Kringles: modules specialized for protein binding. Homology of the gelatin-binding region of fibronectin with the kringle structures of proteases

The collagen-binding site of type-II units of bovine seminal fluid protein PDC-109 and fibronectin

A single type-II domain has been isolated by limited proteolysis of the collagen-binding bovine seminal fluid protein, PDC-109. The 45-residue fragment corresponding to the second type-II domain of the parent molecule was found to have retained affinity for immobilized collagen, indicating that this minidomain carries critical regions of the collagen-binding site. Studies on various fragments of fibronectin have also implicated the two type-II units of this molecule in collagen-binding. In the present work we have found that type-II domains of human fibronectin, expressed in Escherichia coli as beta-galactosidase fusion proteins, bind specifically to immobilized collagen.

Characterization of the cDNA coding for mouse plasminogen and localization of the gene to mouse chromosome 17

A full-length cDNA coding for mouse plasminogen has been isolated and characterized. The cDNA is 2720 bp in length (excluding the poly(A) tail) and contains a 24-bp 5' noncoding region, an open reading frame of 2436 bp, and a 3' noncoding region of 257 bp. The open reading frame codes for 812 amino acids and includes a signal peptide that is likely 19 amino acids in length and the mature protein of 793 amino acids. The calculated Mr of mouse plasminogen is 88,706 excluding carbohydrate. There are two potential N-linked carbohydrate addition sites; one of which is glycosylated in human, bovine, and porcine plasminogens. Mouse plasminogen was found to contain two additional amino acids compared to the human protein. In addition, mouse and human plasminogens were found to be 79 and 76% identical at the protein and DNA levels, respectively. Analysis of the segregation of two allelic forms, Plgb and Plgd, of plasminogen DNA in three sets of recombinant inbred strains has allowed the localization of the mouse plasminogen gene to the proximal end of mouse chromosome 17 within the t complex and close to the locus D17Rp17. The Plg gene is deleted in the semidominant deletion mutant, hair-pintail (Thp).

Characterization of the cDNA coding for mouse prothrombin and localization of the gene on mouse chromosome 2

A series of overlapping cDNAs coding for mouse prothrombin (coagulation factor II) have been isolated and the composite DNA sequence has been determined. The complete prothrombin cDNA is 1,987 bp in length [excluding the poly(A) tail] and codes for 18 bp of 5' untranslated sequence, an open reading frame coding for 618 amino acids, a stop codon, and a 3' untranslated region of 112 bp followed by a poly(A) tail. The translated amino acid sequence predicts a molecular weight of 66,087, which includes 10 residues of gamma-carboxyglutamic acid. There are five potential N-linked glycosylation sites. Mouse prothrombin is 81.4% and 77.3% identical to the human and bovine proteins, respectively. Comparison of the cDNA coding for mouse prothrombin to the human and bovine cDNAs indicates 79.9% and 76.5% identity, respectively. Amino acid residues important for the structure and function of human prothrombin are conserved in the mouse and bovine proteins. In the adult mouse and rat, prothrombin is primarily synthesized in the liver, where is constitutes 0.07% of total mRNA as determined by solution hybridization analysis. The genetic locus for mouse prothrombin, Cf-2, has been mapped using an interspecies backcross and DNA fragment differences between the two species. The prothrombin locus lies on mouse chromosome 2, 1.8 +/- 1.3 map units proximal to the catalase locus. The gene order in this region is Cen-Acra-Cf-2-Cas-1-A-Tel. This localization extends the proximal boundary of the known region of homology between mouse chromosome 2 and human chromosome 11p from Cas-1 about 2 map units toward the centromere.

Solution structure of the kringle 4 domain from human plasminogen by 1H nuclear magnetic resonance spectroscopy and distance geometry

Kringle 4 is an autonomous structural and folding domain within the proenzyme plasminogen. Homologous domains are found throughout the blood clotting and fibrinolytic proteins. In this paper, we present the almost complete assignment of the 1H nuclear magnetic resonance (n.m.r.) spectrum of the kringle 4 domain of human plasminogen. A detailed structural analysis has been completed. The sequential pattern of nuclear Overhauser enhancements indicated little regular secondary structure but rather a series of turns and loops connecting beta-strands. A small stretch of antiparallel beta-sheet was identified between the residues 61 to 63 and 71 to 73 and the close proximity of other strands was determined from two-dimensional nuclear Overhauser enhancement spectra. Slowly exchanging amide (NH) resonances were found to be associated with residues of the beta-sheet and neighbouring strands that support the hydrophobic core of the domain. A total of 526 interproton distance constraints and two hydrogen bonds were specified as input to the distance geometry program DISGEO. Tertiary structures were produced that were consistent with the n.m.r. data. The structures were compared with that of our earlier model based on n.m.r. studies and with that of prothrombin fragment 1 determined crystallographically.

Structure and chromosomal localization of the human thrombospondin gene

Thrombospondin (THBS1) is a large modular glycoprotein component of the extracellular matrix and contains a variety of distinct domains, including three repeating subunits (types I, II, and III) that share homology to an assortment of other proteins. Determination of THBS1 gene structure has revealed that the type I repeat modules are encoded by symmetrical exons and that the heparin-binding domain is encoded by a single exon. To further elucidate the higher level organization of THBS1, the gene was localized to the q11-qter region of chromosome 15.

Co-operative domains in fibronectin

The melting of human plasma fibronectin and its proteolytic fragments has been studied by scanning microcalorimetry to reveal co-operative structural domains in the molecule. It has been established that each of the two similar polypeptide chains of fibronectin has at least 12 structural domains, which differ in stability, size and function. Many of the domains in the N-terminal half of the polypeptide chains appear to be composed of two homologous repeat modules that co-operate to form a single co-operative unit. In the intact fibronectin molecule, the C-terminal regions of both chains seem to interact forming a stable co-operative block.

Stimulation by fibrinogen of the amidolytic activity of single-chain tissue plasminogen activator

The amidolytic steady-state kinetic properties of a series of recombinant tissue plasminogen activators (rt-PA) have been examined in the presence and absence of the positive effectors fibrinogen (Fg) and native soluble (des-A)-fibrin (sFn). Two-chain (tc) native rt-PA displayed a Km value of 0.50 mM and a kcat value of 13.2 s-1 toward the substrate, H-D-Ile-Pro-Arg-p-nitroanilide (S2288) at 37 degrees. When these same assays were conducted in the presence of Fg or sFn, the Km and kcat values remained essentially the same. On the other hand, the activity of single-chain (sc) rt-PA was significantly increased in the presence of Fg or sFn, by approximately 3.4- to 4-fold, due to alterations in both the Km and kcat of the reaction. Similar results were obtained with rt-PA deletion variants, obtained by site-directed mutagenesis. With rt-PA domain-deletion derivatives, consisting of kringles 1 (K1)-2 (K2)-protease (P), and K2-P, the amidolytic activities of the scrt-PA preparations were significantly stimulated (2.0- to 2.5-fold) by Fg and sFn, a property not shared by the corresponding tcrt-PA. On the other hand, neither the single- nor two-chain derivatives of a deletion mutant containing only the finger (F)-growth factor (E)-P domains displayed stimulation by Fg or sFn, results suggestive of the importance of the K2 region in the observed Fg- and Fn-induced stimulations rt-PA amidolytic activity. With one strategically important derivative, a molecule containing the amino acid replacement, Cys264----Gly [(Cys264----Gly)-rt-PA], a change resulting in the loss of covalent attachment of the heavy and light chains of tcrt-PA, the amidolytic activities of neither the single-chain nor the two-chain form of the molecule were stimulated by the presence of the above two positive effectors. With the single-chain form of this same derivative, the kcat of the reaction was extremely low (1.5 s-1), but increased to approximately 50.5 s-1 for the two-chain form, this latter value being nearly 4-fold higher than that of any of the wild-type recombinant rt-PA preparations examined. This suggests that the latent heavy chain of rt-PA inhibits the amidolytic activity found in the trypsin-like P domain.(ABSTRACT TRUNCATED AT 250 WORDS)

A gene homologous to plasminogen located on human chromosome 2q11-p11

A plasminogen probe encoding kringles 1-3 detects homologous loci on human chromosomes 6 and 2 by somatic cell hybrid analysis. Regional localization by in situ hybridization places the loci at 6q26-27 and 2p11-q11. Further analysis by varying washing stringencies of hybridization filters reveals a greater homology with the chromosome 6 locus than with the chromosome 2 locus. These results confirm localization of the plasminogen gene to human chromosome 6q26-27 and indicate that a homologous sequence of unknown identity resides on chromosome 2p11-q11.

A differential scanning calorimetric investigation of the domains of recombinant tissue plasminogen activator

The differential scanning calorimetric (DSC) properties of a series of recombinant tissue plasminogen activators (rt-PA) have been examined. The endotherm obtained for native rt-PA can be deconvoluted into a pair of two-state transitions, which indicates that two separate observable regions of the molecule undergo independent melting. A distinguishing feature of native rt-PA is the dependence of the temperature of the maximum heat capacity of the endothermic transitions (Tm) on the thermal scan rate of the samples, suggesting that a kinetic process, involving conversion of a reversibly denatured to an irreversibly denatured form of the protein, with an energy of activation of approximately 81.5 kcal/mol, characterizes its thermal denaturation. A comparison of the conformational properties of rt-PA preparations which have been produced in different expression systems, as revealed by DSC analysis of their thermal denaturation characteristics, has demonstrated that subtle differences do occur. Similar studies with deletion mutants of rt-PA suggest that the growth factor domain (EGF) plays a role in its overall thermal stability, when the protein also contains the kringle 1 region, and removal of the EGF domain leads to conformational alterations in other areas of the molecule.

Cloning and characterization of a cDNA for rat tissue-type plasminogen activator

Two partly overlapping lambda gt11 cDNA clones coding for the 22S rat tissue-type plasminogen activator (t-PA) mRNA were isolated. The cDNA sequences cover 2445 nucleotides of the mRNA, including a 5' untranslated region of 31 nucleotides, an open reading frame of 1677 nucleotides, a 3' untranslated region of 737 nucleotides, and a poly(A) tail. The open reading frame codes for a 17-amino-acid signal peptide, a propeptide with 12 amino acids, and the mature protein with 530 amino acids. Rat t-PA has 81% and 92% amino acid sequence identity with the human and mouse counterparts and an equal distribution of conserved amino acids, suggesting that the proteins can fold into identical three-dimensional structures. The rat t-PA sequence contains two putative N-glycosylation sites at Asn-120 and Asn-452, while human t-PA has an additional glycosylation site at Asn-187. The site at Asn-187 is glycosylated in the human protein, revealing a different glycosylation pattern between the human and rat proteins.

Detecting distant homologies of mosaic proteins. Analysis of the sequences of thrombomodulin, thrombospondin complement components C9, C8 alpha and C8 beta, vitronectin and plasma cell membrane glycoprotein PC-1

Recognition of homologies may give hints about the structure and function of proteins; therefore, we are developing strategies to aid sequence comparisons. Detecting homology of mosaic proteins is especially difficult since the modules constituting these proteins are usually distantly related and their homology is not readily recognized by conventional computer programs. In the present work we show that the rules of the evolution of mosaic proteins can guide the identification of modules of mosaic proteins and can delineate the group of sequences in which the presence of homologous sequences may be expected. By this approach we can concentrate the search for homology to a limited group of sequences; thus ensuring a more intense and more fruitful search. The power of this approach is illustrated by the fact that it could detect homologies not identified by earlier methods of sequence comparison. In this paper we show that thrombomodulin contains a domain homologous with animal lectins, that complement components C9, C8 alpha and C8 beta have modules homologous with one of the repeat units of thrombospondin and that the somatomedin B module of vitronectin is homologous with the internal repeats of plasma cell membrane glycoprotein PC-1.

Analysis of the aromatic 1H NMR spectrum of chicken plasminogen kringle 4

The intact kringle 4 domain of chicken plasminogen has been characterized by 1H NMR spectroscopy at 300 and 620 MHz in both the presence and absence of epsilon-aminocaproic acid, an antifibrinolytic drug. The study focuses on the aromatic resonances. Comparisons with spectra from human, porcine and bovine kringle 4 homologs indicates a strict conservancy of conformation, reflecting the underlying primary sequence homology, and leads to an unambiguous assignment of all the aromatic resonances, including those of Phe15 and His40 which are unique to the chicken domain. Conclusive evidence is found that the Tyr9 ring fluctuates between two states, one in which it flips fast and other in which it is severely hindered. Similarly, the Tyr64 side chain finds itself in a structurally constrained locus. The Trp62, Tyr64, and Trp72 aromatic resonances are most sensitive to ligand presence, supporting a previously reported model of the kringle 4 lysine-binding site. His40, Phe41, and Tyr74 are also perturbed by ligand indicating proximity to the site. In contrast, the Phe15 aromatic spectrum indicates a rather mobile phenyl ring which is insensitive to ligand presence, thus confirming the lesser importance of the corresponding segment within the first kringle loop in determining kringle structure and/or function.

Two-dimensional 1H-NMR studies of the solution structure of plasminogen kringle 4

Native kringle 4 from human plasminogen has been studied by two-dimensional 1H-NMR methods in order to obtain new structural information about the kringle fold. Two-dimensional scalar correlated spectroscopy (COSY), two-dimensional dipolar correlated spectroscopy (NOESY) and two-dimensional relayed coherance transfer spectroscopy (RCT) experiments were recorded, allowing most resonances arising from the aromatic and methyl-containing residues to be assigned in the spectrum. From an analysis of NOE data, a small segment of double-stranded beta-sheet has been identified near residues Phe63 and Thr64. Further analysis of the NOESY spectrum has allowed detailed study of the conformation of sidechains located in regions near Leu45 and Val69. A model has been constructed of the polypeptide segment comprising residues 40-49 which accounts for the observed NOE interactions.

Proton magnetic resonance study of lysine-binding to the kringle 4 domain of human plasminogen. The structure of the binding site

The binding of L-Lys, D-Lys and epsilon-aminocaproic acid (epsilon ACA) to the kringle 4 domain of human plasminogen has been investigated via one and two-dimensional 1H-nuclear magnetic resonance spectroscopy at 300 and 600 MHz. Ligand-kringle association constants (Ka) were determined assuming single site binding. At 295 K, pH 7.2, D-Lys binds to kringle 4 much more weakly (Ka = 1.2 mM-1) than does L-Lys (Ka = 24.4 mM-1). L-Lys binding to kringle 4 causes the appearance of ring current-shifted high-field resonances within the -1 approximately less than delta approximately less than 0 parts per million range. The ligand origin of these signals has been confirmed by examining the spectra of kringle 4 titrated with deuterated L-Lys. A systematic analysis of ligand-induced shifts on the aromatic resonances of kringle 4 has been carried out on the basis of 300 MHz two-dimensional chemical shift correlated (COSY) and double quantum correlated spectroscopies. Significant differences in the effect of L-Lys and D-Lys binding to kringle 4 have been observed in the aromatic COSY spectrum. In particular, the His31 H4 and Trp72 H2 singlets and the Phe64 multiplets appear to be the most sensitive to the particular enantiomers, indicating that these residues are in proximity to the ligand C alpha center. In contrast, the rest of the indole spectrum of Trp72 and the aromatic resonances of Trp62 and Tyr74, which are affected by ligand presence, are insensitive to the optical nature of the ligand isomer. These results, together with two-dimensional proton Overhauser studies and ligand-kringle saturation transfer experiments reported previously, enabled us to generate a model of the kringle 4 ligand-binding site from the crystallographic co-ordinates of the prothrombin kringle 1. The latter, although lacking recognizable lysine-binding capability, is otherwise structurally homologous to the plasminogen kringles.

The genetic relationships between the kringle domains of human plasminogen, prothrombin, tissue plasminogen activator, urokinase, and coagulation factor XII

A computer-based statistical evaluation of the optimal alignments of the kringle domains of human plasminogen, human prothrombin, human tissue plasminogen activator, human urokinase, and human coagulation Factor XIIa, as well as the putative kringle of human haptoglobin, has been performed. A variety of different alignments has been examined and scores calculated in terms of the number of standard deviations (SD) of a given match from randomness. With the exception of human haptoglobin, it was found that very high alignment scores (8.9-23.0 SD from randomness) were obtained between each of the kringles, with the kringle 1 and kringle 5 regions of human plasminogen displaying the highest similarity, and the S kringle of human prothrombin and the human Factor XII kringle showing the least similarity. The relationships obtained were employed to construct an evolutionary tree for the kringles. The predicted alignments have also allowed nucleotide mutations in these regions to be evaluated more accurately. For those regions for which nucleotide sequences are known, we have employed the maximal alignments from the protein sequences to assess nucleotide sequence similarities. It was found that a range of approximately 40-55% of the nucleotide bases were placed at identical positions in the kringles, with the highest number found in the alignment of the two kringles of human tissue plasminogen activator and the lowest number in the alignment of the S kringle of prothrombin with the second kringle of tissue plasminogen activator. From both protein and nucleotide alignments, we conclude that haptoglobin is not statistically homologous to any other kringle. Secondary structural comparisons of the kringle regions have been predicted by a combination of the Burgess and Chou-Fasman methods. In general, the kringles display a very high number of beta-turns, and very low alpha-helical contents. From analysis of the predicted structures in relationship to the functional properties of these domains, it appears as though many of their functional differences can be related to possible conformational alterations resulting from amino acid substitutions in the kringles.

Fibrinogen-sepharose interaction with prothrombin, prethrombin 1, prethrombin 2 and thrombin

Binding of prothrombin, prethrombin 1, prethrombin 2 and thrombin to fibrinogen-Sepharose was studied. Thrombin and prethrombin 2 bound to fibrinogen-Sepharose, while prethrombin 1 and prothrombin did not. Bound thrombin and prethrombin 2 were recovered from the column by eluting with 0.1 M NaCl/0.05 M Tris-HCl buffer (pH 7.4). The affinity of thrombin and prethrombin 2 to fibrinogen-Sepharose depended on ionic strength and reached a maximum at 50 mm concentration. Prethrombin 2 interacts with fibrinogen as well as thrombin; and prothrombin fragment 1.2 is not important in the formation of this complex. Thus, prethrombin 2, which is a precursor of thrombin without measurable enzymatic activity and which lacks the single cleavage at Arg-322-Ile-323 present in thrombin, has the same or very similar structural conformation as thrombin and has the same macromolecular substrate recognition site. These results confirm the earlier results that active center is not necessary in fibrinogen-thrombin interaction.

Intron-dependent evolution: preferred types of exons and introns

Exon insertions and exon duplications, two major mechanisms of exon shuffling, are shown to involve modules that have introns of the same phase class at both their 5'- and 3'-ends. At the sites of intronic recombinations exon insertions and duplications create new introns which belong to the same phase class as the recipient introns. As a consequence of repeated exon insertions and exon duplications introns of a single phase class predominate in the resulting genes, i.e. gene assembly by exon shuffling is reflected both by this nonrandom intron phase usage and by the correlation between the domain organization of the proteins and exon-intron organization of their genes. Genes that appeared before the eukaryote-prokaryote split do not show these diagnostic signs of exon shuffling. Since ancestral introns (e.g. self-splicing introns) did not favour intronic recombination, exon shuffling may not have been significant in the early part of protein evolution.

Molecular cloning and characterization of a full-length cDNA clone for human plasminogen

A human liver cDNA library enriched for full-length clones was screened for plasminogen cDNA using a synthetic 24-nucleotide probe derived from a reported partial cDNA sequence. 12 positive clones were identified and one of these was characterized in detail. The 2.7 kb insert contains the complete coding region. At 5 positions, it gives residues different from those reported in a previous amino acid sequence analysis of the protein. The present results show an extra Ile at position 65, Gln instead of Glu at positions 53 and 342, Asn at position 88 instead of Asp, and Asp at position 453 rather than Asn. In the 3'-non-coding region an extension of 29 bases is found which does not contain any structure compatible with a known polyadenylation signal. Instead, the consensus signal AATAAA is placed at a distance of 46 bases upstream of the poly(A)-tail.

Proteins, exons and molecular evolution

The discovery of the eukaryotic gene structure has prompted research into the potential relationship between protein structure and function and the corresponding exon/intron patterns. The exon shuffling hypothesis put forward by Gilbert and Blake suggests the encodement of structural and functional protein elements by exons which can recombine to create novel proteins. This provides an explanation for the relatively rapid evolution of proteins from a few primordial molecules. As the number of gene and protein structures increases, evidence of exon shuffling is becoming more apparent and examples are presented both from modern multi-domain proteins and ancient proteins. Recent work into the chemical properties and catalytic functions of RNA have led to hypotheses based upon the early existence of RNA. These theories suggest that the split gene structure originated in the primordial soup as a result of random RNA synthesis. Stable regions of RNA, or exons, were utilised as primitive enzymes. In response to selective pressures for information storage, the activity was directly transferred from the RNA enzymes or ribozymes, to proteins. These short polypeptides fused together to create larger proteins with a wide range of functions. Recent research into RNA processing and exon size, discussed in this review, provides a clearer insight into the evolutionary development of the gene and protein structure.

Warfarin and the biochemistry of the vitamin K dependent proteins

We have reviewed the biochemistry of the normal and warfarinized vitamin K-dependent coagulation proteins, both individually and as members of their respective macromolecular enzyme complexes. Much has been learned in the past 10-15 years about the molecular mechanisms of coagulation. However, we have only scratched the surface and are beset by a number of interesting challenges. Clearly, we must begin to dissect the other vitamin K-dependent macromolecular enzyme complexes as has been done for prothrombinase. Prothrombin differs in fundamental ways from the other vitamin K-dependent proteins and we can anticipate significant differences as well as similarities in the other complexes. The presence of potentially functional domains in the non-catalytic portions of the vitamin K-dependent proteins and their similarities within this family of proteins should prove a fruitful area of future research. Finally, the evaluation of warfarinized vitamin K-dependent proteins, both in fundamental studies with purified systems and using uniquely engineered immunoassays, should shed light on basic molecular mechanisms and modes of monitoring warfarin therapy.

The prothrombin gene and serine proteinase evolution

The gene coding for human prothrombin has been isolated from two human genomic DNA libraries using a human prothrombin cDNA. Present evidence indicates that the gene is approximately 24 kb in length with about 90% of the DNA representing intervening sequence. Thirteen intervening sequences were found to interrupt the region coding for the mRNA into 14 exons. These intervening sequences vary greatly in size and contain at least 11 copies of Alu repetitive DNA. The positions where several of the intervening sequences interrupt the coding region appear to separate functional and structural domains of the protein. A similar placement of intervening sequences in genes coding for proteins homologous to prothrombin has been observed and provides additional evidence that these proteins have evolved from a common ancestor.

1H-NMR spectroscopic manifestations of ligand binding to the kringle 4 domain of human plasminogen

Structural aspects of the binding of the linear ligands N alpha-acetyl-L-lysine (AcLys) and epsilon-aminocaproic acid (epsilon ACA) and of the cyclic analogs trans-(aminomethyl)-cyclohexanecarboxylic acid (AMCHA) and p-benzylaminesulfonic acid (BASA) to the intact plasminogen kringle 4 domain have been investigated by 1H-NMR spectroscopy at 300 and 600 MHz. Ligand binding results in consistent shifts of the His-II (His31), Trp-I (Trp25?), Trp-II (Trp62?), Trp-III (Trp72), Tyr-II (Tyr50), and Phe64 ring signals. BASA tends to induce larger shifts than elicited by the aliphatic ligands, most noticeably on Trp-II and on Trp72, suggesting that the ligand aromatic ring interacts with the two indole groups. Trp-II and, to lesser extent, Trp-I interact with an acidic side chain group, in a manner that is blocked by BASA. BASA binding also perturbs Tyr-II (Tyr50), Tyr-III (Tyr41), and Tyr-IV (Tyr74) over a wide pH range and lowers the pKa* of His31 from approximately 4.8 to approximately 4.6. His-III (His33) responds to BASA and AMCHA but is relatively insensitive to the linear ligands. His33 carries a sterically shielded side chain which, in conjunction with Leu46, Trp-I, Tyr50, and Tyr74, participates in structuring the kringle hydrophobic core, contiguous to the binding site. Pronounced shifts are observed for aliphatic resonances stemming from the kringle-bound molecules of AMCHA, AcLys, and epsilon ACA. It is proposed that the lysine-binding site is mostly supported by the loop that extends from Cys51 through Cys71 and that aromatic residues, which include Trp-II, Trp72, and Phe64, play a major role in interacting with the nonpolar segment of the ligand molecule. The binding site also encompasses Tyr50, Tyr74, His31, and His33 although it is not clear the extent to which these residues interact directly with the ligand.

The kringle 4 domain of chicken plasminogen

The kringle 4 unit of chicken plasminogen is similar to mammalian kringle 4 domains in possessing a lysine-binding site. Chicken kringle 4 shows 73-77% sequence identity with the fourth kringle units of porcine, bovine and human plasminogens. A major difference between mammalian and chicken kringle 4 species is that in the latter a glucosamine-based carbohydrate substituent is linked to asparagine-34. Complexation of this carbohydrate with concanavalin A does not interfere with the binding of kringle 4 to lysine-Sepharose, suggesting that in the kringle-fold the glycosylated region is distant from the entrance of the lysine-binding pocket.

Chemical modification and nuclear magnetic resonance studies on human plasminogen kringle 4. Assignment of tyrosine and histidine resonances to specific residues in the sequence

Modification of kringle 4 with tetranitromethane leads to the selective nitration of tyrosine 40 but on prolonged incubation with reagent, reaction of tyrosine 49 is also observed. Nitration of tyrosines 40 and 49 had no influence on the lysine-Sepharose affinity of kringle 4, indicating that these residues are not important for the functional integrity of the ligand-binding site. Comparison of the NMR spectra of native kringle 4 with those of kringle 4 in which tyrosine 40 or tyrosines 40 and 49 are nitrated permitted the identification of the resonances of these residues. These NMR studies also showed that the chemical modifications caused little perturbation of the three-dimensional structure of the protein. Cross-linking of lysine 35 and tyrosine 40 with 1,3-difluoro-4,6-dinitrobenzene demonstrates that in the kringle-fold the reactive epsilon-amino and phenolic groups of these residues can approach each other to a distance of 0.5 nm. NMR spectra of this kringle 4 species also confirmed the assignment of the resonances to tyrosine 40. NMR spectra of a kringle 4 derivative in which the disulphide bridge between cysteines 1 and 79 has been broken by selective reduction and alkylation showed that the core structure of the kringle-fold and the lysine-binding site are unaltered by this modification. This observation is in agreement with earlier results which showed that the lysine-Sepharose affinity of kringle 4 is not affected by reduction and alkylation of this disulphide bridge. Comparison of the NMR spectra of native and disulphide-cleaved kringle 4 aided in the assignment of resonances to residues adjacent to the site of modification (tyrosine 2 and histidine 3) and permitted the tentative assignment of the resonances of tyrosines 9 and 73.

Macro- and micro-stabilities of the kringle 4 domain from plasminogen. The effect of ligand binding

1H-NMR spectra of kringle 4 from human plasminogen have been recorded over wide pH* and temperature ranges, both in the presence and in the absence of p-benzylaminesulfonic acid (BASA). Several resonances exhibit chemical shift differences between kringle folded and unfolded forms which are sufficiently well resolved to allow for a determination of equilibrium Van't Hoff enthalpies and entropies for unfolding. The interaction with BASA shifts the kringle unfolding temperature from approximately 335 degrees K to approximately 343 degrees K. The pH* range of stability is also wider for the complex than for the free kringle: in the acidic range the pH* of half-unfolding, pHu*, is decreased from 2.8 for the unligated polypeptide to approximately 2.0 in the presence of BASA, while in the basic range pHu*, shifts from approximately 10.8 to 11.5. However, in contrast with what is observed at acidic pH*, unfolding at basic pH* leads to irreversible denaturation and exhibits a sharp, order-disorder transition both in the presence and in the absence of ligand. The structural stabilization conferred by the ligand is accompanied by a drastic reduction of the average rate of 1H-2H exchange in 2H2O under conditions that preclude a major cooperative unfolding. Thus, macro- and micro-stabilities of kringle domains appear to be highly correlated.

Proton Overhauser experiments on kringle 4 from human plasminogen. Implications for the structure of the kringles' hydrophobic core

1H-NMR Overhauser experiments at 300 and 600 MHz have been implemented on the isolated kringle 4 fragment of human plasminogen. This study shows that Leu46 and Leu77 CH3 delta,delta' groups, as well as two threonine CH3 gamma and a methionine S-CH epsilon (probably Met48) groups, are in efficient dipolar contact with histidine and aromatic side-chains. In particular, the experiments reveal that of the two Leu46 CH3 delta,delta' groups, one is in efficient contact with tryptophan (Trp25 and Trp62) indole rings while the other interacts with a tyrosine (probably Tyr41) phenol. Leu46 appears also to be close to an Ala CH3 beta group. Such a hydrophobic cluster appears to be contiguous to Trp72, hence to Arg71, residues that are through to be part of the lysine-binding site. Acid-base titration experiments show that the buried methionine S-CH3 epsilon group senses a neighboring ionizable group of pK*1 = 3.76, suggesting presence of a carboxyl anionic group (probably an aspartic acid side-chain) in the vicinity of the hydrophobic core. A preliminary model is proposed for the overall folding of the kringle polypeptide chain.