Clinical and Experimental Pharmacology and Physiology (2003) 30, 559–564
Muscle Mechanics and Energetics: A Comparative View
MEMBRANES AS METABOLIC PACEMAKERS Paul L Else* and Anthony J Hulbert† Metabolic Research Centre, Departments of *Biomedical and †Biological Sciences, University of Wollongong, Wollongong, New South Wales, Australia
SUMMARY 1. In the present review, we suggest that a few common processes linked to membranes consume the majority of energy used by most organisms. 2. Membranes may act as metabolic pacemakers through changes in lipid composition, altering membrane characteristics and the working environment of membrane proteins. 3. Experiments involving membrane exchanges show predictable changes in protein activities (sodium pump) that are dependent upon the type of membrane used. 4. Potential mechanisms discussed include fluidity, electrical fields, surface area requirements of lipids and peptide–lipid interactions. Key words: body size, development, ectothermy, endothermy, lipids, metabolism, molecular activity, Na+/ K+ATPase, sodium pump.
INTRODUCTION Biology is laden with clever hypotheses and theories to explain the diversity and functioning of living organisms. Paramount in the generation of many of these ideas is the comparison of different species that provides the variation and models for investigation in all areas of biology. In this contribution, we show the value of the comparative approach in understanding the basic mechanisms underlying metabolism, specifically the role of membrane composition. Over the past two decades, our work has concentrated on understanding the bases for metabolism.1,2 During this time, we have used three comparative models of metabolism. The first is ectothermic (cold-blooded) and endothermic (warm-blooded) vertebrates, where body mass- and temperature-matched species commonly display fivefold differences in their weight-specific metabolism at the whole-animal level.3 The second comparative
Correspondence: Associate Professor PL Else, Department of Biomedical Science, University of Wollongong, Northfields Avenue, Wollongong, NSW 2522, Australia. Email:
[email protected] Presented at Muscle Mechanics and Energetics: A Comparative View, Melbourne, October 2002. The papers in these proceedings have been peer reviewed. Received 14 November 2002; revision 5 March 2003; accepted 8 March 2003.
model used has been that of body size, where small and large species within a class, such as mammals (shrew to elephant), show up to 100-fold differences in mass-specific metabolism.4 The final model has been animal development, where, within a species (e.g. rat), two- to threefold differences in mass-specific organismal metabolism are common.5 From the results of our own work using these models and those of many other researchers, three simple ideas have become clear.
IDEAS The first idea is that a few processes consume the majority of energy used by most organisms. No major new cellular processes make up the differences in metabolism between organisms that can be in the order of 100-fold plus. Variation in functions performed by different organisms still involves the same basic energy consuming processes common to all organisms. In essence, there are many processes supported by a few energy consuming processes. For example, the sodium pump, a membrane-bound protein that actively regulates Na+/K+ levels across cell membranes, underpins numerous important and diverse physiological functions (e.g. nerve and muscle function, nutrient uptake, electrolyte reabsorption or osmotic regulation that allows organisms such as the Nautilus to remove sea water from the chambers of its shell). The second idea is that the major energy consuming processes common to organisms account for similar proportions of total metabolism in animals with large differences in their overall rates of metabolism. As an example, the basic cost of living for mammals (using oxygen consumption as an overall measure) may be broken down into a small number of major energy consuming processes as shown in Table 1. Our studies and the work of others (for a review, see Hulbert and Else6) would also support this as a general breakdown of metabolism for most vertebrates. The third idea is that the major energy consuming processes are either directly or indirectly associated with membranes. For some processes, this association is obvious and direct (e.g. Ca2+ pumping and the proton pump leak cycle). For others, such as protein synthesis, the association is indirect, such as needed to move amino acids across membranes or in the functional association that ribosomes often share with membranes of the endoplasmic reticulum.
HYPOTHESIS Combining these ideas we have developed a working hypothesis that we call the ‘membranes as pacemakers of metabolism hypo-
560
PL Else and AJ Hulbert
thesis’. In its simplest form, the hypothesis proposes that variation in animal metabolism is due to membranes setting the overall pace of a small number of common processes. To explore this hypothesis, we have used the proton pump leak cycle of mitochondria7 and the sodium pump leak cycle of the cell membrane,8 which, together, constitute up to 35% of basal energy use. In this paper, we will concentrate primarily on the sodium pump because it is one of the easiest, most identifiable components Table 1 Metabolic energy utilization Energy consuming process Non-mitochondrial oxygen consumption Proton pump-leak cycle across mitochondrial inner and cristae membrane Active transport Sodium pump Calcium pump Protein Synthesis Other assorted processes Adapted from Rolfe and Brown.21
% Basal metabolism 10 20
15 5 20% 30
of metabolism in organisms and is an example of a ubiquitous membrane-bound protein. As a membrane-bound protein, the function of the sodium pump is influenced by the same environment as other primary and secondary active transporters plus channels.
USING THE SODIUM PUMP If we were to measure the maximum activity (Na+/K+-ATPase) and concentration of the sodium pump across a broad range of ecto- and endothermic vertebrates, young and old, small and large mammals, not surprisingly we would encounter substantial variation.8 However, if these same activities are divided by their respective concentrations for each tissue, we would find tissues from ectotherms, young mammals and mammals of larger size tend to display lower (1000–3000 ATP/min) activities, whereas tissues from endotherms, adult and smaller-sized mammals show higher (6000–15 000 ATP/min) molecular activity values when measured under the same conditions. In considering potential explanations for these differences, there are two likely possibilities or combinations thereof. The first is that the sodium pumps are different; the second is that the environment
Fig. 1 Sodium pump molecular activity of kidney and brain microsomal preparations following delipidating detergent (sodium deoxycholate; DOC) treatments and subsequent reconstitution with either original or alternative species heatinactivated membrane. Each symbol represents the mean of two to eight preparations. Error bars, where present, represent the SEM. (Reproduced with permission from Else and Wu.2)
Membranes as metabolic pacemakers is different (i.e properties of the membrane or other factors; for a recent and extensive review of ‘other factors’, see Therien and Blostein9). The sodium pump, as a functional unit, consists of -, - and -subunits. The function of the small peptide unit, which is part of the FYXD gene family of proteins (described after the invariant motif of phenylalanine-X-tyrosine-aspartate found at the beginning of the signature sequence of these small proteins) that are associated with a large variety of membrane-bound proteins, is an area of intense current research and has been linked, along with other functions, with the potential to increase molecular activity.9 The -subunit seems to act as a placement and insertion tag into the cell membrane, whereas the large -subunit is the catalytic unit of the protein.10 Of the four different isozymes currently known, 4 is found only in sperm, 2 and 3 are found, to varying degrees, in an assortment of tissues and 1 is ubiquitous and, in knock-out mice, is the only form that equates with immediate non-viability of the organism.11 As investigations into more novel species continue, the various forms of the sodium pump are also being found in ectotherms, just as they have previously been found in mammals. Currently, there is no evidence to suggest major differences in the molecular activities of different isoforms of the sodium pump. This point is emphasized by the fact that mammalian tissues that express very different mixtures of isoforms share similar molecular activities, whereas in the kidney of ectotherms and endotherms where only the 1 form is expressed, different molecular activities exist.8 Taken together, these facts suggest isoforms are not the primary reason for the molecular activity differences found. In regard to the alternative explanation (i.e. differences in membrane environment), little information is available. What is known is that sodium pump activity, like that of other membrane proteins, is influenced by the lipid environment. However, most studies have tended to concentrate on the immediate annular lipid environment of the membrane proteins rather than on the general membrane environment.12 From our own work examining membrane composition of tissues isolated from ectotherms and endotherms, from old and young and small and large mammals, a number of basic trends in association with membrane lipid composition have emerged. Where examined, similarities seem common in the types and proportion of lipid head groups, the percentage of saturated fatty acids in the membrane and in the average chain length of fatty acids (acyl composition) at approximately 18 carbons. Consistent differences are found in the types of unsaturated fatty acids and overall unsaturation index. Ectotherms, very young mammals and large mammals tend towards more monounsaturated membrane fatty acids. Endotherms, adult mammals and smallsized animals tend to possess membrane fatty acids with higher overall levels of polyunsaturation, particularly docosahexanoeic acid (22 : 6n-3) and higher levels of overall unsaturation. In the few cases where higher levels of polyunsaturation are present but molecular activities are low, there tend to be increased levels of cholesterol or other disturbances in the expected fatty acid profiles of the membranes. The consistent positive correlation between molecular activity and increasing levels of polyunsaturation and unsaturation index means that, in general, endotherms, adult and smaller body sized animals where high molecular activities exist have fatty acids with more abundant C = C double bonds set at increasingly deeper
561
levels within the membrane bilayer. This is because C = C of monounsaturates normally begin at the seventh or ninth C in from the methyl end or middle of the bilayer. The n-6 and n-3 polyunsaturates have C = C starting from the sixth and third carbon in from the methyl end of the chain, respectively. This fact has the interesting consequence that n-3s are the only fatty acids that can place C = C deep into the ‘inner sanctum’ of membranes. The functional consequences of this placement are yet to be fully appreciated (for an overview of this area, see Hulbert and Else13).
FUNCTIONAL LINK Simple correlation and consistent associations cannot prove a functional link between molecular activity and membrane acyl composition; this requires experimentation. The current set of experiments that provide support for this functional link involve ‘cross-over’ experiments.14 These experiments involve taking membranes from tissues of species that display high molecular activities and crossing them over with membrane from the same tissue of a species that shows lower molecular activity and measuring the molecular activities of the affected sodium pumps. These experiments involve detergent treatments that delipidate the preparations as the detergent disperses membrane lipids and lowers molecular activities in both species to similar levels. Enzyme activity is then reconstituted by the addition of heatinactivated membrane (with destroyed sodium pump activity but unaltered fatty acid composition) from either the original source or from that of the other species. Results of some of these experiments using rat and toad are shown in Fig. 1. These experiments show that, when reconstituted with original membrane, the molecular activity of each preparation returns to its original level. However, when the same membrane is reconstituted with membrane from the other species, the movement in molecular activity reflects that of the added membrane. A reduction in pump molecular activity, as occurs when toad membrane is added to a rat sodium pump preparation, is easy to explain as ineffective reconstitution. However, much stronger evidence for the influence of membrane lipids on molecular activity is provided when pumps significantly increase their molecular activity compared with their original level, as occurs when rat membrane is added to toad sodium pumps. Membrane ‘cross-over’ experiments have also been conducted using neonate and adult rat membranes. Results of these experiments (PL Else and BJ Wu, unpubl. obs., 2000) again show molecular activities of each preparation shifting in the direction of the added membrane source (i.e. original membrane restored original activity, neonate membrane reduced the molecular activity of sodium pumps in adult membranes and adult membranes increased the molecular activity of neonate pumps). In the experiments shown in Fig. 1 and those associated with the neonate and adult rats, the changes in molecular activity were only partial shifts of up to 40% in the ectotherm–endotherm cross-overs and up to 60% in the neonate–adult rat cross-overs. These movements are small compared with the 300–400% differences that occur naturally. So the questions become: (i) is the contribution of the membrane to differences in sodium pump molecular activities small; or (ii) are the results being limited by the reconstitution technique in the level of membrane exchange capable of occurring?
562
PL Else and AJ Hulbert
For example, in the case of reconstitution in original membrane, if the lipid exchange is small it is not surprising that the preparations regain their original activity. Likewise, if the exchanges are small for the species cross-over, the benefit or detriment to molecular activities are also likely to be small because the detergent treatment cannot totally remove endogenous lipids because the ability to reconstitute activity is lost in the process. To address this question, a further set of experiments was designed that added a second reconstitution attached to the original technique, where the membrane preparations underwent two reconstitutions with each membrane source. To obtain the quantities of membrane required for these experiments, membrane from kidney of cow and crocodile was used. Cow and crocodile sodium pumps display molecular activities of approximately 3900 and 730 ATP/min, respectively, reflecting both an endotherm–ectotherm difference and a body size difference compared with rat and toad. The results of this experiment (PL Else and BJ Wu, unpubl. obs., 2000) show no change in molecular activity following two reconstitutions in each preparation using original membrane. They also show a clear stepping down of molecular activity when crocodile membrane was added to cow pumps (with a final 380% decrease). More importantly, they also show a clear stepping up of
Table 2 Electrical field strengths of phosphatidylcholine phospholipid of varying acyl composition PC–acyl chains
Temperature (C)
Ψd (mV)
13 : 0 16 : 0 18 : 0 16 : 1 16 : 0/18 : 1n9 18 : 1n9 20 : 1n9 18 : 2n6 18 : 3n3 20 : 4n6
20 45 60 20 20 20 20 20 20 20
293 295 283 182 216 171 128 108 71 62
Phosphatidylcholine was used in all experiments with all double bonds in cis conformation. Dipole potentials (Ψd) and lipid packing density data are estimated from fluorescence excitation of styrylpyridinium dyes. Temperatures used exceeded gel-fluid phase transition points for each phospholipid (data adapted from Clarke16).
pump molecular activity when cow membrane was added to crocodile pumps (a final 200% increase). The results clearly suggest that membrane composition plays a primary role in determining the molecular activity of the sodium pump. Of course, this is a beginning in our understanding of the importance of membranes in determining the final activity of membrane-bound proteins. There are numerous questions that still need to be resolved. For example, what are the critical properties of membranes altered by membrane composition that determine the final molecular activity of membrane-bound protein? Are there specific lipid species with combinations of head and acyl chains that are especially important and what is the role of cholesterol?
MECHANISMS To understand the effect that membranes can have upon the activity of their intrinsic proteins, it will be necessary to provide functional mechanisms. There are a number of differing ideas as to how membranes of differing composition may affect the activity of their proteins. The adjustment of membrane acyl composition that has been found to occur due to temperature, pressure and the ionic strength of solutions etc., has often invoked the requirement to maintain fluidity membrane state in order to facilitate protein function. Although changes in acyl composition do occur in response to environment, adequate adjustment in fluidity can usually be achieved through adjustment in monounsaturates alone (see Hulbert and Else13 for an extended discussion). Therefore, the role or requirement for polyunsaturates in such adjustments is unclear unless, in addition to pure fluidity adjustment, activity adjustment of the membrane proteins is also required. This idea is supported in a study of cold-acclimated trout,15 where the increased activity of erythrocyte K+ influx measured in cold-acclimated compared with control animals was found to be due to increased molecular activity of the sodium pump. The activity of the pumps from the cold-acclimated fish was also higher at both the cold and control (warmer) temperatures compared with those from the control animals. Trout are known to polyunsaturate their membranes in response to cold; therefore, apart from fluidity adjustment, net molecular activity changes also occurred in the poylunsaturated environment. New research is showing that differences in the acyl composition of membranes may result in changes to electrical fields that exist
Fig. 2 Relationship between the molecular activity of the sodium pump and the average molecular area (at a surface pressure of 30 mN/m) of (a) membrane lipids and (b) phospholipids from the kidney and brain microsomes of rats (Rattus norvegicus) and cane toads (Bufo marinus). For (a): R = 0.92 (P < 0.001); for (b): R = 0.82 (P < 0.001). Data are the mean±SEM of six determinations. Each insert graph shows the relationship for the 24 individual data points that contributed to the means (Reproduced with permission from Wu et al.17)
Membranes as metabolic pacemakers within the bilayer. These electrical fields have been shown to vary in strength depending upon the acyl composition.16 As shown in Table 2, lipids with saturated fatty acids show high field strengths (300 mV), whereas membranes made of unsaturated lipids have significantly reduced field strengths (100 mV). How these intramembrane electrical fields influence lipid protein interactions within membranes is unknown, but may provide an alternative explanation to phenomena often attributed to fluidity. Linked to this work are our recent experiments that show that the physical properties contributed by lipids to membranes may also be a factor in determining the molecular activity of proteins.17 When lipids or phospholipids surrounding the sodium pump are isolated, resuspended as monolayers and ‘surface area/pressure’ isotherms measured, the area occupied by the average lipid or phospholipid at a standard pressure or pressure exerted at standard area show highly significant relationships with molecular activities from the original sample. For example, Fig. 2 shows the average molecular area of lipids and phospholipids at a constant surface pressure of 30 mN/m. The overall relationship shows increasing molecular activity with increasing area occupied by lipids. The correlations between molecular activity and the area occupied by lipids in membranes fits with the polyunsaturation level demonstrated by these tissues. The functional link here with membrane electrical field strengths is (as also pointed out by Clarke16) that increasing unsaturated lipids with kinked carbon chains increasingly disperse the head groups of the lipids and, therefore, the electrical field. Figure 2 also shows the well-known condensation effect that cholesterol has upon the surface area of lipids (comparing the area occupied by lipids, which includes cholesterol vs phospholipids that lack cholesterol). This may fit with the tendency of cholesterol to reduce active transport,18 because it would also act to decrease the average area occupied by lipids within a membrane. A final area of new research is the discovery of the -unit, a small single span membrane peptide associated with the sodium pump, part of the FYXD gene protein family that seem closely associated with numerous membrane-bound proteins. Although current work has primarily been associated with establishing the composition, family associations and relative presence in different tissues, some functional work is now also appearing. Interestingly, loosening the association between the - and -units of the sodium pump appears to produce an increase in enzyme activity.19 This can be elicited with mild detergent treatments that have long been known to produce an initial increase in enzyme activity before the drop in activity that occurs with increasing detergent concentration.20 If the -unit applies a ‘brake’ to the activity of the sodium pump in some tissues and mild detergent treatments act to loosen its inhibitory effect, may not different fatty acids associated with membrane lipids have similar ‘loosening’ effects to differing degrees? This idea is interesting considering the broad range of membrane proteins that the FYXD family associate with and adds to the potential list of mechanisms that may allow membrane composition to influence the activity of membrane bound proteins. Finally, it is not possible to conclude this paper without considering the potential role that changes in the molecular activity may have played in the evolution of endothermy. As mentioned previously, in response to cold acclimation many ectotherms polyunsaturate their membranes and some, particularly those
563
specialized at sodium pumping, display high molecular activities (see discussion in Else et al.14). If the evolution of endothermy involved small mammalian and avian ancestors moving into a colder niche environment, this would couple membrane polyunsaturation, higher molecular activities and the potential for increased metabolism. A shift in selection from unsaturation of membranes to accommodate functioning in the cold to the potential benefits of increased metabolism, using the same mechanism, would certainly favour the evolution of endothermy. Although the discussion in the present paper has centred on the sodium pump and membrane composition, we support a broader perspective. The sodium pump is only one example of many membrane-bound proteins and it is interesting to entertain the idea that other proteins may be similarly affected by membrane composition. This would provide support for our proposition that membranes set the overall pace of metabolism.
ACKNOWLEDGEMENTS We acknowledge Ben Jing Wu, whose PhD work contributed much of the data used in this paper. We also thank the Australian Research Council for supporting this work.
REFERENCES 1. Else PL, Hulbert AJ. Comparison of the ‘mammal machine’ and the ‘reptile machine’: Energy production. Am. J. Physiol. 1981; 240: R3–9. 2. Else PL, Wu BJ. Does membrane composition determine the molecular activity of the sodium pump? In: K Taniguchi, S Kaya (eds). Na/ K-Atpase and Related Atpases. Elsevier, Oxford. 2000; 571–4. 3. Hemmingsen AM. Energy metabolism as related to body size and respiratory surfaces and its evolution. Rep. Steno. Meml Hosp. Nord. Insulin Lab. 1960; 9: 1–110. 4. Kleiber M. The Fire of Life. Wiley, New York. 1961. 5. Else PL. Oxygen consumption and sodium pump thermogenesis in a developing mammal. Am. J. Physiol. 1991; 261: R1575–8. 6. Hulbert AJ, Else PL. Membranes underlying the cost of living in animals. Annu. Rev. Physiol. 2000; 62: 207–35. 7. Brookes PS, Buckingham JA, Tenreiro AM, Hulbert AJ, Brand MD. The proton permeability of the inner membrane of liver mitochondria from ectothermic and endothermic vertebrates and from obese rats: Correlations with standard metabolic rate and phospholipid composition. Comp. Biochem. Physiol. B 1998; 119: 325–34. 8. Else PL, Windmill DJ, Markus V. Molecular activity of sodium pumps in endotherms and ectotherms. Am. J. Physiol. 1996; 271: R1287–94. 9. Therien AG, Blostein R. Mechanisms of sodium pump regulation. Am. J. Physiol. Cell Physiol. 2000; 279: C541–66. 10. Kaplan JH. Biochemistry of Na,K-ATPase. Annu. Rev. Biochem. 2002; 71: 511–35. 11. Lingrel JB, Moseley AE, Cougnon MH et al. Functional roles of alpha isoforms of the Na,K-ATPase. Ann. N.Y. Acad. Sci. 2003; 986: 345–59. 12. Kimelberg HK, Papahadjopoulos D. Phospholipid requirements for (Na+-K+)-ATPase activity: Head-group specificity and fatty acid fluidity. Biochim. Biophys. Acta 1972; 282: 277–92. 13. Hulbert AJ, Else PL. Membranes as possible pacemakers of metabolism. J. Theor. Biol. 1999; 199: 257–74. 14. Else PL, Wu BJ. What role for membranes in determining the higher sodium pump molecular activity of mammals compared to ectotherms. J. Comp. Physiol. B 1999; 169: 296–302. 15. Raynard RS, Cossins AR. Homeoviscous adaptation and thermal compensation of sodium pump of trout erythrocytes. Am. J. Physiol. 1991; 260: R916–24.
564
PL Else and AJ Hulbert
16. Clarke RJ. Effect of lipid structure on the dipole potential of phosphatidylcholine bilayers. Biochim. Biophys. Acta 1997; 1327: 269–78. 17. Wu BJ, Else PL, Storlien LH, Hulbert AJ. Molecular activity of Na+/ K+-ATPase from different sources is related to the packing of membrane lipids. J. Exp. Biol. 2001; 204: 4271–80. 18. Bastiaanse EML, Hold KM, Van der Laarse A. The effects of membrane cholesterol content on ion transport processes in plasma membranes. Cardiovasc. Res. 1997; 33: 272–83.
19. Cornelius F, Mahmmoud YA, Christensen HRZ. Modulation of NaKATPase by associated small trnasmembrane regulatory proteins and by lipids. J. Bioenerg. Biomembr. 2001; 33: 415–23. 20. Skou JC, Esmann M. Preparation of membrane-bound and of solubilized (Na+ + K+)-ATPase from rectal glands of Squalus acanthias. Biochim. Biophys. Acta 1979; 508: 436–44. 21. Rolfe DFS, Brown GC. Cellular energy utilization and molecular origin of standard metabolic rate in mammals. Physiol. Rev. 1997; 77: 731–58.
