Báo cáo khoa học: Conservation of the egg envelope digestion mechanism of hatching enzyme in euteleostean fishes pot

Cleavage sites of FHCE and FLCE on egg envelope subunit proteins were determined by comparing the N-termi-nal amino acid sequences of digests with the sequences deduced from the cDNAs fo

of hatching enzyme in euteleostean fishes

Mari Kawaguchi1,2, Shigeki Yasumasu3, Akio Shimizu4, Kaori Sano5, Ichiro Iuchi3and

Mutsumi Nishida1

1 Atmosphere and Ocean Research Institute, The University of Tokyo, Chiba, Japan

2 Research Fellow of the Japan Society for the Promotion of Science (JSPS), Tokyo, Japan

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

4 National Research Institute of Fisheries Science, Fisheries Research Agency, Yokohama, Japan

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

Introduction

Molecular bases of formation, hardening (conversion)

and breakdown of teleostean egg envelope have been

comprehensively studied in medaka Oryzias latipes as

a model animal [1–3] The egg envelope (chorion)

consists of a major thick inner layer and an extremely

thin outer layer The inner layer is constructed of fibrous macromolecules comprising two groups of sub-unit proteins: ZI-1,2 and ZI-3 [4] ZI-1,2 are heteroge-neous glycoproteins derived from the precursor proteins, choriogenin H (ChgH) and choriogenin H

Keywords

chorion; egg envelope; euteleostean fish;

Fundulus heteroclitus; hatching enzyme;

ZP domain

Correspondence

M Nishida, Atmosphere and Ocean

Research Institute, The University of Tokyo,

5-1-5 Kashiwanoha, Kashiwa,

Chiba 277-8564, Japan

Fax: +81 4 7136 6211

Tel: +81 4 7136 6210

E-mail: mnishida@aori.u-tokyo.ac.jp

Database

The nucleotide sequence data reported in this

paper are available in the EMBL ⁄ GenBank ⁄

DDBJ databases under the accession

numbers AB533328 to AB533330

(Received 6 July 2010, revised 1 October

2010, accepted 6 October 2010)

doi:10.1111/j.1742-4658.2010.07907.x

We purified two hatching enzymes, namely high choriolytic enzyme (HCE;

EC 3.4.24.67) and low choriolytic enzyme (LCE; EC 3.4.24.66), from the hatching liquid of Fundulus heteroclitus, which were named Fundulus HCE (FHCE) and Fundulus LCE (FLCE) FHCE swelled the inner layer of egg envelope, and FLCE completely digested the FHCE-swollen envelope In addition, we cloned three Fundulus cDNAs orthologous to cDNAs for the medaka precursors of egg envelope subunit proteins (i.e choriogenins H, H minor and L) from the female liver Cleavage sites of FHCE and FLCE on egg envelope subunit proteins were determined by comparing the N-termi-nal amino acid sequences of digests with the sequences deduced from the cDNAs for egg envelope subunit proteins FHCE and FLCE cleaved differ-ent sites of the subunit proteins FHCE efficidiffer-ently cleaved the Pro-X-Y repeat regions into tripeptides to dodecapeptides to swell the envelope, whereas FLCE cleaved the inside of the zona pellucida domain, the core structure of egg envelope subunit protein, to completely digest the FHCE-swollen envelope A comparison showed that the positions of hatching enzyme cleavage sites on egg envelope subunit proteins were strictly con-served between Fundulus and medaka Finally, we extended such a compar-ison to three other euteleosts (i.e three-spined stickleback, spotted halibut and rainbow trout) and found that the egg envelope digestion mechanism was well conserved among them During evolution, the egg envelope diges-tion by HCE and LCE orthologs was established in the lineage of eu-teleosts, and the mechanism is suggested to be conserved

Abbreviations

ChgH, choriogenin H; ChgHm, choriogenin H minor; ChgL, choriogenin L; FE, fertilized egg envelope; FHCE, Fundulus HCE; FhChgH,

F heteroclitus ChgH; FhChgHm, F heteroclitus ChgHm; FhChgL, F heteroclitus ChgL; FhZPB, F heteroclitus ZPB; FhZPC, F heteroclitus ZPC; FLCE, Fundulus LCE; HCE, high choriolytic enzyme; LCE, low choriolytic enzyme; TFA, trifluoroacetic acid; UFE, unfertilized egg envelope; ZP, zona pellucida.

minor (ChgHm), which are synthesized in the liver of

spawning female, transported through the blood,

pro-cessed at their C-terminal processing sites, and

assem-bled around the egg [5,6] ZI-3 is a homogeneous

glycoprotein derived from another precursor,

chorioge-nin L (ChgL) [7] All of the subunit proteins contain a

zona pellucida (ZP) domain that is the common

struc-ture in all vertebrate egg envelope proteins [8] ChgH

and ChgHm of medaka (precursors of ZI-1,2) are

clas-sified into ZPB, whereas ChgL (precursor of ZI-3) are

classified into ZPC [9]

The unfertilized egg envelope is soft or fragile After

fertilization, the envelope becomes hard and protects

the embryo from the mechanical and chemical stresses

of the environment Egg envelope hardening in

medaka has been suggested to be a result of the

forma-tion of e-(c-glutamyl) lysine cross-links between

sub-unit proteins of the envelope, mainly between inner

layer subunit proteins [10] At the time of hatching of

the embryo, the inner layer is digested by hatching

enzyme [1] The outer layer that remains undigested is

ruptured by movement of the embryo The breakdown

of the inner layer by the enzyme is responsible for

embryo hatching

Medaka hatching enzyme is composed of two

asta-cin family metalloproteases: high choriolytic enzyme

(HCE; choriolysin H; EC 3.4.24.67) and low

choriolyt-ic enzyme (LCE; choriolysisn L; EC 3.4.24.66) [11,12]

At the time of hatching, the two enzymes act

coopera-tively on envelope: HCE swells the inner layer of

enve-lope and LCE completely digests or solubilizes the

HCE-swollen part of the inner layer A previous study

has revealed that HCE and LCE cleave different sites

on the egg envelope subunit proteins in addition to

one common site [13]

Recently, cDNAs for Fundulus heteroclitus orthologs

of HCE and LCE (Fundulus HCE, FHCE; Fundulus

LCE, FLCE) were cloned, and their gene expression

during development was observed by northern blotting

as well as whole-mount in situ hybridization [14] Their

gene structures and expression patterns conserved

those of medaka In the previous study, we separately

purified two isoforms of FHCE (FHCE1 and FHCE2)

By contrast, FLCE was not fully purified In vitro egg

envelope digestion revealed that both the purified

FHCE1 and FHCE2 swell the egg envelope, and the

partially purified FLCE-fraction has the solubilizing

activity of the FHCE1⁄ 2-swollen egg envelope

There-fore, it has been predicted that the mode of their

proteolytic action toward the envelope is conserved

between Fundulus and medaka

In the present study, we first purified FLCE as a

major band by SDS⁄ PAGE Next, we cloned Fundulus

cDNA orthologs for egg envelope protein precursors, ChgH, ChgHm and ChgL The cleavage sites of FHCE1⁄ 2 or FLCE on the egg envelope proteins were determined, and the amino acid sequences around the sites were compared with those of medaka Finally, we extended the comparison to three other euteleosts: three-spined stickleback, spotted halibut and rainbow trout

Results

Purification of FLCE from Fundulus hatching liquid

The purity of the previously obtained FLCE-fraction was not sufficient to determine FLCE-cleavage sites on egg envelope protein Therefore, we developed a new purification method As shown in Fig 1A, the Toyo-pearl HW-50S column chromatography of ammonium sulfate precipitate from hatching liquid revealed two proteolytically active peaks (fractions I and II) As shown in a previous study [14], fractions I and II con-tained FLCE and FHCE1⁄ 2, respectively We employed fraction I for further purification procedures

Fraction I was applied to an S-Sepharose column, and adsorbed proteins were eluted once with 50 mm Tris-buffer containing 0.4 m NaCl The eluate, named fraction IS, was applied to a Source 15S column Most

of the proteins were adsorbed and fractionated mainly into three peaks, named IS-a, IS-b and IS-c (Fig 1B) When caseinolytic specific activity was examined, the highest activity was observed in IS-c (Fig 1C) MALDI-TOF-MS analysis of IS-c showed a major peak of m⁄ z at 23800.9 and the value was well concor-dant with the molecular weight calculated from FLCE cDNA (MW = 23805.65) SDS⁄ PAGE showed that the densities of IS-a, IS-b and IS-c bands at 23 kDa were comparable with the specific activities of the respective fractions (Fig 1C) To confirm LCE activity for fraction IS-c, the envelopes swollen either by FHCE1 or by FHCE2 were incubated with IS-c and observed by microscopy The swollen envelopes were efficiently solubilized by IS-c (data not shown) Thus,

we concluded that the 23 kDa band in fraction IS-c is FLCE

Figure 1D shows the SDS⁄ PAGE patterns of purified FHCE1, FHCE2 and FLCE The electrophoretic mobil-ity of FLCE was slightly higher than those of FHCE1 and FHCE2, and clearly different from them The caseinolytic specific activities of FHCE1, FHCE2 and FLCE were estimated as 16.0, 12.6 and 14.4

DA280min)1Æmg protein)1, respectively, which were similar to each other and approximately one half of

those of medaka HCE (30.2 DA280min)1Æmg protein)1)

and medaka LCE (24.5 DA280min)1Æmg protein)1)

Fundulus orthologs of choriogenin H, H minor

and L

In medaka, it has been reported that choriogenins,

which are precursors of egg envelope subunit proteins,

are synthesized in the liver under the influence of

estro-gen [15] RNAs extracted from the spawning female

liver of Fundulus were used as a template of RT-PCR

and, finally, three kinds of full-length choriogenin cDNAs were cloned According to the phylogenetic analysis, the three Fundulus cDNAs were separately located in the ChgH, ChgHm and ChgL clades (Fig 2), and therefore named FhChgH (F heteroclitus ChgH), FhChgHm (F heteroclitus ChgHm) and FhChgL (F heteroclitus ChgL) cDNAs, respectively Amino acid sequences deduced from FhChgH, FhChgHm and FhChgL cDNAs are shown in Fig 3, together with medaka orthologs OlChgH, OlChgHm and OlChgL All of them possessed a hydrophobic signal peptide at their N-termini The cleavage site of signal peptidase was deduced to be at Ala26⁄ Gln27 for FhChgH, Ala22⁄ Gln23 for FhChgHm and Ala22 ⁄ Gln23 for FhChgL, according to signalp 3.0 software (http:⁄ ⁄ www.cbs.dtu.dk ⁄ services ⁄ SignalP ⁄ ) FhChgH, FhChgHm and FhChgL all had a ZP domain The tre-foil domain was found at the N-terminal side of the ZP domain of FhChgH and FhChgHm The consensus motif for the processing site, such as Arg-Lys-X-fl-Arg, was found near the C-termini of FhChgHm and FhChgL In medaka, the C-terminal regions from those sites are excised before the assembly of water-soluble precursors into the water-insoluble egg envelope [16] The site of FhChgH was Arg-Lys-Gly-Lys Therefore, each processing site is predicted to be at Gly564⁄ Lys565 for FhChgH, Lys423⁄ Arg424 for FhChgHm and Val400⁄ Arg401 for FhChgL One of the characteristics

of OlChgH and OlChgHm is the presence of Pro-X-Y repeat sequences in their N-terminal regions [5,6] Such

Fig 1 Purification of Fundulus hatching enzyme (A) Toyopearl HW-50S column chromatogram of hatching liquid Solid line, A280; dashed line, caseinolytic activity indicated by A280 (B) Elution pattern of fraction IS, which was obtained from fraction I via an S-Sepharose column,

by cation exchange HPLC with a linear gradient of 0–400 m M NaCl (C) Caseinolytic specific activity (DA 280 min)1Æmg protein)1) of fractions IS-a, IS-b and IS-c, as well as their SDS ⁄ PAGE patterns detected by silver staining (D) SDS ⁄ PAGE patterns of purified FHCE1, FHCE2 and FLCE (fraction IS-c), detected by silver staining Numbers on the left refer to the size (kDa) of the molecular markers.

Fig 2 A phylogenetic tree of the ZP domain of choriogenins The

tree was constructed by the maximum likelihood method using the

nucleotide sequences Numbers at the nodes represent bootstrap

values (shown as percentages) Accession numbers: F heteroclitus

(FhChgH, AB533328; FhChgHm, AB533329; FhChgL, AB533330);

O latipes (OlChgH, D89609; OlChgHm, AB025967; OlChgL,

D38630); Oryzias javanicus (OjChgH, AY913759; OjChgL, AY913760);

Oryzias dancena (OdChgH, EF392363; OdChgL, EF392364); Oryzias

sinensis (OsChgL, AY758411); Cyprinodon variegatus (zona radiata-2

CvZR2, AY598615; zona radiata-3 CvZR3, AY598616).

regions were also found in FhChgH and FhChgHm,

and their repetitive units were YPQQPQ(T⁄ K ⁄ Q)PS

and YP(K⁄ N)PQTPPSKPQ for FhChgH and

YPSKP-QQPQQPQ and YPYPSKP-QQPQQPQ for FhChgHm (Fig 3)

Expression of choriogenin genes

Choriogenin gene expression was observed by northern

blotting (Fig 4) Each FhChgH, FhChgHm and

FhChgL probe was hybridized with two transcripts,

and the sizes were approximately 2 and 5 kb

(FhChgH), 1.6 and 5 kb (FhChgHm) and 1.4 and 5 kb

(FhChgL) The sizes of the smaller bands hybridized

with all of the probe (Fig 4, asterisks) were similar to

those of cloned cDNAs: 2037 bp for FhChgH, 1518 bp

for FhChgHm and 1428 bp for FhChgL Therefore, the

smaller bands represented choriogenin genes Each of

the larger bands, obtained from the three probes, was

assumed to be the choriogenin gene-related RNAs,

such as pre-mRNA for choriogenin genes When

com-paring the choriogenin gene signals, strongest

expres-sion was found in the FhChgL gene, followed by the

FhChgH gene This relationship was similar to that of

medaka (i.e strongest in the OlChgL gene, followed

by the OlChgH gene) (Fig 4) Thus, the relative

expression level of choriogenin genes was conserved

between Fundulus and medaka

Cleavage sites of hatching enzyme on unfertilized

egg envelope

One of the goals of the present study was to

deter-mine the cleavage sites of hatching enzyme on egg

envelope proteins The natural substrate of hatching

enzyme is fertilized egg envelope (FE), as described

in the Introduction FE was digested or solubilized

only by the combined action of two enzymes, and

not by any one of the two enzymes, nor by SDS

Alternatively, unfertilized egg envelope (UFE) was

digested by one of the enzymes FHCE1⁄ 2 or FLCE,

and was solubilized by SDS Therefore, UFE was

first used as a substrate to determine the cleavage

sites

When UFE isolated from Fundulus was solubilized

by SDS and applied onto SDS⁄ PAGE, two major bands were found at molecular masses of 60 and

48 kDa (Fig 5B) This pattern was similar to that of medaka UFE [4,13] The 60 and 48 kDa bands are regarded as Fundulus homologs of ZI-1,2 (ZPB groups

of medaka) and ZI-3 (a ZPC group of medaka), respectively, and were designated as FhZPB (F hetero-clitus ZPB) and FhZPC (F heteroclitus ZPC), respec-tively FhZPB is considered to be a group of egg envelope subunit proteins derived from their precur-sors FhChgH and FhChgHm, and FhZPC is an egg envelope subunit protein derived from its precursor FhChgL

Digests of Fundulus UFE by FHCE1 and those by FHCE2 showed the same SDS⁄ PAGE pattern (data not shown), and were observed at 46, 36 and 32 kDa (Fig 5B) An N-terminal amino acid sequence obtained from the 46 kDa digest was NQQQLQTFK and was found from Asn41 of FhChgL (Table 1) The sequence

Fig 3 Alignment of amino acid sequences of choriogenin H (A), H minor (B) and L (C) of Fundulus (FhChgH, FhChgHm and FhChgL) and medaka (OlChgH, OlChgHm and OlChgL) Identical residues are indicated by asterisks below the sequences, and dashes represent gaps The trefoil domain and ZP domain are shown within dark and light gray boxes, respectively Conserved cysteine residues are highlighted in white with a black background Conserved cysteine residues 1–8 of ZP proteins and additional conserved cysteine residues a and b of ZPB proteins are labeled The black arrowheads and black diamonds are HCE and LCE cleavage sites determined using unfertilized egg enve-lopes, respectively The white diamonds represent the cleavage sites determined using hatching liquid The names of the cleavage sites are shown to the left of the marks Four types of dashed ⁄ dotted lines above the sequences indicate the four types of repeating units found in the Pro-X-Y repeat region of FhChgH and FhChgHm White and black triangles indicate putative signal sequence cleavage sites and process-ing sites, respectively Italicized, underlined letters indicate the consensus C-terminal processprocess-ing site, Arg-Lys-X-fl-Arg.

Fig 4 Northern blot analysis of Fundulus or medaka choriogenin gene expression Gene names are shown at the top Bands show-ing choriogenin genes are indicated by asterisks Numbers on the left refer to the size (base) of the molecular markers Gel images of 28S and 18S rRNA are shown at the bottom.

of the 36 kDa digest, APGVPT, was found from Ala230

residing near the N-terminal side of the trefoil domain

of FhChgH (Table 1) Two bands were observed at

32 kDa However, only a single sequence was obtained

from such two bands, and the sequence, TPTET, was

found from Thr77 residing at the N-terminal of the

trefoil domain of FhChgHm (Table 1) The FHCE1⁄ 2

cleavage sites on FhChgH and FhChgL thus determined

were located at positions similar to those of the cleavage

sites of medaka HCE (Fig 3) [13]

The digests by the mixture of FHCE1⁄ 2 and FLCE were observed at 60, 46, 38, 32 and 17 kDa (Fig 5B) The 60 kDa band was that of undigested FhZPB, and the 46 and 32 kDa bands were those of FHCE1⁄ 2-digests We could determine two FLCE sites that were not found in FHCE1⁄ 2 sites An amino acid sequence obtained from the 38 kDa digest was YPVPAATVA and matched the sequence from Tyr74 of FhChgL (Table 1) The N-terminal sequence of the 17 kDa pro-duct was a mixture of two propro-ducts By comparison

Fig 5 The egg envelope digestion pattern of hatching enzyme (A) Schematic presentation of the egg envelope digestion processes by FHCE1 ⁄ 2 and FLCE, together with the respective morphological changes of the fertilized egg envelope of Fundulus FhZPB is the Fundulus ZPB ortholog derived from FhChgH and FhChgHm, and FhZPC is the Fundulus ZPC ortholog derived from FhChgL Black, dark gray and light gray boxes indicate the Pro-X-Y repeat region, trefoil domain and ZP domain, respectively Arrowheads indicate the cleavage sites of FHCE1 ⁄ 2 or FLCE The length between the two arrows in the images indicates the thickness of the egg envelope Scale bar = 0.1 mm (B) SDS ⁄ PAGE pattern of envelopes isolated from Fundulus unfertilized egg (as a control), digests of the envelope by FHCE2, the digests by the mixture of the FHCE1 ⁄ 2 and FLCE, and the digests in Fundulus hatching liquid Numbers on the left refer to the size of the molecular markers (kDa).

Table 1 Cleavage sites of hatching enzymes on egg envelope determined after the incubation of unfertilized egg envelope with FHCE1, FHCE2 or the mixture of FHCE1 ⁄ 2 and FLCE The sites found at natural hatching were determined using hatching liquid.

with the amino acid sequence deduced from cDNA, they

were predicted to be the sequences from Asn395 residing

inside of the ZP domain of FhChgH

(NPPPAVAELG-PIRVA) and from Ala230 located near the N-terminal

side of the trefoil domain of FhChgH

(AP-GVPTPKSxDVEVA) (Table 1) The latter sequence

corresponded to that of the 36 kDa digest in FHCE1⁄

2-digests Therefore, the 36 kDa digest could be further

cleaved by FLCE and divided into two 17 kDa digests

The positions of two FLCE cleavage sites thus

deter-mined were well concordant with LCE sites of medaka

(Fig 3) [13]

To clarify hatching enzyme cleavage sites in natural

hatching, we finally determined the N-terminal

sequence of the digests of FE contained in hatching

liquid As shown in Fig 5B, the SDS⁄ PAGE pattern

of Fundulus hatching liquid was somewhat different

from that of FHCE⁄ FLCE-digests of UFE (i.e the

digests of UFE by the mixture of FHCE1⁄ 2 and

FLCE) However, the cleavage sites determined using

major digests in hatching liquid were essentially

con-cordant with those of UFE as summarized in Table 1

(a) The sequence of the 35 kDa digest was a mixture

of YPVPAATVAV and PVPAATVAVE The former

sequence corresponded to that of the digest cleaved at

the site FhZPC2 of UFE, and the latter was that

cleaved at one amino acid residue from C-terminal side

of the FhZPC2 site (Table 1) (b) The sequence of the

32 kDa digest was TPTETFHTxD, which corresponds

to that of the digest cleaved at the site FhZPB3 of

UFE (Table 1) (c) Two bands were found at 18 and

16 kDa The sequence of the 18 kDa digest was

NPPPAVAELG found from Asn395 in FhChgH, and

the cleavage site Asp394⁄ Asn395 matched with the

FLCE cleavage site, FhZPB2, determined with UFE

(Table 1) The sequence of the 16 kDa digest was

VPTPKSxDVE found from Val233 in FhChgH This

site was located at three amino acid residues from

C-terminal side of the FHCE1⁄ 2 cleavage site,

FhZPB1, determined with UFE Discrepancy between

digests of UFE and those of FE in hatching liquid,

such as minor differences with respect to

electropho-retic mobility and cleavage sites, might result from the

structural difference between UFE and FE, probably

as a result of the existence of e-(c-glutamyl)-lysine

cross-links in FE (d) The 25 kDa band was observed

in the hatching liquid but not in the FHCE⁄

FLCE-digests of UFE The N-terminal sequence of the digest

was TSQAAVIVE and was located inside of

ZP-domain of FhChgL In natural hatching, a part of

the 35 kDa digest was further digested and degraded

into the 25 kDa digest The SDS⁄ PAGE patterns of

the digests of isolated FE by purified FHCE1⁄ 2 and

FLCE were the same as that of the hatching liquid (data not shown) Thus, the results obtained show that the hatching enzyme cleavage sites determined with FE reflect well those determined with UFE

Next, FHCE1⁄ 2 cleavage sites that are present in the Pro-X-Y repeat region were determined The HCE-inducing swelling of FE in medaka releases water-soluble peptides that are excised from the Pro-X-Y repeat region [10] The previous study showed that this region was broken into small peptides that can not be detected by SDS⁄ PAGE Therefore, after UFE was digested with FHCE1 alone, the supernatant was applied to the reverse phase HPLC system Seven major peaks were obtained and subjected to N-terminal sequencing and MALDI-TOF-MS We obtained sequences such as YPQQPQ, YPSKPQ, YPNPQ, YP-KPQ and YPRPQ, suggesting that FHCE1 cleaved the sites locating the tyrosine residue at the P1¢ site and the proline residue at the P2¢ site [17,18] To further study the FHCE1 cleavage sites, all the peaks eluted with chromatography were subjected to MALDI-TOF-MS

As shown in Fig 6A,B, all of the monoisotopic molecu-lar weights thus determined matched the molecumolecu-lar weights calculated from either FhChgH and FhChgHm cDNA In addition, the results obtained were confirmed using a recombinant protein of the Pro-X-Y repeat region of FhChgH, called rec.FhChgH_ProXY After rec.FhChgH_ProXY was digested by FHCE1⁄ 2, the digests were fractionated by reverse phase column chro-matography, and analyzed by MALDI-TOF-MS The result obtained was consistent with the FHCE1-cleavage pattern of the Pro-X-Y repeat region of FhChgH obtained from UFE (Fig 6C) This clearly indicates that the Pro-X-Y repeat region was broken into small pieces, the size of which was three, four, five, six, nine or

12 amino acids in length FHCE1⁄ 2 cleaved a bond between Gln and Tyr of the Pro-X-Y region or between Ser and Tyr, and occasionally also cleaved a bond between Ser and Lys or between Gln and Thr

Estimation of the egg envelope digestion efficiency of FHCE1⁄ 2 and FLCE

The substrate preferences of FHCE1⁄ 2 and FLCE were quantitatively estimated using synthetic peptides The peptide sequences were designed from three FHCE1⁄ 2 cleavage sites (FhZPB1, FhZPB3 and FhZPC1); two FLCE sites (FhZPB2 and FhZPC2); and one site (FhZPC3) determined using hatching liquid The names of the synthetic peptides correspond

to those of the sites (Table 2) In addition, six peptides were designed from the Pro-X-Y repeat region Four

of them (PSYP, PQYP, PQTP and PSKP) were

designed from FHCE1⁄ 2 sites As a control, two (PQQP and PQKP) were designed from the sites that were not cleaved by any type of hatching enzymes As shown in Table 2, FLCE showed high activity toward the peptides for the FLCE sites (FhZPB2 and FhZPC2) but no activity toward all of the peptides designed from the Pro-X-Y repeat region, or low activ-ity toward FHCE1⁄ 2 sites Therefore, FLCE is con-firmed to specifically cleave the N-terminal side of the

ZP domain of FhZPC and the center of ZP domain of FhZPB In addition, FLCE showed high specific activ-ity toward the peptide deduced from hatching liquid (FhZPC3) but FHCE1⁄ 2 had no activity toward the peptide, suggesting that FLCE specifically cleaves the FhZPC3 site on FE, but that FHCE1⁄ 2 do not FHCE1⁄ 2 showed high specific activity toward two peptides designed from the Pro-X-Y repeat region, PQYP and PSYP, confirming that the tyrosine residue

at the P1¢ site is preferred by FHCE1 ⁄ 2 Therefore, FHCE1⁄ 2 have a high specific activity for digesting the center of Pro-Gln⁄ Ser-Tyr-Pro sequence in the Pro-X-Y repeat However, no activity of FHCE1⁄ 2 toward PSKP and PQTP peptides was observed, suggesting that a bond between Ser and Lys or between Gln and Thr in UFE and in rec.FhChgH_ProXY is not so efficiently cleaved by FHCE1⁄ 2 In addition, FHCE1 ⁄ 2 showed

no activity or only low activity toward the peptides designed from FHCE1⁄ 2 cleavage sites (FhZPB1, FhZPB3 and FhZPC1) This difference in substrate

Fig 6 FHCE1 ⁄ 2 cleavage sites found in the Pro-X-Y repeat region.

The cleavage sites in FhChgH (A) and FhChgHm (B) were

deter-mined using the unfertilized egg envelope and recombinant protein,

rec.FhChgH_ProXY (C) Black arrowheads indicate FHCE1 ⁄ 2

cleav-age sites FhZPB1 and FhZPB3 shown next to the white

arrow-heads indicate FHCE1 ⁄ 2 cleavage sites, as described in Fig 3.

Values under the lines indicate observed monoisotopic masses

together with their calculated monoisotopic masses (given in

paren-theses) rec.FhChgH_ProXY possesses additional methionine and

histidine residues at the N- and C-terminus, respectively.

Table 2 Specific activity of FHCE1, FHCE2 and FLCE estimated using synthetic peptide substrates The cleavage site on each pep-tide is indicated by an arrow ND, not detected.

Substrate Sequence

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

FHCE1 ⁄ 2 cleavage sites

FLCE cleavage sites

Cleavage site deduced from hatching liquid

Pro-X-Y region

preference may be a result of structural differences of

the substrate, such as the macromolecular egg envelope

and small peptides, and FHCE1⁄ 2 are able to cleave

these sites only when egg envelope was used as

substrate

Conservation of the egg envelope digestion

mechanism of hatching enzymes in euteleosts

LCE cleavage sites

The present study suggests that the egg envelope

diges-tion mechanism of hatching enzymes is conserved

between two euteleosts, Fundulus and medaka To

extend the comparison from lower to higher euteleosts,

we collected the hatching liquid of several fishes and

determined the N-terminal amino acid sequences of

their digests Figure 7A shows the tricine-SDS⁄ PAGE

patterns of hatching liquid obtained from higher

euteleosts, such as Fundulus, medaka, three-spined

stickleback and spotted halibut, as well as the lower

euteleost, rainbow trout

First, we focused on the three digests in Fundulus

hatching liquid, indicated by *1, *2a and *3a in

Fig 7A When the N-terminal sequences of the bands

in hatching liquid of four fishes were compared with

those of Fundulus, the respective bands were revealed

to correspond well with those of the digests of

Fundu-lus (a) The 35 kDa digest of Fundulus was generated

by cleavage at FhZPC2 The site was found in the

35 kDa product of medaka and spotted halibut or in the 27 kDa of rainbow trout (Fig 7A, *1) (b) A part

of the 35 kDa digest of Fundulus was cleaved at the site FhZPC3 to generate the C-terminal 27 kDa and N-terminal 9 kDa digests (Fig 7A, *2a and *2b) The bands corresponding to them were observed at the 27 and 10 kDa digests of medaka and three-spined stickle-back or at the 27 and 9 kDa digests of spotted halibut, except rainbow trout In three-spined stickleback, no bands around 35 kDa were observed, suggesting that this 35 kDa product is completely digested into the 27 and 10 kDa bands These results suggest that cleavage efficiency at the site corresponding to FhZPC3 site is different from species to species (c) The 18 kDa digest,

as well as the 16 kDa digest, of Fundulus was generated

by cleavage at the site FhZPB2 The site was found in the 17 kDa products in medaka, in the 14 kDa prod-ucts in three-spined stickleback and in the 18 kDa products in spotted halibut and rainbow trout (Fig 7A,

*3a) Alignment of the sequences around the three cleavage sites (Fig 7B,D) suggests that the position of each cleavage site is well conserved in higher euteleosts, and two of three sites also coincide with each other in lower euteleosts

In addition to the digests described above, the

35 kDa digest was observed in rainbow trout hatching liquid (Fig 7A, *4) The sequence analysis revealed

Fig 7 The conservation of the egg envelope digestion mechanism of hatching enzyme in euteleosts (A) Tricine-SDS ⁄ PAGE pattern of hatching liquid for Fundulus, medaka, three-spined stickleback, spotted halibut and rainbow trout The bands comparable to each other are indicated by *1 to *4 (C) The regions of products *1 to *3 are indicated in the schematic presentation of ZPB and ZPC, together with the cleavage sites of HCE and LCE Arrowheads in gray and black indicate cleavage sites of HCE and LCE, respectively Partial amino acid sequence alignment around the LCE cleavage sites on ZPB and ZPC is shown in (B) and (D), respectively The names of the cleavage sites

of Fundulus LCE are shown next to the arrowheads.

that the 35 kDa product was derived from VEPb, one

of two precursors of rainbow trout ZPB The cleavage

site deduced from the 35 kDa digest of rainbow trout

corresponded well with that of the 32 kDa digest

of Fundulus, the digest of FhZPB derived from

FhChgHm At present, the digests derived from two

kinds of ZPBs were only detected in rainbow trout

and Fundulus hatching liquids, whereas the digests

derived from only ChgH orthologs were detected

in the other species This is probably a result of

differences with respect to the content of ZPBs in egg

envelope among species Therefore, major cleavage

patterns and their cleavage sites are conserved among

the euteleosts examined in the present study

HCE cleavage sites

The proline-rich Pro-X-Y repeat region has been also

found in the N-terminal region of precursors of ZPB

of many euteleosts [9] To determine whether the

frag-mentation of the region is a universal feature in

eu-teleosts, we further determined cleavage sites on the

Pro-X-Y repeat region of rainbow trout egg envelope

protein The digests in rainbow trout hatching liquid

were fractionated by reverse phase column

chromatog-raphy, and the small peptides that eluted with a low

acetonitril concentration (30–40%) were collected Six

major peaks were obtained and subjected to

N-termi-nal sequencing The obtained sequences were

WP(A⁄ V), WPPI, WPVQPG, QPPQRPA and

(Q⁄ E)P(L ⁄ F)P(Q ⁄ P)RPA These sequences were found

in the Pro-X-Y repeat regions of VEPa and VEPb

(Fig 8), which are ZPB precursors of rainbow trout

The result clearly indicates that the Pro-X-Y repeat

region of rainbow trout egg envelope proteins is

bro-ken into small pieces, with notable cleavage of the sites

locating the tryptophan, glutamine and glutamic acid

residues at the P1¢ site and the proline residue at the

P2¢ site Considering that the digestion pattern in the

Pro-X-Y repeat region was similar among two of

the higher euteleosts (Fundulus and medaka) and one

of lower euteleosts (rainbow trout), the cleavage

pat-tern of HCE is also suggested to be conserved among

euteleosts

Discussion

The present study investigated the egg envelope

diges-tion mechanism of Fundulus hatching enzyme Three

cDNA orthologs of egg envelope precursor proteins,

ChgH, ChgHm and ChgL cDNAs, were cloned By

comparing the N-terminal amino acid sequences of

HCE- and LCE-digests with the sequences deduced

from cDNAs, the cleavage sites of HCE and LCE on egg envelope subunit proteins were determined The results obtained showed that not only genes of hatch-ing enzymes and egg envelope proteins, but also cleav-age sites of hatching enzymes are well conserved between Fundulus and medaka Below, we discuss the mechanism of egg envelope digestion by hatching enzyme, mainly based on the structural characteristics

of egg envelope protein

In medaka, HCE swells the hardened fertilized egg envelope to convert its compact structure into a loose structure This conversion results from medaka HCE cleaving the Pro-X-Y region of ZI-1,2 in the envelope, leading to the release of small peptide fragments, with notable cleavage of the sites locating tyrosine and asparagine residues at P1¢ site within the repeats [10,13] The released fragments contain e-(c-glutamyl) lysine isopeptide cross-links that are responsible for egg envelope hardening after fertilization [10] Although we did not determine the content of e-(c-glutamyl) lysine isopeptides in the Pro-X-Y region of FhZPB, the content of glutamine (25%) and lysine (7%) in the region resembled that of medaka (gluta-mine, 21%; lysine, 6%) The present study showed that FHCE1⁄ 2 also cleaved the Pro-X-Y region into small fragments (three to twelve amino acids in

Fig 8 Cleavage sites found in Pro-X-Y repeat regions in rainbow trout egg envelope protein Amino acid sequences of the Pro-X-Y repeat regions of rainbow trout ZPB, VEPa (A) and VEPb (B), are shown Broken lines above the sequences indicate repeating units The sequences determined from low molecular weight products in the hatching liquid are underlined The cleavage sites are predicted

to be AflQ, QflW and AflE, and are indicated by arrowheads.