Structural insights into the stability and flexibility of unusual erythroid spectrin repeats

Erythroid spectrin, a major component of the cytoskeletal network of the red cell which contributes to both the stability and the elasticity of the red cell membrane, is composed of two subunits, alpha and beta, each formed by 16-20 tandem repeats. The properties of the repeats and their relative arrangement are thought to be key determinants of spectrin flexibility. Here we report a 2.4 A resolution crystal structure of human erythroid beta-spectrin repeats 8 and 9. This two-repeat fragment is unusual as it exhibits low stability of folding and one of its repeats lacks two tryptophans highly conserved among spectrin repeats. Two key factors responsible for the lower stability and, possibly, its flexibility, are revealed by the structure. A third novel feature of the structure is the relative orientation of the two repeats, which increases the range of possible conformations and provides new insights into atomic models of spectrin flexibility.     

结合AFM探测副伤寒沙门氏茵B感染红细胞膜的生物力学特性

采用原子力显微镜和衍射显微术,在纳米精确尺度探测副伤寒沙门氏菌B（Sp B）感染宿主红细胞（RBC）膜微观结构和力学特性,涉及细胞的形变、膜面内剪切模量和弯曲模量。结合这两种单分子测量技术,利用相关的数学模型表述RBC膜对菌体印B的入侵非常敏感。实验结果显示,不同感染期间的SpB寄生菌体,能够引起宿主RBC膜结构改变,形变能力降低,膜剪切模量和弯曲模量显著增加。这些力学特性的变化影响RBC的输氧和循环功能。实验结果表明,印B具有独特的鞭毛调控系统,入侵的毒性菌体寄生蛋白与血影蛋白网络中的运输蛋白有特异结合位点,导致RBC膜骨架网络、波动力学和细胞内、外基质都产生应激反应,这有可能为理解勋曰感染RBC的发病机理和寄生途径提供一些新的实验思路和分析依据。     

Adhesively-tensed cell membranes: lysis kinetics and atomic force microscopy probing

Membrane tension underlies a range of cell physiological processes. Strong adhesion of the simple red cell is used as a simple model of a spread cell with a finite membrane tension-a state which proves useful for studies of both membrane rupture kinetics and atomic force microscopy (AFM) probing of native structure. In agreement with theories of strong adhesion, the cell takes the form of a spherical cap on a substrate densely coated with poly-L-lysine. The spreading-induced tension, sigma, in the membrane is approximately 1 mN/m, which leads to rupture over many minutes; and sigma is estimated from comparable rupture times in separate micropipette aspiration experiments. Under the sharpened tip of an AFM probe, nano-Newton impingement forces (10-30 nN) are needed to penetrate the tensed erythrocyte membrane, and these forces increase exponentially with tip velocity ( approximately nm/ms). We use the results to clarify how tapping-mode AFM imaging works at high enough tip velocities to avoid rupturing the membrane while progressively compressing it to a approximately 20-nm steric core of lipid and protein. We also demonstrate novel, reproducible AFM imaging of tension-supported membranes in physiological buffer, and we describe a stable, distended network consistent with the spectrin cytoskeleton. Additionally, slow retraction of the AFM tip from the tensed membrane yields tether-extended, multipeak sawtooth patterns of average force approximately 200 pN. In sum we show how adhesive tensioning of the red cell can be used to gain novel insights into native membrane dynamics and structure.     

Cross bonding and stiffening of the red cell membrane

Cross bonding and stiffening of the human red cell membrane was studied using treatments with SH, amino, and carboxyl reagents, oxidizing and denaturing treatments and acidification. Membrane cross bonding was initiated when, after red cell treatment, opposite areas of the cytoplasmic face of the red cell membrane were brought into contact by cell shrinking. Membrane cross bonding was detected by light microscopy when this contact persisted upon swelling the cells in a hypotonic medium. Membrane stiffening was recorded as a decrease in elongation of red cells in the shear field of a viscous dextran solution. No correlation was found between membrane cross bonding and membrane stiffening. The results are explained by the existence of two modifications of spectrin, type I causing solely membrane stiffening, type II causing membrane cross bonding as well as membrane stiffening. The amino and carboxyl reagents caused only type I modification. The other treatments caused both types of modification although with varying proportions. The results support the previously suggested mechanism of membrane cross bonding which involves a rearrangement of spectrin similar to denaturation by heat or urea, a decrease in associations within the membrane skeletal network, and a lateral aggregation of membrane proteins. These changes are proposed to occur by the type II modification. The data further substantiate the membrane stiffening effect of inter- and intra-molecular cross linking of spectrin which is identified with the type I modification. Finally, hypotheses are presented concerning the mechanism of membrane stiffening due to type II modifications of spectrin.     

Shear-response of the spectrin dimer-tetramer equilibrium in the red blood cell membrane

The red cell membrane derives its elasticity and resistance to mechanical stresses from the membrane skeleton, a network composed of spectrin tetramers. These are formed by the head-to-head association of pairs of heterodimers attached at their ends to junctional complexes of several proteins. Here we examine the dynamics of the spectrin dimer-dimer association in the intact membrane. We show that univalent fragments of spectrin, containing the dimer self-association site, will bind to spectrin on the membrane and thereby disrupt the continuity of the protein network. This results in impairment of the mechanical stability of the membrane. When, moreover, the cells are subjected to a continuous low level of shear, even at room temperature, the incorporation of the fragments and the consequent destabilization of the membrane are greatly accentuated. It follows that a modest shearing force, well below that experienced by the red cell in the circulation, is sufficient to sever dimer-dimer links in the network. Our results imply 1) that the membrane accommodates the enormous distortions imposed on it during the passage of the cell through the microvasculature by means of local dissociation of spectrin tetramers to dimers, 2) that the network in situ is in a dynamic state and undergoes a "breathing" action of tetramer dissociation and re-formation.     

Elastic thickness compressibilty of the red cell membrane

We have used an ultrasensitive force probe and optical interferometry to examine the thickness compressibility of the red cell membrane in situ. Pushed into the centers of washed-white red cell ghosts lying on a coverglass, the height of the microsphere-probe tip relative to its closest approach on the adjacent glass surface revealed the apparent material thickness, which began at approximately 90 nm per membrane upon detection of contact (force approximately 1-2 pN). With further impingement, the apparent thickness per membrane diminished over a soft compliant regime that spanned approximately 40 nm and stiffened on approach to approximately 50 nm under forces of approximately 100 pN. The same force-thickness response was obtained on recompression after retraction of the probe, which demonstrated elastic recoverability. Scaled by circumferences of the microspheres, the forces yielded energies of compression per area which exhibited an inverse distance dependence resembling that expected for flexible polymers. Attributed to the spectrin component of the membrane cytoskeleton, the energy density only reached one thermal energy unit (k(B)T) per spectrin tetramer near maximum compression. Hence, we hypothesized that the soft compliant regime probed in the experiments represented the compressibility of the outer region of spectrin loops and that the stiff regime < 50 nm was the response of a compact mesh of spectrin backed by a hardcore structure. To evaluate this hypothesis, we used a random flight theory for the entropic elasticity of polymer loops to model the spectrin network. We also examined the possibility that additional steric repulsion and apparent thickening could arise from membrane thermal-bending excitations. Fixing the energy scale to k(B)T/spectrin tetramer, the combined elastic response of a network of ideal polymer loops plus the membrane steric interaction correlated well with the measured dependence of energy density on distance for a statistical segment length of approximately 5 nm for spectrin (i.e., free chain end-to-end length of approximately 29 nm) and a hardcore limit of approximately 30 nm for underlying structure.     

Occurrence of lipid receptors inferred from brain and erythrocyte spectrins binding NaOH-extracted and protease-treated neuronal and erythrocyte membranes

It was previously shown in model systems that brain spectrin binds membrane phospholipids. In the present study, we analysed binding of isolated brain spectrin and red blood cell spectrin to red blood or neuronal membranes which had been treated as follows: (1). extracted with low ionic-strength solution, (2). the above membranes extracted with 0.1 M NaOH, and (3). membranes treated as above, followed by protease treatment and re-extraction with 0.1 M NaOH. It was found that isolated, NaOH-extracted, protease-treated neuronal and red blood cell membranes bind brain and red blood cell spectrin with moderate affinities similar to those obtained in model phospholipid membrane-spectrin interaction experiments. Moreover, this binding was competitively inhibited by liposomes prepared from membrane lipids. The presented results indicate the occurrence of receptor sites for spectrins that are extraction- and protease-resistant, therefore most probably of lipidic nature, in native membranes.     

Ultracentrifugal analysis of the junction complexes of the red cell membrane cytoskeletal network: application to hereditary spherocytosis and metabolically depleted cells

A method has been developed for the assessment of the number of spectrin dimer units associated with each actin protofilament junction, in the membrane cytoskeletal network (i.e. the degree of branching) of the red cell. Ghosts are first exposed to elevated temperature at low ionic strength to dissociate some 65% of the spectrin tetramers (that link the network junctions) into dimers, without causing their release from the actin filaments. Non-ionic detergent is then added to solubilize the membrane itself with its intrinsic proteins, so as to liberate the cytoskeletal material, and the mixture is immediately examined in the analytical ultracentrifuge. The predominant components observed are isolated junctions (20 S), free spectrin dimers and the residual undissociated cytoskeletal material, with very minor components, probably corresponding to multiple junctions, linked by spectrin tetramers. The junction boundary is homogeneous within the accuracy of measurement and is taken to correspond to a complex containing six spectrin dimers, known to predominate in situ. About 17% of the total network is liberated in this form and 12% as free spectrin dimers. In hereditary spherocytosis both the size of the junction complex (as reflected by its sedimentation coefficient) and the proportion of the complex and of free spectrin liberated are indistinguishable from normal values. We conclude that the reported deficit of spectrin in hereditary spherocytosis is not reflected by a lower degree of branching of the network, and, if the membrane area is not correspondingly reduced, this must mean that the junctions are more widely spaced and the spectrin tetramers therefore more extended. In metabolically depleted cells, in which the cytoskeletal proteins are known to be extensively dephosphorylated, there is no change in the sedimentation pattern and thus no detectable loss of spectrin from the junctions or weakening in the cohesion of the cytoskeletal network.     

Physical-chemical studies of spectrin

In recent years considerable progress has been made in the understanding of the structure and function of the red blood cell membrane. The protein spectrin, of high molecular weight and propensity for self-association, appears to play a major role, in concert with actin, in maintaining the shape and integrity of the membrane. A study of the physical-chemical properties of spectrin, and its size, shape, self-association pattern, and its interaction with other components, leads to a plausible model for the way this protein performs its biological role. The evidence from the structure and interactions of spectrin suggests a structure which is relatively symmetrical yet highly expanded, and which allows extensive, two-dimensional network formation with actin. In these respects, the structure of spectrin is quite different from that of myosin, to which it has often been likened.     

Spectrin properties and the elasticity of the red blood cell membrane skeleton

Two models of spectrin elasticity are developed and compared to experimental measurements of the red blood cell (RBC) membrane shear modulus through the use of an elastic finite element model of the RBC membrane skeleton. The two molecular models of spectrin are: (i) An entropic spring model of spectrin as a flexible chain. This is a model proposed by several previous authors. (ii) An elastic model of a helical coiled-coil which expands by increasing helical pitch. In previous papers, we have computed the relationship between the stiffness of a single spectrin molecule (K) and the shear modulus of a network (μ), and have shown that this behavior is strongly dependent upon network topology. For realistic network models of the RBC membrane skeleton, we equate μ to micropipette measurements of RBCs and predict K for spectrin that is consistent with the coiled-coil molecular model. The value of spectrin stiffness derived from the entropic molecular model would need to be at least 30 times greater to match the experimental results. Thus, the conclusion of this study is that a helical coiled-coil model for spectrin is more realistic than a purely entropic model.     

Structure of the erythrocyte membrane skeleton as observed by atomic force microscopy.

The structure of the membrane skeleton on the cytoplasmic surface of the erythrocyte plasma membrane was observed in dried human erythrocyte ghosts by atomic force microscopy (AFM), taking advantage of its high sensitivity to small height variations in surfaces. The majority of the membrane skeleton can be imaged, even on the extracellular surface of the membrane. Various fixation and drying methods were examined for preparation of ghost membrane samples for AFM observation, and it was found that freeze-drying (freezing by rapid immersion in a cryogen) of unfixed specimens was a fast and simple way to obtain consistently good results for observation without removing the membrane or extending the membrane skeleton. Observation of the membrane skeleton at the external surface of the cell was possible mainly because the bilayer portion of the membrane sank into the cell during the drying process. The average mesh size of the spectrin network observed at the extracellular and cytoplasmic surfaces of the plasma membrane was 4800 and 3000 nm2, respectively, which indicates that spectrin forms a three-dimensionally folded meshwork, and that 80% of spectrin can be observed at the extracellular surface of the plasma membrane.     

Spectrin, red cell shape and deformability. II. The antagonistic action of spectrin and sialic acid residues in determining membrane curvature in genetic spectrin deficiency in mice

In a companion paper, the shapes of spectrin deficient mouse erythrocytes were described; in contrast to previous assumptions, spherules with tethered microvesicles rather than true "spherocytes" were found. Thence, spectrin deficient mouse erythrocytes are endowed with an excess of surface area for the given volume but the membrane is assuming a highly positive curvature. Observations during and after the action of enzymes cleaving the red cell surface charge (Neuraminidase, Trypsin, Chymotrypsin) showed that the previously positive membrane curvature, as well as the tendency of the membrane to flow into fingerlike protrusions was completely abolished. The erythrocytes of the spectrin deficient, desialylated mouse erythrocytes assumed a variety of shapes, often discocytic or even stomatocytic, i.e. their membrane presented with negative curvature. However, while these desialylated membranes could be easily deformed (elongated) by shear flow they did not recoil elastically into any definitive configuration after removal of the deforming forces. It is concluded from these observations that spectrin (acting on the inner interface between membrane and cytoplasm) and sialic acid residues (acting on the outer interface between membrane and plasma) exert antagonizing effects on membrane curvature and membrane bending elasticity. Sialic acid residues, strongly charged and situated on the outer side of the cell, produce positive membrane curvature; this observation can most readily be explained by assuming that this mechanical effect is caused by repulsive coulombic forces expanding the outer half of the bilayer. To explain the effect of the spectrin-complex in counteracting positive or in producing negative membrane curvature, a similar expansive coulombic force acting between the highly charged residues has been postulated. Thence, a model for explaining the overall elastic behaviour of the normal mammalian red cell is developed which is based on the assumption of elastic interactions of proteinacous membrane components coupled to the lipid bilayer of the membrane.     

Composition of the ternary protein complex of the red cell membrane cytoskeleton

The red cell membrane skeletal network is constructed from actin, spectrin and protein 4.1 in a molar ratio of actin subunits/spectrin heterodimer/protein 4.1 of 2:1:1. This represents saturation of the actin filaments, since incubation with extraneous spectrin and protein 4.1 leads to no binding of additional spectrin, either to the inner surface of ghost membranes or to lipid-free membrane cytoskeletons. Partial extraction of spectrin from the membrane is accompanied by release of actin under all conditions. Regardless of the proportion of spectrin extracted, the molar ratio of spectrin dimers/actin subunits is constant at 1:2. This is not the result of release or cooperative breakdown of whole lattice junctions from the network, for the number of actin filaments, judged by capacity to nucleate polymerisation of added G-actin, remains unchanged even when as much as 60% of the total spectrin has been lost. A similar 1:2:1 stoichiometry characterises the complex formed when G-actin is allowed to polymerise in the presence of varying amounts of spectrin and protein 4.1. When this complex is treated with the depolymerising agent, 1 M guanidine hydrochloride, it breaks down into smaller units of the same stoichiometry. After cross-linking these can be recovered from a gel-filtration column. Complexes prepared starting from G-actin appear to be much more stable than those formed when spectrin and protein 4.1 are bound to F-actin.     

Mechanochemistry of protein 4.1's spectrin-actin-binding domain: ternary complex interactions, membrane binding, network integration, structural strengthening

Mechanical strength of the red cell membrane is dependent on ternary interactions among the skeletal proteins, spectrin, actin, and protein 4.1. Protein 4.1's spectrin-actin-binding (SAB) domain is specified by an alternatively spliced exon encoding 21 amino acid (aa) and a constitutive exon encoding 59 aa. A series of truncated SAB peptides were engineered to define the sequences involved in spectrin-actin interactions, and also membrane strength. Analysis of in vitro supramolecular assemblies showed that gelation activity of SAB peptides correlates with their ability to recruit a critical amount of spectrin into the complex to cross-link actin filaments. Also, several SAB peptides appeared to exhibit a weak, cooperative actin-binding activity which mapped to the first 26 residues of the constitutive 59 aa. Fluorescence-imaged microdeformation was used to show SAB peptide integration into the elastic skeletal network of spectrin, actin, and protein 4.1. In situ membrane-binding and membrane-strengthening abilities of the SAB peptides correlated with their in vitro gelation activity. The findings imply that sites for strong spectrin binding include both the alternative 21-aa cassette and a conserved region near the middle of the 59 aa. However, it is shown that only weak SAB affinity is necessary for physiologically relevant action. Alternatively spliced exons can thus translate into strong modulation of specific protein interactions, economizing protein function in the cell without, in and of themselves, imparting unique function.     

Calculation of a Gap restoration in the membrane skeleton of the red blood cell: possible role for myosin II in local repair

Human red blood cells contain all of the elements involved in the formation of nonmuscle actomyosin II complexes (V. M. Fowler. 1986. J. Cell. Biochem. 31:1-9; 1996. Curr. Opin. Cell Biol. 8:86-96). No clear function has yet been attributed to these complexes. Using a mathematical model for the structure of the red blood cell spectrin skeleton (M. J. Saxton. 1992. J. Theor. Biol. 155:517-536), we have explored a possible role for myosin II bipolar minifilaments in the restoration of the membrane skeleton, which may be locally damaged by major mechanical or chemical stress. We propose that the establishment of stable links between distant antiparallel actin protofilaments after a local myosin II activation may initiate the repair of the disrupted area. We show that it is possible to define conditions in which the calculated number of myosin II minifilaments bound to actin protofilaments is consistent with the estimated number of myosin II minifilaments present in the red blood cells. A clear restoration effect can be observed when more than 50% of the spectrin polymers of a defined area are disrupted. It corresponds to a significant increase in the spectrin density in the protein free region of the membrane. This may be involved in a more complex repair process of the red blood cell membrane, which includes the vesiculation of the bilayer and the compaction of the disassembled spectrin network.     

Native ultrastructure of the red cell cytoskeleton by cryo-electron tomography

Erythrocytes possess a spectrin-based cytoskeleton that provides elasticity and mechanical stability necessary to survive the shear forces within the microvasculature. The architecture of this membrane skeleton and the nature of its intermolecular contacts determine the mechanical properties of the skeleton and confer the characteristic biconcave shape of red cells. We have used cryo-electron tomography to evaluate the three-dimensional topology in intact, unexpanded membrane skeletons from mouse erythrocytes frozen in physiological buffer. The tomograms reveal a complex network of spectrin filaments converging at actin-based nodes and a gradual decrease in both the density and the thickness of the network from the center to the edge of the cell. The average contour length of spectrin filaments connecting junctional complexes is 46 ± 15 nm, indicating that the spectrin heterotetramer in the native membrane skeleton is a fraction of its fully extended length (∼190 nm). Higher-order oligomers of spectrin were prevalent, with hexamers and octamers seen between virtually every junctional complex in the network. Based on comparisons with expanded skeletons, we propose that the oligomeric state of spectrin is in a dynamic equilibrium that facilitates remodeling of the network as the cell changes shape in response to shear stress.     

Influence of the membrane undercoat on filipin perturbation of the red blood cell membrane

Filipin, a polyene antibiotic, interacts with beta-hydroxy sterols such as cholesterol in most cell membranes, forming bumps and pits that are visible by electron microscopy of freeze-fracture replicas. The markedly reduced perturbability of the red blood cell (RBC) membrane, compared to other cells, has been attributed to the constraining influence of the red cell membrane skeleton, the undercoat composed of spectrin, actin, and protein 4.1. To test the influence of the membrane skeleton on filipin-induced perturbation of the RBC membrane, we studied the interaction of filipin with red cells that were inherently devoid of spectrin and RBC in which spectrin had been crosslinked or denatured. These spectrin-deficient, crosslinked, and denatured cells have a fivefold increase in the number of filipin-induced perturbations as compared to control cells, despite equivalent membrane cholesterol content. These findings confirm that the spectrin-based membrane skeleton strongly influences the organization of the membrane so as to limit perturbation by filipin:cholesterol interaction and that for membranes in which the cholesterol content is known, filipin is a useful probe for testing the avidity of spectrin-based cytoskeletal attachment.     

Membrane protein organization in ATP-depleted and irreversibly sickled red cells

It has been proposed that the spectrin-actin submembrane network participates in control of red cell shape and deformability. We have examined ATP- and calcium-dependent changes in organization of spectrin in the membrane employing cross-linking of the nearest membrane protein neighbors by spontaneous or catalyzed (CuSO₄, O-phenanthroline) intermolecular disulfide couplings and two-dimensional sodium dodecyl sulfate polyacrylamide gel electrophoresis. Cross-linking of fresh red cells resulted in the formation of spectrin and actin dimers and tetramers. ATP-depleted red cells differed from fresh cells in the presence of an additional reducible polymer of MW > 1 × 10⁶ selectively enriched in spectrin. This polymer formed spontaneously when red cells were depleted of ATP under aerobic conditions. After anaerobic ATP depletion, the polymer formed in ghosts after cross-linking by catalytic oxidation. Polymerization was prevented by maintenance of ATP and coincided with an ATP-dependent discocyte-echinocyte transformation. This suggests that, in ATP-depleted red cells, spectrin is rearranged to establish closer contacts, and that this may contribute to the discocyte-echinocyte transformation. The introduction of greater than 0.5 mM Ca⁺⁺ into ghosts by inclusion in hemolysis buffer or into fresh red cells (but not ATP-depleted red cells) by treatment with ionophore A23187 spontaneously produced a nonreducible polymer which others have attributed to transamidative cross-linking of spectrin, band 3, and other proteins. Spontaneous formation of both polymer types (reducible in aerobically ATP-depleted red cells and nonreducible in fresh, Ca⁺⁺ enriched red cells) resulted in stabilization (“autocatalytic fixation”) of spheroechinocytic shape. Irreversibly sickled cells, which have increased calcium and decreased ATP, and exhibit a permanent membrane deformation, failed to form any of the above polymers. This suggests that in contrast to normal cells depleted of ATP in vitro, fixation of ISC shape in vivo is not related to Ca- and ATP-dependent membrane protein polymerization. However, ISCs had an increased propensity to form the reducible, spectrin-rich polymer during a subsequent metabolic depletion in vitro. This was associated with transformation of ISCs into spheroechinocytes. Similar echinocytic ISCs were found to constitute 5–10% of the densest fractions of freshly separated ISCs. ISCs then exhibit sphero-echniocyte transformation, both in vitro and in vivo. We propose that this is due to spectrin reorganization that presumably results from the progressively increasing calcium and decreasing ATP of ISCs. These data provide evidence of altered spectrin organization in membranes of ATP-depleted, calcium-enriched red cells in vitro and in vivo.     

An examination of the soluble oligomeric complexes extracted from the red cell membrane and their relation to the membrane cytoskeleton

A part of the spectrin extracted from red cell membranes at low ionic strength occurs in the form of a high-molecular weight oligomeric complex with actin and proteins 4.1 and 4.9. When the extraction is performed at 35 degrees, the spectrin is present in this complex as the dimer, all higher forms being dissociated. We have been unable to establish any correlation between the fraction of the spectrin thus complexed and the metabolic state of the cell. At least a large part of the complex appears to be a defined monodisperse species, sedimenting at 31S. The actin is present as short protofilaments. The average number of spectrin molecules associated with each molecule of complex has been studied by cytochalasin binding and electron microscopy. The complexes present the appearance in the electron microscope of spiders, in which the legs are spectrin dimers, attached to a globular element, containing by inference, actin and proteins 4.1 and 4.9; they are active in nucleating the polymerization of G-actin. The complexes are extremely stable, being resistant to dissociation under the conditions of the deoxyribonuclease assay, even after treatment with trypsin to degrade the actin-associated proteins. It is suggested that the complexes represent intact junctions of the membrane cytoskeletal network. Relevant structural features of the network are revealed by electron microscopy. The results lead to inferences concerning the mechanism of dissociation of the network from the membrane.     

Red Blood Cell Shape and Fluctuations: Cytoskeleton Confinement and ATP Activity

We review recent theoretical work that analyzes experimental measurements of the shape and fluctuations of red blood cells. Particular emphasis is placed on the role of the cytoskeleton and cell elasticity and we contrast the situation of elastic cells with that of fluid-filled vesicles. In red blood cells (RBCs), the cytoskeleton consists of a two-dimensional network of spectrin proteins. Our analysis of the wave vector and frequency dependence of the fluctuation spectrum of RBCs indicates that the spectrin network acts as a confining potential that reduces the fluctuations of the lipid bilayer membrane. However, since the cytoskeleton is only sparsely connected to the bilayer, one cannot regard the composite cytoskeleton membrane as a polymerized object with a shear modulus. The sensitivity of RBC fluctuations and shapes to ATP concentration may reflect the transient defects induced in the cytoskeleton network by ATP.