Analysis of xysA, a gene from Streptomyces halstedii JM8 that encodes a 45-kilodalton modular xylanase, Xys1.

The gene xysA from Streptomyces halstedii JM8 encodes a protein of 461 amino acids (Xys1) which is secreted into the culture supernatant as a protein of 45 kDa (Xys1L). Later, this form is proteolytically processed after residue D-362 to produce the protein Xys1S, which conserves the same xylanolytic activity. The cleavage removes a domain of 99 amino acids that shows similarity to bacterial cellulose binding domains and that allows the protein Xys1L to bind to crystalline cellulose (Avicel). Expression of this monocistronic gene is affected by the carbon source present in the culture medium, xylan being the best inducer. By using an anti-Xys1L serum, we have been able to detect xylanases similar in size to Xys1L and Xys1S in most of the different Streptomyces species analyzed, suggesting the ubiquity of these types of xylanases and their processing mechanism.     

<Emphasis Type="Italic">Streptomyces lividans</Emphasis> and <Emphasis Type="Italic">Brevibacterium lactofermentum</Emphasis> as heterologous hosts for the production of X<Subscript>22</Subscript> xylanase from <Emphasis Type="Italic">Aspergillus nidulans</Emphasis>

The Aspergillus nidulans gene xlnA coding for the fungal xylanase X₂₂ has been cloned and expressed in two heterologous bacterial hosts: Streptomyces lividans and Brevibacterium lactofermentum. Streptomyces strains yielded 10 units/ml of xylanase when the protein was produced with its own signal peptide, and 19 units/ml when its signal peptide was replaced by the one for xylanase Xys1 from Streptomyces halstedii. B. lactofermentum was also able to produce xylanase X₂₂, affording 6 units/ml upon using either the Aspergillus xlnA signal peptide or Streptomyces xysA. These production values are higher than those previously reported for the heterologous expression of the A. nidulans xlnA gene in Saccharomyces cerevisiae (1 unit/ml). Moreover, the X₂₂ enzyme produced by Streptomyces lividans showed oenological properties, indicating that this Streptomyces recombinant strain is a good candidate for the production of this enzyme at the industrial scale.     

Recombinant human lecithin‐cholesterol acyltransferase Fc fusion: Analysis of N‐ and O‐linked glycans and identification and elimination of a xylose‐based O‐linked tetrasaccharide core in the linker region

Recombinant human lecithin‐cholesterol acyltransferase Fc fusion (huLCAT‐Fc) is a chimeric protein produced by fusing human Fc to the C‐terminus of the human enzyme via a linker sequence. The huLCAT‐Fc homodimer contains five N‐linked glycosylation sites per monomer. The heterogeneity and site‐specific distribution of the various glycans were examined using enzymatic digestion and LC‐MS/MS, followed by automatic processing. Almost all of the N‐linked glycans in human LCAT are fucosylated and sialylated. The predominant LCAT N‐linked glycoforms are biantennary glycans, followed by triantennary sugars, whereas the level of tetraantennary glycans is much lower. Glycans at the Fc N‐linked site exclusively contain typical asialobiantennary structures. HuLCAT‐Fc was also confirmed to have mucin‐type glycans attached at T₄₀₇ and S₄₀₉. When LCAT‐Fc fusions were constructed using a G‐S‐G‐G‐G‐G linker, an unexpected +632 Da xylose‐based glycosaminoglycan (GAG) tetrasaccharide core of Xyl‐Gal‐Gal‐GlcA was attached to S₄₁₈. Several minor intermediate species including Xyl, Xyl‐Gal, Xyl‐Gal‐Gal, and a phosphorylated GAG core were also present. The mucin‐type O‐linked glycans can be effectively released by sialidase and O‐glycanase; however, the GAG could only be removed and localized using chemical alkaline β‐elimination and targeted LC‐MS/MS. E₄₁₆ (the C‐terminus of LCAT) combined with the linker sequence is likely serving as a substrate for peptide O‐xylosyltransferase. HuLCAT‐Fc shares some homology with the proposed consensus site near the linker sequence, in particular, the residues underlined PPP E₄₁₆GS₄₁₈G G G GDK. GAG incorporation can be eliminated through engineering by shifting the linker Ser residue downstream in the linker sequence.     

Heterologous expression of tomato glycoside hydrolase family 3 α‐L‐arabinofuranosidase/β‐xylosidases in tobacco suspension cultured cells and synergic action of a family 51 isozyme under antisense suppression of the enzyme

Four cDNA clones (SlArf/Xyl1‐4) encoding α‐l ‐arabinofuranosidase/β‐xylosidase belonging to glycoside hydrolase family 3 were obtained from tomato (Solanum lycopersicum) fruit. SlArf/Xyl1 was expressed in various organs. Its level was particularly high in flower and leaves but low in fruit. SlArf/Xyl3 was highly expressed in flower. On the contrary, SlArf/Xyl2 and 4 were expressed in early developmental stage in various organs. Comparison with SlArf/Xyl4, SlArf/Xyl2 expression was observed in earlier stages. The active recombinant proteins were obtained by using BY‐2 tobacco (Nicotiana tabacum) suspension cultured cells. The SlArf/Xyl1 and 2 recombinant proteins showed a bi‐functional activity of α‐l ‐arabinofuranosidase/β‐xylosidase while the SlArf/Xyl4 protein possessed a β‐xylosidase activity predominantly. Neither enzyme activities were detected for the SlArf/Xyl3 protein under the same conditions. Although SlArf/Xyl2 possessed a bi‐functional activity, it preferentially hydrolyzed arabinosyl residues from tomato hemicellulosic polysaccharides. Antisense suppression of SlArf/Xyl2 resulted in no apparent changes in the enzyme activities, monosaccharide composition or fruit phenotype. Increment of a family 51 α‐l ‐arabinofuranosidase expression rather than that of family 3 resulted in a restoring the activity in SlArf/Xyl2‐suppressed fruit. The ability of recombinant SlArf/Xyl2 to hydrolyze both arabinan and arabinoxylan is nearly identical to that of α‐l ‐arabinofuranosidases belonging to family 51. Our results suggested that BY‐2 cells are a useful expression system for obtaining active cell wall hydrolyzing enzymes. In addition, an α‐l ‐arabinofuranosidase activity derived from SlArf/Xyl2 would be essential in young organ development and the action of the enzyme could be restored by the other enzyme belonging to a different family under a defective condition.     

An aminoacylase activity from Streptomyces ambofaciens catalyzes the acylation of lysine on α‐position and peptides on N‐terminal position

The presence of aminoacylase activities was investigated in a crude extract of Streptomyces ambofaciens ATCC23877. First activities catalyzing the hydrolysis of N‐α or ε‐acetyl‐L‐lysine were identified. Furthermore, the acylation of lysine and different peptides was studied and compared with results obtained with lipase B of Candida antarctica (CALB). Different regioselectivities were demonstrated for the two classes of enzymes. CALB was able to catalyze acylation only on the ε‐position whereas the crude extract from S. ambofaciens possessed the rare ability to catalyze the N‐acylation on the α‐position of the lysine or of the amino‐acid in N‐terminal position of peptides. Two genes, SAM23877_1485 and SAM23877_1734, were identified in the genome of Streptomyces ambofaciens ATCC23877 whose products show similarities with the previously identified aminoacylases from Streptomyces mobaraensis. The proteins encoded by these two genes were responsible for the major aminoacylase hydrolytic activities. Furthermore, we show that the hydrolysis of N‐α‐acetyl‐L‐lysine could be attributed to the product of SAM23877_1734 gene.     

Crystal structures of the extracellular domains of the CrRLK1L receptor‐like kinases ANXUR1 and ANXUR2

Catharanthus roseus Receptor‐Like Kinase 1‐like (CrRLK1L) proteins contain two tandem malectin‐like modules in their extracellular domains (ECDs) and function in diverse signaling pathways in plants. Malectin is a carbohydrate‐binding protein in animals and recognizes a number of diglucosides; however, it remains unclear how the two malectin‐like domains in the CrRLK1L proteins sense the ligand molecule. In this study, we reveal the crystal structures of the ECDs of ANXUR1 and ANXUR2, two CrRLK1L members in Arabidopsis thaliana that have critical functions in controlling pollen tube rupture during the fertilization process. We show that the two malectin‐like domains in these proteins pack together to form a rigid architecture. Unlike animal malectin, these malectin‐like domains lack residues involved in binding to the diglucosides, suggesting that they have a distinct ligand‐binding mechanism. A cleft is observed between the two malectin‐like domains, which might function as a potential ligand‐binding pocket.     

Chromophore release from kraft pulp by purified streptomyces roseiscleroticus xylanases

Treating kraft pulps with the crude xylanase from Streptomyces roseiscleroticus followed by alkali extraction reduces the kappa number in a linear manner with enzyme doses up to about 3 IU/gm of oven-dry pulp. The enzyme complex consists of four isoenzymes designated Xyl1, Xyl2, Xyl3 and Xyl4. Each can release chromophores when used alone and each can facilitate alkali extraction to reduce the kappa number, but their relative abilities are different. Of the four isozymes, Xyl4 releases the least color and 237-nm-absorbing material whereas Xyl3 releases the most. Xyl4 best enhances the ability of alkali to reduce the kappa number. The UV absorption spectrum of the material released by alkali extraction differs significantly from the spectral characteristics of that released during enzyme treatment. The alkali-solubilized material has a maximum absorptivity at 265 nm and relatively little absorptivity at 237 nm. The material released during enzyme treatment absorbs strongly at 205 and 237 nm. UV/VIS spectroscopy of the enzyme- or alkali-released material does not show a characteristics lignin peak at 280 nm, nor does it reveal any notable peaks in the visible region. Analysis of the material released by enzyme treatment revealed more than 40 product peaks after fractionation by reversed-phase HPLC. We observed many products with strong UV absorption. These were relatively hydrophilic. Fewer products absorbed in the visible region. These were more hydrophobic. All four isoenzymes exhibit endo-action patterns; none forms xylose from oat-spelt xylan. The action patterns fell into two groups: endo-1 enzymes (Xyl1 and Xyl3) formed xylotriose (X₃) and other lower oligosaccharides as the predominant products; endo-2 enzymes (Xyl2 and Xyl4) formed roughly equimolar amounts of X₃, xylotetraose (X₄), and xylopentaose (X₅), and tended to leave larger amounts of undigested higher oligosaccharides.     

Mutagenesis of cellulase EGZ for studying the general protein secretory pathway in Erwinia chrysanthemi

Extracellular secretion of endoglucanase Z (EGZ) from Erwinia chrysanthemi is mediated by the so-called Out general secretion pathway and, presumably, involves recognition of EGZ-carried structural information by one or more of the Out proteins. Investigating the relationships between structure and secretability of EGZ was the purpose of the present work. EGZ is made of two independent domains, located at the N-and C-proximal sides, separated by a Ser/Thr-rich region, which are responsible for catalysis and cellulose-binding, respectively. The existence of a secretion region (‘targeting signal’) was investigated by studying the secretability of modified EGZ derivatives. These resulted from deletion or peptide insertion and were designed by using the domain organization cited above as a guide. Catalytic and/or cellulose-binding tests showed that all proteins exhibited at least a functional EGZ domain while immunoblot analyses confirmed that neither the insertions nor the deletions led to grossly misfolded proteins. In contrast, all of the proteins lost their secretability in E. chrysanthemi. This suggested that at least two secretion motifs existed, one lying within each functional domain. The role of the Ser/Thr-rich linker region was subsequently tested. Accordingly, two proteins containing a linker region whose length was increased by the addition of 8 and 18 additional residues and one protein lacking the linker region were studied. All three exhibited endoglucanase activity and cellulose-binding ability, confirming the independence of the domains within the context of EGZ/polysaccharide interaction. In contrast, none was secreted by E. chrysanthemi. Collectively, our results with EGZ (i) suggest the occurrence of multiple secretion-related sites either acting sequentially or forming a single three-dimensional secretion signal, (ii) show that secretability is not determined by either one of the two functional domains alone, and (iii) reveal that the linker region plays a role in secretion. We propose that all EGZ derivatives were impaired in the recognition step, the nature of which is discussed.     

Purification and characterization of two sugarcane bagasse-absorbable thermophilic xylanases from the mesophilic <Emphasis Type="Italic">Cellulomonas flavigena</Emphasis>

We report the purification and characterization of two thermophilic xylanases from the mesophilic bacteria Cellulomonas flavigena grown on sugarcane bagasse (SCB) as the only carbon source. Extracellular xylanase activity produced by C. flavigena was found both free in the culture supernatant and associated with residual SCB. To identify some of the molecules responsible for the xylanase activity in the substrate-bound fraction, residual SCB was treated with 3 M guanidine hydrochloride and then with 6 M urea. Further analysis of the eluted material led to the identification of two xylanases Xyl36 (36 kDa) and Xyl53 (53 kDa). The pI for Xyl36 was 5.0, while the pI for Xyl53 was 4.5. Xyl36 had a K _m value of 1.95 mg/ml, while Xyl53 had a K _m value of 0.78 mg/ml. In addition to SCB, Xyl36 and Xyl53 were also able to bind to insoluble oat spelt xylan and Avicel, as shown by substrate-binding assays. Xyl36 and Xyl53 showed optimal activity at pH 6.5, and at optimal temperature 65 and 55°C, respectively. Xyl36 and Xyl53 retained 24 and 35%, respectively, of their original activity after 8 h of incubation at their optimal temperature. As far as we know, this is the first study on the thermostability properties of purified xylanases from microorganisms belonging to the genus Cellulomonas.     

A new xylanase from Streptomyces megasporus DSM 41476 with high yield of xylobiose

A new xylanase gene, xynBM4, was cloned from Streptomyces megasporus DSM 41476 and expressed in Pichia pastoris. The full-length gene consists of 1,443 bp and encodes 480 amino acids including a putative 49-residue signal peptide. The deduced amino acid sequence of xynBM4 shows the highest identity of 66.3% to the xylanase Xys1L from Streptomyces halstedii JM8. The purified recombinant XYNBM4 had a high specific activity of 350.7 U mg^-1 towards soluble wheat arabinoxylan, exhibited optimal activity at pH 6.0 and 57°C, showed broad pH adaptability (>75% of the maximum activity at pH 2.5–9.0), was resistant to neutral proteases and most chemicals, and produced simple products. The hydrolysis products of birchwood xylan and corncob xylan were predominantly xylobiose (76.9 and 90.8%, respectively) and no xylose. These characteristics suggest that XYNBM4 has potential in various applications, especially in the food industry.     

The presence of multidomain linkers determines the bundle-shape structure of the phycobilisome of the cyanobacterium Gloeobacter violaceus PCC 7421

The complete genome sequence of Gloeobacter violaceus [Nakamura et al. (2003a, b) DNA Res 10:37–45, 181–201] allows us to understand better the structure of the phycobilisomes (PBS) of this cyanobacterium. Genomic analysis revealed peculiarities in these PBS: the presence of genes for two multidomain linker proteins, a core membrane linker with four repetitive sequences (REP domains), the absence of rod core linkers, two sets of phycocyanin (PC) α and β subunits, two copies of a rod PC associated linker (CpcC), and two rod cap associated linkers (CpcD). Also, there is one ferredoxin–NADP⁺ oxidoreductase with only two domains. The PBS proteins were investigated by gel electrophoresis, amino acid sequencing and peptide mass fingerprinting (PMF). The two unique multidomain linkers contain three REP domains with high similarity and these were found to be in tandem and were separated by dissimilar Arms. One of these, with a mass of 81 kDa, is found in heavy PBS fragments rich in PC. We propose that it links six PC hexamers in two parallel rows in the rods. The other unique linker has a mass of 91 kDa and is easily released from the heavy fragments of PBS. We propose that this links the rods to the core. The presence of these multidomain linkers could explain the bundle shaped rods of the PBS. The presence of 4 REP domains in the core membrane linker protein (129 kDa) was established by PMF. This core linker may hold together 16 AP trimers of the pentacylindrical core, or alternatively, a tetracylindrical core of the PBS of G. violaceus.     

Critical and optimal Ig domains for promotion of neurite outgrowth by L1/Ng‐CAM

Mammalian L1 and avian Ng‐CAM are homologous neural cell adhesion molecules (CAMs) that promote neurite outgrowth and cell adhesion in most neurons. Previous attempts to map these activities to discrete regions in the CAMs have suggested the involvement of a variety of different domains. However, these studies mainly used bacterially expressed proteins that were much less active on a molar basis than the native molecules. To define regions that are critical for maximal neurite outgrowth, we constructed and tested a panel of eukaryotically expressed proteins containing various extracellular segments of human L1 (hL1) or Ng‐CAM. Our results indicate that Ig domains 1–4 of hL1 are critical for homophilic binding and neurite outgrowth; however this segment is less potent than the entire extracellular region. Optimal neurite outgrowth activity was seen with proteins containing all six Ig domains of hL1 or Ng‐CAM. The adhesive properties of hL1 fragments correlated tightly with their neurite outgrowth activities, suggesting that these two processes are closely linked. These results suggest that Ig domains 1–4 form a structural cassette responsible for hL1 homophilic binding, while Ig domains 1–6 represent a functional region for optimal promotion of neurite outgrowth in vitro and possibly in vivo. © 2000 John Wiley & Sons, Inc. J Neurobiol 42: 287–302, 2000     

Cinnamic acid production using <Emphasis Type="Italic">Streptomyces lividans</Emphasis> expressing phenylalanine ammonia lyase

Cinnamic acid production was demonstrated using Streptomyces as a host. A gene encoding phenylalanine ammonia lyase (PAL) from Streptomyces maritimus was introduced into Streptomyces lividans, and its expression was confirmed by Western blot analysis. After 4 days cultivation using glucose as carbon source, the maximal level of cinnamic acid reached 210 mg/L. When glycerol (30 g/L) was used as carbon source, the maximal level of produced cinnamic acid reached 450 mg/L. In addition, using raw starch, xylose or xylan as carbon source, the maximal level of cinnamic acid reached 460, 300, and 130 mg/L, respectively. We demonstrated that S. lividans has great potential to produce cinnamic acid as well as other aromatic compounds.     

A xylanase with high pH stability from Streptomyces sp. S27 and its carbohydrate-binding module with/without linker-region-truncated versions

A xylanase gene, xynAS27, was isolated from a genomic library of Streptomyces sp. S27. The full-length gene consists of 1,434 bp and encodes 477 amino acids, including a putative signal peptide of 41 residues at its N-terminus. The mature xylanase comprises two functional domains, a family 10 glycoside hydrolase, and a family 13 carbohydrate-binding module (CBM), which were joined by a short Gly/Pro-rich linker region. The intact, the CBM-truncated and the CBM-linker-truncated versions of the mature proteins were expressed in Escherichia coli BL21 (DE3), purified to electrophoretic homogeneity and subsequently characterized. XynAS27 showed high pH stability over the pH range 2.2–12.0. XynAS27 may be a compelling tool for the food industry because it generates xylobiose (85% w/w) as the main product of xylan hydrolysis. The truncated versions showed less pH and thermal stability, and less affinity and hydrolytic activity to insoluble substrate than the intact one. These results indicate that the CBM of XynAS27 plays a key role in the hydrolysis of insoluble substrate, and the CBM and linker region are also important for the enzyme stability, and the linker region contributes more.     

粘虫颗粒体病毒增效蛋白基因片段优化及功能

【目的】研究转宿主粘虫颗粒体病毒(Pseudaletia unipuncta granulovirus,Pu GV-Ps)增效蛋白基因截短片段优化及其增效作用,探索增效蛋白基因的合理利用途径。【方法】生物信息学分析增效蛋白结构域,构建增效蛋白基因截短片段原核表达载体,分析目的基因片段表达产物的表达水平、围食膜蛋白降解效能和增强活性,进一步明确Pu GV-Ps增效蛋白基因的功能区域。【结果】Pu GV-Ps增效蛋白含有M60-like结构域、锌离子催化域和糖蛋白结合域,并包含13个潜在的糖基化位点。以此为依据设计P69(短截M60-like结构域)和P77(短截糖蛋白结合域)2个截短片段,构建了表达载体p ET15b-P69和p ET15bP77,原核表达量明显高于全长基因P104。表达产物纯化蛋白围食膜降解活性表明,P69对斜纹夜蛾围食膜大分子蛋白降解程度高于P77,但两者均低于P104。病毒增强苏云金杆菌(Bt)实验表明,截短片段的表达产物提高了Bt对小菜蛾的毒力,但增强活性显著低于P104。【结论】研究结果表明,Pu GV-Ps增效蛋白基因N端M60-like结构域和C端糖蛋白结合域对其增效作用的发挥都具有一定功能,这些结构对维持增效蛋白的构象也发挥了一定的作用,截短片段P69有利于保持Pu GV-Ps增效蛋白的活性、提高表达水平。该研究结果对增效蛋白的工业化生产具有一定的指导意义。     

Biochemical and genetic characterization of ChiA,the major enzyme component for the solubilization of chitin by<Emphasis Type="Italic"> Cellulomonas uda</Emphasis>

Cellulomonas uda efficiently solubilized chitinous substrates with a simple chitinase system composed of an endochitinase, designated ChiA, which hydrolyzed insoluble substrates into long-chain chitooligosaccharides, and an as yet uncharacterized exochitinase activity. ChiA, isolated from culture supernatant fluids, was found to be a glycosylated endochitinase with an apparent molecular mass of approximately 70 kDa and pI of 8.5. The gene encoding ChiA was cloned in Escherichia coli and sequenced, revealing an open reading frame of 1,716 bp encoding a 571-amino-acid protein with a predicted molecular mass of 59.2 kDa. The region upstream of chiA included a conserved –35 hexamer flanked by two direct repeats analogous to those found in many Streptomyces chitinase promoters, and thought to function as binding sequences for regulatory proteins. Analysis of the deduced amino acid sequence showed a modular protein consisting of a signal peptide at its N terminus, a family 2 carbohydrate-binding module (CBM2) that was closely related to the substrate-binding domains of glycosyl hydrolases from distantly related bacteria, and a family 18 glycosyl hydrolase catalytic module related to Streptomyces chitinases. In contrast to the fibronectin type III domains of Streptomyces chitinases, the linker region between modules in ChiA consisted of a long proline- and threonine-rich module, thought to contribute to the glycosylation and flexibility of the mature protein.Abbreviations CBM Carbohydrate-binding module - P-T Proline- and threonine-rich domain - Fn3 Type III repetitive sequences of fibronectin domain - PKD Polycystic kidney disease I domain     

Expression, Purification, and Biophysical Characterization of the BRCT Domain of Human DNA Ligase IIIα

The C-terminal regions of several DNA repair and cell cycle checkpoint proteins are homologous to the breast-cancer-associated BRCA-1 protein C-terminal region. These regions, known as BRCT domains, have been found to mediate important protein-protein interactions. We produced the BRCT domain of DNA ligase IIIα (L3[86]) for biophysical and structural characterization. A glutathione S-transferase (GST) fusion with the L3[86] domain (residues 837–922 of ligase IIIα) was expressed in Escherichia coli and purified by glutathione affinity chromatography. The GST fusion protein was removed by thrombin digestion and further purification steps. Using this method, ¹⁵N-labeled and ¹³C/¹⁵N-double-labeled L3[86] proteins were prepared to enable a full determination of structure and dynamics using heteronuclear NMR spectroscopy. To obtain evidence of binding activity to the distal BRCT of the repair protein XRCC1 (X1BRCTb), as well as to provide insight into the interaction between these two BRCT binding partners, the corresponding BRCT heterocomplexes were also prepared and studied. Changes in the secondary structures (amount of helix and sheet components) of the two constituents were not observed upon complex formation. However, the melting temperature of the complex was significantly higher relative to the values obtained for the L3[86] or X1BRCTb proteins alone. This increased thermostability imparted by the interaction between the two BRCT domains may explain why cells require XRCC1 to maintain ligase IIIα activity.     

Polymorphism of lectin genes in <Emphasis Type="Italic">Lathyrus</Emphasis> plants

The carbohydrate-binding sequences of the lectin genes from spring vetchling Lathyrus vernus (L.) Bernh., marsh vetchling L. palustris (L.), and Gmelin’s vetchling L. gmelinii (Fitsch) (Fabaceae) were determined. Computer-aided analysis revealed substantial differences between nucleotide and predicted amino acid sequences of the lectin gene regions examined in each of the three vetchling species tested. In the phylogenetic trees based on sequence similarity of carbohydrate-biding regions of legume lectins, the sequences examined formed a compact cluster with the lectin genes of the plants belonging to the tribe Fabeae. In each plant, L. vernus, L. palustris, and L. gmelinii, three different lectin-encoding genes were detected. Most of the substitutions were identified within the gene sequence responsible for coding the carbohydrate-binding protein regions. This finding may explain different affinity of these lectins to different carbohydrates, and as a consequence, can affect the plant host specificity upon development of symbiosis with rhizobium bacteria.     

Gene Structure and Molecular Phylogeny of the Linker Chains from the Giant Annelid Hexagonal Bilayer Hemoglobins

Giant extracellular hexagonal bilayer hemoglobin (HBL-Hb), found only in annelids, is an ∼3500-kDa heteropolymeric structure involved in oxygen transport. The HBL-Hbs are comprised of globin and linker chains, the latter being required for the assembly of the quaternary structure. The linker chains, varying in size from 225 to 283 amino acids, have a conserved cysteine-rich domain within their N-terminal moiety that is homologous to the cysteine-rich modules constituting the ligand binding domain of the low-density lipoprotein receptor (LDLR) protein family found in many metazoans. We have investigated the gene structure of linkers from Arenicola marina, Alvinella pompejana, Nereis diversicolor, Lumbricus terrestris, and Riftia pachyptila. We found, contrary to the results obtained earlier with linker genes from N. diversicolor and L. terrestris, that in all of the foregoing cases, the linker LDL-A module is flanked by two phase 1 introns, as in the human LDLR gene, with two more introns in the 3′ side whose positions varied with the species. In addition, we obtained 13 linker cDNAs that have been determined experimentally or found in the EST database LumbriBASE. A molecular phylogenetic analysis of the linker primary sequences demonstrated that they cluster into two distinct families of linker proteins. We propose that the common gene ancestor to annelid linker genes exhibited a four-intron and five-exon structure and gave rise to the two families subsequent to a duplication event.