Considerations in the isolation of rat liver nuclear matrix,nuclear envelope,and pore complex lamina

A number of recent studies have demonstrated a salt-, nuclease, and detergent-resistant subnuclear structure termed the nuclear protein matrix which consists of a fibrogranular intranuclear network, residual components of the nucleolus, and a peripheral lamina. Other workers, however, have shown that somewhat similar methods result in the isolation of the peripheral lamina devoid of the intranuclear components. In this report we demonstrate that seemingly slight changes in the isolation procedure cause major changes in the morphology of the residual structures obtained. When freshly purified rat liver nuclei were digested with DNase I and RNase A and then extracted with buffers of low magnesium ion concentration (LS buffer) and high ionic strength (HS buffer), the resulting structures isolated prior to or after Triton X-100 extraction lacked the extensive intranuclear network and the easily identifiable residual nucleoli present in the nuclear protein matrix. Systematic modification of this extraction procedure revealed that morphologically identifiable residual nucleoli were present when digestion with RNase A followed extraction with HS buffer but were absent when the order of these steps was reversed. The removal of the nucleolus by RNase A and HS buffer correlated with the removal of nuclear RNA by the same treatments. These coordinate events could not be prevented by treatment with protease inhibitors but were prevented by treatment of the RNase A with diethylpyrocarbonate, an RNase inhibitor. The extensive intranuclear network seen in the nuclear protein matrix was sparse or absent when residual structures were prepared from DNase- and RNase-treated nuclei under conditions which minimized the oxidation of protein sulfhydryl groups. In contrast, an extensive non-chromatin intranuclear network was seen if the formation of intermolecular protein disulfide bonds was promoted by extraction of nuclei with cationic detergents, by overnight incubation, or by treatment with oxidizing agents like sodium tetrathionate prior to nuclease digestion and subsequent extraction. By varying the order of extraction steps and the extent of disulfide cross-linking, it is possible to isolate from a single batch of nuclei residual structures with a wide range of morphologies and compositions.     

Quantitative immunofluorescence and immunoelectron microscopy of the topoisomerase IIα associated with nuclear matrices from wild-type and drug-resistant Chinese hamster ovary cell lines

Topo IIα is considered an important constituent of the nuclear matrix, serving as a fastener of DNA loops to the underlying filamentous scaffolding network. To further define a mechanism of drug resistance to topo II poisons, we studied the quantity of topo IIα associated with the nuclear matrix in drug-resistant SMR₁₆ and parental cells in the presence and absence of VP-16. Nuclear matrices were prepared from nuclei isolated in EDTA buffer, followed by nuclease digestion with DNase II in the absence of RNase treatment and extraction with 2 M NaCl. Whole-mount spreading of residual structures permits, by means of isoform-specific antibody and colloidal-gold secondary antibodies, an estimate of the amount of topo IIα in individual nuclear matrices. There are significant variations in topo IIα amounts between individual nuclear matrices due to the cell cycle distribution. The parental cell line contained eight to ten times more nuclear matrix–associated topo IIα than the resistant cell line matrices. Nuclear matrix–associated topo IIα from wild-type and resistant cell lines correlated well with the immunofluorescent staining of the enzyme in nuclei of intact cells. The amount of DNA associated with residual nuclear structures was five times greater in the resistant cell line. This quantity of DNA was not proportional to the quantity of topo IIα in the same matrix; in fact they were inversely related. In situ whole-mount nuclear matrix preparations were obtained from cells grown on grids and confirmed the results from labeling of isolated residual structures. J. Cell. Biochem. 67:112–130, 1997. © 1997 Wiley-Liss, Inc.     

Flow cytometric studies of the nuclear matrix

We have devised a method to measure the protein and nucleic acid content of the nuclear matrix using flow cytometry. Nuclear matrices were prepared from nuclei by DNase I digestion followed by 3 M NaCl extraction. The resulting particles were stained with fluorescein isothiocyanate (FITC) for protein and propidium iodide (PI) for double-stranded nucleic acids, and fluorescence as well as forward angle light scatter was detected. The matrices were also subjected to additional chemical or enzymatic perturbations, and changes in the above parameters were measured. Results showed that matrices from heat-shocked cells not only retained the majority of heat-induced excess nuclear protein, but also exhibited higher PI signals than controls after RNase A digestion. This observation did not hold if RNase A digestion preceded high-salt extraction, suggesting that a salt-extractable moiety had been replaced or altered by heat so that double-stranded RNA was protected from the nucleolytic attack. The residual PI fluorescence in matrices from heated cells bore a linear relationship to the increased protein content in those matrices, indicating that the excess protein sequesters matrix-associated RNA. Polyacrylamide gel electrophoresis of matrix polypeptides revealed increased amounts of many proteins as a result of heat as well as the appearance of several new proteins, one of which comigrates with the HSP72/73 heat-shock proteins. The results of these studies show that flow cytometry can be used to study the nuclear matrix and is capable of detecting changes that result from alterations in its protein composition.     

The type of DNA attachment sites recovered from nuclear matrix depends on isolation procedure used

A large variety of DNA sequences have been described in nuclear matrix attachment regions. It could be most likely a result of the different methods used for their isolation. The idea about how different types of known DNA sequences (strongly attached to the nuclear matrix, weakly attached, or not attached) directly participate in anchoring DNA loops to the nuclear matrices isolated by different experimental procedures was tested in this study. Matrix-attached (M) and matrix-independent or loop (L) fractions as well as nuclear matrices were isolated using extractions of nuclei with 25 mM lithium 3,5-diiodosalicylate (LIS), 2 M NaCl, 0.65 M ammonium sulphate containing buffers followed by DNase I/RNase A digestion, or according to so designated conventional method. Using PCR-based and in vitro binding assays it was established that LIS and ammonium sulphate extractions gave similar results for the type of attachment of sequences investigated. The harsh extraction with 2 M NaCl or the conventional procedure led to some rearrangements in the attachment of DNA loops. As a result a big part of matrix attached sequences were found detached in the loop fractions. However, the in vitro binding abilities of the MARs to the nuclear matrices isolated by different methods did not change.     

Demonstration of a DNase-sensitive network associated with the nuclear pore complexes in rat liver nuclei

Rat liver nuclei have been studied by transmission electron microscopy after resuspension in a phosphate-buffered salt solution containing SO²⁻ ₄ as the quantitatively dominant anion. Owing to the high solubility of chromatin in the presence of SO²⁻ ₄ instead of Cl⁻ at isotonicity, nuclei are depleted for chromatin by DNase I digestion in this buffer, eliminating the need for high-salt extraction. This shows that at least 75% of the nuclear pore complexes are associated with fibrogranular structures, which ramify as a network throughout the nucleus, interconnecting the nuclear lamina, interchromatin granule clusters and nucleoli. Perichromatin granules are located in this material proximal to the nuclear pore complexes. Most of the chromatin is removed without major impact on the network, but below a level of 25% residual chromatin there is a considerable reduction of this material, and only about 15% of the connections to the nuclear pore complexes are resistant to digestion with DNase I or streptodornase A and B. The percentage of nuclear pore complexes connected to the network is further reduced by salt extraction and RNase treatment. These results suggest that DNA is an integral part of the network, which presumably plays a role in nucleo-cytoplasmic transport of RNA and protein. Received: 1 September 1998; in revised form: 17 December 1998 / Accepted: 17 December 1998     

Cytological identification and characterization of the nuclear matrix

The presence of a matrix system in the nucleus has been cytologically demonstrated in intact untreated and α-amanitin-treated HeLa and rat kangaroo cells. The matrix is Feulgen negative and is stained very faintly with toluidine blue, whereas it can be brightly stained with haemalum or light green. It is highly susceptible to pepsin digestion but resistant to RNase. Ultrastructurally the matrix consists of 50–60 Å thick fibrils. Some granules, which may be ribonuclear protein granules are also found associated with the fibrils. The matrix is essentially a network of acidic protein.     

Biochemical and morphological changes in the nuclear matrix prepared from apoptotic HL‐60 cells: Effect of different stabilizing procedures

Apoptotic cell death is characterized by deep morphological changes that take place in the nucleus. It is unclear whether modifications also occur in the nuclear matrix, a mainly proteinaceous structure that conceivably acts as a nuclear framework. We have investigated whether biochemical and morphological alterations of the nuclear matrix prepared from apoptotic HL‐60 cells were dependent on the manipulations to which isolated nuclei were subjected before DNase I digestion and 2 M NaCl extraction. Our results showed that the stabilizing procedures employed to preserve the inner fibrogranular network and nucleolar remnants of the matrix (i.e., a 37°C incubation; exposure to sodium tetrathionate at 4°C; exposure to sodium tetrathionate at 37°C) had no effect on the protein recovery of apoptotic nuclear matrices, which was always approximately two‐ to fivefold less than in control matrices. Moreover, one‐ and two‐dimensional gel analysis of nuclear matrix proteins showed that, in apoptotic samples, striking quantitative changes were present, as compared with controls. Once again, these changes were seen irrespective of the stabilizing procedures employed. Also, transmission electron microscope analysis showed similar morphological alterations in all types of apoptotic nuclear matrices. By contrast, the immunofluorescent distribution of the 240‐kDa NuMA protein seen in apoptotic samples was more sensitive to the stabilizing treatments. Our results indicate that the biochemical and morphological changes of the apoptotic nuclear matrix are largely independent of the isolation protocols and strengthen the contention that destruction of the nuclear matrix network is one of the key events leading to apoptotic nuclear destruction. J. Cell. Biochem. 74:99–110, 1999. © 1999 Wiley‐Liss, Inc.     

Nuclear matrix-like filaments and fibrogranular complexes form through the rearrangement of specific nuclear ribonucleoproteins

The behavior of nuclear pre-mRNA-binding proteins after their nuclease and/or salt-induced release from RNA was investigated. After RNase digestion or salt extraction, two proteins that initially exist as tetramers (A2)(3)B1 in isolated heterogeneous nuclear ribonucleoprotein (hnRNP) complexes quantitatively reassociated to form regular helical filaments ranging in length from 100 nm to >10 microm. In highly magnified preparations prepared for scanning transmission electron microscopy, single filaments have diameters near 18 nm. In conventional negatively stained preparations viewed at low magnification, the diameters of the thinnest filaments range from 7 to 10 nm. At protein concentrations of >0.1 mg/ml, the filaments rapidly aggregated to form thicker filamentous networks that look like the fibrogranular structures termed the "nuclear matrix." Like the residual material seen in nuclear matrix preparations, the hnRNP filaments were insoluble in 2 M NaCl. Filament formation is associated with, and may be dependent on, disulfide bridge formation between the hnRNP proteins. The reducing agent 2-mercaptoethanol significantly attenuates filament assembly, and the residual material that forms is ultrastructurally distinct from the 7- to 10-nm fibers. In addition to the protein rearrangement leading to filament formation, nearly one-third of the protein present in chromatin-clarified nuclear extracts was converted to salt-insoluble material within 1 min of digestion with RNase. These observations are consistent with the possibility that the residual material termed the nuclear matrix may be enriched in, if not formed by, denatured proteins that function in pre-mRNA packaging, processing, and transport.     

Heterogeneous nuclear RNA-protein fibers in chromatin-depleted nuclei

The heterogeneous nuclear RNA-protein (hnRNP) fibers in HeLa cell nuclei are visualized by a nuclear subfractionation technique which removes 96% of the chromatin in a single step and 99% in a two-step elution but leaves the bulk of the hnRNA complexed with the remnant nuclear structure or lamina. Both steady-state and newly synthesized (approximately 15-s label) hnRNA are associated with the remnant nuclei to about the same extent. This association does not appear to depend on the presence of chromatin and exists in addition to any possible association of hnRNP with chromatin itself. Electron microscopy of partially purified nuclear hnRNA complexes shows that the hnRNP fibers form a ribonucleoprotein network throughout the nucleus, whose integrity is dependent on the RNA. Autoradiography confirms that hnRNA is a constituent of the fibers. The RNA network visualized in these remnant nuclei may be similar to RNA networks seen in intact cells. The hnRNA molecules appear to be associated with the nuclear lamina, at least in part, by unusual hnRNA sequences. More than half of the recovered poly(A) and double-stranded hnRNA regions remains associated with the nuclear structures or the laminae after digestion with RNase and elution with 0.4 M ammonium sulfate. In contrast, the majority of oligo(A), another ribonuclease resistant segment, is released together with most of the partially digested but still acid-precipitable single- stranded hnRNA and the hnRNP proteins not eluted by the ammonium sulfate alone. These special RNA regions appear to be tightly bound and may serve as points of attachment of the hnRNA to nuclear substructures. It is suggested that hnRNA metabolism does not take place in a soluble nucleoplasmic compartment but on organized structures firmly bound to the nuclear structure.     

Synaptonemal complexes are integral components of the isolated mouse spermatocyte nuclear matrix

Synaptonemal complexes (SCs) have been isolated as integral components of the nuclear matrix from purified mouse pachytene spermatocytes. These nuclear synaptonemal complex-matrices are prepared by extracting Triton X-100-treated nuclei with low (0.2 M) and high (1.0 or 2.0 M) NaCl, DNase I, and RNase A to remove 85% of the nuclear proteins, 97% of the RNA, and 99% of the DNA. Studies with the light and electron microscopes indicate that these matrices, while lacking a distinct lamina, contain nuclear pores interconnected by a fiber network, residual nucleoli, and interchromatin fibers. In addition, the pachytene spermatocyte matrices contain residual XY heterochromatin and the principal components of the SCs, including two lateral elements, a central element, a presumptive centromere, and attachment plaques. These SCs are preserved within the matrix and retain their structural association with the pore-fiber complex, even when subjected to strong dissociating conditions. Nuclear matrices from pachytene spermatocytes and spermatids (steps 1-8), when analyzed by SDS PAGE, contain an array of polypeptides distinct from those of mouse liver nuclear matrices. Proteins of spermatogenic matrices range in Mr from 8,000 to approximately 150,000. The prominent lamina proteins (Mr approximately 60,000-70,000) of somatic nuclear matrices are either absent or represent only a minor part of the spermatogenic matrix. The polypeptide composition of the pachytene spermatocyte and spermatid matrices are similar, although minor quantitative and qualitative differences are evident. These observations suggest that the SC constituents may consist of a heterogeneous group of proteins present in low proportion relative to total matrix proteins, or they may be retained, but in a different form, within the spermatid matrix.     

Core filaments of the nuclear matrix

The nuclear matrix is concealed by a much larger mass of chromatin, which can be removed selectively by digesting nuclei with DNase I followed by elution of chromatin with 0.25 M ammonium sulfate. This mild procedure removes chromatin almost completely and preserves nuclear matrix morphology. The complete nuclear matrix consists of a nuclear lamina with an interior matrix composed of thick, polymorphic fibers and large masses that resemble remnant nucleoli. Further extraction of the nuclear matrices of HeLa or MCF-7 cells with 2 M sodium chloride uncovered a network of core filaments. A few dark masses remained enmeshed in the filament network and may be remnants of the nuclear matrix thick fibers and nucleoli. The highly branched core filaments had diameters of 9 and 13 nm measured relative to the intermediate filaments. They may serve as the core structure around which the matrix is constructed. The core filaments retained 70% of nuclear RNA. This RNA consisted both of ribosomal RNA precursors and of very high molecular weight hnRNA with a modal size of 20 kb. Treatment with RNase A removed the core filaments. When 2 M sodium chloride was used directly to remove chromatin after DNase I digestion without a preceding 0.25 M ammonium sulfate extraction, the core filaments were not revealed. Instead, the nuclear interior was filled with amorphous masses that may cover the filaments. This reflected a requirement for a stepwise increase in ionic strength because gradual addition of sodium chloride to a final concentration of 2 M without an 0.25 M ammonium sulfate extraction uncovered core filaments.     

Reversibly contractile nuclear matrix. Its isolation, structure, and composition

From Tetrahymena macronuclei we have isolated a reversibly contractile nucleo-skeleton, i.e., an "expanded" nuclear matrix which reversibly contracts when the total concentration of the bivalent cations, Ca and Mg (3:2), is decreased to 5 mM or increased to 125 mM. During contraction the average diameter of the expanded matrix becomes reduced by about 24%; this corresponds to a volume contraction of about 55%. The reversible contraction of the nuclear matrix does not depend on ATP and cannot be inhibited by salygran. The expanded matrix is obtained by removing carefully from the macronuclei 89.7% of the phospholipid, 99.6% of the DNA, 98.5% of the RNA, and 74.8% of the protein by treatment with Triton X-100 and digestion with DNase and RNase followed by an extraction with 2 M NaCl. Electron microscopy reveals, within the expanded matrix, residual equivalents to the structures characteristic for macronuclei: (a) a residual nuclear envelope with nuclear pore complexes; (b) residual nucleoli at the periphery; (c) a fibrillar internal network. The expanded matrix is essentially composed of proteins (96.2%) and traces of DNA (0.8%), RNA (0.5%), phospholipid (1.6%), and carbohydrates (0.9%). The last, which have been determined by gas chromatography, contain glucose, mannose, and an unidentified sugar in the ratio 1:5.4:5.7. The ratio of acidic to basic amino acids of the expanded matrix is 1.55. Sodium dodecyl sulfate (SDS) gel electrophoresis reveals a predominant protein with a mol wt of 18,000 which is apparently involved in the reversible contractile process. The mechanism of this reversible contraction of the expanded matrix remains to be elucidated, but it differs both from actin-myosin contraction systems and from the contractile spasmoneme system in vorticellids.     

Composition, structure and various properties of DNA-protein complexes located at the sites of DNA loop attachment to the nuclear skeleton

The intimate structure of the complexes located at the sites of DNA loops attachment to the nuclear skeleton was analysed. It is shown that: there are at least three components of the attachment site complex: DNA, protein, RNA; protein moiety consists of 7-8 species with Mr 70-17 kDa. Their association with DNA is resistant to ionic detergents, high salt and urea treatments. The DNA-protein complex is also resistant to the SDS-pronase-phenol deproteinisation procedure; the buoyant density of the complex is the same as DNA density. RNase digestion at low ionic strength reduces density of the complex while the same treatment at 0,4 M NaCl has no effect; DNA-protein complexes isolated with urea-high salt treatment are visualised as globular particles 25-35 nm in diameter with DNA loops attached. These particles were not observed after detergent treatment although protein composition of the complex remained the same.     

Modification of nuclear matrix proteins by ADP-ribosylation. Association of nuclear ADP-ribosyltransferase with the nuclear matrix

Nuclear matrices were isolated by treatment of isolated HeLa cell nuclei with high DNase I, pancreatic RNase and salt concentrations. ADP-ribosylated nuclear matrix proteins were identified by electrophoresis, blotting and autoradiography. In one experimental approach nuclear matrix proteins were labeled by exposure of permeabilized cells to the labeled precursor [32P]NAD. Alternatively, the cellular proteins were prelabeled with [35S]methionine and the ADP-ribosylated nuclear matrix proteins separated by aminophenyl boronate column chromatography. By both methods bands of modified proteins, though with differing intensities, were detected at 41, 43, 46, 51, 60, 64, 69, 73, 116, 140, 220 and 300 kDa. Approximately 2% of the total nuclear ADP-ribosyltransferase activity, but only 0.07% of the nuclear DNA, was tightly associated with the isolated nuclear matrix. The matrix-associated enzyme catalyzes the incorporation of [32P]ADP-ribose into acid-insoluble products of molecular mass 116 kDa and above, in a 3-aminobenzamide-inhibited, time-dependent reaction. The possible function of ADP-ribosylation of nuclear matrix proteins and of the attachment of ADP-ribosyltransferase to the nuclear matrix in the regulation of matrix-associated biochemical processes is discussed.     

Adrenal nuclear matrix isolation. A morphologic and biochemical study

Summary— The nuclear matrix of adrenal cells was isolated by using the methods proposed by Commerford et al and Kaufmann et al for the liver nuclear matrix isolation. Both methods permitted, to the best of our knowledge for the first time, to prepare the nuclear matrix of a steroidogenic cell and therefore to study some regulatory mechanisms governing steroidogenesis. Commerford et al's method retains nuclear envelope and so produces a higher contamination; Kaufmann et al's method presents a higher purity since the nuclear envelope was removed by Triton X-100. No RNase digestion has been employed for the isolation of the residual nuclear matrices. Both methods however, permit the isolation of fractions with a good morphology, retaining a reticular nucleolus, interchromatinic granules, and a fibrogranular scaffold extending from the nucleolus to the nuclear lamina. The major peptides detected by 1-D SDS-PAGE were 123, 56, 46 and 41 kDa; with both methods protein profiles were similar. Identification of proteins by immunodetection reveals lamins A and C, 80 and 65 kDa respectively; no labeling was found for actin (45 kDa) and vimentin (57 kDa). In short, adrenal nuclear matrix was isolated, Kaufmann et al's method being the method of choice.     

Phospholipase C digestion induces the removal of nuclear RNA: A cytochemical quantitative study

Summary It has been reported that the incubation of isolated rat liver nuclear matrices with phospholipase C causes the digestion of the matrix-bound phospholipids and the release of most matrix-linked RNAs (Coccoet al., 1980). In this paper, the presence of phospholipids in nuclear substructures and the effects of their removal by phospholipase C digestion have been investigated by means of enzyme—colloidal gold cytochemistry. The nuclear phospholipids appear to be localized in the interchromatin areas and in the nucleolus and are virtually absent in the heterochromatin, when labelled with phospholipase C—colloidal gold. The double labelling test with ribonuclease A and phospholipase C conjugated with gold particles of different diameters shows that the nuclear phospholipids are co-localized with RNA-containing structures. The enzymatic digestion of phospholipids on thin sections of either isolated nuclei or pancreas embedded in LR White resin results in the decrease of the RNase-A colloidal gold labelling of nuclear RNA-containing structures, but not of the rough endoplasmic reticulum. The data confirm the presence of phospholipids in the nucleus in the absence of possible translocation due to isolation procedures and strengthen the hypothesis that they are involved in interactions between nucleic acids and proteins of the nuclear matrix.     

Effect of size and location of the oligosaccharide chain on protease degradation of bovine pancreatic ribonuclease

We have investigated the effect of size and location of the oligosaccharide chain on protease degradation of bovine pancreatic ribonuclease. The sensitivity of nonglycosylated RNase A to trypsin and chymotrypsin was compared with three glycosylated species of RNase B which differed with respect to the size of the carbohydrate chain. Two forms of glycosylated RNase B were isolated by concanavalin A-Sepharose affinity chromatography, and each was shown to contain a single carbohydrate chain composed of GlcNAc2Man1 (RNase B") or GlcNAc2Man5-8 (RNase B). A third form (RNase B'), with oligosaccharide composed of GlcNAc2Man4, was prepared by partial digestion of RNase B with alpha-mannosidase. Fully glycosylated RNase B was found to be 6-10 times more resistant to trypsin digestion than nonglycosylated RNase A. RNase B' and B", with intermediate chain sizes, were 3.0- and 1.3-fold more resistant to trypsin digestion than RNase A, respectively. With chymotrypsin, however, differences in rates of digestion were much less marked, with a maximum difference of 3-fold between RNase A and B. In addition, we found that the specificity of the primary trypsin (Arg 33-Asp 34 bond) or chymotrypsin (Tyr 25-Cys 26 bond) cleavage site was not affected by the presence or size of the oligosaccharide chain. These results are consistent with the view that the size of the oligosaccharide chain and its proximity to the primary or rate-limiting cleavage site are important for expression of the carbohydrate protection against proteolytic degradation, which thus appears to be mediated by steric hindrance.     

The nuclear matrix of duck erythroblasts is associated with globin mRNA coding sequences but not with the major proteins of 40S nuclear RNP

A residual protein matrix has been prepared from avian erythroblast nuclei by extensive extraction with salines and detergent and subsequent digestion with high concentrations of RNase and DNase. Ultrastructural examination reveals considerable internal structure, the most prominent feature being the remains of the nucleoli embedded in a network of fibres of fairly uniform diameter of 50 Å. The proteins which make up this structure have been examined by two-dimensional electrophoresis and are shown to consist of a characteristic set of about 30, mainly acidic components, including four prominent species of 43 000, 52 000, 66 000 and 68 000 molecular weight (MW). In parallel preparations of the nuclear matrix digested with DNase alone, much of the nuclear RNA is found associated with the residual structure, including globin-coding sequences. These results correlate well with the ultrastructural appearance of DNase-digested matrix preparations which show that superimposed on the 50 Å fibrous network is a 200–300 Å granular component, the combined fibrillo-granular structure resembling the interchromatin RNP previously identified in situ. However, the proteins of the DNase-digested matrix seen by two-dimensional electrophoresis are indistinguishable from the proteins of matrix preparations digested with both DNase and RNase. Furthermore, two-dimensional comparison between the proteins of the DNase-digested matrix and purified 40S nuclear RNP particles shows that the bulk of the proteins found associated with nuclear RNA in vitro are extracted during matrix preparation, and only two, with MWs of 43 000 and 73 000, remain. The latter species co-migrates with the poly(A)-binding protein.