THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269, No. 48, Issue of December 2, pp. 30181-30186, 1994 Printed in U.S.A.

Structural Characterization of an Ascaris Myoglobin* (Received for publication, July 19, 1994, and in revised form, September 10, 1994)

Mark. L. BlaxterS, Jacques R. Vanfleterensn, Jiazhi Xiall, and Luc Moensll** From the $Wellcome Centre for Parasitic Infections, Department of Biology, Imperial College of Science, Technology and Medicine, London SW7 ZBB, United Kingdom, the §Department of Morphology, Systematics and Ecology, University of Ghent, B.9000 Ghent, Belgium, and the IDepartment of Biochemistry, University of Antwerp, B-2610 Antwerpen (Wilrijk), Belgium

Globin was purified from the body wall of adults of the parasitic nematode Ascaris mum. Internal peptide frag- ments were sequenced and cDNAs encoding a polypep- tide of 154 amino acids isolated by polymerase chain reaction. The polypeptide lacks a signal sequence, iden- tifying it as a cytosolic myoglobin-like species. The na- tive protein is a dimer. The predicted amino acid se- quence shares several unusual substitutions with other nematode globins. Like the abundant pseudocoelomic A. suum hemoglobin it has a ’&I- at B10 and a Gln at E7, substitutions thought to be determinants of high affin- ity. However, the 10-fold lower oxygen affinity of body wall globin suggests that in this molecule Tyr(B10) does not form an additional hydrogen bond with the heme bound oxygen. Evolutionary analysis of the nematode globins suggests that the monodomain myoglobin-like molecules and the two-domain hemoglobin-like mol- ecules diverged about 500 million years ago, well before the divergence of the ascarid genera Ascaris and Pseudoterranova. The absence of introns in the A. suum myoglobin, in contrast to other nematode globin genes, is consistent with the hypothesis that during evolution intron elimination was the predominant event.

~ ~~ ~~

Ascaris suum is a parasitic nematode species the adult stage of which lives in the microaerobic environment of the pig in- testine. The worms contain substantial quantities of globin in the pseudocoelomic (perienteric, extracellular) fluid and the body wall tissue. The pseudocoelomic globin has been studied in detail. The native molecule consists of eight identical mono- mers of 40.6 kDa, each containing two tandemly linked globin domains. These are held together by a barrel of salt bridges between the side chains of Glu, Lys, and His residues in the COOH-terminal polar zipper of each monomer (1-4). The ex- tracellular hemoglobin from the related ascarid species Pseudo- terranova decipiens has a very similar primary structure (5, 6 ) and might be expected to assume a similar configuration.

Body wall globin is distinct in several ways. It is localized in cellular compartments belonging to the hypodermis (epider- mis), the dorsal, ventral and lateral cords, the nerve ring, and body wall muscle (7,8). It is found as single-domain subunits of 17 kDa which form dimers of 35-37-kDa native molecular mass. This isoform has been described from many species (9)

* This work was supported by the Wellcome Trust Award 038294 the Belgian Fund for Joint Basic Research Grant 2.0023.94. The costs of

charges. This article must therefore be hereby marked “advertisement” publication of this article were defrayed in part by the payment of page

in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 11 Research Director with the National Fund for Scientific Research. ** To whom correspondence should be addressed: Dept. of Biochem-

istry, University of Antwerp, Universiteitsplein 1, B-2610 Wilrijk, Belgium. Tel.: 32-03-8202323; Fax: 32-03-8202248; E-mail: luc.moens@ reks.uia.ac.be.

~~

and representatives have been cloned from the strongylid nematodes Nippostrongylus brasiliensis (10) and Dicho- strongylus colubriformis (11) and the rhabditid Caenorhabditis elegans (4, 12, 13). Electrophoretic and chromatographic analy- ses ofAscaris body wall globin extracts suggested the presence of two distinct molecular species of 35-37 kDa (14-18). In the strongylids two forms are found in distinct compartments (the body wall tissue and the cuticle) and the cuticle isoforms have an N-terminal extension leader peptide which directs secretion to this extracellular structure (9, 10, 13).

The pseudocoelomic and body wall globins detected in Ascaris both display unusual physiological features. For ex- ample, the oxygen affinity of the hemoglobin and myoglobin- like molecules are about 300 and 5-fold higher than their mam- malian counterparts, respectively, as measured by their dissociation constants (19, 20) (Table I). It follows that the pseudocoelomic globin is unable to deliver oxygen to the body wall globin and thus that there can be no transport of oxygen by facilitated diffusion into the pseudocoelomic fluid and the tis- sues it bathes. The pseudocoelomic globin may not have a res- piratory function. A number of alternative functions have been advanced e.g. delivery of hematin to the eggs (8), a n oxygen sink to protect the metabolic activity, which is essentially an- aerobic, from oxygen (91, a n osmotic function (9) and a function in sterol metabolism (21), but essentially the problem remains unresolved. In contrast, there is plausible evidence that the body wall globin of Ascaris or at least one of its constituent isoforms supplies oxygen needed for muscle activity, since de- oxygenation of live worms results in reversible cessation of movement (7).

We describe here the isolation, cloning and sequence of an A. suum body wall globin in order to shed some light on the relationship of structure to oxygen affinity in ascarid globins.

MATERIALS AND METHODS Purification of Body Wall Globin-Live specimens of A, suum were

collected from a local slaughterhouse. An extract enriched in body wall globin was prepared essentially as described by Okazaki et al. (16) and Wittenberg et al. (17), precipitated at 40-80% saturated ammonium sulfate, and size fractionated by gel filtration on Sephadex G-75 (1 x 75-cm column) equilibrated and run in 5 mM sodium phosphate buffer, pH 7.0. The red globin fraction was adsorbed onto the top of a 3 x 13-cm column of DE52 equilibrated with 5 mM sodium phosphate buffer, pH 7.0, and was eluted with a linear gradient from 0 to 200 mM NaCl in the same buffer. Purified globin was focused on a 7.5% polyacrylamide gel containing 2% ampholytes, pH 3.5-10, according to Drysdale et al. (22). Purification was monitored by one- and two-dimensional PAGE‘ under native (23) and denaturing (24, 25) conditions.

Determination of Internal Peptide Sequences-Purified globin was cleaved at aspartic acid residues in 2% formic acid at 120 “C for 4 h (26). The resulting peptides were separated by narrow-bore HPLC on a

The abbreviations used are: PAGE, polyacrylamide gel electrophore- sis; PCR, polymerase chain reaction; HPLC, high performance liquid chromatography.

30181

30182 Ascaris Myoglobin TABLE I

A. suum myoglobin (Mb) and hemoglobin (Hb), compared Association rates of oxygen and carbon monoxide for

to sperm whale myoglobin

Species KG," (0,) Kt,r(O,) Kl, (2;) Ref

p i - l s- I s" n n p.w" s"

Ascaris Hb Wild type 1.5 0.0041 2.7 17 19 Domain 1; Tyr (B10) - Leu 9.0 5.0 555 0.75 38 Domain 1; Tyr (B10) - Phe 40.3 2.0 50 2.7 38

Ascaris Mb 1.2 0.23 192 19 Physeter Mb 14 11.9 850 0.51 20

Vydac C-4 column (2.1 x 100 mm). Proteins and peptides were se- quenced using a model 471/140B Sequenator (Perkin-Elmer-Applied Biosystems) operated as recommended by the manufacturer.

Derivation of Degenerate Oligonucleotides-Sequenced peptide frag- ments were aligned to a set of nematode globin sequences and those having likely similarity were selected for primer design. Three degen- erate primers were made containing synthetic EcoRI sites at their 5' ends: PAHl2F (CGAATTCAYTAYTTYAARGGNAAYGA, a 26-mer with 128 redundancies), corresponding to the sense strand predicted by pep- tide fragment pahl2 (FHYFKGNE); PAHl2R (CGGAATTCRTTNC- CYTTRAARATRTG, a 26-mer with 128 redundancies), corresponding to the antisense strand predicted by peptide fragment pahl2; and PAH17R (CGGAATTCGCRTANGCYTTRAAYTTNGCNGG, a 31-mer with 1028 redundancies), corresponding to the antisense strand pre- dicted by peptide fragment pahl7 (AKFKAYA). The primers were pre- dicted to give poor or no amplification from the pseudocoelomic globin gene due to 3' mismatches.

Cloning of the Ascaris Myoglobin Gene from a Muscle cDNA Library-An A. mum adult muscle cDNA library in the Not1 site of A-Zap (provided by Dr. Tim Geary of Upjohn Research Laboratories, Kalamazoo, MI) was screened by PCR. Briefly, 10 pl of the cDNA library (stock a t 10"' plaque-forming units/ml) were denatured a t 95 "C for 5 min and centrifuged a t 12,000 x g , and 1 pl was taken to a 50-pl PCR reaction mix (20 mM (NH,)$O,, 75 mM Tris-HC1, pH 9.0,0.1% Tween 20, 1.5 mM MgCI,, and 0.1 unit of Thermoprime polymerase (Advanced Biotechnologies)) containing 50 ng each of ( a ) either the M13/pUC uni- versal left or universal right primer and ( b ) the specific primer. The PCR was carried out with 35 cycles of 94 "C for 1 min, 55 "C for 2 min and 72 "C for 2 min. Positive fragments were tested for the presence of nested primer sites, and then blunt end-cloned into pBluescript KS (27) (as EcoRI cleavage revealed the presence of an internal site). Recombi- nants were confirmed by PCR and the DNA sequence obtained from double-stranded templates using T7 polymerase (28) (Pharmacia Bio- tech Inc.). To obtain the 5' end of the transcript, a specific primer (AMDN, see Fig. 5) was synthesized and used in library PCR as above with a primer corresponding to the nematode universal spliced leader, SL-I, which is present at the 5' end of a majority ofAscaris messenger RNAs (29, 30). The sequence was determined from both strands from multiple independent clones and a contiguous 570-base pair sequence assembled.

Genomic DNA PCR-To isolate genomic DNA fragments which would be predicted to contain introns by comparison with other nematode globin genes, specific primers were synthesized (AM5, AM3, AMUP, and AMDN, see Fig. 5) and used in PCR reactions with 100 ng of A. suum adult genomic DNA as target. Reactions were analyzed on 1% agarose gels and Southern blotted with a ''*P-labeled probe (31) made from a cDNA clone spanning the COOH-terminal60% of the body wall globin to identify globin gene-containing fragments.

Alignment and 'Dee Construction-Globin amino acid sequences were aligned by use of the Bashford et al. (32) template I1 and a non- vertebrate template: which are both based on tertiary structure con- siderations. An evolutionary tree was constructed by neighbor-joining (34) as implemented in the software package TREECON (35), which was adapted for the analysis of amino acids, including bootstrapping.

RESULTS AND DISCUSSION

Purification of Body Wall Globin-Gel filtration of salt frac- tionated body wall extract resolved a heme containing protein

L. Moens, J. Vanfleteren, Y. Van de Peer, K. Peeters, 0. Kapp, J. Czelusniak, M. Goodman, M. Blaxter, and S. Vinogradov, submitted for publication.

2.5 A

0 0

46 60 74 88 102 Fraction nr

~ w . 1 0 ~ I3

1 2 3 4 F - FIG. 1. Size-exclusion chromatography of A. nuurn body wall

globin. A , Gelfiltration was performed on a Sephadex G-75 column (1 x 100 cm) in 5 mM sodium phosphate buffer pH 7. Fractions were measured at 280 (H) and 412 nm ("+) respectively. M , mark- ers are myoglobin, ovalbumin, and bovine serum albumin. B , 15% SDS- PAGE; lane 1, markers; lane 2, crude extract; lanes 3-5, samples as indicated by arrows in A .

fraction of approximately 3 5 4 0 kDa (Fig. lA). This fraction emerged from a DE32 ion exchange resin as a single band a t 110 mM NaCl with no additional purification (results not shown). Molecular mass estimation by SDS-PAGE in the pres- ence of 2-mercaptoethanol yielded a mass of 15-17 kDa (Fig. 1B) with traces of a 3 5 4 0 kDa band. This strongly suggests that the native body wall globin consists of 2 monomers of 15-17 kDa held together by disulfide bonds. The native molec- ular mass estimate is in good agreement with those derived for native body wall globin by Hamada et al. (15) and Okazaki et al. (16) from the sedimentation coefficient ( s ~ ~ , , ~ = 3.1). In contrast, Smith and Morrison (36) determined a sedimentation coeffi- cient s,,,~~ = 1.55, which corresponds to the molecular mass of the monomer. The body wall globin monomer has the usual size of intracellular globins such as vertebrate myoglobin as well as the nematode body wall globins (including the cuticular iso- forms) known so far (10, 13). Pseudocoelomic globin monomers are twice this size, being didomain polypeptides (1, 2, 4).

The red globin fraction obtained by gel filtration and ion exchange chromatography was subjected to two-dimensional electrophoresis using PAGE under native conditions, compara- ble to the technique used by Hamada et al. (15) and Waring et al. (23) in the first dimension and SDS-PAGE in the second dimension. 11 major polypeptide spots ranging from 11 to 37 kDa were resolved (Fig. 2). Expecting a recognizable globin sequence, all spots were subjected to NH,-terminal sequencing. Six of these represented contaminating proteins, whereas four were blocked at the N terminus (Table 11). The remaining rela- tively abundant polypeptide spot (11 in Fig. 2 and Table 11) yielded an amino-terminal sequence identical to a polypeptide previously sequenced as a major contaminating hemoprotein

Ascaris Myoglobin 30183

~ w . 1 0 ~ > +

94 68 43

30

21

14

FIG. 2. Two-dimensional electrophoresis of purified body wall globin. Native protein was separated in the first dimension according to charge heterogeneity (23). Second dimension was 15% SDS-PAGE.

co-purifying with cytochrome c (13). The sequence can be ten- tatively aligned with the nematode globins using the tripeptide Gln-Gly-Gln as a landmark. This motif is highly conserved a t E6-E8 in nematode globins (see Fig. 7). The estimated mass of approximately 11-13 kDa matches that predicted for a trun- cated globin starting a t a t position E,. Could this polypeptide represent the carboxyl-terminal region of a truncated globin molecule? Its abundance and the relatively low proteolytic ac- tivity ofAscaris extracts argues against it being a degradation product. The sequence does not derive from either the pseudocoelomic or the body wall globin. Further work will be needed to clarify its nature.

Although the Ascaris myoglobin must be one of the major spots in the pattern shown in Fig. 2 it was not identifiable by its NH,-terminal sequence. This suggest that the NH, terminus is blocked (Fig. 2, spots 4 and 8-10, and Table 11). Native isoelec- tric focusing was choosen as an alternative method for sepa- rating the myoglobin preparation based on its high resolution and on the fact that heme containing proteins can be easily detected visually (22). Isoelectric focusing of purified CO- derivatized body wall globin resolved two major and four minor heme containing bands spanning the pH 5.0-6.5 interval (Fig. 3).

The patterns of Figs. 2 and 3 cannot be correlated with each other as the techniques used are too different. However, both approaches, native electrophoresis (Fig. 2, first dimension) and isoelectric focusing (Fig. 3), show a charge heterogeneity as previously reported for body wall globin prepared by Okazaki et al. (161, Hamada et al. ( E ) , and Wittenberg et al. (17).

Amino-terminal sequence determination of approximately 1 nmol of globin extracted from the focused bands 1 and 2 (Fig. 3) yielded no phenylthiohydantoin derivative signal, suggesting that both isoforms were blocked and that they were essentially free of contaminating polypeptides with accessible termini. In- ternal fragments produced by limited acid hydrolysis of both subfractions produced dissimilar reversed-phase HPLC pro- files (Fig. 4). Several peptides yielded identical sequences, how- ever, and still other peptide fragments derived from both bands were mixtures of similar though not identical amino acid sequences.

cDNA and Genomic Structure of a Body Wall Globin Gene- Primers PAHl2F and PAH17R derived from the sequences of the best aligned peptides from bands 1 and 2 proteolysis gave strong products when used with vector-derived primers in PCR from an A. suum adult muscle cDNA library. Multiple clones from both PAH12F-vector primer and PAH17R-vector primer reactions were sequenced: only a single sequence class of clones was found, suggesting that they all derive from a single gene.

TARIX I1 Amino-terminal sequence of polypeptides marked in Fig. 2

Identification was done by data base searches.

No. Sequence Similarity to:

1

2

3

4

5

6

7

8

9

10

11

LRQQAVAIKGKLLDGTAPARN- VRVTKL

GVGGPMLTLSNGRQMPQVGLG

VQIPSVKLATGADLPLFGLGT

I n a c c e s s i b l e f o r Edman d e g r a d a t i o n GYFPLETMLNYH

SYELHDFMTEQQKQQIRSM

ELEQYREYLQFMKDKQYK

I n a c c e s s i b l e f o r Edman d e g r a d a t i o n I n a c c e s s i b l e for Edman d e g r a d a t i o n I n a c c e s s i b l e f o r Edman d e g r a d a t i o n LSAQGQQYVEEVKKLLGPATV- KAILAIRDDRSL

Hypothetical 15.7- kDa protein chro- mosome 111, C. elegans

Probable reductase, Leismania

Aldose reductase, bovine

HMG-CoA synthetic homolog, peptide

Inositol trip recep- C. elegans

tor, Drosophila Ribonucleo protein, YML3 yeast

Possible globin-like (Fig. 7)

li; 3 A

t - 5 ‘7 6 I I !

+ FIG. 3. Separation of native body wall globin by isoelectric

focusing under non-denaturing conditions. Purified body wall glo- bin was converted to the CO derivative and separated by 7.5% PAGE (rods measuring 0.6 x 13.5 cm) in a 3.5-10 pH gradient for 4 h a t 400 V (22). Globin isoforms are revealed by their natural reddish color.

These fragments gave a contiguous sequence which was pre- dicted to be missing the NH,-terminal 2-8 amino acids of the complete open reading frame (Fig. 5). To obtain this sequence (as the NH,-terminal of bands 1 and 2 was found to be blocked), the nematode spliced leader was used as a 5‘ primer with specific oligonucleotides predicted from the sequence of the clones to isolate a single size class of fragments from the library which gave unique sequence and completed the 5’ end of the message (Fig. 5). While not strict proof of the presence of spliced leader at the 5‘ end of the Ascaris body wall globin

30184 Ascaris Myoglobin

A 60 70 80 * AMs*-* * * * 90 100

LGAG TCT GTT CAG TGC GI+ ACC TGC GAG AAA ACC ATC GCC AAT GGC ACT E S V Q C G T C E K T I A N G T

110 * * 120 130 * * 1 4 0 150

GAA TTC TAC GCA CTA CTA TTC GAT AAG CAT CCG GAC TTA CGC- * 902 *

E F Y A L L F D K H P D L R H Y

160 902 *"+ * 170 *AMUP+ 180 - 190

I TTC AAG GGT +T GAA TTG ACG GGA GCC GAT GTG AAG AGC GAT * *

F K G N E N L T G A D V K K S D

L 0 1 2 3 4

-1 M I N x 10

FIG. 4. Reversed-phase chromatography of globin peptide fragments. Protein focused into bands I and 2 (Fig. 3) was extracted and subjected to partial acid hydrolysis and the resulting peptide mix- tures were resolved on a 2.1 x 150-mm Vydac C-4 column with a linear gradient from 0 to 75% acetonitrile in 0.1% trifluoroacetic acid a t 0.25 mumin, for 45 min.

message, the isolation of clones from a library constructed by standard methods3 and the finding of spliced leader at the 5' end of other nematode globin mRNAs strongly suggests that this is not an artifact. There are 4 bases of 5"untranslated region before the initiation methionine: such a short 5'-un- translated region has been noted in other nematode monodo- main globins. The open reading frame extends for 153 codons and is followed by a 71-base 3"untranslated region containing a potential polyadenylation signal sequence (nucleotides 514- 517).

Primers derived from the cDNA sequence and designed to flank the sites of intron insertion in other nematode globin genes were used in PCR reactions on genomic DNA to deter- mine the presence, site and size of introns in the globin gene. All primer pairs gave reaction products of identical size to that predicted from the cDNA, strongly suggesting that the A. suum body wall globin gene has no introns (see Fig. 6). These prod- ucts, and not others which derived from products seen when the primers were used singly, hybridized to BwMb cDNA probes (not shown). Two separate preparations of A. mum genomic DNA gave the same result. Southern blot analysis also argues strongly for a single gene, as only one or two (EcoRI; there is an internal EcoRI site) fragments were detected (not shown).

The absence of introns in this globin gene is not inconsistent with the hypothesis that during evolution intron elimination was the predominant event (37).

Structural Aspects of the Deduced Amino Acid Sequence- The deduced amino acid sequence can be reliably aligned with

T. Geary, unpublished results

200 210 220 230 2 4 0 * * * * * + *AMDN* CAC TTC AAA AAG CAA GGA CAG AGA TTG CTA G GCA TGT CAT GTA TTG H F K K Q G Q R L L L A C H V L

250 260 270 280 79n * * _ _ ~

-T CTG GAA AAT GAT ~CCA GCA AGT TTC AAG GCA TAT] GCA CGA GAA * c* P17 * * *

A H L E N D P A S F K A Y A R E

300 310 330 320 * * I t 340

* * ATT GTC GAT CCC CAT CTT AGA ATG AGC GTT CAT CTC GAA CCT AAA CTC I V D P H L R M S V H L E P K L

350 360 370 * * I f

380 390

TGG AGT GAA TTC TGG CCC ATC TGG CTC GAC TAT CTG TCA ACA AAG GAG W S E F W P I W L D Y L S T K E

* *

400 410 420 * * * *

430

AGC GTT GAC GAT GCG ACA AAG AAC GCA TGG CTC GCA CTT GGC AAG AAG S V D D A T K N A W L A L G K K

* *

440 450 460 * * 470

* * 480

* * TTC TCC GAT GAA TGC CTC GAC CAC CTC AAA AAT CTT GGT CAA CCA CAC F S D E C L D H L K N L G Q P H

490 500 510 520 530 * * * * * * - * m 3 * * *

TAA ACAATCTTTC TTCATTCAGC ATT+GTAAC ACGCGTGCAT TTCT~PAAATT ter

540 550 560 * * * * TATGTTATAA AGACTAGCGT CA (poly A)

FIG. 5. cDNA and predicted amino acid sequence of Ascaris myoglobin. The complete cDNA sequence is shown. The open reading frame is translated below the DNA sequence. The primers used in PCR are boxed. The nematode spliced leader ( S L - I ) is also indicated.

the other nematode globin sequences know to date because of the conservation of several landmarks such as the obligatory Phe(CDl), His(F8), and Pro(C2) which determines the folding of the BC corner, Gly(B6) and Gly(ES), which position the crossover of the Band E helices, and Trp(H8), which is typi- cal of non-vertebrate globins (Fig. 7). The conservation of the volume and hydrophobicity of the solvent accessible and non- accessible residues, including the heme contacts can be ex- pressed by penalty scores against vertebrate and non-verte- brate globin templates (32).' The scores listed in Table I11 also clearly confirm classification of this sequence as a true non- vertebrate globin. We designate this polypeptide a myoglobin for the following reasons. First its mass matches that of canon- ical myoglobin, second it lacks a leader sequence, suggesting that it resides in the cytoplasm, and third it was cloned from a muscle cDNA library.

Several unusual substitutions, some of which are character- istic of nematode globins are discussed here. The presence of Cys(A8) as well as the absence of an aromatic residue at A12 and the five residue long AB corner are typical of nematode globins. Only annelid globins also have long AB corners. The exceptionally high oxygen affinity of the A. suum pseudocoelo- mic globin has been ascribed to the combination of Tyr(BlO), Gln(E7), and Ile(E11). More specifically, a second hydrogen bond was thought to be formed between the OH group of

Ascaris Myoglobin 30185

Tyr(B10) and bound oxygen (38). The myoglobin sequence re- ported here has Tyr(BlO), Gln(E7), and Leu(El1) (Fig. 71, yet its oxygen affinity is 60-fold lower (Table I). The discrepancy may be explained by assuming that other residues are essential to correctly position the B and E helices to enable hydrogen bonding between Tyr(B10) and O,, as already suggested by De Baere et al. (38). When the phenolic OH group is too far away, no extra hydrogen bonding will be possible. Alternatively, when the aromatic ring is too close, normal oxygen binding will be impaired, resulting in lower oxygen affinity, as observed in engineered globin (3840) . Since tight crossing of the B and E helices is favored by Gly at B6 and E8, the heme ligand must be displaced by about 1 A to provide the required space (38). Pos- sibly, substitution of Gly(FG2) in the Ascaris hemoglobin by Ser in myoglobin has this effect.

The occurrence of Pro a t E7 (a heme-lining position) is un- usual but probably has no adverse effect. Other heme contacts are conserved, e.g. Trp(G5), Trp(G9), and "rp(G12) form a hy- drophobic environment enclosing the heme pocket.

lo00 900 800 700 600 500 400

300

200

l o o

M 1 2 3 4 5 6 7

* * * *

+

+ +

FIG. 6. PCR assessment of intron presence in the Ascaris myo- globin gene. PCRs were performed on two independent preparations of A. suum genomic DNA as described using (lanes 1 and 2 ) primer A M 5 and AMDN (lanes 3 and 4 ) AMUP and AM3 (lanes 5 and 6) AMUP and AMDN or (lane 7) AMUP alone and analyzed on a 1.5% agarose gel. Molecular size markers ( M ) are in base pairs. Products marked with * were produced by AMUP alone, those marked with an arrow hybridized to the A. suum cDNA.

TABLE 111 Penalty scores of the globin sequences aligned in Fig. 7 against a

non-vertebrate template and Bashford et al. template I I (32) Abbreviations as in Fig. 7. Non-vertebrate data from L. Moens, J.

Vanfleteren, Y. Van de Peer, K. Peeters, 0. Kapp, J. Czelusniak, M. Goodman, M. Blaxter, and S. Vinogradov, submitted for publication.

Motif Total

A BC E FG H N V E " V E R b N V E V E R N V E V E R N V E V E R N V E V E R

~~~~ NVE VER

SW 0.0 0.0 0.0 0.0 0.2 0.5 0.0 0.0 0.0 0.0 0.2 0.5 LL 0.0 0.0 0.0 1.0 1.2 0.0 0.0 0.0 1.0 1.0 2.2 2.0 CE 2.0 2.0 0.5 1.7 0.0 3.4 1.0 3.9 0.0 2.4 3.5 13.4 TC 1.0 1.5 0.7 1.9 0.5 2.7 0.7 4.2 0.0 1.4 2.9 10.2 NBb 0.5 1.5 0.0 1.2 0.5 2.7 0.7 3.3 0.0 1.4 1.7 10.1 NBc 0.5 0.5 1.0 1.9 0.0 2.2 0.7 3.1 1.0 1.7 3.2 9.4 PD1 1.0 1.0 0.0 0.7 2.0 4.9 0.0 2.4 0.0 0.7 3.0 9.7 PD2 0.5 1.5 0.0 0.7 2.5 4.2 0.0 2.4 0.0 1.4 3.0 10.2 AS1 1.0 1.0 0.0 0.2 1.0 3.9 0.0 2.4 0.0 0.7 2.0 8.2 AS2 0.5 0.5 0.0 0.7 1.5 4.9 0.0 4.4 1.0 1.9 3.0 12.4 ASMb 0.0 0.0 0.0 1.2 0.0 2.9 0.7 4.1 1.0 2.4 1.7 10.6

(I NVE, non-vertebrate. * W R , vertebrate.

The combination of Trp(G9) with Gly(H12) in A. suum hemo- globin was previously interpreted as a coupled substitution, whereby the small side group of Gly(H12) provides the neces- sary space for the bulky side chain of Trp(G9) (1). It now ap- pears that this particular configuration is a plesiomorphic and generally conserved feature of nematode globins. The occupa- tion of G12 by Trp is also unusual but is only seen in Ascaris myoglobin and in both globins from Nippostrongylus (Fig. 7).

A strongly hydrophobic patch is formed by the residues oc- cupying E l l through E15. In Scapharca hemoglobin two iden- tical subunits interact in part through hydrophobic residues exposed a t their E-F surfaces (33, 41). Possibly, dimer forma- tion in Ascaris myoglobin also proceeds through hydrophobic interaction at the E surfaces, with additional salt bridges be- tween positively and negatively charged residues that are more or less regularly spaced in the E-G region of the molecule. The heterogeneity of body wall globin observed during peptide se- quencing might reflect the presence of heterodimers formed between different globin isoforms. I t should be stressed in this respect that the extract was made from whole body wall tissue which consists of cuticle, hypodermal, neural and muscle tis- sue. While we detected only a single gene, divergent globins may have been missed in our screening.

Evolutionary and Phylogenetic Aspects-The inferred evolu- tionary relationships of the globins aligned in Fig. 7 are repre- sented in Fig. 8. The evolutionary events predicted by the re-

bdll22a bd 1112:lla bd2l2O bd2/22 b d 2 1 9

bd1/16~:12a bdl/llb ."" ~ .". < ."."".... bd1/9 bdlll6b:lllb

I,C crt c f r 2

FIG. 7. Alignment of nematode globins with sperm whale myoglobin and lupin leghemoglobin. Physeter catodon (SW), Lupinus luteus I1 (LL) , Scapharca inequivalvis (Sca) , rl decipiens domain 1/2 ( P D l / Z ) , A. suum domain 112 (AS1 / 2 ) , C. elegans (CE) , T colubriformis (TC) , N. brasiliensis body isoform (NBb), N. brasiliensis cuticular isoform (NBc), A. suum myoglobin (ASMb), A. suum body wall band 112 (ASBWbdl.I2), A. suum cytochrome c fraction 2 ( A S Cyt c Fr 2 ) (16). The myoglobin fold (Mb fold) displays the template I of Bashford et al. (29). The proxlmal and distal heme contacts are marked as D and P.

30186 Ascaris Myoglobin

Distance 0.1

H

N. brasiliensis b

T. colubriformis

N. brasiliensis c

C. elegans

A. suum Mb

A. suum Hb dom2

P. decipiens dom2

A. Suum Hb doml

P. decipiens doml

L. iuieus

P. cathodon

FIG. 8. Unrooted neighbor-joining tree constructed from the globin sequences aligned in Fig. 7. The evolutionary distance sepa- rating any pair of species is given by summing the lengths of the con- necting branches along the horizontal axis, using the scale on top, which

for superimposed events. Deletionshsertions were ignored. Bootstrap represents 10 amino acid replacements per 100 amino acids, corrected

values higher than 50% are indicated a t each node.

sulting tree substantiate and slightly modify scenarios suggested previously (9, 13). The A. suum myoglobin clearly roots within a clade of other monodomain nematode globins. The duplication of the monodomain globin to give both intra- cellular and extracellular (cuticular) forms is still peculiar to the strongylids. However, the heterogeneity observed during peptide sequencing of body wall globins and the presence of potentially distinct globins in other tissues ofAscaris raises the possibility that additional gene duplications of monodomain globins may have occurred in the ascarid lineage.

The pseudocoelomic globins constitute an evolutionarily dis- tinct group of sequences whose divergence from the mono- domain myoglobin-like molecules suggests that the two groups diverged through gene duplication about 500 million years ago (13). The duplication which created the two domains of this globin preceded the divergence of Pseudoterranova and Ascaris and is dated at 300 million years ago (13). Given the phylogeny derived from the monodomain globins this would predict that pseudocoelomic hemoglobin-like proteins should be present in the strongyle (and perhaps rhabditid) lineage too, but such globins have not yet been described. A concerted search for such molecules seems justified. If indeed the strongylids and rhab- ditids (and other related groups) do not express pseudocoelomic globin homologues two scenarios can be envisaged. One is to propose that the other lineages have lost the pseudocoelomic globin gene, and that its absence is a derived (apomorphic) character. The other is to hypothesize that the pseudocoelomic globins have been under a radically different and saltatory selective pressure than the monodomain globins. This pressure could arise from two requirements. The function of the pseudocoelomic globin is unknown beyond the unlikely func- tion as an oxygen carrier in the gut microaerobic environment. They may act as either oxygen sinks (to keep an essential and strictly anaerobic reaction oxygen free) or as enzymatic factors (in an aerobic reaction that derives its oxygen from the pseudocoelomic globin directly rather than by diffusion), or they may have still other functions. These novel functions may have driven the selection of the distinct pseudocoelomic globin domain identity. Second, due to the extracellular location of the hemoglobin there was a selective pressure on the molecule to acquire a high M , in order to avoid elimination by excretory processes. This high M , was achieved by the fusion of the tan- demly duplicated domains and by the addition of a polar zipper

extension of the carboxyl terminus resulting in the aggregation of the globin chains into an homo-octamer (1,3). Again, distinct selection based on the sequence requirements of multimeriza- tion may have driven the sequence to a novel part of evolution- ary space which may confound molecular clock-based phyloge- netic comparisons.

Nematodes have a variable number of introns in the genes encoding globin domains. The gene encoding the pseudocoelo- mic globin has retained three introns in each globin domain (21, whereas the myoglobin gene has none. This is not inconsistent with the hypothesis that during evolution intron elimination was the predominant event (37).

Acknowledgments-We thank Dr. Tim Geary for the supply of the cDNA library and Dr. J. Moore for Ascaris genomic DNA.

1.

2.

3.

4. 5.

6.

7. 8. 9.

10. 11.

12.

13.

14.

15.

16.

17.

18.

REFERENCES De Baere, I., Liu, L., Moens, L., Van Beeumen, J., Gielens, C., Richelle, J.,

Trotman, C., Finch, J., Gerstein, M. & Perutz, M. (1992) Proc. Natl. Acad. Sei. U. S. A. 89,4638-4642

Sherman, D. R., Kloek, A. P., Krishoan, B. R., Guinn, B. & Goldberg, D. E. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 1169~11700

Perutz, M. F., Staden, R., Moens, L. & De Baere, I. (1993) Curr. Biol. 3, 249-251

Kloek, A. P., Sherman, D. R. & Goldberg, D. (1993) Gene (Amst. 129,215-221 Dixon, B., Walker, B., Kimmins, W. & Pohajdak, B. (1991) Proc. Natl. Acad.

Dixon, B., Walker, B., Kimmins, W. & Pohajdak, B. (1992) J. Mol. Euol. 35,

Davenport, H. E. (1949) Proc. R. SOC. Land. E Biol. Sci. 136,255-270 Smith, M. H. & Lee, D. L. (1963) Proc. R. SOC. Lond. B Biol. Sci. 157,234-257 Blaxter, M. L. (1993) Parasitol Today 9, 353-360 Blaxter, M. L. (1994) Mol. Biochem. Parasitol., in press Frenkel, M. J., Dopheide, T. A. A., Wagland, B. M. & Ward, C. W. (1992) Mol.

Mansell, J. B., Timms, K., Tate. W. P. T., Moens, L. & Trotman, C. N. A. (1993)

Vanfleteren, J., Van De Peer, Y., Blaxter, M. L., Tweedie, S. A. R., Trotman, C.,

Hamada, K., Okazaki, T., Shukuya, R. & Kaziro, K. (1962) J. Biochem. (Tokyo)

Hamada, K., Okazaki, T., Shukuya, R. & Kaziro, K. (1963) J. Biochen. (Tokyo)

Okazaki, T., Wittenberg, B. A., Briehl, R. W. & Wittenberg, J. B. (1967)

Sci. U. S. A. 88, 5655-5659

131-136

Biochem. Parasit. 50, 27-36

Biochem. Mol. Biol. Int. 30,643-647

Lu, L., Van Hauwaert, M. L. & Moena, L. Mol. Phyl. Euol. 3, 92-101

52,290-296

53.479488

Biochim. Biophys. Acta 140,258-265 Wittenberg, J. B., Bergerson, F. J., Applehy, C. A,, and Turner, G. L. (1974) J .

Wittenberg, J. B. & Wittenberg, B. A. (1981) Methods Enzymol. 76, 2 9 4 2 19. Gibson, Q. H. & Smith, M. H. (1965) Proc. R. SOC. Lond. B Biol. Sei. 163,

20. Gibson, Q. H., Olson, J. S., McKinnie, R. E. & Rohlfs. R. J. (1986) J. Biol.

Biol. Chem. 249, 40574066

206214

Chem. 261,10228-10239 . .

21. Sherman, D. R., Guinn, B., Perdok, M. M. & Goldberc. D. E. (1992) Science

26.

27. 28.

29.

30. 31. 32. 33. 34. 35. 36. 37.

38.

39.

40.

41.

258, 1930-1932

42-50

Yl

22. Drysdale, J. W., Righetti, P. & Bum, H. F. (1971) Biochim. Biophys. Acta 229,

23. Waring, G., Poon, M. C. & Bowen, S. T. (1970) Znt. J. Biochern. 1,537-540

25. O'Farrell, P. H. (1975) J. Biol. Chem. 250,40074021 24. Laemmli, U. K. (1970) Nature 227,680-685

Inglis, A,, McKern, N., Roxburgh, C. & Strike, P. (1980) in Methods in Peptide and Protein Sequence Analysis (Birr, C., ed) pp. 329-343, Elsevier Science

Starr, L. & Quaranta, V. (1992) BioTechniques 13, 612-618 Williams, S. A,, Slatko, B. E., Moran, L. S. & DeSimone, S. M. (1986) Bio

Nilsen, T. W., Shambaugh, J., Denker, J., Chuhh, G., Faser, C., Putnam, L. &

Bektesh, S., van Doren, K. & Hirsh, D. (1988) Genes & Deu. 2, 1277-1283 Feinberg, A. P. & Vogelstein, B. (1983) Anal. Biochem. 132 6 1 3 Bashford, D., Chothia, C. & Lesk, A. A. (1987) J. Mol. Biol. 196, 199-216 Royer, W. E. (1994) J. Mol. Biol. 235, 657-681 Saitou, N. & Nei, M. (1989) Mol. Biol. Euol. 4, 406-425 Van de Peer, Y. & De Wachter, R. (1992) Comput. Appl . Biosci. 9, 177-182 Smith, M. H. & Morrison, M. (1963) Biochirn. Biophys. Acta 71,364-370 Moens, L., Vanfleteren, J., De Baere, I., Jellie, A. M., Tate, W. & Trotman, C.

De Baere, I . , Perutz, M. F., Kiger, L., Marden, M. C. & Poyart, C. (1994) Proc. (1992) FEBS Lett. 312, 105-109

Gibson, Q. H., Regan, R., Olson, J. S., Carver, T., Dixon, B., Pohajdak, B., Natl. Acad. Sci. U. S. A. 91, 1594-1597

Carver, T. E., Brantley, R. E., Jr., Singleton, E. W., Arduini, R. M., Quillin, M. Sharma, P. & Vinogradov, S. (1993) J. Bioi. Chem. 268,16993-16998

Royer, W. E., Jr., Hendrickson, W. A. & Chiancone, E. (1989) J. Biol. Chem. L., Phillips, G. N., Jr., & Olson, J. S. (1992) J. Biol. Chem. 267,14443-14450

264,21052-21061

Publishers, Amsterdam

Techniques 4,138-147

Bennett, K. (1989) Mol. Cell. Biol. 9, 35433547