Effects Of Acyl Chain Composition On Membrane Protein Function

Numerous studies have been published describing modulation of membrane protein function by changes in the degree of unsaturation of the phospholipid acyl chains. It is convenient to divide these studies into two types, those that investigated membrane protein function in natural membranes and those that examined the function of purified membrane proteins reconstituted with defined phospholipids. The literature abounds with examinations of the effects of diets or drugs on a wide range of physiological and behavioral outcomes. In this chapter, we are concerned with the relatively few studies that have examined the isolated function of one or more specific receptors and performed an analysis of the composition of the receptor's host membrane.

Several different receptor systems in the heart have been examined in this kind of detail. The amount of 22:6n-3-containing phospholipid in the sarcolemmal membranes of rats was elevated by injections of hydrocortisone, and this was accompanied by a downregulation of ^-adrenergic receptors (Skuladottir et al., 1993). This finding is supported by the results of a detailed examination of 22:6n3-supplemented cardiomyoctes (Grynberg et al., 1995). Fatty acid enrichment produced cells in which 20% of the total fatty acids were 22:6n3. The a-adrenergic system was unaffected, but the ^-adrenergic receptors had a decreased affinity for ligand. In a study of streptozotocin-induced diabetes mellitus, the sarcoplasic reticulum membranes of treated rats showed a loss of arachi-donic acid, 20:4n6, acyl chains and an increase in 22:6n-3 (Kuwahara et al., 1997). This was accompanied by decreased sodium/potassium-ATPase activity and a reduction in calcium uptake.

A number of different types of membrane protein have been examined in reconstituted lipid vesicles containing high levels of acyl chain unsaturation, 20:4n6 or 22:6n3. These include sarcoplasmic reticulum calcium ATPase (Matthews et al., 1993), protein kinase C (PKC) (Giorgione et al., 1995; Slater et al., 1994;), gramicidin (Cox et al., 1992), and rhodopsin (Brown, 1994; Gibson & Brown, 1993; Litman & Mitchell, 1996b; Mitchell et al., 1990; Mitchell et al., 1992; O'Brien et al., 1977; Wiedmann et al., 1988). Higher levels of acyl chain unsaturation promote membrane protein function in all of these membrane proteins except calcium ATPase. Calcium ATPase was reconstituted into PCs with a 16:0 acyl chain at the sn-1 position and 18:0, 18:1n9, 18:2n6, 20:4n6, or 22:6n3 acyl chain at the sn-2 position. Enzyme function obtained with 18:1n9 or 18:2n6 at the sn-2 position was more than 10 times higher than that obtained with 20:4n6 or 22:6n3 at the sn-2 position (Matthews et al., 1993). In the other studies cited above, the highest level of protein function was obtained in the most highly unsaturated bilayer examined. However, there is no agreement among these studies regarding the physical property of the bilayer and the associated forces, which promote membrane protein function in highly unsaturated bilayers. The following subsections describe studies, wherein the authors have attributed their observations relative to membrane protein function to bilayer thickness, curvature stress, or acyl chain packing properties (e.g., acyl chain packing free volume).

3.1. Bilayer Thickness

Relatively large changes in membrane thickness have been demonstrated to alter the function of integral membrane proteins. An example of the magnitude of the change in membrane thickness needed to alter protein function is provided by studies of the sarcoplasmic reticulum calcium ATPase. Activity of this integral membrane protein in bilay-ers with symmetrically substituted, monounsaturated acyl chains with 16, 18, or 20 carbons is nearly constant. However, when the acyl chains are shortened to 14 carbons or lengthened to 22 carbons, activity is reduced by more than a factor of 3 (Lee, 1998).

This indicates that extreme changes in membrane thickness can alter protein function. However, the small changes in membrane thickness caused by physiologically relevant changes in acyl chain unsaturation are well within the range of constant calcium ATPase activity and seem unlikely to alter the function of other integral membrane proteins.

3.2. Curvature Strain

The activity of several different membrane proteins has been correlated with the propensity of the lipids composing the membrane to form nonlamellar structures or the curvature strain of the host bilayer (Epand, 1998; Li et al., 1995). Examples of alteration of membrane protein function via variable acyl chain composition, which were interpreted as changes in curvature stress, are provided by mitochondrial ubiquinol-cytochrome-c reductase and H+ ATPase. When these protein are reconstituted into liposomes of di18:1n9-cis PC and di18:1n9-trans PE, their activity is increased as the percentage of di18:1n9-cis PE is raised, whereas the addition of di18:1n9-trans PE had no effect (Li et al., 1995). This demonstrated that the effect was not the result of the addition PE headgroups to the bilayer and suggested that it was related to the greater propensity of di18:1n9-cis to form a nonbilayer phase.

Epand and co-workers found that 22:6n-3 acyl chains produced the highest level of PKC function when incorporated into PE, but not PC, and the activity was correlated with increased partitioning of PKC to the membrane (Giorgione et al., 1995). Stubbs and co-workers found that PKC activity was optimal when the membrane had a combination of headgroup spacing and bilayer curvature, which could be obtained with a mixture of PEs and PCs containing a 22:6n3 acyl chain (Slater et al., 1994). The rate at which gramicidin converted from a nonchannel to a channel-forming conformation was highest in PC membranes, which contained 22:6n3 acyl chains or PE phospholipids, and it was proposed that this is the result of increased curvature stress (Cox et al., 1992).

A recent detailed study of calcium-ATPase function and phospholipid motion concluded that PE headgroups promote the activity of calcium ATPase by specific noncovalent interactions, rather than by bilayer curvature stress, as had been previously proposed (Hunter et al., 1999). In biological membranes, which contain both PE and PC and a range of acyl chain compositions, the effects of polyunsaturated acyl chains and PE headgroups on curvature stress can be difficult to assess, because of the complex influence the acyl chains of one phospholipid species exert on those of other species (Holte et al., 1996; Separovic & Gawrisch, 1996). Although quantitative correlation of protein function with curvature stress derived from acyl chain unsaturation remains difficult, there is ample evidence to suggest that the curvature stress induced by high polyunsatu-rated acyl chains could be functionally significant. The challenge is to measure both alterations in function and curvature stress in the same system, so as to allow a direct correlation between these membrane properties.

Rhodopsin, the light receptor in the G-protein-coupled visual transduction system, has been studied extensively in reconstituted systems. Light converts rhodopsin's antagonist, 11-cis retinal, to the agonist, all-trans retinal, resulting in the formation of activated receptor. The conformation of activated receptor, which binds the visual G protein, metarhodopsin II (MII), exists in equilibrium with an inactive conformation, metar-hodopsin I (MI). Thus, the extent of functional activation is given by the equilibrium between MI and MII. The general observation is that the formation of MII is highest in bilayers composed of phospholipids with 22:6n3 acyl chains. In a series of experiments,

Brown and co-workers demonstrated that components that produce curvature strain could replace 22:6n3 acyl chains without compromising the extent of formation of MII. The ability of PE headgroups to support optimal rhodopsin function was analyzed in mixtures of di18:1 PC and di18:1 PE. MII formation in bilayers composed of 75% di18:1 PE and 25% di18:1 PC was found to be equivalent to that observed in bilayers where 50% of the acyl chains are 22:6n-3 (Brown, 1994). In a second set of measurements, it was found that diphytanoyl PC was as effective as di22:6 PC in promoting MII formation in PC/PE/PS mixtures, which mimicked the headgroup composition of the native rod outer segment disk membrane (Brown, 1994). Based on these findings, Brown has proposed that MII formation is facilitated by a lipid bilayer, which has curvature strain because the formation of MII results in a release of the curvature strain in the bilayer adjacent to rhodopsin (Brown, 1994). Although these studies provide a strong inference relative to the role of curvature strain in modulating MII formation, this interpretation suffers from a lack of direct measurements of curvature strain on the reconstituted membrane systems used in this study.

3.3. Acyl Chain Packing

An alternative explanation of the promotion of MII formation comes from direct measurements of acyl chain packing in rhodopsin-containing bilayers composed of a series PCs with varying levels of acyl chain unsaturation and cholesterol (Litman & Mitchell, 1996b; Mitchell et al., 1990; Mitchell et al., 1992). Acyl chain packing was characterized by analyzing the decay of fluorescence anisotropy of the membrane probe DPH in terms of the orientation distribution of DPH in the bilayer. The orientation distribution of DPH was summarized by a parameter, Fv, which is a measure of the difference between the DPH orientation distribution in the sterically restricted space of the bilayer and that which would be observed for an unrestricted free-tumbling DPH molecule (Mitchell & Litman, 1998a; Straume & Litman, 1987). Fv is positively correlated with the acyl chain packing free volume. Both MII formation and, Fv were measured in bilayers composed of a variety of PCs with and without cholesterol.

Examples of the effects of acyl chain composition and cholesterol on the MI-MII equilibrium constant, Keq, are summarized in Fig. 4. The equilibrium constant for the formation of MII from its inactive precursor MI, Keq, was measured in these rhodopsin-containing vesicles. For each acyl chain composition, Keq was determined to be linearly correlated with Fv in a manner that was independent of cholesterol content (Mitchell et al., 1990; Mitchell et al., 1992). These correlations demonstrate that each acyl chain composition produces a unique correlation between MII formation and acyl chain packing that is not altered by cholesterol. The bars in Fig. 4 demonstrate that high levels of acyl chain unsaturation enhance MII formation. The linear correlations between Keq and Fv demonstrate that this enhancement is related to the higher degree of disorder in acyl chain packing, resulting in increased acyl chain packing free volume. An unexplored question is whether the enhancement of MII formation by PE phospholipid groups is also correlated with acyl chain packing disorder.

In visual signal transduction, activation proceeds from the receptor, rhodopsin, to the effector, phosphodiesterase (PDE), via the visual G protein, Gt. Each MII sequentially binds and activates up to 100 Gt, thus MII-Gt binding initiates the first stage of signal amplification in the visual pathway. Litman et al. (2001) have studied the phospholipid acyl chain dependence of the kinetics of formation of both the MII conformation and the

Fig. 4. Examples of the effects of acyl chain composition (white bars) and cholesterol (gray bars) on Keq for the MI-MII equilibrium of photolyzed rhodopsin at 37°C. Higher values of Keq correspond to higher equilibrium concentrations of MII, the state of photolyzed rhodopsin that participates in visual signal transduction by binding the visual G protein.

Fig. 4. Examples of the effects of acyl chain composition (white bars) and cholesterol (gray bars) on Keq for the MI-MII equilibrium of photolyzed rhodopsin at 37°C. Higher values of Keq correspond to higher equilibrium concentrations of MII, the state of photolyzed rhodopsin that participates in visual signal transduction by binding the visual G protein.

MII-Gt complex. The temporal nature of the interaction of MII and Gt is characterized by the ratio of the rate of formation of the MII-Gt complex divided by the formation rate of MII. This ratio varied from 1.39 to 4.95 in native disk membranes and 18:0,18:1 PC bilayers, respectively. In 18:0,22:6 PC bilayers, the ratio had an intermediate value of 3.46.

An important feature of 22:6n-3-containing bilayers is the ability to buffer the inhibitory effects of cholesterol. The inclusion of 30 mol% cholesterol in an 18:0,22:6 PC bilayer had relatively little effect on MII coupling to Gt, whereas this level of cholesterol in 18:0,18:1 PC bilayers resulted in a ratio of 9.1 and an increase in lag time for complex formation of 6.5-fold, relative to native disk membranes. Complex formation involves MII and Gt diffusion in the surface of the membrane. The increased lag time suggests that the diffusion process is dramatically slowed by the presence of cholesterol in 18:0,18:1 PC, whereas this process is relatively unaffected by cholesterol in 18:0,22:6 PC bilayers. A delay in the coupling of MII with Gt decreases the response time of the pathway. In addition to the kinetics of MII-Gt complex formation, a reduced binding affinity of MII to Gt was observed in 18:0,18:1 PC relative to 18:0,22:6 PC bilayers (Niu, Mitchell, and Litman, unpublished results). The addition of cholesterol reduced the binding affinity of MII for Gt to a greater degree in 18:0,18:1 PC bilayer than in 18:0,22:6 PC bilayers. Thus, signal amplification along the pathway will be reduced in less unsaturated bilayers. These data demonstrate explicitly that 22:6n-3-containing phospholipids can buffer the inhibitory effects of cholesterol in a signaling pathway and highlight the potential importance of 22:6n-3 acyl chains in optimizing both the response time and magnitude of response in signaling pathways.

In the visual pathway, the activity of the PDE is a measure of the integrated pathway activity. Litman et al. (2001) studied the phospholipid acyl chain dependence of the light-stimulated PDE activity. This study was carried out in reconstituted systems, which included Gt, PDE, and rhodopsin in unilamellar vesicles, whose phospholipid composition was either 16:0,18:1 PC or 16:0,22:6 PC. Each rhodopsin absorbing a photon is analogous to an agonist-bound receptor. A level of 1 in 1000 rhodopsin molecules activated by light produced 59% of the activity obtained in native disk membranes for rhodopsin in 16:0,22:6 PC bilayers and only 26% of disk activity for rhodopsin in 16:0,18:1 PC bilayers. Under conditions of saturating stimulation, 97% of the activity of native disk-membrane response was observed in the 16:0,22:6 PC vesicles system, whereas only 50% of the disk activity was seen in 16:0,18:1 PC vesicles. Here again, the system properties are optimized in 22:6n-3-containing bilayers.

The presence of lateral domains in di22:6 PC bilayers demonstrates an additional mechanism whereby 22:6n-3-containing bilayers can enhance signaling processes. If, in addition to rhodopsin, Gt and PDE also show preferential partitioning into regions rich in 22:6n-3, then lateral domain formation will increase the efficiency of association of these proteins by reducing the diffusion pathway for their interaction and increasing their effective concentration in the region of the microdomains.

Delicious Diabetic Recipes

Delicious Diabetic Recipes

This brilliant guide will teach you how to cook all those delicious recipes for people who have diabetes.

Get My Free Ebook

Post a comment