Báo cáo khoa học: Purification and characterization of zebrafish hatching enzyme – an evolutionary aspect of the mechanism of egg envelope digestion pot

The cross-species digestion using enzymes and substrates of zebrafish and medaka revealed that both ZHE1 and MHCE cleaved the same sites of the egg envelope proteins of two species, sugge

enzyme – an evolutionary aspect of the mechanism of

egg envelope digestion

Kaori Sano1, Keiji Inohaya2, Mari Kawaguchi3, Norio Yoshizaki4, Ichiro Iuchi5and

Shigeki Yasumasu5

1 Graduate Program of Biological Science, Graduate School of Science and Technology, Sophia University, Tokyo, Japan

2 Department of Biological Information, Tokyo Institute of Technology, Yokohama, Japan

3 Ocean Reseach Institute, The University of Tokyo, Japan

4 Department of Biological Diversity, Faculty of Agriculture, Gifu University, Japan

5 Department of Materials and Life Sciences, Faculty of Science and Technology, Sophia University, Tokyo, Japan

Hatching enzyme is an enzyme that digests an egg

envelope at the time of embryo hatching Fish

hatch-ing enzymes have been purified from several fish

species [1–5] Among them, the hatching enzyme of

medaka Oryzias latipes has been extensively studied, and its study field was extended not only to character-ization of the enzyme itself, but also to the mechanism

of its egg envelope digestion [6,7] The hatching of

Keywords

astacin family; egg envelope; hatching

enzyme; molecular evolution

Correspondence

S Yasumasu, Department of Materials and

Life Sciences, Faculty of Science and

Technology, Sophia University, 7-1 Kioi-cho,

Chiyoda-ku, Tokyo 102-8554, Japan

Fax / Tel: +81 3 3238 3393

E-mail: s-yasuma@hoffman.cc.sophia.ac.jp

(Received 17 June 2008, revised

22 September 2008, accepted 2

October 2008)

doi:10.1111/j.1742-4658.2008.06722.x

There are two hatching enzyme homologues in the zebrafish genome: zebrafish hatching enzyme ZHE1 and ZHE2 Northern blot and RT-PCR analysis revealed that ZHE1 was mainly expressed in pre-hatching embryos, whereas ZHE2 was rarely expressed This was consistent with the results obtained in an experiment conducted at the protein level, which demonstrated that one kind of hatching enzyme, ZHE1, was able to be purified from the hatching liquid Therefore, the hatching of zebrafish embryo is performed by a single enzyme, different from the finding that the medaka hatching enzyme is an enzyme system composed of two enzymes, medaka high choriolytic enzyme (MHCE) and medaka low chorio-lytic enzyme (MLCE), which cooperatively digest the egg envelope The six ZHE1-cleaving sites were located in the N-terminal regions of egg envelope subunit proteins, ZP2 and ZP3, but not in the internal regions, such as the

ZP domains The digestion manner of ZHE1 appears to be highly analo-gous to that of MHCE, which partially digests the egg envelope and swells the envelope The cross-species digestion using enzymes and substrates of zebrafish and medaka revealed that both ZHE1 and MHCE cleaved the same sites of the egg envelope proteins of two species, suggesting that the substrate specificity of ZHE1 is quite similar to that of MHCE However, MLCE did not show such similarity Because HCE and LCE are the result

of gene duplication in the evolutionary pathway of Teleostei, the present study suggests that ZHE1 and MHCE maintain the character of an ancestral hatching enzyme, and that MLCE acquires a new function, such

as promoting the complete digestion of the egg envelope swollen by MHCE

Abbreviations

MCA, 7-amino-4-methylcoumarin; MHCE, medaka high choriolytic enzyme; MLCE, medaka low choriolytic enzyme; ZHE, zebrafish hatching enzyme; ZPD, ZP domain.

medaka embryos is performed by two enzymes, high

choriolytic enzyme, choriolysin H (HCE; EC

3.4.24.67) and low choriolytic enzyme, choriolysin L

(LCE; EC 3.4.24.66), which cooperatively digest egg

envelope Two enzymes have been separately purified

from hatching liquid [4,5] HCE swells the egg

enve-lope by its limited proteolytic action, whereas LCE

efficiently digests the HCE-swollen envelope and

solu-bilizes it completely We have named this digesting

system the ‘HCE-LCE system’ cDNA cloning analysis

revealed that both enzymes belong to the astacin

fam-ily and comprise 200 amino acid residues in mature

enzyme portions with 55% identity in amino acid

sequence [8] In addition, two hatching enzymes have

been purified from killifish Fundulus heteroclitus

embryos, and hatching has been demonstrated to be

performed by the HCE-LCE system [9] Two types of

enzymes homologous to medaka HCE (MHCE) and

medaka LCE (MLCE) were cloned from other

euteleostean fishes, such as fugu Takifugu rubripes,

spotted green pufferfish Tetraodon nigroviridis and ayu

Plecoglossus altivelis altivelis [10] Thus, the hatching

of euteleostean fishes can be performed by the

HCE-LCE system

According to the phylogenetic tree based on the

mitochondorial DNA of Teleostei, Osteoglossomorpha

first branched off from an ancestor, followed by

Elopomorpha, and then branched paraphyletically to

Otocephala and Euteleostei [11–14] The cDNA

cloning analysis using Japanese eel Anguilla japonica

belonging to Elopomorpha revealed that several

hatch-ing enzyme cDNAs were present, their nucleotide

sequences were similar to each other, and all formed a

monophyletic clade in the phylogenetic tree of fish

hatching enzymes [15] These results suggest that the

hatching of eel embryos is performed by a single type

of enzyme Therefore, HCE and LCE are considered

to have been produced by a gene duplication event

after Elopomorpha had diverged [16]

At present, and in contrast with such genetic

information, information at the protein level is

restricted to euteleostean fishes and is not available

for fishes belonging to Elopomorpha and

Otocepha-la In the present study, we purified hatching enzyme

from embryos of zebrafish Danio rerio belonging to

Cypriniformes in Otocephala, analyzed the

mecha-nism of its egg envelope digestion and compared it

with that of medaka hatching enzyme Finally, the

evolution of the mechanism of egg envelope

diges-tion is discussed on the basis of the manner of the

reciprocal or cross-species egg envelope digestion

using enzymes and substrates of both species:

zebra-fish and medaka

Results Expression of zebrafish hatching enzyme ZHE1 and ZHE2 genes

It has been reported that two cDNAs, ZHE1 and ZHE2, are cloned from the RNA of prehatching embryos [17] According to the zebrafish genome pro-ject, three orthologues, ZHE1a, ZHE1b and ZHE2, were clustered in the genome [16] The amino acid sequence of ZHE1a is 99% identical to that of ZHE1b, and 60.8% identical to that of ZHE2 Therefore, we considered that two types of hatching enzyme genes are present in the zebrafish genome

First, we observed the expression of ZHE1 and ZHE2 genes by northern blot analysis (Fig 1A) The expression of ZHE1 was detected in embryos at 11.5 h, and a strong signal was observed at 24 h After hatching, no expression was observed The size of the band ( 1 kbp) was in agreement with that deduced from ZHE1 cDNA (800 bp) By contrast, no signal for ZHE2 gene expression was detected at any of the developmental stages (Fig 1A)

Next, RT-PCR, a method more sensitive than nor-thern blot analysis, was used to detect expression using RNA of 24 h embryos An amplified band for ZHE1 transcript became visible at the 19th cycle of PCR, whereas only a faint band for the ZHE2 transcript was observed at the 28th cycle (Fig 1B) The result suggests that the amount of ZHE1 transcript is quite different from that of ZHE2 The amount of ZHE2 transcript is considered to be much lower than that of ZHE1 Taken together with the results of the northern blot analysis, the ZHE2 gene in developing embryos is considered to be expressed to a very small extent Finally, the expression of ZHE genes was observed

by whole mount in situ hybridization ZHE1 gene expression was observed in the hatching gland cells located on the yolk sac of 24 h embryos (Fig 1C) However, no positive signal for ZHE2 was observed in the same staining condition Several more days of incubation with a substrate solution showed only a weak ZHE2 signal in the cells (Fig 1C) Thus, the ZHE1 gene, and not the ZHE2 gene, is predominantly expressed in zebrafish embryos

Purification of ZHE The hatching liquid (i.e culture medium after embryos hatched out) was used to purify ZHE First, we applied the concentrated hatching liquid onto a Super-dex 75 10⁄ 300 GL column in the HPLC system (Fig 2A) Most of the protein was eluted just after the

void volume, and a single peak of caseinolytic activity

was eluted near the bed volume After dialysis against

a 25 mm Tris–HCl buffer (pH 7.5), the fraction having

caseinolytic activity was applied onto a Source 15S

column in the HPLC system and eluted with a linear

gradient of 0–1 m NaCl (Fig 2B) Most of the activity

was retained in the column and eluted at the

concen-tration of approximately 0.35 m NaCl as a sharp single

peak SDS⁄ PAGE of the active fraction gave a single

band, with an estimated molecular mass of 23 kDa

(Fig 3) A partial amino acid sequence from the

N-ter-minus of the 23 kDa protein was NALIXE, which

matched with the sequence from the N-terminus of

mature protein portion deduced from ZHE1 cDNA,

but not from ZHE2 Thus, a single enzyme, ZHE1, was contained in the hatching liquid This is consistent with the results of the gene expression analysis: ZHE1

0 0.01

0.2

0.1

Time (min)

0.02

0

1

A280

A280

0.5

0

Time (min)

0.1 0.2 0.3

0.2 0.4 0.6 0.8

B

A

A280

A280

Fig 2 Purification of the hatching enzyme of zebrafish (A) Super-dex 75 10 ⁄ 300 GL column chromatogram of hatching liquid The solid line indicates A280 and the dotted line shows caseinolytic activity (B) Source-15S column chromatogram of the caseinolytic active fractions obtained from the Superdex 75 10 ⁄ 300 GL column chromatograqphy The sample was eluted with a line gradient from 0–1 M NaCl (broken line) The solid line indicates the protein amount measured at A280 The dotted line indicates the caseinolytic activity.

Fig 3 SDS ⁄ PAGE patterns of rec ZHE1 (lane 1) and purified ZHE1 (lane 2) The gel was stained with silver Numbers on the left refer to the sizes of molecular markers.

A

B

C

Fig 1 Expression analysis of ZHE1 and ZHE2 genes (A) The result

of the northern blot analysis probed with ZHE1 and ZHE2 cDNAs.

Total RNAs were isolated from 11.5 h embryos (lane 1), 24 h

embryos (lane 2) or embryos after hatching (lane 3) Arrowheads

indicate the positions of 28S and 18S rRNA (B) Expression of

ZHE1 and ZHE2 was analyzed by RT-PCR using RNA isolated from

24 h embryos b-actin was used as a control (C) The result of

whole mount in situ hybridization probed with ZHE1 and ZHE2

cDNAs The color precipitation was developed for 2 h (ZHE1) or

several days (ZHE2) Arrowheads in (C) indicate positive signals

observed in hatching gland cells Scale bars = 200 lm.

is mainly expressed in the developing embryo, but very

little ZHE2 is expressed

Generation of recombinant ZHE1

Recombinant ZHE1 (rec ZHE1) was generated by an

Escherichia coli expression system using pET3c as an

expression vector, and the active enzyme was obtained

by the astacin-refolding method with slight

modifica-tions [18] The specific caseinolytic activity of rec ZHE1

(900 min)1Æmg)1 protein) was higher than that of

purified medaka hatching enzymes (800 min)1Æmg)1

protein for MHCE, 540 min)1Æmg)1protein for MLCE)

[4,5] The result suggests that almost all rec ZHE1

molecules were correctly refolded and had activity By

contrast, rec ZHE2 failed in the refolding

rec ZHE1 was completely inhibited by 1 mm

EDTA, but not by 10 mm diisopropylfluorophosphate

or 10 mm iodoacetic acid, consistent with the fact that

fish hatching enzymes generally belong to the

metallo-protease family The substrate specificity of rec ZHE1

was determined using various

7-amino-4-methylcou-marin (MCA) peptides (Table 1) ZHE1 cleaved the

peptide bonds at the C-terminal side of Arg, Tyr, Asn,

Trp, Ala, Asp, Phe and Gly, suggesting that ZHE1 has

broad substrate specificity One of the most suitable

substrates was Z-Phe-Arg-MCA, and the specific

activ-ity was 27.02 nmolÆ30 min)1Æmg)1 protein The

sub-strate specificity of rec ZHE1 was similar to that of

the protease contained in hatching liquid This result

supported the findings of the purification indicating that only a single enzyme, ZHE1, was contained in hatching liquid

Changes of fertilized egg envelopes treated with recombinant ZHE1

Figure 4B shows an egg envelope after hatching At the natural hatching of zebrafish embryo, the egg enve-lope was not completely solubilized, but was softened and ruptured by the contractile movement of the embryo When isolated egg envelopes were incubated with rec ZHE1, no marked structural changes could

be observed under a binocular microscope (Fig 4C) Using electron microscopy, we observed changes of the fine structure of envelope Figure 4D shows the structure of an intact egg envelope, which was composed

of a thick inner layer and a thin outer layer The inner layer comprised a lamellar structure with microvillous

Table 1 The specific activity of rec ZHE1 examined by various

MCA substrates The activity of hatching liquid was normalized by

caseinolytic activity per 1 lg of rec ZHE1 ND, not detected.

MCA substrate

Specific activity (nmolÆ30 min)1Ælg)1enzyme)

rec ZHE1

Hatching liquid

A

D

B

E

Fig 4 Morphological changes of egg envelope The isolated egg envelope of zebrafish before hatching (A), after hatching (B) and digested by rec ZHE1 (C) were observed using a binocular micro-scope Scale bars = 200 lm (D–F) Electron microscopic observa-tion of zebrafish egg envelopes Secobserva-tions of the egg envelope isolated from pre-hatching embryo (D), the egg envelope after hatching (E) and the egg envelope digested by rec ZHE1 (F) Arrowheads indicate outer layers Scale bars = 1 lm.

channels After incubation with rec ZHE1, the fibrous

structure of the inner layer became evident, and its

thickness was increased two-fold more than that of the

intact envelope Figure 4E shows an egg envelope after

natural hatching Its fine structure was similar to that of

the egg envelope incubated with rec ZHE1 (Fig 4F)

Taken together with the result of the purification, the

single enzyme, ZHE1, is suggested to act on the egg

envelope at the time of natural hatching

Digestion of unfertilized egg envelope by ZHE1

It is well known that the egg envelope becomes

hard-ened after fertilization The hardening of the envelope

is considered to be achieved by the polymerization of

egg envelope subunits The polymerization is due to

the formation of e-(c-glutamyl) lysine isopeptide

cross-links by transglutaminase [19–21] Such cross-cross-links

make it difficult to clearly determine the sites of egg

envelope cleaved by ZHE1 Therefore, initially, an

unfertilized egg envelope was used as a substrate

The zebrafish egg envelope is known to be mainly

constructed by two glycoproteins, ZP2 (44 kDa) and

ZP3 (49 kDa), which were visualized by the

SDS⁄ PAGE analysis of unfertilized egg envelopes

(Fig 5, lane 1) The isolated unfertilized egg envelopes

were digested by rec ZHE1 and analyzed by

SDS⁄ PAGE After incubation for 2 min, bands with

molecular masses of 43 and 39 kDa were observed in

addition to undigested bands of ZP2 and ZP3 After

incubation for 10 min, three major bands with

molecu-lar masses of 43, 39 and 36.5 kDa were observed

(Fig 5, lane 2) After further incubation (60 min), the

39 kDa band disappeared, and only two bands with the same mobility as the 43 and 36.5 kDa products were detected (Fig 5, lane 3) These results indicate that the 43 and 39 kDa products appear first, and the

39 kDa product is then further digested and shifted to the 36.5 kDa product

To determine rec ZHE1-cleaving sites in egg enve-lope subunits, we analyzed the sequence of each prod-uct from its N-terminus The sequences were compared with those deduced from ZP2 and ZP3 cDNA [22,23]

It is well known that ZP2 is composed of an N-terminal region (95 amino acids), an internal trefoil domain and a C-terminal ZP domain ( 260 amino acids) including eight consensus Cys residues (Fig 6A) ZP3 is composed of an N-terminal region (45 amino acids) and a C-terminal ZP domain (Fig 6B)

The detected N-terminal amino acid sequence of the

43 kDa product was APEPFT, which matched with the sequence from Ala80 of ZP3, and we therefore deduced that the cleaving site is Gln79⁄ Ala80 (Fig 6,

Fig 6 Amino acid sequences of ZP2 and ZP3 deduced from their cDNAs The arrows and capital letters (sites A to F) indicate the cleaving sites of ZHE1 determined from the N-terminal amino acid sequences of the ZHE1 digests Circled Q indicates a glutamine residue that is presumed to form a e-(c-glutamyl) lysine cross-link

by the sequence analysis ZP domains and trifoiled domain are indi-cated in light gray and dark gray boxes, respectively Predicted N-glycosylation site is underlined Black and white triangles indicate putative signal sequence cleaving sites and predicted C-terminal processing sites, respectively.

Fig 5 SDS ⁄ PAGE patterns of unfertilized egg envelopes digested

by rec ZHE1 The envelopes isolated from unfertilized egg of

zebrafish (lane 1) were incubated with rec ZHE1 for 2 min (lane 2),

10 min (lane 3) and 40 min (lane 4) Numbers on the right show

the molecular masses of the major bands.

site E) The molecular mass of the 43 kDa product

was somewhat larger than the molecular mass

pre-dicted from ZP3 cDNA (39 070.60; from Ala80 to

lle431; Fig 6) Because the amino acid sequence of

ZP3 contains one of the consensus sequences for

N-glycosylation site, such a difference is considered to

be due to the existence of a sugar chain The

N-termi-nal amino acid sequences of the 39 and 36.5 kDa

prod-ucts were DYLIKEIVQP and VEEVVVK, respectively,

and these matched with the sequences from Asp48 and

Val67 deduced from ZP2 cDNA, respectively

There-fore, the cleaving sites are Ser47⁄ Asp48 and Arg66 ⁄

Val67 (Fig 6, sites A and B) The molecular masses of

the 39 and 36.5 kDa products of ZP2 were consistent

with those calculated from ZP2 cDNA (39 107.06 from

Asp48 to Arg405; 36 902.50 from Val67 to Arg405)

Digestion of fertilized egg envelope by ZHE1

Next, a fertilized egg envelope was digested by

rec ZHE1 As a control, the SDS extract of intact egg

envelopes was analyzed by SDS⁄ PAGE Several bands

with a mobility that did not correspond to that of ZP2

and ZP3 were observed (Fig 7, lane 1) These bands

are considered to be proteins that are secreted from

cortical alveoli and adhere to the envelope at

fertiliza-tion The hardened, fertilized egg envelopes are not

considered to be solubilized by SDS

The egg envelope is not soluble but became swollen

by rec ZHE1 This swollen envelope was dissolved

into SDS and analyzed by SDS⁄ PAGE SDS ⁄ PAGE

of the fertilized egg envelope digested by rec ZHE1 for 50 min gave three bands (150, 43 and 36.5 kDa), which were not found in the control (Fig 7, lane 2) The N-terminal amino acid sequence of the 43 kDa product (APEPFT) was identical to that of the 43 kDa product obtained in the unfertilized egg envelope digestion, suggesting that rec ZHE1 cleaves the com-mon sites of unfertilized and fertilized egg envelope (Fig 6, site E)

Amino acid sequence analysis of the 36.5 kDa product revealed that this sequence was a mixture of two peptide sequences One of them was VEEV-VVKAGPVDK and matched that of the 36.5 kDa product from ZP2 in unfertilized egg envelope digests (Fig 6, site B) The other sequence, APLDLXE, did not correspond to any cleaving site obtained in the unfertilized egg envelope digestion However, this sequence was found in the sequence from Ala68 of ZP3 (Gln67⁄ Ala68; Fig 6, site D) This cleaving site was located 12 amino acid residues upstream from the cleaving site obtained from the 43 kDa product of ZP3 (site E) Therefore, the finding that the 36.5 kDa prod-uct is a mixture of two peptides from ZP2 and ZP3 suggested that the 36.5 kDa product of ZP2 obtained from unfertilized digest binds the 12 amino acid resi-dues fragment (from site D to site E) of ZP3 via an e-(c-glutamyl) lysine cross-link

Further analysis revealed that 150 kDa product also contained two amino acid sequences identical to those

of the 36.5 kDa product, VEEVVVKAGPVDK and APLDLXE The sequence APLDLXE is quite similar

to the sequence APLDLQE of ZP3 deduced from cDNA However, the sixth glutamine residue (Q of APLDLQE; Fig 6, circle) was not detected in sequenc-ing of the 36.5 and 150 kDa product by Edman degra-dation There is evidence that Edman degradation did not release amino acid residues at the e-(c-glutamyl) lysine cross-linked position [24] Although further investigation is necessary, we conclude that Gln73 in ZP3 is one of the glutamine acceptor sites for e-(c-glut-amyl) lysine cross-link formation The lysine donor site presumed to exist in the ZP2 sequence of the 36.5 and

150 kDa product was not determined in the present study The 150 kDa proteins disappeared after further digestion (90 min; Fig 7, lane 3), and this analysis identified a new cleaving site, Gln79⁄ Ala80 peptide bond (Fig 6, site E) in ZP3, which is identical to the site found in the 43 kDa product Therefore, this digestion probably resulted in further digestion of the

150 kDa products into the 36.5 and 43 kDa product All the results suggest that rec ZHE1 cleaves the N-terminal portions of ZP2 and ZP3 and eliminates

Fig 7 SDS ⁄ PAGE patterns of fertilized egg envelopes Lane 1,

SDS extract of intact fertilized egg envelopes; lanes 2 and 3, SDS

extract of egg envelopes digested by rec ZHE1 for 50 and 90 min,

respectively; lane 4, SDS extract of egg envelopes after hatching.

Numbers on the left show the molecular masses of the major

bands.

the tight bindings between subunits by cleaving out the

small portion of peptides including e-(c-glutamyl)

lysine isopeptide cross-links

We compared the cleaving sites determined from egg

envelopes after natural hatching (post-hatching) with

those artificially digested by rec ZHE1 As shown in

Fig 7 (lane 4), the SDS⁄ PAGE pattern of the proteins

of the post-hatching egg envelopes was similar to that

digested by rec ZHE1 for 90 min The 43 kDa band

became weaker than that of the 50 min incubation,

suggesting further digestion of the 43 kDa product

Furthermore, their 36.5 kDa bands became broader

than that for the 50 min incubation The sequence

analyses of the 36.5 kDa bands revealed that two

sequences of further digested products were detected,

in addition to those of 36.5 kDa product obtained

from the 50 min incubation One was a minor

sequence, QPASPG, which was found to locate at

Q106 of ZP3; the cleaving site is K105⁄ Q106 (Fig 6,

site F), suggesting that the 43 kDa product of ZP3 is

further digested and decreases its molecular mass to

approximately 36.5 kDa The other was AGPVDK

(from Ala74 of ZP2; Fig 6, site C), which was shifted

seven amino acid residues to the C-terminal from the

site B Thus, the cleaving sites obtained from the

90 min incubation and the post-hatching egg envelopes

are considered to contain the sites that can be cleaved,

although inefficiently, by ZHE1 Considering that the

perivitelline space where hatching enzyme is secreted is

only a small area, a rather considerably high

concen-tration of ZHE1 appears to act on egg envelope, and

therefore the ZHE1-cleaving sites at natural hatching

are suggested to include not only its preferred sites,

but also inefficient cleaving sites for ZHE1

Specific activity of ZHE1 judged by synthetic

peptide substrates

The cleaving efficiency of ZHE1 was quantitatively

estimated with synthetic peptide substrates that were

designed from the determined ZHE1-cleaving sites

The specific activities of rec ZHE1 toward five

pep-tides (Fig 6, sites A, B, C, D and E) were determined

The most efficient substrate was site A peptide and the

second most efficient was site E peptide (Table 2) Sites

A and E corresponded to the N-termini of the 39 kDa

product of ZP2 and the 43 kDa product of ZP3

observed in the 2 min ZHE1 digestion of unfertilized

egg envelopes, respectively By contrast, the specific

activities toward site B and D peptides were much

lower than those toward the former two (5.86% and

2.47% of site A peptide, respectively) Therefore, these

values well reflected the results of the egg envelope

digestion experiment According to the fertilized egg envelope digestion experiment, the e-(c-glutamyl) lysine isopeptide cross-links formed between ZP2 and ZP3 subunits are considered to be eliminated by cleaving of site E Thus, site E is conjectured to be a key cleaving site leading to a conformational change that results in swelling of the egg envelope Therefore, it is reasonable

to consider that site E is one of the efficient cleaving sites for ZHE1 The cleaving activity at site C, which was detected in the 90 min digestion of the fertilized egg envelope and considered as an inefficient cleaving site, was not easily detected in this condition

Species specificity of digestion by hatching enzyme

As described earlier, zebrafish hatching is performed

by a single enzyme, ZHE1 Different from zebrafish, hatching of medaka is performed by a two enzyme sys-tem To obtain an evolutionary aspect of the mecha-nism of egg envelope digestion by hatching enzyme, we changed the substrate–enzyme combination between zebrafish and medaka and performed the cross-species digestion experiment using unfertilized egg envelopes

as substrate First, the unfertilized egg envelopes of zebrafish were digested either by ZHE1, MHCE or MLCE, and their SDS⁄ PAGE patterns were com-pared SDS⁄ PAGE of the MHCE digest after a 10 and

40 min incubation gave two bands (43 and 39 kDa), and an additional band (36.5 kDa) was observed after

a 120 min incubation (Fig 8A, lanes 3–5) These corre-sponded to three bands obtained from the ZHE1 digest after a 10 min incubation (Fig 8A, lane 2) The N-terminal sequence analyses of three digests revealed that each of the MHCE-cleaving sites on zebrafish egg envelope was the same as the three ZHE1-cleaving

Table 2 The specific activity of ZHE1, MHCE and MLCE examined

by synthetic peptide substrates The cleaving site of each peptide

is indicated by an arrow ND, not detected.

Peptide

Specific activity (nmolÆ30 min)1Ælg)1enzyme)

sites (Fig 6, sites A, E and B) The results suggest that

ZHE1 and MHCE have the same substrate specificity

toward zebrafish unfertilized egg envelopes, although

the digestion efficiency of MHCE appears to be

some-what lower than that of ZHE1 By contrast, MLCE

digested the zebrafish egg envelopes and produced two

bands, the mobilities of which corresponded to 43 and

39 kDa ZHE1 digests (Fig 8A, lanes 6–8) However,

the cleaving efficiency of MLCE appears to be less

than that of MHCE because a considerable amount of

undigested bands remained in the digest after a 10 min

incubation, and the 36.5 kDa band was not easily

detected even after a 120 min incubation The sequence

analyses revealed that the N-terminal sequence of

43 kDa product digested by MLCE was identical to

that of site E cleaved by ZHE1 (Fig 6) However, the

site in the 39 kDa product was not identical to the

ZHE1-cleaving site of the 39 kDa product (site A),

and the cleaving site was shifted one amino acid

resi-due to the N-terminal side from site A Therefore, the

specificity of MLCE toward zebrafish egg envelope is

similar to, but somewhat different from, that of

ZHE1

By contrast, the medaka unfertilized egg envelope

was digested by ZHE1 The SDS⁄ PAGE pattern was

similar to that by MHCE (Fig 8B, lanes 10 and 11)

The N-terminal sequences of the 47 and 35 kDa

prod-ucts in ZHE1 digest matched with those of MHCE

digests, suggesting that ZHE1-cleaving specificity

toward medaka unfertilized egg envelope is similar to

that of MHCE, and not MLCE (Fig 8B, lane 12)

Comparison of specific activities of ZHE1, MHCE and MLCE judged by synthetic peptide substrates

Cleaving efficiencies of ZHE1, MHCE and MLCE were quantitatively estimated using synthetic peptide substrates (Fig 9) For the zebrafish egg envelope, the peptides designed from sites A, B and E (Fig 6) were employed As mentioned earlier, the best substrate for ZHE1 was a site A peptide and the second best was a site E peptide, and the specific activity toward the site

B peptide is lower than one tenth of that toward the site A peptide In respective peptide substrates, the values of the specific activity of MHCE were similar

to those of ZHE1

By contrast, the specific activity of MLCE was much lower than those of ZHE1 and MHCE As was true in the egg envelope digestion experiment, the cleaving sites of site A peptides of MLCE did not coincide with those of ZHE1 and MHCE However, the ratios of the specific activities of MLCE toward the three substrates were similar to that of ZHE1 and MHCE In summary, the substrate specificity of MHCE toward peptides for zebrafish egg envelopes is quite similar to that of ZHE1, whereas that of MLCE is similar to a certain extent

The medaka egg envelope is known to consist of the subunits proteins having a ZP domain (i.e ZI-1,2 and ZI-3) that are homologous to zebrafish ZP2 and ZP3, respectively [25,26] One of the obvious differences of the subunit protein between medaka and zebrafish is

Fig 8 Cross-species digestion using hatching enzyme and the egg envelope of zebrafish and medaka (A) Zebrafish unfertilized egg enve-lopes (lane 1) were incubated with either ZHE1 (lane 2), MHCE (lanes 3–5) or MLCE (lanes 6–8) After incubation for 10 min (lanes 2, 3 and 6), 40 min (lanes 4 and 7) and 120 min (lanes 5 and 8), each digest was separated by SDS ⁄ PAGE (B) Medaka unfertilized egg envelopes (lane 9) were incubated with either ZHE1 (lane 10), MHCE (lane 11) or MLCE (lane 12) for 15 min Each digest was separated by SDS ⁄ PAGE Numbers on the left are the sizes (kDa) of molecular markers, and those on the right are the molecular masses for major bands.

that ZI1,2 possesses Pro-X-Y repeat sequences in its

N-terminal region, which are not found in that of

zebrafish ZP2 MHCE and MLCE are known to have

different cleaving specificity toward the medaka egg

envelope MHCE mainly cleaves Pro-X-Y repeat

sequences present in the N-terminal region of ZI-1,2

[7] By contrast, the most efficient cleaving site of

MLCE is in the center of the ZP domain of ZI-1,2

Therefore, two peptide substrates were designed from the amino acid sequences of Pro-X-Y repeat region, named Pro-X-Y-1 and Pro-X-Y-2, for MHCE, and another peptide substrate was designed from the amino acid sequence of ZP domain, named ZP domain (ZPD)-center for MLCE As we expected, MHCE effi-ciently cleaved Pro-X-Y-1 and Pro-X-Y-2, whereas MLCE efficiently cleaved ZPD center (Fig 9) In addi-tion, ZHE1 well cleaved Pro-X-Y-1 and Pro-X-Y-2, the sites for MHCE, and their specific activity values were approximately 80% of those of MHCE, whereas ZHE1 did not easily cleave the ZPD-center, the site for MLCE The results suggest that ZHE1 has the MHCE-like activity toward medaka egg envelope; this was consistent with the results obtained in the diges-tion experiment using unfertilized egg envelopes It is interesting to note that ZHE1 cleaves Pro-X-Y sequences that are not present in the subunit proteins

of zebrafish egg envelope

Around the cleaving sites of ZHE1 and MHCE, we were unable to find a common or consensus amino acid sequence between zebrafish and medaka This is sup-ported by the finding that ZHE1 has broad substrate specificity, as judged by the MCA substrate experiment

Discussion Gene expression analyses revealed that ZHE1, one of two zebrafish hatching enzyme genes, was mainly expressed, whereas ZHE2 was scarcely expressed This was supported by the result that only a single enzyme, ZHE1, was purified from hatching liquid In addition, the fine morphology of fertilized egg envelope digested

by rec ZHE1 was similar to that after natural hatch-ing Thus, only one enzyme, ZHE1, is suggested to be essential for hatching of zebrafish embryo, and ZHE2 does not contribute to the hatching

We have suggested that the ZHE1 and ZHE2 genes were produced by gene duplication and subsequent diversification during the evolutionary process to zebrafish [16] The whole mount in situ hybridization revealed that ZHE2 transcript was expressed specifi-cally, but weakly, in the hatching gland cells At an earlier period of evolution, ZHE2 is inferred to have worked as a hatching enzyme and to have lost its abil-ity of egg envelope digestion during its further evolu-tionary process, namely a mutation(s) in the amino acid sequence of ZHE2 changed its substrate specificity

as a proteolytic enzyme and, eventually, ZHE2 become uninvolved in the egg envelope digestion Subse-quently, the amount of its expression would be decreased by accumulation of a mutation(s) in a regu-latory region(s) responsible for gene expression

100

50

0

100

50

0

0

25

50

0

25

50

0

25

50

10

5

0

Site-A

Site-B

Site-E Pro

XY-1

Pro XY-2

ZPD -Center

Pro XY-1

Pro XY-2

ZPD -Center

Pro XY-1

Pro XY-2

ZPD -Center

Site-A

Site-B

Site-E

Site-A

Site-B

Site-E

ZHE1

MHCE

MLCE

Fig 9 Specific activity of ZHE1, MHCE and MLCE examined by

synthetic peptide substrates Names of the synthetic peptides are

indicated at the bottom of the figures Sites A, B and C indicate

the ZHE1-cleaving sites on the zebrafish egg envelope Pro XY-1

and Pro XY-2 indicate MHCE-cleaving sites, whereas ZPD-center is

the MLCE-cleaving site on the medaka egg envelope.

According to molecular phylogenetic analysis of fish

hatching enzyme genes, hatching enzyme originally

consisted of a single type of enzyme, and HCE and

LCE were produced by duplication and diversification

of the gene [16] As comparing the egg envelope

diges-tion mechanism between zebrafish and medaka, we will

discuss the evolution of hatching enzyme function

In medaka egg envelope digestion, it has been

reported that MHCE mainly cleaves Pro-X-Y repeat

sequences located at the N-terminal region of ZI1,2

and releases small peptides containing most of

the e-(c-glutamyl) lysine isopeptide cross-links [7] The

present study revealed that ZHE1 also cleaved the

N-terminal regions of egg envelope subunits where

most of cross-links are located, and swelled the egg

envelope Therefore, the manner of egg envelope

diges-tion is analogous between ZHE1 and MHCE

The cross-species digestion experiments and the

experiments using synthetic peptide substrates revealed

that ZHE1 and MHCE cleaved the same sites on both

zebrafish and medaka egg envelopes with a similar

effi-ciency ZHE1 swelled the medaka egg envelope but did

not solublize its swollen envelope (data not shown)

Such an agreement is surprising when it is considered

that zebrafish and medaka diverged 140 million years

ago [27] From an evolutionary aspect on egg envelope

digestion, ZHE1 and MHCE are presumed to maintain

the substrate specificity of a common ancestral hatching

enzyme By contrast, the cleaving specificity of MLCE

toward the zebrafish egg envelope was similar to those

of ZHE1 and MHCE, but its cleaving efficiency was

approximately ten-fold lower Furthermore, MLCE had

another efficient cleaving site, the center of ZP domain,

where ZHE1 and MHCE hardly cleave These results

imply that MLCE changed its substrate specificity to

one different from that of an ancestral enzyme, although

its substrate specificity still remains to a certain extent

Therefore, we consider that the single

enzyme-depen-dent swelling of the egg envelope in zebrafish is closely

related to an original or ancestral form of egg envelope

digestion, and the HCE-LCE system comprises a more

developed form HCE and ZHE1 would inherit the

character of the ancestral enzyme with respect to the

swelling of egg envelope After gene duplication and

diversification, LCE would be produced by changing its

substrate specificity and would acquire a new function,

the digestion of the swollen egg envelope

On comparing amino acid sequences between

zebra-fish and medaka egg envelope subunits, we see that the

identity of the ZP domains was approximately 60%

(ZP3⁄ ZI3 = 55%; ZP2 ⁄ ZI1,2 = 65%); however, there

was no similarity in their N-terminal regions in which

the cleaving sites for ZHE1 or MHCE are located

Hatching enzyme recognition sites on the egg envelope are suggested to have changed with a relatively higher substitution rate during evolution By contrast, one of the present studies using MCA substrates showed that ZHE1 had broad substrate specificity MHCE also had broad substrate specificity [4] In addition, some stud-ies report that astacin and meprin A, members of the same astacin family as hatching enzyme, have broad substrate specificity [28,29], suggesting that the sub-strate specificity of proteases belonging to this family

is not so strict Therefore, due to such a character common to the astacin family proteases, fish hatching enzymes could flexibly adapt the changes in amino acid residues around the cleaving sites on the N-termi-nal regions that had a relatively higher substitution rate, and the manner of egg envelope digestion was conserved between ZHE1 and MHCE

During evolution, mutations would be independently generated in the genes of egg envelope and hatching enzyme Some mutations of the two genes would be selected and accumulated under a common pressure with respect to egg envelope digestion Such evolution

of an enzyme and substrate is one typical of the phe-nomena called ‘molecular co-evolution’ Therefore, the cleaving site recognition of both enzymes would be established under a rule that makes it possible to co-evolve hatching enzyme and egg envelope subunit protein To understand such a rule, it is necessary to obtain more information from other fish species, such

as the Japanese eel belonging to Elopomorpha that is sister to the common ancestor of zebrafish and medaka The present study is the first approach aiming

to fully understand the molecular mechanism of co-evolution of hatching enzyme and egg envelope

A further study is now in progress

Experimental procedures Fish

Wild-type embryos of the Ab strain of zebrafish were used Embryos were obtained from natural mating and cultured

were transferred into a beaker with a small amount of

allowed to hatch When 80–90% of the embryo hatched out, the culture medium, now termed hatching liquid, was

Northern blot analysis Ten microgram of total RNA extracted from embryos at 11.5 or 24 h, or after hatching, were electrophoresed on a