºC tanδ decreased with increasing temperature in all cases, showing specific gelation induced by heat, consistent with that observed at low frequency (50 ºC and 75 ºC) and compatible with the cheeses softening at high temperature. From high amplitude oscillatory shear stress sweeps (LAOStress) analysis, the nonlinear viscoelastic properties of Afuega’l Pitu cheese from different manufacturers (F1–F9) were analyzed at different frequencies (0.628 rad/s, 6.28 rad/s and 31.4 rad/s) and temperatures (20 ºC, 50 ºC and 75 ºC). The elastic Lissajous curves (strain vs stress) were slightly distorted with increasing shear stress and temperature, particularly at 75 ºC. The narrow area enclosed in these curves accounted for the dominant elastic character of the samples. Also, the solid-like character was reinforced with increasing shear stress at high temperatures, resulting in a minor area enclosed by the Lissajous curves at 75 ºC and low frequency (0.628 rad/s). The viscoelastic compliances (J' and J'') indicated a progressive softening of the cheese network during the stress sweep for all frequencies and temperatures. The extension of the intra-cycle nonlinear viscoelastic behaviour was small (J'3 =J'1 ≥ 0.01). On the other hand, J'L=J'M (≤1) indicated a nonlinear intra-cycle behaviour that showed a tendency to stress-stiffening, except for experiments at 31.4 rad/s (stress-softening). The extension and type of nonlinear behaviour were slightly intensified with increasing shear stress and temperature. To study the microstructural, thermoviscoelastic and biochemical properties of Afuega’l Pitu cheese atroncau blancu (B) and roxu (R) varieties during ripening, four batches of each variety of cheese, manufactured by a PDO cheese factory (recognized and awarded cheeses in different competitions), were analyzed. Three batches were ripened in an external chamber under the same controlled conditions (ripening 1) and one batch was ripened in the cheese factory itself (ripening 2); for each variety of cheese. Samples were taken at 3, 15, 30 and 60 days of ripening. The compositional, physicochemical, proteolytic and lipolytic parameters were determined. The fat content (> 45 g/100 g TS), protein content and NaCl (<3 g/100 g TS) were similar between B and R, and did not change during ripening. The lactose content was higher in B than in R at 3 days, and it was gradually degraded throughout the ripening without completely disappear, in both varieties, after 60 days. The different environmental conditions in the ripening rooms marked significant changes in the evolution of the moisture content that increased by ~50% (ripening 1) or up to ~70% (ripening 2). These environmental conditions can also be related to the different intensity of proteolytic and lipolytic phenomena during ripening. The pH, nitrogen fractions and the fat acidity increased during ripening 1, more notably in B than in R samples. Cryomicroscopy images showed a protein matrix evolution towards a less porous microstructure, which increased the thickening of the casein strands. Large non-globular fat particles distributed among the protein matrix were also observed for variety R. This could be a consequence of kneading, specific to its manufacturing process, to mix the partially dehydrated curd with paprika and salt. In contrast, samples B exhibited more homogeneous microstructures with small fat globules. These microstructural characteristics, in B, improved the viscoelastic parameters such as stability (σcrit) and conformational flexibility of the network (γcrit). In general, σcrit and G* increased, while γcrit decreased up to 60 days. Improved stiffness (higher G*), structural stability (high σcrit) and more shear sensitive (or physically structured) cheese network (lower γcrit) were related to microstructural changes during ripening. Throughout ripening 1 vs 2, lower gel firmness was achieved in both varieties, which could be related to the higher proteolysis and lipolysis observed and the higher pH. Structural trends were verified by high positive correlations between γcrit-MPR, and negative between σcrit-MPR and G*-MPR. The mechanical spectra were also determined, in the LVE region, at 20 ºC, 50 ºC and 75 ºC, observing that all samples behaved as true gels (G' > G''), regardless of temperature. These modules were fitted to the angular frequency (ω) by the power-law according to G' = G'0 · ω^n' (Eq. 1) and G^'' = G'' · ω^n'' (Eq. 2). The most desirable gel properties, i.e, low values of tanδ = G''/G' and less dependence of both modules with the frequency —–low exponents n' (Eq. 1) and n'' (Eq. 2)—, were observed at 20 ºC for B30 and R30. The viscoelastic parameters (G'0 and G''0) increased with the ripening time, regardless of temperature, whereas at fixed ripening time were lower at higher temperatures. At 75 ºC, all samples exhibited a higher frequency dependence of G'' vs G' (n'' >> n'), implying shear-induced gelation at low frequencies (high oscillation times) that was consistent with data in thermal profiles from 70 ºC to 90 ºC. The thermal profiles, obtained by dynamic thermomechanical analysis, showed casein matrices with solid-like character (G' > G'') from 20 ºC to 90 ºC, regardless of the ripening time. The nonlinear properties (LAOStress tests) of Afuega’l Pitu cheese during ripening were qualitatively analyzed through the interpretation of the elastic curves of Lissajous. A small distortion of the curves was observed in all cases at the largest shear stresses while maintaining a narrow enclosed area (dominant elastic character of the samples), regardless of ripening time, frequency and temperature. Thus, the transition into the nonlinear region was mainly evidenced through the rotation of the elastic (non-normalized) Lissajous curve, suggesting a predominant contribution of the J'1 and J''1 relative to J'3 and J''3 to the nonlinear viscoelastic behaviour. At 75 ºC and 0.628 rad/s, beyond the LVE range, the area enclosed by the Lissajous curve was reduced in samples at 30 days of ripening, indicating a reinforcement of the solid-like character.