Search

Negatively stained complexes of Panulirus interruptus (spiny lobster) hemocyanin with two different monoclonal antibodies, named E and J, were studied by electron microscopy and image processing. The attachment site of the antibodies to the hexameric hemocyanin molecule was deduced from two perpendicular views of hemocyanin/antibody complexes, in which either the threefold axis or one of the twofold axes was oriented perpendicular to the supporting film. Images of complexes in these orientations were searched with reference images simulated from the known X-ray structure of P. interruptus hemocyanin. The two sites were further characterized by combining our results from electron microscopy with structural data obtained by X-ray diffraction and other methods. These two antibodies recognize different non-overlapping epitopes. The epitope for clone E is located on domain 3 at the surface of the beta barrel and consists of certain loops, which form connections between beta-strand structures. The epitope for clone J is situated on domain 1 at the surface of an alpha-helical region and consists mainly of certain alpha-helices connecting loops. The orientation of the hemocyanin hexamers in the two complexes is very different, as is demonstrated most clearly when they form chains. Clone E forms complexes with the threefold axes perpendicular to the chain direction, while for clone J the threefold axes seem to be parallel to the main direction. The angle between the Fab part of an IgG molecule and the threefold axis of the hexamer is 60 +/- 5 degrees for clone E and 35 +/- 7 degrees for clone J. This observation is clearly related to the difference in orientation of the hexamers for the two complexes.

Arthropod hemocyanins (Hcs) are regular assemblies of 1, 2, 4, 6 or 8 hexameric protein molecules. They transport oxygen and can bind it in a cooperative manner. The hexameric X-ray structures of Panulirus interruptus (spiny lobster) and of Limulus polyphemus (horseshoe crab) subunit II Hc were solved recently by the groups of Hol (Groningen, The Netherlands) and Magnus (Cleveland, U.S.A.). They related cooperativity to a rotational movement of a domain within a subunit and of the two trimers mutually inside the hexamer. In our study a model was derived for the structure and related to the function of the four-hexameric Hc from the tarantula Eurypelma californicum by combining data from electron microscopy and image processing, from the X-ray diffraction studies mentioned earlier and from amino acid sequence studies. Interhexameric contacts were determined at the level of secondary structure elements and in some cases of single amino acids. Loops, undefined in the X-ray structures of the hexamers, were often involved in these contacts. In one case the contact was formed between four parallel alpha-helices, two from each hexamer. Based on these findings a mechanism is proposed for the transmission of cooperativity between the hexamers, in which the concept of "helical friction" plays a key role.

A NAD-dependent, oxygen-labile alcohol dehydrogenase was purified from Desulfovibrio gigas. It was decameric, with subunits of M(r) 43,000. The best substrates were ethanol (Km, 0.15 mM) and 1-propanol (Km, 0.28 mM). N-terminal amino acid sequence analysis showed that the enzyme belongs to the same family of alcohol dehydrogenases as Zymomonas mobilis ADH2 and Bacillus methanolicus MDH.

The quaternary protein structure of two methanol:N,N'-dimethyl-4-nitrosoaniline (NDMA) oxidoreductases purified from Amycolatopsis methanolica and Mycobacterium gastri MB19 was analyzed by electron microscopy and image processing. The enzymes are decameric proteins (displaying fivefold symmetry) with estimated molecular masses of 490 to 500 kDa based on their subunit molecular masses of 49 to 50 kDa. Both methanol:NDMA oxidoreductases possess a tightly but noncovalently bound NADP(H) cofactor at an NADPH-to-subunit molar ratio of 0.7. These cofactors are redox active toward alcohol and aldehyde substrates. Both enzymes contain significant amounts of Zn2+ and Mg2+ ions. The primary amino acid sequences of the A. methanolica and M. gastri MB19 methanol:NDMA oxidoreductases share a high degree of identity, as indicated by N-terminal sequence analysis (63% identity among the first 27 N-terminal amino acids), internal peptide sequence analysis, and overall amino acid composition. The amino acid sequence analysis also revealed significant similarity to a decameric methanol dehydrogenase of Bacillus methanolicus C1.

The architecture of alcohol oxidase crystalloids occurring in vivo in the peroxisomes of methylotrophic yeasts was deduced from electron micrographs of similar crystals of the Hansenula polymorpha enzyme grown in vitro. Three characteristic views of the crystal are observed, as well as single layers in the very early stages of crystal formation. The crystal is concluded to be cubical, with every octameric molecule making the same contacts with four neighbors in one plane, at right angles to its fourfold axis. The unit cell contains six octamers, in three mutually orthogonal orientations, and two large holes, which can accommodate other peroxisomal proteins involved in methanol metabolism. The crystal contains channels, connecting the holes, which allow the diffusion of relatively large molecules through the crystal. Crystal formation depends on just one contact per subunit, which may explain the fragility of the crystals.

Methanol dehydrogenase from the thermotolerant Bacillus sp. C1 was studied by electron microscopy and image processing. Two main projections can be distinguished: one exhibits 5-fold symmetry and has a diameter of 15 nm, the other is rectangular with sides of 15 and 9 nm. Subsequent image processing showed that the 5-fold view possesses mirror symmetry. The rectangular views can be divided into two separate classes, one of which has 2-fold rotational symmetry. It is concluded that methanol dehydrogenase is a decameric molecule, and a tentative model is presented. The estimated molecular weight is 430,000, based on a subunit molecular weight of 43,000. The enzyme contains one zinc and one to two magnesium ions per subunit. N-terminal amino acid sequence analysis revealed substantial similarity with alcohol dehydrogenases from Saccharomyces cerevisiae, Zymomonas mobilis, Clostridium acetobutylicum, and Escherichia coli, which contain iron or zinc but no magnesium. In view of the aberrant structural and kinetic properties, it is proposed to distinguish the enzyme from common alcohol dehydrogenases (EC 1.1.1.1) by using the name NAD-dependent methanol dehydrogenase.

The octameric protein alcohol oxidase from the yeast Hansenula polymorpha was studied by electron microscopy and image analysis. Two-dimensional crystals were formed by applying the protein, in a phosphate buffer containing poly(ethylene glycol) and EDTA, to a carbon-coated formvar film which had been glow-discharged in pentylamine at least several hours earlier. The crystals show p4 symmetry and have a unit cell of 12.5 X 12.5 nm2, containing one molecule. Image analysis of the crystals and of single molecules yielded two different views. From these it can be deduced that the subunits have an elongated shape and form two layers of four, stacked face to face. A tentative model of the structure is presented.

The Mo-Fe protein of nitrogenase from both Azotobacter vinelandii and Klebsiella pneumoniae (Av1 and Kp1, respectively) consists of four subunits of similar, but not identical, relative molecular mass. The hydrodynamic properties of Av1 (sedimentation and diffusion coefficient) and its total relative molecular mass are very similar to those of Kp1 and catalase from bovine liver, a tetramer of four identical subunits. By electron microscopy the Av1, Kp1 and catalase tetramers are seen as protein particles of diameter 9.0-10.0 nm; no details of the subunit structure can be observed. Av1 (but not Kp1) forms regular polymers of variable length at low ionic strength in the presence of MgCl2. The structure of these polymers, of diameter 21.2 nm, is complex. Optical diffraction studies give a smallest repeating distance of 8.4 nm (corresponding to the diameter of the Av1 tetramer) and indicate a four-start helix. The latter structure is incompatible with a flat, square subunit arrangement of the Av1 tetramer as proposed by Stasny et al. [(1974) J. Cell. Biol. 60, 311-316]. We propose, therefore, that the subunit arrangement of the Av1 tetramer is of the tetrahedral type. This has also been proposed for the catalase tetramer from optical diffraction studies of electron micrographs of catalase tubes indicating a 222 symmetry [Kiselev, D. A., De Rosier, N. J. and Klug, A. (1968) J. Mol. Biol. 35, 561-566]. Our proposal is in agreement with the recent finding that Av1 protein crystals belong to the P2(1) space group [Weiniger, M. S. and Mortenson, L. E. (1982) Proc. Natl Acad Sci. USA, 79, 378-380].

Partial alkaline dissociation of 24 S (12-meric), 35 S (24-meric) and 60 S (48-meric) hemocyanin from various arthropods was studied by polyacrylamide gradient gel electrophoresis and, in some cases, by electron microscopy. If there are no stable dimers among the subunits, dissociation starts by cleavage of interhexamer bonds, leading to intermediates which are hexamers or multiples of hexamers. Whenever a hemocyanin contained stable dimers, inter-hexamer bonds were also very stable as indicated by the formation of 30 S (19-meric) or 18 S (7-eric) intermediates as primary products. In such cases, inter-hexamer bonds could be cleaved by treatment with reducing agents or with 4M urea; correspondingly, these agents also cleaved the respective dimers into the constituent polypeptide chains. It is concluded that in all cases the dimeric subunits function as bridges between hexamers.

1. Pyridine nucleotide transhydrogenase from Azotobacter vinelandii was purified with a scaled-up procedure. In a typical purification 500 ml cell-free extract from 200 g cells is loaded on an Ado-2',5'-P2--Sepharose 4B affinity column (20 ml bed volume). After washing, the enzyme is desorbed with 2'AMP at neutral pH and further purified by Sephadex G-200 gel chromatography. The enzyme (10--12 mg) is obtained in 40--60% yield and is homogeneous as judged by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulphate. 2. The homogeneity of the purified enzyme is also apparent from electron microscopy studies, where the enzyme appears as a polydisperse set of polymers without contaminating structures and from fluorescence lifetime studies by the method of single-photon counting. The flavin fluorescence appears to decay with a single lifetime tau = 2.5 ns. The polymeric nature of transhydrogenase can be aptly demonstrated by density gradient centrifugation in the presence of KBr. After centrifuging for 50 h at 160 000 X g and 10 degrees C the enzyme is concentrated in a narrow fluorescent band with buoyant density rho b = 1.305 g cm-3. 3. The arrangement of subunits in the transhydrogenase polymer has been derived from optical diffraction studies of electron micrographs. The polymers are built up from a linear assembly of tetramers. Four subunits are placed in a rhomb with sides of 13.5 mm and an angle of 45 degrees (135 degrees) between the sides. A second tetramer is located staggered on top of the first one. Since a variety of other studies have indicated that the polymers dissociate into octamers under alkaline conditions [Voordouw, G. et al. (1979 Eur. J. Biochem. 98,447--454] we conclude that this smallest functional unit is build up from two tetramers.

Limulus polyphemus hemocyanin is a 3.3 x 10(6)-Mr protein containing 48 subunits in an assemblage of eight hexamers. The molecule can be dissociated into monomers and dimers at pH 8.9 in the presence of 0.01 M EDTA. These subunits are heterogeneous and can be separated into five zones (I--V) by DEAE-Sephadex chromatography. Reassembly experiments were carried out with varied subunit mixtures, based on different combinations of the five chromatographic zones, in order to study the structural role of the diverse subunits in the eight-hexamer molecule. The reassembly products were analysed by electron microscopy and ultracentrifugation. No structural role for zone I could be found. Zone V and possibly zone II are needed to form structures larger than hexamers. Absence of zone III causes irregular aggregation of hexamers. Zone IV and perhaps zone II are needed to make eight-hexamer molecules from four-hexamer molecules. From these results we conclude that there is a high degree of subunit specificity in the inter-subunit contacts in the native Limulus hemocyanin molecule.

1. Electron micrographs of the linear mtDNA from Tetrahymena pyriformis strain GL show linear molecules with a duplex 'eye' of variable size in the middle. This indicates that replication of this DNA starts near the middle of the molecule and proceeds bidirectionally to the ends, as previously shown for the mtDNA of strain ST (Arnberg, A.C., Van Bruggen, E.F.J., Clegg, R.A., Upholt, W.B. and Borst, P. (1974) Biochim. Biophys. Acta 361, 266-276). The mtDNAs of these two strains have little base sequence homology beyond the ribosomal RNA cistron (Goldbach, R.W., Bollen-De Boer, J.E., Van Bruggen, E.F.J. and Borst, P. (1978) Biochim. Biophys. Acta 521, 187-197). 2. Electron micrographs of mtDNA from strain ST, spread under non-denaturing conditions, contain only molecules with fully duplex ends. mtDNA spread under conditions of early denaturation contains duplex loops on one end (40% of all molecules) or both ends (37%). The loops are stable to partial denaturation and vary in size from 0.15 to approximately 1.0 micron, most loops measuring 0.25--0.40 micron. No loops are formed with single-stranded DNA under analogous conditions and we conclude from this result that loop formation is based on the presence of straight, rather than inverted, duplications near the ends. 3. When full-length 3H-labelled mtDNA from strain ST, 32P-labelled at the 5'-termini with T4 polynucleotide kinase, was sedimented in alkaline sucrose gradients, greater than 70% of the 3H and less than 30% of the 32P cosedimented with full-length molecules; the remaining 32P sedimented heterogeneously and predominantly with the DNA less than 10% the size of intact single strands. Brief incubations of full-length mtDNA with DNA polymerase I from Escherichia coli and labelled dNTPs at 15 degrees C did not lead to preferential labelling of terminal EcoRI fragments of the DNA. From these results we infer that the DNA contains nicks or gaps near the termini and that these are not bordered by free 3'-OH groups. 4. A model is presented in which straight sequence repetitions at the termini of Tetrahymena pyriformis mtDNA are involved in the later stages of replication. This model can also account for the pronounced terminal heterogeneity previously observed in this DNA.

We have used freeze fracture electron microscopy to study the distribution of membrane proteins in the cytoplasmic membrane of Escherichia coli W3110. While these proteins were distributed randomly at the growth temperature (37 degrees C), there was extensive protein lipid segregation when the temperature was lowered, resulting in bare patches containing no visible particles (protein), and areas of tightly packed or aggregated particles. To understand the segregation process, we have separated the bare patches from the particle rich membrane areas. Lysis of spheroplasts at 0 degrees C leads to cytoplasmic membrane fragments with different amounts of membrane particles per unit area; such fragments have been separated on isopycnic sucrose gradients. The bare patches occurred as low density membranes which were completely devoid of particles. They were compared to normal density cytoplasmic membranes with respect to fatty acid composition, protein distribution as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and their content of several cytoplasmic membrane marker enzymes. The phospholipid to protein ratio of low density membranes was five times greater than that of normal membranes; unsaturated fatty acids were more abundant in the low density membranes. Most proteins had disappeared from the low density membranes. One protein, which had an apparent molecular weight of 26000 on sodium dodecyl sulfate gels appeared to be concentrated in the low density membranes; it accounted for about 50% of the total protein found in this membrane fraction. Of the cytoplasmic membrane markers tested, NADH oxidase and succinate dehydrogenase were excluded, while D-lactate dehydrogenase remained, and even appeared to be concentrated in the low density membranes. These results indicate that while most membrane proteins are associated with the fluid portion of the bilayer, some proteins evidently associate preferentially with phospholipids in the gel or frozen state.

We have analysed the structure of the mtDNAs of six amicronucleate Tetrahymena pyriformis strains, belonging to at least four phenosets, as defined by Borden et al. (Borden, D., Whitt, G.S. and Nanney, D.L. (1973) J. Protozool. 20, 693--700). 2. The mtDNAs of all strains are linear, but they differ in size, in their fragmentation by endonuclease EcoRI and in overall sequence; less than 20% sequence homology was found by DNA-DNA hybridization in all combinations tested, except for the mtDNAs from strains T and ST which are indistinguishable. 3. In spite of these marked sequence differences the mtDNAs of all strains share two structural peculiarities: ragged (gnawed) duplex ends and a duplication-inversion, which varies in length between 0.3 and 1.2 micrometer, depending on the strain. In four strains the duplication-inversion is terminal, allowing formation of single-stranded DNA circles with a duplex tail; in two strains it is subterminal. 4. The ragged ends and sub-terminal position of the duplication-inversion in some of the Tetrahymena mtDNAs do not fit any of the current models for the replication of linear mtDNAs.

11S and 18S fractions of yeast mitochondrial RNAs, isolated by electrophoresis through agarose gels, have been found by electron microscopy to contain approximately 50% circular molecules. Circles in the 11S fraction have a contour length of 0.36 +/- 0.02 micron, which is approximately equal to the length of the majority of linear molecules also present. Circles in the 18S fraction have an average length of 0.78 +/- 0.11 micron. The size distribution is broader than for the 11S fraction, and we cannot exclude the possibility that more than one size class may be present. The 11S circular RNA forms circular R loops and RNA-DNA hybrids with DNA fragments of the oxi 3 region of mtDNA, which contains the structural gene for subunit 1 of cytochrome oxidase. As judged from the electron micrographs, the complete RNA participates in hybrid formation and the sequences coding for it appear to be continuous. Both 11S and 18S circles withstand treatment with DNAase and pronase. They are not eliminated by treatment with 1 M glyoxal in 50% formamide for 1 hr at 50 degrees C. We conclude that they are covalently closed. The function of the circular RNAs is unknown. They may be active as mRNAs, storage forms, or arise in a cut-and-splice process which generates mRNAs from longer transcripts.

The 'microheterogeneity model' (R.J. Siezen and R. van Driel, Biochim. Biophys. Acta 295 (1973) 131) and the 'incompetent whole molecule model' (G. Kegeles, Arch. Biochem. Biophys. 180 (1977) 530) for the dissociation of Helix pomatia alpha-hemocyanin whole molecules to half molecules were tested experimentally, using ultracentrifugation and stopped-flow dilution analysis. Results of differential sedimentation experiments followed by stopped flow analysis of separated fractions of 60 S and 100 S molecules were not entirely as predicted by the incompetent model, the agreement depending on the pH and ionic strength of analysis. A considerable amount of stopped-flow dilution response could be attributed to material sedimenting between 60 and 100 S. This material appears to be the main equilibrating fraction, and its amount is considerably larger than that predicted by the microheterogeneity model. Increased hydrostatic pressure was found to enhance this fraction, whereas fixation or low ionic strength reduced or eliminated this fraction. Nonequilibrium components of 30, 50 and 80 S were detected and partially purified by differential sedimentation.

Electron-microscopic studies revealed that two types of subunits of Panulirus interruptus haemocyanin crystallize in different ways. Homohexamers of subunit a give close-packed two-dimensional crystals whereas homohexamers of subunit c form open two-dimensional arrays. We applied computer-image analysis to these arrays and studied the differences in crystallization properties by combining the electron-microscopic data with amino acid sequence information and the X-ray diffraction model of subunit a.

1. Pyridine nucleotide transhydrogenase of Azotobacter vinelandii purified by affinity chromatography consists of a mixture of polydisperse rods at neutral pH. No other structures are seen by electron microscopy. 2. At high pH (8.5--9.0) the rods depolymerize. Complete depolymerization can be achieved in 0.1 M Tris-Cl pH 9.0. The depolymerized enzyme has a molecular weight of 421000 (sedimentation equilibrium), its sedimentation coefficient s20, w = 15 S and its Stokes' radius Rs = 7 nm. Since gel electrophoresis in the presence of sodium dodecyl sulphate shows that transhydrogenase consists of a single polypeptide chain of molecular weight (54 +/- 2) X 10(3) it follows that the depolymerized enzyme has an octameric quaternary structure. We propose that this octamer serves as the functional monomeric unit ('unimer') from which the polymeric form of transhydrogenase is constructed. 3. Gel filtration and sucrose gradient centrifugation studies of cell-free extracts from A. vinelandii show the unimer to be the predominant active species.

Digestion of beta c-hemocyanin yielded tubular polymers as well as so-called collars. Analysis of the collar fraction revealed that it consisted of two fragments, one having a relative molecular mass of approx. 125000 and the other a relative molecular mass of approx. 65000. They were separated using ion-exchange chromatography. The large fragment dissociated around pH 9.5 into two components having a molecular mass of approx. 65000. Both fragments bound oxygen and displayed heterotropic interactions. The large fragment bound oxygen with a Hill coefficient less than unity under certain conditions. The fragments differed also in amino acid composition, sugar content, spectral parameters, association-dissociation phenomena and ageing. by comparison with fragments obtained after limited proteolysis of tenth molecules, their location in the polypeptide chain was established.

1. We have constructed a physical map of the mtDNA of Tetrahymena pyriformis strain ST using the restriction endonucleases EcoRI, PstI, SacI, HindIII and HhaI. 2. Hybridization of mitochondrial 21 S and 14 S ribosomal RNA to restriction fragments of strain ST mtDNA shows that this DNA contains two 21-S and only one 14-S ribosomal RNA genes. By S1 nuclease treatment of briefly renatured single-stranded DNA the terminal duplication-inversion previously detected in this DNA (Arnberg et al. (1975) Biochim. Biophys. Acta 383, 359--369) has been isolated and shown to contain both 21-S ribosomal RNA genes. 14 S ribosomal RNA hybridizes to a region in the central part of the DNA, about 8000 nucleotides or 20% of the total DNA length apart from the nearest 21 S ribosomal RNA gene. 3. We have confirmed this position of the three ribosomal RNA genes by electron microscopical analysis of DNA . RNA hybrid molecules and R-loop molecules. 4. Hybridization of 21 S ribosomal RNA with duplex mtDNA digested either with phage lambda-induced exonuclease or exonuclease III of Escherichia coli, shows that the 21-S ribosomal RNA genes are located on the 5'-ends of each DNA strand. Electron microscopy of denaturated mtDNA hybridized with a mixture of 14-S and 21-S ribosomal RNAs show that the 14 S ribosomal RNA gene has the same polarity as the nearest 21 S ribosomal RNA gene. 5. Tetrahymena mtDNA is (after Saccharomyces mtDNA) the second mtDNA in which the two ribosomal RNA cistrons are far apart and the first mtDNA in which one of the ribosomal RNA cistrons is duplicated.

A model for the quaternary structure of Escherichia coli RNA polymerase (nucleosidetriphosphate:RNA nucleotidyltransferase, EC 2.7.7.6) is presented. It is based on results from classification of profiles of enzyme molecules, and from application of immuno electron microscopy. Classification of molecules, prepared with the single carbon layer technique, was first achieved for images recorded in dark field with the scanning transmission electron microscope and later on for images recorded in bright-field transmission electron microscopy. It results in five approximately equally sized groups, containing about 80% of the core enzyme profiles. Holoenzyme profiles can be grouped into the same classes, and have approximately the same dimensions (9 nm X 16 nm). Based on the shapes and sizes of the classified profiles, a tentative model for core enzyme has been constructed. Correlation of shadow projections of this model, with the distributions of attachment sites of antibodies against alpha, beta, beta' and sigma over the profiles, has led to models for core and holoenzyme in which the subunits are localized. The model is compared with literature data on the quaternary structure of RNA polymerase.

The binding of carbon monoxide to alpha and beta-hemocyanin from the snail Helix pomatia was studied under equilibrium conditions. Homotropic interactions upon carbon monoxide binding were much weaker than upon the binding of oxygen. Heterotropic interactions (Bohr effect and calcium-ion effect), however, were just as strong as in the case of the binding of oxygen. For alpha-hemocyanin a linkage has been observed between the binding of carbon monoxide and a change in quaternary structure of the protein.

Our experience in the growth of two-dimensional crystals of different proteins is presented. Polyethylene glycol was used to produce two-dimensional arrays of haemocyanin from O. vulgaris and of cholera toxin. The arrays showed a hexagonal close-packed structure of only randomly oriented molecules. The increase in protein concentration probably occurred too quickly to allow complete crystallization. Different two-dimensional arrays of hexameric haemocyanin molecules (from P. interruptus) were obtained by microdialysis through the specimen supporting film. A comparison was made with X-ray data. Two-dimensional tetrameric arrays of molecules, possibly rhodopsin, were seen in samples of bovine retinal rod outer segments in the presence of ammonium sulphate. Two-dimensional crystals of complex I (from bovine mitochondria) were prepared by dialysis in the presence of ammonium sulphate. A three-dimensional reconstruction was made from two tilt-series by computer filtration using the direct SIRT procedure. Finally, the possibility of computer crystallization using correlation techniques in combination with correspondence analysis is discussed.

1. We have studied denatured Tetrahymena mtDNA by electron microscopy using the formamide technique. 2. After denaturation all DNA is single stranded, but within a few minutes single-stranded circles with a duplex tail are formed. 3. The duplex tail is 1.3 mum long, i.e. 8 percent of the length of native mtDNA, and it often contains a small single-stranded eye. 4. Digestion of the duplex DNA with exonuclease III of Escherichia coli abolishes its ability to form circles and duplex tails after denaturation. 5. Renaturation of denatured mtDNA leads to the formation of duplex circles with single-stranded section and/or duplex tails. In addition, a minority of duplex circles without apparent tails is formed, but these circles contain a small ambiguous section. 6. We conclude that this mtDNA contains a long terminal duplication-inversion, that could be involved in the replication of this linear mtDNA.

A simple and efficient procedure for the construction of bifunctional molecules is described and their use in a variety of applications documented. This procedure is based on our observation that mouse IgG1 monoclonal antibodies, when mixed with equimolar amounts of a high-affinity rat monoclonal antibody specific for mouse IgG1, yield uniform cyclic tetramolecular complexes each consisting of two mouse and two rat antibodies as shown by gel electrophoresis and electron microscopy. When solutions of two mouse antibodies (e.g. a and b) are mixed prior to the formation of complexes with the rat antibody, stable bispecific (a X b) complexes together with monospecific (a X a and b X b) complexes are obtained. Bispecific complexes prepared in this way were able to efficiently bind peroxidase to cell surface antigens, and to bind red blood cells to selected nucleated cell types present in heterogeneous populations. Tetrameric antibody complexes are more easily prepared than bispecific antibodies or bifunctional antibodies produced by transfection of myelomas with recombinant genes. They also have the advantage that the antigen-binding properties of the bivalent monoclonal antibodies are not compromised. Tetrameric antibody complexes thus represent a powerful new type of cross-linking reagent that may have a wide spectrum of applications in biology and medicine.

1. Isolated mtDNA from Tetrahymena pyriformis strain GL is a linear duplex molecule with an average molecular weight of 32.6 - 10(6) and without internal gaps or breaks. Denaturation of this DNA results in single strands with a duplex hairpin at one end. The length of this hairpin varies between 0 and 5 micrometer within one preparation. 2. Uder renaturation conditions the single strands with hairpins are able to circularize in two ways, depending on the length of the hairpin. Circularization is also observed after partial digestion with exonuclease III of native strain GL mtDNA. 3. All these data fit a model (see Fig.2) in which the DNA is heterogeneous in length at both ends. At the left end a 10-micrometer duplication-inversion is present; part of this duplication-inversion is complementary to a region at the right end of the molecule. 4. The analogy between the structural peculiarities of strain GL mtDNA and of some linear viral DNAs is stressed.

We investigated the structure of two-dimensional crystals from bovine heart mitochondrial NADH: ubiquinone oxidoreductase. A detailed description of uranyl acetate-stained crystals demonstrated that they are composed of fragments in a spatial arrangement according to space group P4212 [J. Brink, S. Hovmöller, C.I. Ragan, M.W.J. Cleeter, E.J. Boekema and E.F.J. van Bruggen, European J. Biochem. 166 (1987) 287]. To gain more structural information on the crystal structure and to assess the effects of various negative stains on the structure preservation and appearance, we examined stained crystals by means of electron microscopy and image analysis. The space group P4212 appeared to be present for several stains tested, i.e. ammonium molybdate, uranyl acetate, uranyl nitrate and uranyl sulphate. Use of phosphotungstic acid and silicotungstate resulted in a reduction of symmetry to pseudo-P4212 or p4. Use of sodium tungstate led to a considerable loss of resolution to 3.8 nm at best, whereas otherwise 1.5 to 1.9 nm could be demonstrated. The lattice vectors were not affected by the stains; they were determined as a = b = 14.9 +/- 0.25 nm with gamma = 89.8 degrees +/- 0.6 degrees. Image analysis showed the presence of similar structures with the molybdate and uranyl compounds. Differences were observed in the case of the tungstate type of stains. Furthermore, the analysis revealed the complete absence of the four small pores of 2.0 nm diameter in the unit cell. This effect was observed irrespective of the type of stain and supporting film, and could be ascribed only to the glow-discharge treatment of the supporting film. The observed difference must be caused by changed interactions between the protein, stain and supporting film. Application of correspondence analysis and clustering algorithms to the various reconstructed images of the crystals showed that they could be separated into several clusters. Each of these clusters corresponded on the average to only one type of stain, whereas a further division according to the specific uranyl compounds was observed. This study therefore shows that under identical preparation conditions subtle differences between individual stains can be detected.

The practical usefulness of a STEM (Scanning Transmission Electron Microscope) for the study of the structure of biological macromolecules has been investigated using a STEM attachment connected to a TEM (Transmission Electron Microscope), which in one case was equipped with a tungsten hairpin cathode, and in the other case with a field emission gun. The point to point resolution has been determined. Results obtained in STEM dark field from light negatively stained specimens are compared with results obtained in TEM bright field from normal negatively stained specimens. In addition unstained molecules have been visualized. Some remarks are made about preparation methods suitable for STEM.

Two-dimensional crystalline arrays of NADH:Q oxidoreductase preparations have been obtained by microdiffusion of protein dissolved in detergent against a 15 mM sodium acetate buffer of pH 5.5 containing 10% (w/v) ammonium sulphate. Electron microscopy was used to study the structure of negatively stained crystals. Computer-reconstructed images were obtained by the Fourier peak filtering method. The crystals have p4 symmetry and a square unit cell with dimensions of 15.2 +/- 0.5 nm. The four asymmetric units in the unit cell form a single tetrameric molecule with a dimension in the third direction of 8.2 nm. It is concluded on the basis of the estimated molecular mass that each tetramer cannot contain more than only one FMN molecule. This implies that the tetramers possibly are only a part of Complex I, since there is much evidence that one functional enzyme molecule of Complex I contains two FMN molecules.

A series of overlapping recombinant clones, which cover the vitellogenin gene, has been isolated from a phage-lambda linked chicken gene library. The DNA of the overlapping clones spans 28 kb of contiguous DNA sequences in the chicken genome. Electron microscopic analysis of hybrids between vitellogenin mRNA and the genomic clones indicates that the chicken vitellogenin gene has a length of approximately 22 kb, about 3.8 times the size of the mRNA. The mRNA sequence is interrupted by at least 33 intervening sequences (introns). Comparison with the vitellogenin gene A2 from Xenopus laevis (Wahli et al., 1980, Cell 20: 107-117) indicates conservation of the number and length of the exons during evolution. Heteroduplex analysis reveals a short stretch of sequence homology between the genes from chicken and frog.

Squilla mantis hemocyanin is composed of two hexameric subunits but has electron microscopic profiles different from other bis-hexameric hemocyanins, e.g. Astacus and Homarus. We distinguished three different electron microscopic profiles of S. mantis hemocyanin: two sideviews and a topview. These profiles were studied using computer image alignment and correspondence analysis [Van Heel, M. and Frank, J. (1981) Ultramicroscopy 6, 187 - 194]. With the results of this analysis we were able to build a three-dimensional model for the quaternary structure of this hemocyanin. In this model the two hexamers are stacked in such a way that their hexagonal surfaces overlap to about 60% of their width. In the overlap area four subunits are arranged in two different interhexameric pairs, each forming a bridging area between the two hexamers.

Electron microscopic techniques have been employed to investigate the ability of didodecylphosphate vesicles (diameter approx. 900 A) to fuse in the presence of Ca2+. As revealed by negative staining, Ca2+ induces extensive fusion and large vesicles with diameters up to 7000 A are formed. In a process secondary to fusion, the fused vesicles display a tendency to flatten and are subsequently transformed into extended tubular structures. Freeze-fracture electron microscopy, in conjunction with 31P NMR and selected area electron diffraction measurements indicate that the tubes are packed in a hexagonal (HII) array and that the amphiphiles are converted from the lamellar to the hexagonal HII phase. The relationship between membrane fusion and the lamellar-to-hexagonal phase transition is discussed in terms of formation and abundance of transiently stable inverted micellar intermediates at contact regions between two interacting membranes. A model for the conversion of the (vesicular) lamellar into the (tubular) hexagonal HII phase is presented, taking into account the molecular shape of the amphiphile. The relevance of using simple synthetic amphiphiles as models for phospholipid bilayers and complex biomembrane behavior is briefly discussed.

An electron microscopical study was made of the influenza virus, type B/Hong Kong, in the unstained, frozen, hydrated state after quench-freezing in cooled liquid ethane. The results are compared with data from negatively stained specimens. It is shown that cryo-electron microscopy confirms and extends the data obtained by conventional methods. In particular, the virus is shown to be circular in projection with no indication of icosahedral symmetry, the lipid membrane is clearly resolved as a bi-layer and it is demonstrated that the distribution of material within the bi-layer is non-uniform, with a shell of more electron dense material surrounding a less dense central region. Neuraminidase spikes are not clearly distinguished from haemaglutinin spikes. The diameter of the complete B/Hong Kong virus was estimated from cryo-micrographs as 1270(+/- 70) A. Some preliminary data for influenza virus type A/X31 are presented.

Upon trypsinolysis Helix pomatia beta-hemocyanin forms long tubular structures, which appear to be linear polymers of hemocyanin molecules from which the collar structure has been removed. Polyacrylamide gel electrophoresis in the presence of sodium dodecylsulfate shows that only few peptide bonds are hydrolyzed by trypsin. The structure of the polymers has been investigated by electron microscopy, combined with optical diffraction. Preliminary X-ray diffraction data are presented. Functional properties of the polymers are similar to those of the native protein. Both show a calciumion-dependent co-operativity of oxygen binding and a Bohr effect. The results suggest that the collar of a hemocyanin molecule has no special function in the process of (co-operative) oxygen binding, different from that of the wall of the molecule.

alpha-Hemocyanin from the Roman snail Helix pomatia is composed of polypeptide chains with a molecular weight of 360000 +/- 30000. The cylindrically shaped hemocyanin molecule contains 20 of these large chains. The polypeptide chain has been split into components with molecular weights of: 210000, 154000, 147000, 112000, 120000, 98000, 55000, and 50000, by gentle proteolysis with enzymes of different specificities. Most of the fragments have molecular weights which are about 50000 or a multiple of 50000. Departure from these values, as found in the 112000 and 120000 fragments, is probably caused by the high carbohydrates content of these components. A mixture of these fragments has the same oxygen binding properties as the nondigested protein. Subtilisin converts the hemocyanin polypeptide chain, under appropriate conditions, almost completely into fragments of 50000 and 55000 daltons with conservation of the oxygen binding properties of the nondigested protein. We conclude from these studies that the polypeptide chain of Helix pomatia alpha-hemocyanin is folded into about seven compact tertiary structures, which are covalently interconnected. This chain of structural domains has been visualized. (Siezen and Van Bruggen (1974), J. Mol. Biol. 90, 77-89) by electron microscopy, which shows 1/20 hemocyanin molecules to be flexible structures consisting of 7-8 apparently spherical units of 55-60 A diameter.

beta-Haemocyanin molecules consist of 20 very large polypeptide chains. These chains are composed of eight structural domains. So-called 'collar' domains can be removed by trypsinolysis of the native cylindrical molecule, resulting in an association of the remaining hollow cylinders into large tubular polymers. Dissociation of the tubular polymers gives one single- and four multi-domain fragments. The role of these fragments in the reassembly process of these tubular polymers was investigated. The two-domain fragment could form tubular polymers. The other domain fragments were not able to form tubular polymers unless in the presence of the two-domain fragment. Tubular polymers with enlarged diameter and ribbon-like structures were observed in the reassembly products when the one-domain fragment was omitted.

Purified ribosomal DNA from Saccharomyces carlsbergensis contains a small proportion of circular DNA molecules with a contour length of 3 micron or integral multiples thereof. Hybridization of yeast ribosomal DNA with 26 S rRNA, using the R-loop technique, reveals that these circular molecules contain sequences complementary to yeast ribosomal RNA. We suggest that these extrachromosomal rRNA genes may be intermediates in the amplification of rRNA genes in yeast.

Hemocyanin of Helix pomatia is a respiratory protein with a molecular weight of 9 times 10-6; it contains 180 oxygen binding sites. The reaction of hemocyanin with the bifunctional reagent dimethyl suberimido ester, which reacts with amino groups, has been studied. Up to 75 per cent of the amino groups can be modified without inactivation of oxygen binding sites or dissociation of the protein, It appears that hemocyanin can be fixed in a state with low oxygen affinity by modification of the deoxy protein, and in a state with high oxygen affinity by modification of the oxy protein. Using conditions under which native hemocyanin binds oxygen cooperatively (Hill coefficient 2.9), modification of deoxy- and oxyhemocyanin yields derivatives with different oxygen affinities (P50 equals 10 and 2.2 mm, respectively). Both the deoxy and oxy derivatives show strongly reduced cooperativity (Hill coefficients 1.4 and 1.1, respectively). Modification of oxy- and deoxyhemocyanin subunits (molecular weight one-tenth of the native protein), which bind oxygen noncooperatively, results in derivatives with oxygen binding properties identical with those of unmodified subunits. Parallel experiments have been carried out with a unifunctional reagent, methyl acetoimido ester. Modification of partially oxygenated hemocyanin under conditions at which the protein binds oxygen cooperatively yeilds derivatives with redued cooperativity (Hill coefficents 1.1-1.2) and an oxygen affinity depending on the oxygen saturation whivh modification had been carried out. The results are consistent with a simple two-state model for the cooperativity of oxygen binding by these giant hemocyanin molecules.