What amino acids is most likely to be found on the outside of a properly folded protein in the cytoplasm?

Globular Proteins

Globular proteins are named for their approximately spherical shapes and are the most abundant proteins in nature. The globular proteins exist in an enormous variety of three-dimensional structures. Nearly all globular proteins contain substantial numbers of α-helices and β-sheets folded into a compact structure that is stabilized by both polar and nonpolar interactions (Fig. 4.14). The globular three-dimensional structure forms spontaneously and is maintained as a result of interactions among the side chains of the amino acids. Most often, the hydrophobic amino acid side chains are buried, closely packed, in the interior of a globular protein, out of contact with water. Hydrophilic amino acid side chains lie on the surface of the globular proteins exposed to the water. Consequently, globular proteins are usually very soluble in aqueous solutions. The diversity of protein structures reflects the remarkable variety of functions performed by the globular proteins: binding, catalysis, regulation, transport, immunity, cellular signaling, and more.

Proteins range in molecular weight from 1000 to more than 1 million daltons (Da), but the folded size of a globular protein is not necessary correlated to its molecular weight. Proteins composed of about 250 amino acids or less often have a simple, compact globular shape. Larger globular proteins are usually made up of two or more recognizable and distinct structures, termed domains or modules. These are compact, folded protein structures that are usually stable by themselves in aqueous solution. Typical domain structures consist of hydrophobic cores with hydrophilic surfaces. Individual domains often possess unique functional behaviors and often perform unique functions within the larger protein in which they are found.

3 Disulfide Bonds

Most globular proteins contain internal covalent cross-links in the form of disulfide bonds (cystine residues). There is relatively free rotation about the disulfide bond in simple compounds, such as dimethyl disulfide. However, more highly substituted versions, including the free amino acid cystine, experience a considerable steric hindrance to rotation, amounting to as much as 3–10 kcal/mol (Winnewisser et al., 1968). Because of this high barrier to internal rotation, most disulfides, including those found in polypeptides and proteins, exist in only one of two possible rotational isomeric forms. These two geometries are shown in Fig. 6. The two bonds joining Rx and R2 to the sulfur atoms, along with the disulfide bond itself, define two planes whose relative orientation can be described conveniently by the dihedral angle (ϕ) between them. In most disulfides this angle is close to ±90°. In the case of proteins, R1 and R2 will not be equivalent, and the two rotamer forms will be intrinsically asymmetric. This asymmetry is preserved by the nonrotating nature of the S–S bond. Accordingly, the electronic transitions of the disulfide bonds, which produce weak absorption bands between 250 and 300 nm, are also capable of optical activity.

From studies carried out on small disulfide-containing compounds (Carmac and Neubert, 1967; Dodson and Nelson, 1968; Coleman and Blout, 1968; Ito and Takagi, 1970; Casey and Martin, 1972; Ludescher and Schwyzer, 1971; Nagarajan and Woody, 1973), it is now known that the disulfide produces very broad, weak CD bands, devoid of vibrational structure. The maxima of these bands appear anywhere between 250 and 300 nm, depending on the value of the dihedral angle. It is also known that in many cases the sign of the disulfide CD band can be used to determine the chirality or rotamer form of the bond in question.

At present, little is known about the contributions of disulfide bonds to the total optical activity of the larger globular proteins. The relatively low content of these chromophores and their characteristically weak CD bands, devoid of vibrational fine structure, make their recognition and direct measurement in such materials very difficult. In many instances it has been assumed that their contribution is negligible, although in others where the disulfide content is unusually high and/or the aromatic chromophore content is low, this is known not to be the case (Horwitz et al., 1970; Breslow, 1970; Bewley et al., 1972a, 1974; Puett, 1972; Menendez-Botet and Breslow, 1975; Holladay and Puett, 1975).

Having completed this brief description of CD, this is perhaps a good point to interject the following, quite legitimate, question: Why pursue these CD studies? An X-ray crystallographic study would provide an enormously more detailed picture of the structures of the hormones. As stated, the point is completely valid. However, there are several other points to be considered. First, with the notable exceptions of insulin (Blundell et al., 1971a, b) and glucagon (Sasaki et al., 1975) very few polypeptide or protein hormones, and none of the pituitary hormones, have provided crystals of X-ray quality. This is probably a technical difficulty, and such crystals will possibly be available in the near future. There is a second more important point, however, which is perhaps not just a technical problem. Although a crystallographic picture of these hormones would undoubtedly provide an enormous amount of important information, it might prove very difficult or impossible to confidently extrapolate this basically “static” information to include the more dynamic properties, such as the interactions between hormones and antibodies, or hormones and receptors, which we wish most to understand and ultimately control. Investigation of the dynamic properties of these interactions will almost certainly involve studies under solution conditions. It is in these types of studies where techniques such as CD will reach their full potential. The true understanding of hormone action will, of course, come from a condensation of information from all types of studies including crystallographic, conformations in solution, and the various biological studies. Thus X-ray and CD studies are adjunct to one another rather than competitive or redundant in helping to understand biological systems.

Troponin

Troponins are globular proteins composed of three polypeptide chains: TpC (mol. wt 18 000), TpI (mol. wt 24 000), and TpT (mol. wt 37 000). One troponin molecule sits astride each tropomyosin molecule, a short distance from one end of the molecule. One of the troponin peptides, troponin C, plays an important role in the binding of Ca2+. The nerve impulse causes a release of Ca2+ from the sarcoplasmic reticulum that then binds to troponin C causing a configurational change in the protein. This initiates a series of events eventually leading to ATP hydrolysis and muscle contraction (see Sections 31.4, 31.5 and 31.6).

Abstract

Enzymes are globular proteins with catalytic activity with wide use in the food industry. Formulations developed to produce such biocatalysts as insoluble particles through enzyme immobilization have been developed to enhance their applicability and overcome some shortcomings toward cost-effective commercial processes. When properly performed, immobilization can enhance enzyme stability, specificity, and activity. Moreover, immobilization enables biocatalyst reuse or continuous use, widens the setup of biocatalytic systems, and eases downstream processing. Immobilization can be performed either by attaching/entrapping the enzyme to/inside a carrier or by cross-linking enzyme molecules to provide a large and insoluble network. Hence, the properties of the immobilized enzyme formulations can be related to the immobilization procedure and to the carrier. Accordingly, a vast diversity of immobilized enzyme formulations and setups has been implemented in the food industry. This chapter provides an overview on recent developments and trends on the design and application of immobilized enzyme formulations for food production.

Histone acetylation

Histones are globular proteins that undergo posttranslational modifications and alter their interaction with DNA and other nuclear proteins (15). H3 and H4 histones have tails that are coiled around nucleosomes, which are functionally modified by the addition of acetyl and methyl groups (Fig. 9.2B). Histone acetylation and methylation are the important posttranslational mechanisms involved in determining cell lineage fate. Histone deacetylases (HDACs) remove acetyl residues from target histones or non-histone proteins resulting in the condensation of chromatin structure to discourage TF binding. In contrast, histone acetylation results in loose packing of nucleosomes enabling TF binding and gene expression. Acetylated histones have been identified to be important in regulating CD4+ T cell function [59]. These marks have been detected in promoter regions of IFN-γ and IL4 genes and are in part responsible for determining lineage-specific differentiation to Th2 and Th2 cells [60].

Enzymes that methylate or acetylate histones have an important role in the state of chromatin, nucleosome processes, and DNA repair [61, 62]. Acetylation of histones by HDACs is a reversible process with important potential in regulating gene expression, but also as a possible therapeutic in immune-mediated inflammatory diseases. Some HDAC inhibitors have been extensively studied to increase the transcription of genes involved in regulating T cell inflammatory processes and autoimmunity [63]. The function of HDAC inhibitors and genetic deficiencies of HDAC members impact the outcome of autoimmune and autoinflammatory diseases, which was the subject of a recent review [64].

4.2.2 Fibrin

Fibrin is a globular protein derived from humans and animals that forms the basis of the blood clotting process [151]. Produced largely by liver tissue, fibrin, and its precursor molecule fibrinogen have both been studied and applied extensively in biomedical applications due to its simple isolation and ECM similarity [4, 152]. Fibrinogen is a large protein of approximately 350 kDa consisting of two identical subunits, each consisting of three polypeptide chains [151]. Many commercial forms of fibrinogen exist isolated from animal sources including murine, bovine, and primate sources—and purified human fibrinogen is a sealant approved for use by the FDA. In vivo, fibrin is a protein that forms spontaneously when fibrinogen is cleaved by thrombin when tissue injury occurs and is degraded by plasmin when tissue repair takes place [4, 153, 154]. Physical and chemical properties of fibrin polymer products can be tuned by controlling pH, ionic strength, or concentration of the fibrinogen precursor during the polymerization process [4, 155]. Due to fibrin’s animal and human sourcing, scaffolds are biocompatible, resorbable, and naturally cell adhesive—supporting cell infiltration of a variety of cell types including those that are involved in forming natural ECM [4, 156]. Because of these features, fibrin has been used extensively in deliverables, clinical tissue engineering applications, and both direct and indirect gene therapy applications [4, 95, 157–159].

Regulation of cytoskeletal dynamics

Actin is a globular protein that participates in microfilament formation for cellular division, phagocytosis, chemotaxis, and adhesion. The cytoskeletal structure is highly responsive to changes in the exposome. This involves the production of ROS to mediate eustress signals for the control of G-actin polymerization and microfilament disassembly (Hurd, Degennaro, & Lehmann, 2012). Interest in how S-glutathionylation switches modulate cell functions in response to changes in redox buffering capacity was ignited by a study demonstrating that G-actin polymerization is regulated by GSH conjugation reactions (Wang et al., 2001). Epidermal growth factor signaling induced the S-glutathionylation of G-actin in cultured A431 cells, which reduced the rate of actin polymerization by ~ sixfold (Wang et al., 2001). This was induced by increased ROS production by NADPH oxidase (NOX), resulting in glutathione pool oxidation and G-actin modification (Fig. 3) (Dalle-Donne, Giustarini, Rossi, Colombo, & Milzani, 2003). G-actin S-glutathionylation occurs on Cys374, which is reversed by GRX1, restoring actin polymerization (Wang et al., 2001). Later studies found that the reversible modification of G-actin is integral for immune cell function. Disruption of GRX1 impairs neutrophil chemotaxis, adhesion, and particle phagocytosis via the maintenance of G-actin in a glutathionylated state (Fig. 3) (Sakai et al., 2012). The antimicrobial defense function of neutrophils also depends on the formation of neutrophil extracellular traps (NETS). Intriguingly, NET formation and disassembly was found to be controlled by glutathione pool oxidation and GRX1 (Fig. 3) (Stojkov et al., 2017).

Cytoplasmic S-glutathionylation reactions also regulate microtubule and neurofilament dynamics (Wilson & Gonzalez-Billault, 2015). Histological analyses of nervous tissue showed that a number of cytoskeletal proteins are basally S-glutathionylated in the spinal cord, cerebellum, and prefrontal cortex (Sparaco et al., 2006). Cytoskeletal proteins found to be modified included actin, tubulin, and neurofilaments (Sparaco et al., 2006). In nervous tissue, redox signals are important for axonal guidance and cone development, processes that are vital for neural growth and development. This is achieved through the NOX-mediated oxidation and reduction of cysteine switches in G-actin and the regulation of the semaphorin/plexin signaling platform (Wilson & Gonzalez-Billault, 2015). Tubulins are also subjected to S-glutathionylation in response to increases in ROS and other oxidants (Landino, Moynihan, Todd, & Kennett, 2004; Wilson & Gonzalez-Billault, 2015). Exposure of brain tubulin to oxidants results in its S-glutathionylation, hindering polymerization (Landino et al., 2004). GRX1 deglutathionylates tubulin restoring the rate of polymerization (Landino et al., 2004). Regulation of microtubule polymerization and dynamics in response to changes in cytoplasmic redox buffering capacity also controls neurite growth. Persistent S-glutathionylation of tubulins has been implicated in neurite degeneration due to an inability to maintain cytoskeletal dynamics. Additionally, Friedrich’s ataxia is characterized by increased tubulin S-glutathionylation in spinal cord motor neurons (Sparaco et al., 2006). Collectively, reversible S-glutathionylation of cytoskeletal proteins is required to modulate cell responses to different physiological cues and disruption of these pathways can affect immune cell function and potentially compromise neural health and development.

B The Membrane Matrix: Lipoprotein Subunits or Continuum?

Much attention has been given in recent years to the hypothesis that membranes are organized as an assembly of distinct repeating lipoprotein subunits, as contrasted to a continuum such as the Davson-Danielli-Robertson model proposes. (These lipoprotein subunits of a membrane are not to be confused with the subunits of the individual proteins of membranes discussed in the preceding section.) The question not only is of intrinsic interest to membrane structure, but it bears directly on the problem of membrane biosynthesis (see section V,D).

The experimental evidence which has been adduced in favor of a lipoprotein subunit assemblage as a general mode of organization of membranes has been thoroughly reviewed by Stoeckenius and Engelman (1969) and will not be detailed here. We agree with their conclusion that the evidence at present is not convincing. The great heterogeneity of membrane proteins and lipids requires that if lipoprotein subunits did constitute a membrane, either they would have to be very large to be identical, or they would have to be correspondingly heterogeneous both chemically and structurally. If the latter were the case, it is unclear what would determine how such heterogeneous subunits would be arranged and assembled. Furthermore, the existence of discrete lipoprotein subunits would seem to require that the synthesis of lipid and membrane protein be reasonably synchronized and that the ratio of lipid to protein in a particular membrane be fairly constant. Neither of these expectations has been borne out in recent experiments with mycoplasma (Kahane and Razin, 1969). The membranes of chloramphenicol-treated cells (with protein synthesis stopped) had a significantly lower buoyant density than those of untreated cells, showing that a substantial increase in the ratio of lipid to protein had been produced in the membrane with no apparent effect on membrane function.

The lipid–globular protein mosaic model clearly has structural features of both a continuum and (at least superficially) a lipoprotein subunit assembly. In particular, if the protein of a membrane were of only one or a few kinds [as, for example, in the outer segments of visual receptor rods (cf. Bownds and Gaide-Huguenin, 1970) or in viral membranes], its arrangement in a lipid–globular protein mosaic might exhibit a fairly uniform periodic structure in the plane of the membrane, whereas in a more heterogeneous membrane such as that of the red blood cell, the mosaic arrangement of the many different integral proteins of the membrane, with perhaps variable amounts of intervening lipid, might blur out any appearance of periodicity. On the other hand, there is nothing intrinsic to the lipid–globular protein mosaic model that requires the existence of discrete lipoprotein subunits in the membrane. The fact that treatment of membranes with detergents usually results in the appearance of small lipid–protein aggregates has sometimes been used to suggest that such aggregates are the membrane subunits. One would predict, however, that detergent treatment of a membrane organized as a continuum of a lipid–globular protein mosaic would split the membrane within the lipid portions of the mosaic and produce small aggregates such as are experimentally observed.

If the lipid–globular protein mosaic is indeed organized as a continuum, is it one of proteins embedded in a lipid continuum or of lipids embedded in a protein continuum (i.e., which component provides the matrix of the membrane structure)? The schematic cross-sectional representation of the lipid–globular protein mosaic model in Fig. 3c deliberately avoids the issue, which must be joined when the third dimension of the model is specified. In favor of the view that the lipid provides the matrix of the continuum are the following arguments. (a) It accommodates more easily the results of Kahane and Razin (1969) mentioned above that the lipid–protein ratio of a mycoplasma membrane can be significantly increased without apparent alteration of membrane properties; since most of the lipid in the bilayer portion of the mosaic does not interact strongly with the protein (section IV,F), extra lipid could be inserted without much effect. If protein–protein contacts provided the matrix of the continuum, however, and the lipids filled the spaces between, there might be greater difficulty in accommodating a significant additional amount of lipid into the membrane, if it were uniformly distributed throughout. (b) It allows many different arrangements of grossly heterogeneous integral membrane proteins in the plane of the membrane.* If protein–protein contacts provided the matrix, these contacts would presumably be specific, and a highly organized and uniform arrangement of the membrane proteins might result. The greater the heterogeneity of the membrane proteins, the more difficult it is to envision a regular repeating protein matrix.

The electron microscopic experiments discussed in section IV,A, which revealed no marked alteration in the railroad-track appearance of fixed and embedded membranes if their lipids were previously extracted, might appear to be more consistent with a protein than with a lipid matrix for the mosaic structure. However, the possibility that some lateral shrinkage occurred in the specimens, as might be expected if the lipid provided the matrix, has not been ruled out. Also, bilayer-type phase changes observed for the lipids in intact membranes do not necessarily imply that the lipids form the continuum, since, as is discussed in section IV,G, the size of the cooperative unit involved in the particular phase change is not known. It is not inconceivable that a bounded unit containing of the order of 100 lipid molecules might“melt” at a temperature not noticeably different from that of an infinite bilayer, if melting involved a cooperative increase in a uniaxial rotational motion of the long fatty acid chains.

We are inclined to the view that if the membrane is a continuous lipid–globular protein mosaic, its matrix is provided by lipid, but we acknowledge that there is no firm evidence at present to rule out the possibility that the matrix is provided by protein.

3.48.1 Introduction

Enzymes are specialized (globular) proteins with catalytic features, which occur in all living organisms and accelerate biochemical reactions necessary to support life. Enzymes are also ubiquitous in fresh and processed foods, even though they are often deactivated; as it happens with other dietary proteins, enzymes are degraded by hydrolysis after ingestion and eventually metabolized. Enzymes present in traditional human diets have not been associated with relevant forms of toxicity, so they are considered as intrinsically safe [1].

A major fraction of industrial enzymes available at present are used in the food industry, where they have found a wide variety of applications [2]; this, in particular, is the case of hydrolases, a large family that includes carbohydrases, proteases, and lipases, usually employed as food additives – which amount to at least 75% of all enzymes on the market. Proteases dominate, in turn, the industrial enzyme market, with a market share of ∼60% – owing chiefly to their application in detergents. Enzymes have been successfully used in food processing to replace steps that encompass harsh chemical or physical conditions (eg, temperature, pressure, or presence of chemicals); hence, they contribute to sustainable industrial production, besides guaranteeing a safer and more nutritious food supply.

Most enzymes have been produced at the commercial scale via submerged cultures and have found relevant uses in the beverage and bakery industries, the production of specialty dairy products, and the processing of starch – besides their role as antimicrobial agents in the preservation of several foods. There is, however, a growing trend to manufacture enzymes via solid-state fermentation techniques – not only because enzyme titers can be higher than in submerged cultures, but also because of the major public concerns with the upgrade of agricultural wastes, which can thus serve as suitable (and low-cost) substrates for the host microorganisms. On the other hand, solid and liquid fermentation affect biomass formation and enzyme biosynthesis in different modes; for example, molds grown on solid-state substrates produce more spores and fruiting bodies, which, in turn, have a physiological influence upon enzyme production therein.

The industrial production of enzymes for use in food processing dates back to 1874, when the Danish scientist Christian Hansen extracted rennin (also known as chymosin) from calf stomachs, for later use to clot milk in cheese manufacturing. Chymosin is nowadays produced by several transformed microorganisms, containing the bovine prochymosin gene introduced through recombinant deoxyribonucleic acid (rDNA) techniques. Bovine chymosin expressed in Escherichia coli K-12 became, in fact, the first recombinant enzyme approved for use in food by the Food and Drug Administration (FDA) in the United States [1]. The major industrial association dealing with enzymes worldwide, Association of Manufacturers and Formulators of Enzyme Products (AMFEP), currently lists ∼160 enzymes manufactured for use specifically in the food industry, at least 36 of which are produced via genetically modified microorganisms (GMOs). Note that the application of purified enzymes from microbial origin is a rather recent development, which dates back to the late first half of the 20th century [2].

Improvements in bioprocessing and biocatalyst development – especially with the advent of genetic engineering, have greatly increased the availability of food enzymes and have also permitted tailoring of their properties to meet the (sometimes very) distinct requirements of food matrices and food process. By judicious selection of host microorganisms as recombinant strains containing the DNA sequences that code for a specific enzyme amino acid sequences, more efficient synthesis of said enzymes will be possible, often in the absence of undesirable enzymes or other microbial metabolites. In addition, it also became possible to engineer enzymes with increased heat stability and improved compatibility to other medium components – at the expense of introducing deliberate changes in the original enzyme amino acid sequences, targeted at specific base pairs of the complementary DNA.

Enzyme preparations, commercialized for food protection, typically contain not only the enzymes of interest, but also several other compounds, namely diluents, preservatives, and stabilizers. These extra materials are usually well-known substances that have been previously cleared for use in foods and which perform useful specific functions. Enzyme preparations may also contain other enzymes and metabolites synthesized by the production organisms, as well as residues of the raw materials used in the fermentation broth and of the solvents used in the isolation and purification of the enzyme of interest. All these materials are expected to be of a purity that is consistent with good manufacturing practices (GMPs) [1].

The expansion of the portfolio of application of food enzymes as food-processing aids, and as part of the formulation of processed foods, has obviously captured the attention of regulators in the most developed countries. Hence, new food enzymes require a prior market authorization by FDA in the United States, or the European Food Safety Authority (EFSA) in Europe; furthermore, several European Union (EU) member states have national legislations focusing on food enzymes and all serve the main purpose of ensuring the safety of enzyme preparations for the final consumers, so they often include specifications for purity and activity, and are enforced by national regulatory boards. In certain cases, occupational health issues that may arise during manufacturing and handling of food-grade enzymes are also part of the underlying legislation [1].

In principle, identical safety measures should be considered with regard to food-grade enzymes derived from either native microorganisms or GMOs. The key issue is the safety assessment of the production organism – in particular, whether it holds any pathogenic or toxigenic potential. Although, no formally known pathogenic or toxigenic organisms have intentionally been used to produce enzymes intended to contact foods, certain fungi that have traditionally been used as sources of specific enzymes have meanwhile been found to also produce low levels of toxic secondary metabolites under the fermentation conditions employed for optimized synthesis of the desired enzymes; hence, special care has to be exercised in these cases, especially with the purification steps downstream [1].

This chapter provides an overview of the food-grade enzymes of highest relevance – to either aid in processing or be included in the final product. Hence, all steps from manufacture to safety and legislation are tackled, covering production, transformation, and utilization in foods.

History and methods