Comparative analysis of the sequences of the three collagen chains α1(I), α2 and α1(III): Functional and genetic aspects

The amino acid sequences of type I collagen containing α1(I) and α2 chains at a ratio of 2:1, and of type III collagen consisting of α1 (III) chains are known. A statistical analysis of the sequences of these α chains is presented. The inter-chain comparison showed a high level of homology between the three α chains. The interactive amino acids, such as the polar charged and part of the hydrophobic residues responsible for the assembly of the molecules, are strongly conserved. The intra-chain analysis revealed that the α chains are divided into four related D units, each with a length of 234 residues. Between the D units within a chain the polar residues show a higher variability than the hydrophobic amino acids.Besides the D units, other periodicities such as $\text{D}\text{3}$ (78 residues), $\text{D}\text{6}$ (39 residues), $\text{solD}\text{11}$ (21 residues) and $\text{solD}\text{13}$ (18 residues) were observed, particularly in α1 (I) and α1 (III). The D unit is a functional repeat that is formed by the interactive polar charged and hydrophobic residues and which determines the aggregation of the molecules. The $\text{solD}\text{3}$ unit is mainly pronounced by the non-interactive residues such as proline and alanine and appears to be a reminiscence of a primordial gene. The smaller periodic repeating units may be considered as additional genetic units or as structural units, which determine the triplehelical pitch and thus the lateral aggregation of the molecules.In contrast to α1 (I) and α1 (III), the α2 chain shows less regularity in its internal structure.     

The determination of the helical screw angle of a helical particle from its diffraction pattern

A Fourier analysis of the distributions of different types of amino acids in the sequence of tropomyosin shows strong 14th-order peaks in the profiles of both negatively charged and non-polar amino acids, with a period of $\text{19}\text{2}\text{3}$ residues and an overall repeat length of 275 ± 2 amino acids, which is shorter than the sequence length of 284 amino acids. Both peaks are statistically significant and confirm Parry's work (1974, 1975b). The regularities are analysed in terms of an assumed supercoil structure in which two α-helices lie parallel and in register to form a supercoil with a pitch of 137 Å. These molecules are then assumed to overlap end-to-end by eight to nine amino acids so that the periodicity is continuous along an extended filament of linked tropomyosin molecules. The periodic features are stronger in the outer surface of the molecule away from the core of the supercoil. The sequence divides into 14 bands which each have a narrow zone of net positive charge and a broader negatively charged zone. Overlapping every positive zone is a hydrophobic zone which always has at least one non-polar group on the outer surface. Anomalies in the charge distribution are found near the molecular ends and close to Cys190. These are attributed to the end-to-end overlap site and the troponin binding site.In the thin filament the 137 Å pitch supercoil would make seven half-twists relative to the twisted actin helix along a 385 Å length, so that a pair of adjacent bands would be oriented equivalently with respect to a pair of actins 28 Å apart. We therefore suggest that the bands (each containing one zone of each type) should be divided alternately into two series, α and β. Every pair of bands is $\text{39}\text{1}\text{3}$ residues long and each of the seven pairs corresponds with one segment of the 42-residue gene duplication repeat observed previously in the sequence. The disparity between the periods of 42 and $\text{39}\text{1}\text{3}$ is overcome by deletions and insertions. The $\text{39}\text{1}\text{3}$-residue periodicity is not simply a consequence of the supercoil structure or gene duplication but is probably a result of adaptation to the spatial periodicity of the actin helix in muscle. Although the α and β bands are alike in general, they differ systematically in detail and the α bands are more regular than the β.We propose that the seven α and seven β bands are alternative sets of sites which bind equivalently to complementary groups of sites on seven actins in the “relaxed” and “active” states of muscle, respectively. In each band the negative zone probably attaches to actin by magnesium bridges and the hydrophobic zone by direct contacts with the narrow outer edge of the supercoil. Since the supercoil twists 90 ° relative to actin on passing between adjacent α and β bands, a quarter rotation of the whole tropomyosin molecule would detach one set of seven sites and attach the other, allowing a highly co-operative switch mechanism.     

Sequence regularities and packing of collagen molecules

Fourier analysis of sequences along edges of the type I collagen molecule constructed from two α1(I) and one α2 chains shows that the molecule is two-sided if the supercoil pitch of the α chains along the molecular axis, P, is 39 residues ($\text{D}\text{6}$, where D = 234 residues or 67 nm). One side has alternating charged and hydrophobic regions with spacings of $\text{D}\text{6}$, while the other side has an excess of hydrophobic residues with a spacing of $\text{D}\text{11}$. These characteristics arise from sequence regularities in the α chains and the geometric relationship between the chains. The pattern is marginally strongest with α2 as chain 1. The $\text{D}\text{6}$ sides could form the inside of a helical microfibril where contacts between molecules would fall P apart along the α chains. The $\text{D}\text{11}$ sides could form the outside of the microfibril where contacts between microfibrils would be spaced apart by the α chain supercoil along the microfibril axis, P′. If the microfibril is a 5₄ helix of D-staggered collagen molecules with a left-handed supercoil of pitch $\text{20D}\text{11}$, P′ is close to $\text{2D}\text{11}$ (43 residues). $\text{2D}\text{11}$ subsets in the α chains give rise to the $\text{D}\text{11}$ spacing along the molecule. The microfibril has 4₁ screw symmetry satisfying X-ray diffraction evidence that microfibrils pack in a tetragonal unit cell.This model is the same as proposed previously by us (Trus & Piez, 1976: Piez & Trus, 1977) except that P = 39 rather than 30 residues. Contrary to our earlier assumption, P = 39 residues is within the range allowed by X-ray diffraction measurements. The present results favor P = 39 since it relates regularities in the α chain sequences to helical parameters in a direct way. Furthermore, model studies show that geometric arguments which support P = 30 are equally strong at P = 39 residues.     

Dipole moments of random coil polypeptides

The Rotational Isomeric States model is applied to calculate dipole moments of polypeptides of the twenty natural α-amino acids in the random coil state. Dipole moments of each repeat unit (μ_i), are evaluated using a quantum mechanics procedure. Dipole moment ratios ($\text{D}{}_{\mathrm{x}}\text{=}\text{〈μ}{}^2\text{〉}\text{x}\text{μ}{}_{\mathrm{i}}{}^2\text{,}\text{x = number of repeat units}$) of homopolypeptides are calculated and extrapolated to x →?. With a few exceptions, D_? = 0.36 ± 0.1. Ten actual proteins and three enzymes are also studied; their dipole ratios (D_x′ =〈μ〉/x) range from 7.34 to 10.57 in 10^?59 C² m² (6.6–9.5 D²). Diffferences in the values of D_x′ are due mainly to the different contributions, μ_i, of the amino acid residues contained in each polymer, whereas the sequence of amino acids has a very minor effect.     

Structural studies on the extracellular polysaccharide of the red alga, Porphyridium cruentum

Partial acid hydrolyzates of the extracellular polysaccharide from Porphyridiunm cruentum yield three disaccharides and two uronic acids. These constitute all of the uronic acid in the polymer. The novel disaccharides are 3-O-(α-$\text{D}$-glucopyranosyl- uronic acid)-$\text{L}$-galactose, 3-O-(2-O-methyl-ca-glucopyranosyluronic acid)-$\text{D}$- galactose, and 3-0-(2-0-methyl-a-$\text{D}$-glucopyranosyluronic acid)-$\text{D}$-glucose. The polyanion of high molecular weight contains $\text{D}$- and $\text{L}$-galactose, xylose, $\text{D}$-glucose, $\text{D}$-glucuronic acid and 2-O-methyl-D-glucuronic acid, and sulfate in molar ratio (relative to $\text{D}$-glucose) of 2.12:2.42:1.00:1.22:2.61. Preliminary periodate-oxidation studies suggest that the hexose and uronic acids are joined to other residues by ( 1→3) glycosidic linkages. About one-half of the xylose residues are (1→3)-linked.     

Physical measurements on the lipid-containing bacteriophage φ6

Bacteriophage φ6 has been studied by small-angle X-ray scattering, intensity-fluctuation spectroscopy, analytical ultracentrifugation, and spectroscopy. The sedimentation coefficient (s₂₀⁰, w) is 375 S, the diffusion coefficient (D₂₀⁰, w) is 2.66 · 10^?8 cm²/s. Using the Svedberg equation and an estimate of the partial specific volume, the M_r is 1.49 ± 0.32 · 10⁸.A simple model which describes φ6, is a central sphere consisting of RNA and protein of radius 330 Å and an outer shell of low electron density 40 Å thick. The RNA may form five concentric shells in the region $\text{r = 140?290}\text{A}\text{?}$     

Analysis and prediction of the packing of α-helices against a β-sheet in the tertiary structure of globular proteins

The packing of α-helices and β-sheets in six $\text{α}\text{β}$ proteins (e.g. flavodoxin) has been analysed. The results provide the basis for a computer algorithm to predict the tertiary structure of an $\text{α}\text{β}$ protein from its amino acid sequence and actual assignment of secondary structure.The packing of an individual α-helix against a β-sheet generally involves two adjacent ± 4 rows of non-polar residues on the α-helix at the positions i, i + 4, i + 8, i + 1, i + 5, i + 9. The pattern of interacting β-sheet residues results from the twisted nature of the sheet surface and the attendant rotation of the side-chains. At a more detailed level, four of the α-helical residues (i + 1, i + 4, i + 5 and i + 8) form a diamond that surrounds one particular β-sheet residue, generally isoleucine, leucine or valine. In general, the α-helix sits 10 Å above the sheet and lies parallel to the strand direction.The prediction follows a combinational approach. First, a list of possible β-sheet structures (10⁶ to 10¹⁴) is constructed by the generation of all β-sheet topologies and β-strand alignments. This list is reduced by constraints on topology and the location of non-polar residues to mediate the sheet/helix packing, and then rank-ordered on the extent of hydrogen bonding. This algorithm was uniformly applied to 16 $\text{α}\text{β}$ domains in 13 proteins. For every structure, one member of the reduced list was close to the crystal structure; the root-mean-square deviation between equivalenced C^α atoms averaged 5.6 Å for 100 residues. For the $\text{α}\text{β}$ proteins with pure parallel β-sheets, the total number of structures comparable to or better than the native in terms of hydrogen bonds was between 1 and 148. For proteins with mixed β-sheets, the worst case is glyceraldehyde-3-phosphate dehydrogenase, where as many as 3800 structures would have to be sampled. The evolutionary significance of these results as well as the potential use of a combinatorial approach to the protein folding problem are discussed.     

Evolution of the three overlapping gene systems in G4 and phi X174.

Bacteriophage G4 has the same $\text{A}\text{B}$ and $\text{D}\text{E}$ overlapping gene systems as φX174 and together with the A and $\text{C}\text{K}$ overlapping gene system (Shaw et al., 1978), 7 of the 11 G4 and φX174 genes are involved in overlaps. The nucleotide differences between G4 and φX174 in the overlapping portions of the A, C and D genes are 23%, 27% and 21%, respectively, compared with 32%, 36% and 34% in the non-overlapping portions of the same genes. The amino acid differences between the G4 and φX174 overlapping B, K and E proteins, are 44%, 39% and 44%, respectively, compared with 28%, 26% and 16% in the regions of genes A, A and C, and D which contain genes B, K and E. These results suggest that the nucleotide sequences of overlapping genes evolve at almost the same rate as in non-overlapping genes, and that this is made possible by a lower amino acid sequence stringency of one of the pairs of proteins. The overlapping $\text{D}\text{E}$ and A and $\text{C}\text{K}$ gene systems may have originated by taking advantage of a high incidence of T nucleotides in the second codon position to produce a hydrophobic protein and the $\text{A}\text{B}$ gene system may have evolved by read-through of the A gene into the B gene. From the nucleotide sequences, other overlapping genes appear to be possible in these bacteriophages.     

X-ray diffraction analysis of dehydrated myelin

Improved X-ray diffraction data from dry nerve myelin are presented. In addition to the spacings of approx. 150 Å, 60 Å, 44 Å and 34.6 Å, which have been previously reported, we identify a 14 Å series. The data suggests that the hydrocarbon chains in the single bilayer ($\text{≈ 60}\text{A}\text{?}$) is ordered, whereas in the double bilayer ($\text{≈ 150}\text{A}\text{?}$) and in the fluid phase ($\text{≈ 44}\text{A}\text{?}$) it is disordered. It is shown that cholesterol ($\text{≈34.6}\text{A}\text{?}$) exists as a bilayer, and the 14 Å series is probably another cholesterol phase.     

Crystallization of alfalfa mosaic virus coat protein as a T = 1 aggregate

Reassembled alfalfa mosaic virus coat protein was partially digested with trypsin to remove the first 26 amino acids (Bol et al., 1974). These particles are empty icosahedral protein shells built with 60 alfalfa mosaic virus protein subunits. This aggregate has been crystallized in two different crystal forms, one of which diffracts X-rays to at least 3.4 Å resolution. The type I crystals (space group $\text{P6}{}_3\text{, a = 200}\text{A}\text{?}\text{, c = 314}\text{A}\text{?}$) contain two particles per cell separated by 195 Å with each sitting on a 3-fold axis. The type II crystals contain three particles per cell in space group P3₁or P3₂ ($\text{a = 201}\text{A}\text{?}\text{, c = 485}\text{A}\text{?}$). Other T = 1 viral particles have very similar diameters.     

Crystallographic studies of immunoglobulins: crystallization of the Fc fragment of rabbit IgG with and without cleavage of the inter-chain disulphide bridge

The Fc fragment was prepared from rabbit immunoglobulin G by digestion with papain, both with the inter-chain disulphide bond intact, and after reduction and alkylation. These two types of Fc crystallized in different, yet related forms, each with one dimer in the asymmetric unit. The covalently linked dimer crystallized in space group $\text{P2}{}_1\text{; a = 68.85 ± 0.05}\text{A}\text{?}\text{, b = 72.50 ± 0.05}\text{A}\text{?}\text{, c = 60.40 ± 0.05}\text{A}\text{?}$ and β = 105.1 ± 0.2 °. The reduced, non-covalently linked dimer also crystallized in space group $\text{P2}{}_1\text{; a = 81.55 ± 0.05}\text{A}\text{?}\text{, b = 55.65 ± 0.05}\text{A}\text{?}\text{, c = 68.85 ± 0.05}\text{A}\text{?}$ and β = 1051 ± 0.2 °. A non-crystallographic 2-fold axis relating the two identical polypeptide chains is clearly visible in the h0l projection of the second crystal form.     

Neutron small angle scattering of the Mo-Fe protein (nitrogenase) from Clostridium pasteurianum

Neutron small angle scattering measurements of solutions of the Mo-Fe protein from $\text{C. pasteurianum}$ have yielded the following results. The molecular weight of the protein is 208,000 ± 10,000, in agreement with figures obtained by other methods. The radius of gyration is 39.8 ± 0.7 Å in H₂O, and 37.6 ± 0.3 Å in D₂O. The experimental scattering curves have been compared with the calculated scattering curves of simple homogeneous bodies. It is concluded that the MoFe protein from $\text{C. pasteurianum}$ is a non spherical particle having an axial ratio of 2:1, and that it probably has little, if any, solvent containing cavities.     

Laser-light scattering investigation of the micellar properties of gangliosides

The micellar properties of gangliosides in water solutions were investigated by quasielastic light scattering measurements. G_M1 and G_D1a gangliosides were isolated from calf brain, purified to more than 99% and dissolved in 0.025 M Tris—HCI buffer (pH 6.8) at 37°C. The average intensity of scattered light and the intensity correlation function were measured by an apparatus including a 5145 Å argon laser and a real-time digital correlator. The scattered intensity data allowed the derivation of an upper limit to the critical micelle concentration (c₀) and the evaluation of the molecular weight (M) of the micelle. The intensity correlation function gave the diffusion coefficient D, and hence the hydrodynamic radius R_H, and also contained information on the polydispersity of the sample. We find c_o < 1 × 10^?6 M for both G_M1 and G_D1a, M = 532 000 ± 50 000 and $\text{R}{}_{\mathrm{<mtext>H</mtext>}}\text{= 63.9 ± 2}\text{A}\text{?}$ for G_M1, and M = 417 000 ± 40 000 and $\text{R}{}_{\mathrm{<mtext>H</mtext>}}\text{= 59.5 ± 2}\text{A}\text{?}$ for G_D1a. The mixture 3:1 of the two gangliosides gave intermediate values for all examined parameters. The presence of cations, within the physiological concentration range. and, in particular of Ca²⁺, did not influence significantly the values of c_o and the main features of the micelle.     

Effects of lighting programs on onset of the ovulatory season in mares

Using 240 pony mares, lighting regimens were tested for their efficiency in hastening the onset of the ovulatory season. The mean number of days from January 1 to first ovulation was used as the end point. No advantage was gained by beginning a fixed lighting regimen ($\text{15.5L}\text{8.5D}$, hours light/hours dark) November 1 (66 ±8) versus December 1 (65 ±9), but beginning on January 1 was less efficient (98 ±8; controls, 132 ±5; P<0.05). In another experiment, daily three-hour interruptions of either the light phase (67 ±10) or the dark phase (71 ±11) did not significantly retard the effectiveness of a fixed regimen of $\text{15L}\text{9D}$ (54 ±5; controls, 142 ±6). A $\text{15L}\text{9D}$ regimen every other day (natural day length on alternate days) resulted in an interval (85 ±7) that was shorter (P<0.05) than for the controls and longer (not significant) than for the daily $\text{15L}\text{9D}$ regimen. When used with natural day length, a one-hour pulse of light in the evening (15 hours after sunrise) was not effective (141 ±6); a one-hour pulse in the morning 9.5 hours after sunset) was only partially effective (117 ±6). In another experiment, the interval was reduced (P<0.05) in a group with one hour of light fixed at 4:00 a.m. with natural day length (85 ±8; $\text{15L}\text{9D}$, 75 ±7; controls, 126 ±9). Results indicated that a fixed one-hour pulse of light at 4 a.m., used with natural day length, may provide an acceptable level of stimulation.     

The linkage of teleost skin keratan sulfate to protein

The linkage of teleost skin keratan sulfate to protein was investigated. Afeter its exhaustive digestion with pronase, peptidokeratan sulfate was obtained with aspartic acid as the predominant amino acid. The N-terminal of the amino acid residues of the preparation was dansylated, and the carbohydrate-peptide linkage fragment was isolated, with the aid of fluorescence, by sequential digestion with Flavobacterium endo-β-galactosidase, β-galactosidase, β-N-acetylhexosaminidase and endo-β-N-acetylglucosaminidase D, followed by Bio-Gel p-4 column chromatography. The structure of the dansylated fragment thus obtained was identified dansylated asparaginyl $\text{N-}\text{acetyl-}\text{D}\text{-glucosamine}$. Treatment of the dansylated keratan sulfate peptide with almond glycopeptidase, which specially cleaves thet asparaginyl $\text{N-}\text{acetyl-}\text{D}\text{-glucosamine}$ linkage in the glycoproteins, also showed asparaginyl $\text{N-}\text{aceytyl-}\text{D}\text{-glucosamine}$ linkage to be in the core region of this keratan sulfate. We conclude that teleost skin keratan sulfate is bound to protein via an N-glycosyl linkage between $\text{N-}\text{acetyl-}\text{D}\text{-glucosamine}$ and asparagine. The keratan sulfate core apparently consist of trimannosyl-N,N′-diacetylchitobiose units, considering the specificity of endo-β-N-acetylglucosaminidase D.     

Structure of a bacteriochlorophyll-protein from the green photosynthetic bacterium Chlorobium limicola: crystallographic evidence for a trimer

Crystals of the bacteriochlorophyll-protein from the green photosynthetic bacterium Chlorobium limicola, previously described by J. M. Olson et al. (1969), have been re-examined by X-ray diffraction. The space group is P6₃, as reported in the earlier work, but revised cell dimensions, $\text{a = b = 112.4 ± 0.4}\text{A}\text{?}\text{, c = 98.4 ± 0.4}\text{A}\text{?}$, were obtained, leading to a unit cell volume one third of that reported previously. Correction of this error leads to the conclusion that the bacteriochlorophyll-protein complex must be a trimer consisting of three identical subunits arranged about a crystallographic symmetry axis. Also a new trigonal crystal form of the bacteriochlorophyll-protein has been obtained, and is consistent only with a molecule composed of three identical or near-identical subunits. Models of the molecular packing for both crystal forms are presented.The molecular weight of the bacteriochlorophyll-protein complex, determined from crystal density measurements, is (1.53 ± 0.23) × 10⁵, and the overall molecular dimensions are about 55 Å along the trimer axis, and 83 Å at right angles to this. There are probably seven bacteriochlorophyll molecules in each subunit.     

Crystallographic data of the selenoenzyme glutathione peroxidase

Glutathione peroxidase prepared from bovine erythrocytes yields small, but well-ordered plate-like crystals. X-ray investigation shows them to belong to monoclinic space group C2. Unit cell dimensions are: $\text{a = 90.4}\text{A}\text{?}\text{, b = 109.5}\text{A}\text{?}\text{, c = 58.6}\text{A}\text{?}$, β = 99 ° ± 15 min. The crystal density is ?_c = 1.36 ± 0.02 g.cm^?3. Consequently, the asymmetric unit of the crystal cell is occupied by one tetrameric molecule of M_r 84,000. Matthew's (1968) parameter Γ_M is calculated to be 1.71 ų/dalton.     

The content of water and potassium in fat cells

The distribution spaces at equilibrium for ³H₂O, [¹⁴C]urea and $\text{3-O-[}{}^{14}\text{C]-}\text{methylglucose}$ were measured in white fat cells using centrifugation through silicone oil at 2500 × g; no significant differences were observed. l-[¹⁴C] Glucose added immediately before the centrifugation was used as a marker for the extracellular water space. The calculated intracellular water content of the cells after the centrifugation through oil (e.g. ³H₂O space minus l-[¹⁴C] glucose space) is an unbiased measure of the water content of the cells in suspension as judged by the following criteria: (1) The intracellular distribution space for $\text{3-O-[}{}^{14}\text{C}\text{]}$methylglucose at equilibrium (methylglucose space minus l-glucose space) was not different from that calculated from a methylglucose wash-out curve. (2) The intracellular content of l-[¹⁴C]glucose (half time of efflux about 60 min) in cells preloaded during incubation of the tissue with collagenase was not different in cells recovered by (a) centrifugation through oil at 2500 × g, (b) centrifugation through oil at 600 × g, (c) centrifugation at 600 × g in the absence of oil and (d) filtration on Millipore filters.The intracellular content of water determined on cells from single rats weighing 120–150 g was 2.75 ± 0.55 μl/100 μl fat cells (± S.D., n = 30). The intracellular content of potassium, determined on cells from the same rats, was 252 ± 62 nmols/100 μl fat cells (± S.D., n = 30). The concentration of potassium in the intracellular water was calculated as 104 ± 15 mM (± S.D., n = 30).     

Unit cell transormations in yeast phenylalanine transfer RNA crystals

The orthorhombic unit cell of crystalline yeast phenylalanine transfer RNA has dimensions $\text{a = 33}\text{A}\text{?}$, $\text{b = 56}\text{A}\text{?}$ and $\text{c = 161}\text{A}\text{?}$. When the mother liquor dries partially, a series of transformations takes place in which the a and b axes change very little but the c axis decreases abruptly first to 128 Å and then to 109 Å. In a closely related orthorhombic cell in a different space group the c axis is 104 Å. Although there is some loss in resolution in these smaller unit cells, the over-all distribution of scattering intensity does not change substantially. This suggests that the tRNA molecules can slide together along the c axis without a substantial change in internal structure.