Analysis by Raman spectroscopy of the conformational structure of whey proteins constituting fouling deposits during the processing in a heat exchanger

https://doi.org/10.1016/j.jfoodeng.2011.12.005Get rights and content

Abstract

Whey protein fouling deposits generated on the hot wall downstream a plate heat exchanger were analyzed by micro Raman spectroscopy (MRS) carried out in the 800–1800 cm−1 range. Deposits were formed using a model beta-lactoglobulin (BLG) fouling solution which was made using a whey protein isolate powder (89 wt.% in BLG) and a known amount of calcium. Thermal denaturation of the fouling solution was also analyzed by MRS as well as isolated BLG aggregates obtained by microfiltration of heated solutions. Specific Raman signatures of aggregates were identified, which were not detected in the Raman spectra of denatured (i.e. unfolded BLG molecule) solutions. MRS analyses at different depths of the deposit reveal a loss of α-helix structures, as observed in denatured BLG solutions, without the detection of aggregate signatures. For the range of calcium content investigated (from 97 to 160 mg l−1), no effect of calcium ions on the molecular conformation of BLG within the deposit was shown. Of great significance, results suggest that, for our set of operating conditions used, the mass distribution of the fouling deposit in a plate heat exchanger is primarily controlled by the distribution of the unfolded protein generated by the denaturation process.

Highlights

► Micro Raman spectroscopy analysis of the molecular structure of beta-lactoglobulin. ► Raman signatures of native, denatured (unfolded) and aggregated beta-lactoglobulin. ► Whey protein deposits reveal a loss of α-helix structures. ► Beta-lactoglobulin state is not aggregated at the deposit – stainless steel interface. ► Calcium ion does not change the beta-lactoglobulin molecular conformation within deposit.

Introduction

Fouling of heat exchangers in the dairy industry (mostly plate heat exchangers, PHEs) is still a serious issue as it reduces heat transfer efficiency, increases pressure drop and requires numerous cleaning-in-place procedures. For a temperature below 110 °C, the deposit is largely a “milk film” deposit or a type A (protein) fouling as classified by Burton (1968), consisting of 50–70 wt.% proteins, 30–40 wt.% minerals and 4–8 wt.% fat (Fickak et al., 2011). It has been largely shown in the literature that BLG plays a predominant role in milk fouling deposits; indeed BLG is the milk protein most sensitive to heating, which makes up only 10% of raw milk protein but can constitute half of the protein deposit content (Changani et al., 1997, Lalande et al., 1984, Lalande et al., 1985).

The scientific community has examined the BLG conformational structure and its link with temperature for several decades (Mulvihill and Donovan, 1987). At room temperature and at pH 7, the secondary structure of native BLG is mainly composed of β-sheets (54%) and α-helices (17%) (Seo et al., 2010). Literature data show that an increase in temperature initiates a heat denaturation process which leads to different BLG protein conformations in addition to the native one (N), i.e. unfolded (U) and aggregated (A) state (Tolkach and Kulozik, 2007). Today it is well established that when heated in a neutral aqueous solution, globular molecules of BLG unfold with a loss of their tertiary and secondary structure. Such unfolding of the protein exposes reactive sulphydryl (–SH) groups which are normally contained within the core of protein and permits the formation of disulfide bridges with different denatured proteins present in the solution (for example casein, α-lactalbumin), resulting in the formation of aggregates (Lalande et al., 1989).

Calorimetric and spectroscopic investigations of BLG thermal denaturation, under different conditions of temperature, pH and protein concentration were carried out. Among them, one can mention those of Nonaka et al., 1993, Ikeda and Li-Chan, 2004, Ngarize et al., 2004, Seo et al., 2010. These studies conclude that a heat treatment induces changes in the BLG secondary structure. There is notably a trend toward an increase in β-sheet structures at the expense of α-helices structures with a simultaneous decrease of turn structure when heating from 70 to 90 °C. Using Raman scattering in the region of amide modes (800–1800 cm−1), the mechanism of BLG thermal denaturation in the 20–100 °C temperature range could be described as a two-step process (Seo et al., 2010). The first step corresponds to the dissociation of dimers associated with the softening of the tertiary structure (observed between room temperature and 65 °C). At the end of this step (around 65 °C), the so-called ‘‘molten globule state’’ is obtained prior to conformational changes of the secondary structure taking place between 65 and 95 °C (Seo et al., 2010). The detailed thermal behavior of BLG at pH >6.8 and at temperatures up to 150 °C has been recently described by the review paper of de Wit (2009).

Although the connection between the thermal stability and conformational changes of BLG is well established, the role of the denaturated and aggregated species on the build-up of the fouling deposit onto a heat exchange surface is still unclear. Indeed, few studies take sides unreservedly whether BLG fouling is controlled by BLG aggregation rather than by denaturation. Trends reported in the literature are not really conclusive as regards which species is involved in protein deposits, as underlined by Bansal and Chen, 2006a, Bansal and Chen, 2006b. Moreover, it is not obvious from the analysis of mechanistic models dealing with the prediction of fouling deposits in the literature to guess which species (unfolded or aggregated) shapes the fouling deposit. For example, Mahdi et al., 2009, Jun and Puri, 2007 assumed that only the aggregated protein is deposited on the wall, which is not functional for the model of De Jong et al., 1992, De Jong et al., 1993 for which it is hypothesized that the main mechanism in fouling process is a reaction-controlled adsorption of unfolded BLG. More detailed information regarding the involvement of protein denaturation and protein aggregation as governing reaction in fouling mechanisms is given in the review paper of Bansal and Chen (2006a) to which the reader can refer.

Protein adsorption at the stainless steel interface is considered to be the initial stage of substantial milk fouling. Following protein adsorption, protein–protein interactions prevail to build-up milk fouling on stainless steel surface used for heat exchange. However, as reported by Liu and Chen (2010), this field of knowledge has been rarely investigated at the molecular level since the pioneering work of Kim and Lund (1998) who have illustrated that protein adsorption onto stainless steel could be relevant to explain the deposition mechanism during thermal processing of milk.

It is noticed that the experimental mass balance realized on the denatured protein performed between the inlet and outlet of a heat exchanger is also confusing as regards the role played by the different species (unfolded or aggregated) on fouling growth. As pointed out by Lalande et al., 1985, Belmar-Beiny et al., 1993, only a small fraction of unfolded proteins sticked on each exchange surface. Precisely, Delplace et al., 1994, Delplace et al., 1997 observed that only 3.6% of irreversibly unfolded BLG was involved in the deposit formation.

In view of the literature data, a controversial question arises regarding which form of BLG (unfolded or aggregated one) constitutes the first deposit layer close to the hot wall. There is also a paucity of knowledge as regards the effect of calcium ions upon the conformational structure of protein contained in the fouling deposit. Indeed no data are available in this field although it has been observed in the past (Bansal and Chen, 2006a, Guérin et al., 2007) and more recently (Petit et al., in press) that the calcium content enhances deposition and influences the stability of the BLG thermal denaturation.

More information on the conformational structure of the whey protein molecules (especially BLG) which are present within a dairy protein deposit is needed in order to elucidate the fouling mechanisms of PHEs. The molecular structure and conformation changes of proteins can be analyzed by Raman spectroscopy (RS) which is a technique capable of providing information with respect to peptide backbone conformation as well as environment of certain side chains (Nonaka et al., 1993, Li-Chan, 1996, Chi et al., 1998, Jung et al., 2000, Ngarize et al., 2004, Ikeda and Li-Chan, 2004). Chi et al. (1998) ascertained that the protein secondary structure can be determined from the analysis of the amide I, II and III bands and the Cα–H amide bending vibration. The amide (peptide) bond of proteins has several distinct vibrational modes, of which the amide I and III bands are the most useful for the study of secondary structure (Li-Chan, 1996). Jung et al. (2000) have demonstrated the potential of two-dimensional (2D) Raman correlation spectra and its complementarity with 2D infrared correlation spectra in the amide III region (1330–1200 cm−1) to investigate the secondary structure of BLG in a buffer solution at pH 6.6. Another advantage is that RS is a non destructive analytical technique capable of gaining data at high protein concentrations sufficient to cause gelation, i.e. as those encountered in a protein fouling deposit (Ikeda and Li-Chan, 2004). Ivleva et al. (2009) pointed out that micro Raman spectroscopy (MRS) or Raman microscopy has also the advantage of providing fingerprint spectra with a spatial resolution of an optical microscope (with axial resolutions of about 5 μm in confocal mode).

Considering the dependence of the denaturation process on pH, protein concentration, thermal history, and the controversial description of the denaturation mechanism (Seo et al., 2010), a denaturation curve (represented by the temperature dependence of the amide I band frequency in the 25–100 °C temperature range) was carried out. For the denaturation curve, a 6 wt.% whey protein isolate (WPI) solution with a calcium content of 264 mg l−1 dissolved in H2O (pH 6.8) was used in order to characterize the heat-induced denaturation process. Additionally, prior to the Raman analysis of deposits, the Raman fingerprint of native, unfolded and aggregated BLG was collected for a comparison purpose with deposit spectra.

The main objective of this study was to determine, by means of MRS, the nature of BLG (i.e. native, denatured and aggregated BLG) at the deposit – stainless steel surface interface and within the thickness of the deposit. This was carried out from the identification of the Raman signatures obtained during investigations of the BLG thermal denaturation and from the analysis of aggregates in the 800–1800 cm−1 spectral range. A second objective consisted in evaluating a possible effect of calcium ions on the molecular structure of BLG within the deposit. A 6 wt.% WPI (purity in BLG ∼89 wt.%) solution was selected as a model fouling solution for which the heat induced denaturation/aggregation process of BLG has been previously determined via a kinetic model (Petit et al., in press). Additional deposits were generated using a 0.5 wt.% WPI solution with a calcium content ranging from 97 to 160 mg l−1 in order to investigate the role of calcium ions. Raman spectra of deposits in the amide I and amide III bands region (800–1800 cm−1) were collected to measure possible changes in the secondary structure of BLG within the successive layers of the deposit.

Section snippets

Experimental equipment

Fouling experiments were conducted using the pilot plant represented schematically in Fig. 1. The fouling rig was composed of three distinct zones: a pre-heating zone composed of a tubular joule effect heater, noted JEH (Actijoule, Actini, 30 kW), a heating zone composed of a PHE (VICARB, model V2, Alpha-Laval, France) and a holding zone which was made up of a stainless steel pipe of 35 mm internal diameter, including a detachable fouling test surface which was set aside for MRS analysis. PHE

Investigation of the amide modes in the 800–1800 cm−1 region. Assignment of major bands of BLG

At room temperature and at the temperature of the holding zone in fouling runs (85 °C), Raman spectra of the fouling solution (6 wt.% BLG in H2O plus 264 mg l−1 calcium, pH 6.8) in the 800–1800 cm−1 region are plotted in Fig. 3 for the assignment of Raman bands. Spectra were baselined and area normalized for a clear presentation of the bands. It is well accepted that the secondary protein structure can be analyzed in the amide I and amide III regions between 1550 and 1750 cm−1 and 1200–1350 cm−1,

Conclusions

Micro Raman spectroscopy was used to determine the molecular conformation of BLG at the deposit – stainless steel surface interface from the knowledge of the Raman signatures of BLG solutions. Raman signatures were recorded upon heating toward the thermal unfolded state and in aggregates, in the 800–1800 cm−1 frequency range. A whey protein isolate was selected as a model fouling solution. Deposits were generated by means of a fouling-rinsing sequence performed at pilot scale using a WPI

Acknowledgements

This work was carried out within the framework of an ARCIR (“Actions de Recherche Concertées d’Initiative Régionale”) program (European regional development fund 09 310 243 and 03 310 241), which joined non elucidated scientific issues of the Globule project (ANR-08-ALIA-08).

References (39)

  • Y. Mahdi et al.

    A dynamic model for milk fouling in a plate heat exchanger

    Applied Mathematical Modelling

    (2009)
  • J. Petit et al.

    Influence of calcium on β-lactoglobulin denaturation kinetics: implications in unfolding and aggregation mechanisms

    Journal of Dairy Science

    (2011)
  • S. Platt et al.

    Cleaning of membranes fouled by proteins to evaluate the importance of fully developed flow

    Desalination

    (2007)
  • B. Bansal et al.

    A critical review of milk fouling in heat exchangers

    Comprehensive Reviews in Food Science and Food Safety

    (2006)
  • Bansal, B. and Chen, X.D. 2006b. Deposition of electrode surfaces during thermal processing of milk in an ohmic heater....
  • H. Burton

    Deposit of whole milk in treatment plants – a review and discussion

    Journal of Dairy Research

    (1968)
  • Z. Chi et al.

    UV resonance Raman-selective amide vibrational enhancement: quantitative methodology for determining protein secondary structure

    Biochemistry

    (1998)
  • B. Czarnik-Matusewicz et al.

    Two-dimensional attenuated total reflection/infrared correlation spectroscopy of adsorption-induced and concentration-dependent spectral variations of β-lactoglobulin in aqueous solutions

    Journal of Physical Chemistry B

    (2000)
  • P. De Jong et al.

    Fouling of heat treatment equipment in relation to denaturation of β-lactoglobulin

    Journal of the Society of Dairy Technology

    (1992)
  • Cited by (48)

    View all citing articles on Scopus
    View full text