Analysis of alphavirus polypeptides by two dimensional polyacrylamide gel electrophoresis

Semliki Forest, Sindbis and Chikungunya viruses were grown and radio-labeled with [³H]-amino acids in Vero cells. Analysis of virus infected cell lysates by two dimensional polyacrylamide gel electrophoresis resulted in detection of polypeptides of molecular, weights corresponding to those of E1, P62, ns60, ns70/72 for Semliki Forest virus, the C, E1, 6K, 14K, PE2, P97, ns60, ns82 for Sindbis virus and E1. P62, P97, ns70/72 for Chikungunya virus. Charge and molecular weight heterogeneity in the precursor polypeptide P62 of Semliki Forest virus was detected. Structural poly-peptides e.g. E1 and E2 of Semliki Forest virus and C, E1, E2 of Sindbis virus and E1 of Chikungunya virus were detected when purified radiolabeled virus preparations were analyzed by two dimensional polyacrylamide gel-electrophoresis. Membrane glycoprotein E1 and E2 of Semliki Forest and E1 of Sindbis and Chikungunya viruses exhibited charge heterogeneity. In contrast to the marked difference in isoelectric points of E1 and E2 of Sindbis virus; E1 and E2 of Semliki Forest virus had almost identical isoelectric points.     

A tracking marker for the first dimension of the two-dimensional gel electrophoresis of ribosomal proteins

Protein content of different subcellular fractions from chick brain is compared by using Lowry, TCA—Lowry and Bradford assay methods. Caution is urged in application of Bradford's method to general assay for protein concentration in subcellular fractions.     

Inhibition of chloroplast protein synthesis by lincomycin and 2-(4-methyl-2,6- dinitroanilino)-N-methylpropionamide

Selective effects of lincomysin and cycloheximide in detached shoots of Pisum sativum on the synthesis of photosystem I and II proteins, and a chloroplast membrane protein of molecular weight 32000, confirm results obtained from studies of protein synthesis by isolated chloroplasts. A model is proposed in which one role of chloroplast ribosomes is to synthesize membrane proteins required for the immobilization of chloroplast components, such as photosystem I protein, which are synthesized by cytoplasmic ribosomes. 2-(4-Methyl-2,6-dinitroanilino)-N-methylpropionamide rapidly inhibits the synthesis of both the large and small subunits of Fraction I protein in greening detached pea shoots. This observation can be reconciled with the site of synthesis of the large subunit being in the chloroplast by a model which proposes that the small subunit is a positive initiation factor for the synthesis or translation of the messenger RNA for the large subunit.     

Adducin Is an In Vivo Substrate for Protein Kinase C: Phosphorylation in the MARCKS-related Domain Inhibits Activity in Promoting Spectrin–Actin Complexes and Occurs in Many Cells,Including Dendritic Spines of Neurons

Adducin is a heteromeric protein with subunits containing a COOH-terminal myristoylated alanine-rich C kinase substrate (MARCKS)-related domain that caps and preferentially recruits spectrin to the fast-growing ends of actin filaments. The basic MARCKS-related domain, present in α, β, and γ adducin subunits, binds calmodulin and contains the major phosphorylation site for protein kinase C (PKC). This report presents the first evidence that phosphorylation of the MARCKS-related domain modifies in vitro and in vivo activities of adducin involving actin and spectrin, and we demonstrate that adducin is a prominent in vivo substrate for PKC or other phorbol 12-myristate 13-acetate (PMA)-activated kinases in multiple cell types, including neurons. PKC phosphorylation of native and recombinant adducin inhibited actin capping measured using pyrene-actin polymerization and abolished activity of adducin in recruiting spectrin to ends and sides of actin filaments. A polyclonal antibody specific to the phosphorylated state of the RTPS-serine, which is the major PKC phosphorylation site in the MARCKS-related domain, was used to evaluate phosphorylation of adducin in cells. Reactivity with phosphoadducin antibody in immunoblots increased twofold in rat hippocampal slices, eight- to ninefold in human embryonal kidney (HEK 293) cells, threefold in MDCK cells, and greater than 10-fold in human erythrocytes after treatments with PMA, but not with forskolin. Thus, the RTPS-serine of adducin is an in vivo phosphorylation site for PKC or other PMA-activated kinases but not for cAMP-dependent protein kinase in a variety of cell types. Physiological consequences of the two PKC phosphorylation sites in the MARCKS-related domain were investigated by stably transfecting MDCK cells with either wild-type or PKC-unphosphorylatable S716A/S726A mutant α adducin. The mutant α adducin was no longer concentrated at the cell membrane at sites of cell–cell contact, and instead it was distributed as a cytoplasmic punctate pattern. Moreover, the cells expressing the mutant α adducin exhibited increased levels of cytoplasmic spectrin, which was colocalized with the mutant α adducin in a punctate pattern. Immunofluorescence with the phosphoadducin-specific antibody revealed the RTPS-serine phosphorylation of adducin in postsynaptic areas in the developing rat hippocampus. High levels of the phosphoadducin were detected in the dendritic spines of cultured hippocampal neurons. Spectrin also was a component of dendritic spines, although at distinct sites from the ones containing phosphoadducin. These data demonstrate that adducin is a significant in vivo substrate for PKC or other PMA-activated kinases in a variety of cells, and that phosphorylation of adducin occurs in dendritic spines that are believed to respond to external signals by changes in morphology and reorganization of cytoskeletal structures.     

Protein kinase C prevents oligodendrocyte differentiation: modulation of actin cytoskeleton and cognate polarized membrane traffic

In a previous study, we showed that activation of protein kinase C (PKC) prevents oligodendrocyte differentiation at the pro‐oligodendrocyte stage. The present study was undertaken to identify downstream targets of PKC action in oligodendrocyte progenitor cells. Activation of PKC induced the predominant phosphorylation of an 80‐kD protein, identified as myristoylated alanine‐rich C‐kinase substrate (MARCKS). Upon phosphorylation, MARCKS is translocated from the plasma membrane to the cytosol. Furthermore, PKC activation perturbed the organization of the actin cytoskeleton, causing a redistribution of actin filaments to the submembranous or cortical actin cytoskeleton. As a consequence, transport of a protein traffic marker, the vesicular stomatitis virus glycoprotein, from the trans‐Golgi network to the plasma membrane becomes perturbed. The effect of disruption of the actin filament network by cytochalasin D perfectly matched the effect of PKC. These data thus favor the existence of a causal relationship between actin rearrangement and docking and/or fusion of proteins to the plasma membrane. Interestingly, neither in control cells nor in PKC‐activated cells did another protein traffic marker, influenza hemagglutinin (HA), reach the cell surface. However, an eminent and specific accumulation of HA just underneath the plasma membrane became apparent upon PKC activation. Yet, this effect could not be simulated by cytochalasin D treatment. Therefore, these observations imply that although MARCKS represents a prominent PKC target site in regulating differentiation, another target involves the differential control of cognate polarized trafficking pathways, which are apparently operating in oligodendrocyte progenitor cells. © 1999 John Wiley & Sons, Inc. J Neurobiol 41: 385–398, 1999     

Two-dimensional electrophoresis of soybean root plasma membrane proteins solubilized by SDS and other detergents

Proteins solubilized from enriched soybean root plasma membrane with sodium dodecyl sulphate (SDS) and selected non-denaturing detergents (octyl-β-d-glucopyranoside, Zwittergent 312, Zwittergent 314, Zonyl FSK, and Nonidet P-40) were electrophoresed in two-dimensions by standard procedures. The basic electrophoretogram ‘fingerprint’ was similar for all detergents tested. However, differences in the total number of polypeptides resolved and the presence or absence of certain polypeptides on specific two-dimensional gels indicated some selectivity. Of all detergents tested, SDS solubilized the most polypeptides (ca 95) and provided the best resolution. The other detergents solubilized 50–80 polypeptides with varying resolution. Of those tested, octyl-β-d-glucopyranoside consistently provided the best balance between the number of polypeptides resolved (ca 70) and the level of resolution. The results suggest that selected detergents may prove useful in plant plasma membrane studies which require non-denaturing conditions.     

Alteration of Phosphoinositide Degradation by Cytosolic and Membrane-Bound Phospholipases after Forebrain Ischemia - Reperfussion in Gerbil: Effects of Amyloid Beta Peptide

The reperfusion of previously ischemic brain is associated with exacerbation of cellular injury. Reperfusion occasionally potentates release of intracellular enzymes, influx of Ca²⁺, breakdown of membrane phospholipids, accumulation of amyloid precursor protein or amyloid -(like) proteins, and apolipoprotein E. In this study, the effect of reperfusion injury on the activity of cerebral cortex enzymes acting on phosphatidyl [³H] inositol (PI) and [^l4C-arachidonoyl] PI was investigated. Moreover the effect of amyloid 25–35 on PI degradation by phospholipase(s) of normoxic brain and subjected to ichemia-reperfussion injury was determined. Brain ischemia in gerbils (Meriones unguiculatus) was induced by ligation of both common carotid arteries for 5 min and then brains were perfused for 15 min, 2 h and 7 days. Statistically significant activation of enzyme(s) involved in phosphatidylinositol degradation in gerbils subjected to ischemia-reperfusion injury was observed. Nearly all gerbils showed a higher activity of cytosolic PI phos-pholipase C (PLC) at 15 min after ischemia. Concomitantly, the significant enhancement of the level of DAG and AA radioactivity at this short reperfusion time confirmed the active PI degradation by phospholipase(s) in cerebral cortex and hippocampus. After a prolonged reperfusion time of 7 days after ischemia, both cytosolic and membrane-bound forms of PI-PLC were activated. The question arises if alteration of membranes by the degradation of phospholipids occurring after an ischemic episode potentates the effect of A on membrane-bound enzymes. A neuro-toxic fragment of amyloid, A 25–35, incubated in the presence of endogenous Ca²⁺, increased significantly the PI-PLC activity of normoxic brain. In its non-aggregated form, A 25–35 activates PI-PLC but in the aggregated form the enzymatic activity decreased. Thus, A 25–35 exerts a similar effect on the membrane-bound PI-PLC from normoxic brain or subjected to ischemia reperfussion injury. We conclude that the degradation of phosphatidylinositol by cytosolic phosphoinositide-phospholipase C may contribute to the pathophysiology of delayed neuronal death following cerebral ischemia. Thus, a specific inhibitor of this enzyme(s) may offer therapeutic strategies to protect the brain from damage triggered by ischemia. Ischemia-reperfusion injury had no effect on A-evoked alterations of synaptic plasma membrane-bound PI-PLC.     

Mucin synthesis and secretion by cultured tracheal cells: effects of collagen gel substratum thickness

Summary We previously demonstrated that human tracheobronchial epithelial (TBE) cells synthesize mucin and form mucous granules in culture when they are maintained on a collagen gel (CG) substratum, but not on a plastic tissue culture surface or a thin collagen-coated surface (Wu et al., Am. J. Respir. Cell Mol. Biol., 3:467–478; 1990). This observation led us to examine the effects of CG thickness on cell growth and differentiation in primary human/monkey TBE cell cultures. Using the same CG preparation, culture dishes with different thicknesses of CG substratum were prepared. In general, equivalent degrees of cell attachment and proliferation were observed in all cultures maintained on a collagen gel, independent of the thicknesses of CG substratum. However, a greater degree of mucin synthesis and secretion by the cells was observed as the thickness of the CG substratum was increased. Cultures maintained on a thick collagen gel (1 mm) exhibited greater apical membrane complexity, more pseudostratification, and more mucous granules than did cultures maintained on a thin CG substratum. The optimal culture surface for airway mucous cell differentiation contains more than 1-mm thickness of collagen gel substratum.     

Protein prenylation in an insect cell-free protein synthesis system and identification of products by mass spectrometry

To evaluate the ability of an insect cell-free protein synthesis system to carry out proper protein prenylation, several CAIX (X indicates any C-terminal amino acid) sequences were introduced into the C-terminus of truncated human gelsolin (tGelsolin). Tryptic digests of these mutant proteins were analyzed by MALDI-TOF MS and MALDI-quadrupole-IT-TOF MS. The results indicated that the insect cell-free protein synthesis system possesses both farnesyltransferase (FTase) and geranylgeranyltransferase (GGTase) I, as is the case of the rabbit reticulocyte lysate system. The C-terminal amino acid sequence requirements for protein prenylation in this system showed high similarity to those observed in rat prenyltransferases. In the case of rhoC, which is a natural geranylgeranylated protein, it was found that it could serve as a substrate for both prenyltransferases in the presence of either farnesyl or geranylgeranyl pyrophosphate, whereas geranylgeranylation was only observed when both prenyl pyrophosphates were added to the in vitro translation reaction mixture. Thus, a combination of the cell-free protein synthesis system with MS is an effective strategy to analyze protein prenylation.     

Specific Proteolysis of Neuronal Protein GAP-43 by Calpain: Characterization,Regulation, and Physiological Role

The mechanism of specific proteolysis of the neuronal protein GAP-43 in axonal terminals has been investigated. In synaptic terminals in vivo and in synaptosomes in vitro GAP-43 is cleaved only at the single peptide bond formed by Ser41; this is within the main effector domain of GAP-43. Proteolysis at this site involves the cysteine calcium-dependent neutral protease calpain. The following experimental evidences support this conclusion: 1) calcium-dependent proteolysis of GAP-43 in synaptosomes is insensitive to selective inhibitor of micro-calpain (PD151746), but it is completely blocked by micro- and m-calpain inhibitor PD150606; 2) GAP-43 proteolysis in the calcium ionophore A23187-treated synaptosomes is activated by millimolar concentration of calcium ions; 3) the pattern of fragmentation of purified GAP-43 by m-calpain (but not by micro-calpain) is identical to that observed in synaptic terminals in vivo. GAP-43 phosphorylated at Ser41 by protein kinase C (PKC) is resistant to the cleavage by calpain. In addition, calmodulin binding to GAP-43 decreases the rate of calpain-mediated GAP-43 proteolysis. Our results indicate that m-calpain-mediated GAP-43 proteolysis regulated by PKC and calmodulin is of physiological relevance, particularly in axonal growth cone guidance. We suggest that the function of the N-terminal fragment of GAP-43 (residues 1-40) formed during cleavage by m-calpain consists in activation of neuronal heterotrimeric GTP-binding protein G(o); this results in growth cone turning in response to repulsive signals.     

Protein transport in plant cells: in and out of the Golgi

In plant cells, the Golgi apparatus is the key organelle for polysaccharide and glycolipid synthesis, protein glycosylation and protein sorting towards various cellular compartments. Protein import from the endoplasmic reticulum (ER) is a highly dynamic process, and new data suggest that transport, at least of soluble proteins, occurs via bulk flow. In this Botanical Briefing, we review the latest data on ER/Golgi inter-relations and the models for transport between the two organelles. Whether vesicles are involved in this transport event or if direct ER-Golgi connections exist are questions that are open to discussion. Whereas the majority of proteins pass through the Golgi on their way to other cell destinations, either by vesicular shuttles or through maturation of cisternae from the cis- to the trans-face, a number of membrane proteins reside in the different Golgi cisternae. Experimental evidence suggests that the length of the transmembrane domain is of crucial importance for the retention of proteins within the Golgi. In non-dividing cells, protein transport out of the Golgi is either directed towards the plasma membrane/cell wall (secretion) or to the vacuolar system. The latter comprises the lytic vacuole and protein storage vacuoles. In general, transport to either of these from the Golgi depends on different sorting signals and receptors and is mediated by clathrin-coated and dense vesicles, respectively. Being at the heart of the secretory pathway, the Golgi (transiently) accommodates regulatory proteins of secretion (e.g. SNAREs and small GTPases), of which many have been cloned in plants over the last decade. In this context, we present a list of regulatory proteins, along with structural and processing proteins, that have been located to the Golgi and the 'trans-Golgi network' by microscopy.     

Protein synthesis in halophytes: The influence of potassium,sodium and magnesium in vitro

The amino acid (³⁵S-methionine) incorporating activity of an in vitro wheat germ translation system was found to be maximal in 80 to 125 mol m^–3 K with 2 to 4 mol m^–3 Mg both as the acetate. Substitution of Na for K, or chloride for acetate at concentrations above 80 mol m^–3 inhibited incorporation. When the K acetate concentration was raised to 200 mol m^–3, no incorporation of radioactive methionine occurred.Translation by polysomes extracted from leaf tissue of S. maritima, supplemented with postribosomal supernatant from wheat germ, showed activity which was optimal in the presence of 225 mol m^–3 K acetate and 8 mol m^–3 Mg acetate. However, the translation system was not directly comparable with the wheat germ system, as studies with an initiation inhibitor, aurintricarboxylic acid, suggested that the S. maritima system was essentially elongation-dependent, while initiation occurred in the wheat germ system.Elongation-dependent polysomal preparations were extracted from leaves of the glycophytes Pisum sativum, Triticum aestivum, Oryza sativa and Hordeum vulgare, and from the halophytes Atriplex isatidea and Inula crithmoides. Translation by polysomes from the salt-tolerant plants was optimal at higher K and Mg concentrations, than by polysomes from the glycophytes. Furthermore, NaCl was better able partially to substitute for the role of K in polysomal preparations from halophytes than glycophytes.     

Protein synthesis by rice coleoptiles during prolonged anoxia: implications for glycolysis, growth and energy utilization

BACKGROUND AND AIMS: Anoxia-tolerant plant tissues synthesize a number of proteins during anoxia, in addition to the 'classical anaerobic proteins' involved in glycolysis and fermentation. The present study used a model system of rice coleoptile tips to elucidate patterns of protein synthesis in this anoxia-tolerant plant tissue. METHODS: Coleoptile tips 7-11 mm long were excised from intact seedlings exposed to anoxia, or excised from hypoxically pre-treated seedlings and then exposed to anoxia for 72 h. Total proteins or 35S-labelled proteins were extracted, separated using two-dimensional isoelectric focusing/SDS-polyacrylamide gel electrophoresis and analysed using mass spectrometry. KEY RESULTS: The coleoptile tips excised after intact seedlings had been exposed to anoxia for 72 h had a similar proteome to tips that were first excised and then exposed to anoxia. After 72 h anoxia, Bowman-Birk trypsin inhibitors and a glycine-rich RNA-binding protein decreased in abundance, whereas a nucleoside diphosphate kinase and several proteins with unknown functions were strongly enhanced. Using [35S]methionine as label, proteins synthesized at high levels in anoxia, and also in aeration, included a nucleoside diphosphate kinase, a glycine-rich RNA-binding protein, a putative elicitor-inducible protein and a putative actin-depolymerizing factor. Proteins synthesized predominately in anoxia included a pyruvate orthophosphate dikinase (PPDK), alcohol dehydrogenase 1 and 2, fructose 1,6-bisphosphate aldolase and a protein of unknown function. CONCLUSION: The induction of PPDK in anoxic rice coleoptiles might, in combination with pyruvate kinase (PK), enable operation of a 'substrate cycle' producing PPi from ATP. Production of PPi would (a) direct energy to crucial transport processes across the tonoplast (i.e. the H+-PPiase); (b) be required for sucrose hydrolysis via sucrose synthase; and (c) enable acceleration of glycolysis, via pyrophosphate:fructose 6-phosphate 1-phosphotransferase (PFP) acting in parallel with phosphofructokinase (PFK), thus enhancing ATP production in anoxic rice coleoptiles; ATP production would need to be increased if there was a substantial requirement for PPi.     

Outer membrane vesicles of the VA-MENGOC-BC vaccine against serogroup B of Neisseria meningitidis: Analysis of protein components by two-dimensional gel electrophoresis and mass spectrometry

Neisseria meningitidis is a Gram-negative bacterium responsible for significant mortality worldwide. While effective polysaccharides-based vaccines exist against serogroups A, C, W135, and Y, no similar vaccine is suitable for children under 4 years against disease caused by serogroup B strains. Therefore, major vaccine efforts against this serogroup are based on outer membrane vesicles (OMVs), containing major outer membrane proteins. The OMV-based vaccine produced by the Finlay Institute in Cuba (VA-MENGOC-BC) contributed to the rapid decline of the epidemic in this Caribbean island. While the content of major proteins in this vaccine has been discussed, no detailed work of an outer membrane proteomic map of this, or any other, commercially available OMV-derived product has been published so far. Since OMVs exhibit a large bias toward a few major proteins and usually contain a high content of lipids, establishing the adequate conditions for high resolution, 2-DE of this kind of preparation was definitely a technical challenge. In this work, 2-DE and MS have been used to generate a proteomic map of this product, detailing the presence of 31 different proteins, and it allows the identification of new putative protective protein components it contains.     

Intermediate filament proteins immunologically related to desmin in astrocytes: A study of chicken spinal cord by two-dimensional gel electrophoresis and immunoblotting

Co-migration experiments by two-dimensional SDS-PAGE using chicken spinal cord extracts and desmin purified from chicken gizzard showed that desmin is not present in spinal cord. However, by the immunoblotting procedure, desmin antibodies recognized 3 spinal cord antigens with different molecular weights and isoelectric points than desmin and the glial fibrillary acidic (GFA) protein. These antigens which also reacted with GFA protein antibodies were not identified in chicken gizzard extracts. The reactivity of the antigens with a monoclonal antibody recognizing an epitope common to most intermediate filament proteins (1) suggests that immunostaining of astrocytes with desmin antibodies (2, 3) is due to the presence of new intermediate filament proteins immunologically related to desmin.Special issue dedicated to Dr. Paola S. Timiras     

Yeast Mitoribosome Large Subunit Assembly Proceeds by Hierarchical Incorporation of Protein Clusters and Modules on the Inner Membrane

1. Download high-res image (300KB)
2. Download full-size image

     

Kinetical Analysis of Tumor Cell Death-Inducing Mechanism by Polymorphonuclear Leukocyte-Derived Calprotectin: Involvement of Protein Synthesis and Generation of Reactive Oxygen Species in Target Cells

We have previously shown that calprotectin, the most abundant cytosolic protein existing in polymorphonuclear leukocytes (PMNs), induces apoptotic cell death in various tumor cells, suggesting that calprotectin is an effector molecule against tumor cells in PMNs. To explore the cell death-inducing mechanism of the factor, we examined the involvement of target protein synthesis and generation of reactive oxygen species (ROS) in the reaction. Calprotectin induced cell death in MM46 mouse mammary carcinoma cells after a 14-16 hr lag time. When the factor was removed from the medium up to about 12 hr after culturing, the effect was diminished. The induction of cell death by calprotectin was markedly inhibited by the presence of the RNA synthesis inhibitor actinomycin D or the protein synthesis inhibitor cycloheximide. However, the addition of these inhibitors after 12 hr of culturing was unable to inhibit the reaction. Up to 12 hr of culturing, the net protein synthesis of MM46 cells was augmented by the presence of calprotectin, but thereafter was impaired. The induction of cell death was also inhibited by the antioxidative reagents N-acetyl-l -cysteine (NAC) or propyl gallate. The addition of NAC even 15 hr later significantly attenuated the calprotectin effect. Flow cytometry analysis showed that calprotectin began to increase the ROS content in MM46 cells after 8-12 hr of culturing, and that the increase was abrogated by the antioxidants. Thus, protein synthesis and ROS generation may be essential elements in the early or later phases of the cell death-inducing reaction of calprotectin, respectively.     

Red algal LHC I genes have similarities with both Chl a/b- and a/c-binding proteins: A 21 kDa polypeptide encoded by LhcaR2 is one of the six LHC I polypeptides

Polypeptides from the PS I holocomplex of the red alga P. cruentum, purified for microsequencing, confirmed that six LHC I polypeptides from SDS-PAGE are distinct apoproteins. Analysis of a cDNA clone, designated as LhcaR2, from a cDNA library, indicates that it shares major structural features with the recently cloned first red algal gene LhcaR1. The LhcaR2 is believed to encode the 21.0 kDa polypeptide of the LHC I complex from comparison of the deduced amino acid sequence and the microsequences of several tryptic digest fragments from the isolated polypeptide. As in chlorophytic and chromophytic LHCs, the essential residues for Chl-binding and helix stabilization in helices 1 and 3 are highly conserved. Relatedness between rhodophytes and the chlorophytes is also inferred from sequence conservation in the N-flanking regions of helices 1 and 3. Conversely, helix 2 exhibited the highest similarity between LHC sequences of Chl a/c-binding chromophytes and the Chl a-binding rhodophytes, with 11 of 22 residues identical or conservatively substituted. Moreover, whereas in chlorophytes, the Q and E Chl-binding residues are separated by seven amino acid residues, they are always separated by 8 residues in rhodophytes and chromophytes. Superimposition of the predicted LhcaR2 sequence with the LHC II model [Kühlbrandt et al. (1994) Nature 367: 614–621] shows the same structural features except shortened connecting sequence between helices 1 and 2 on the lumenal side. The chimeric nature of rhodophyte genes, with both chromophytic and chlorophytic features, leads to the suggestion that they reflect attributes of an intermediate stage in LHC apoprotein evolution.