Tài liệu Báo cáo khoa học: Type I antifreeze proteins expressed in snailfish skin are identical to their plasma counterparts doc

Blood plasma from Atlantic Liparis atlanticus and dusky Liparis gibbus snailfish contain type I AFPs that are significantly larger than all previously described type I AFPs.. In this study

identical to their plasma counterparts

Robert P Evans and Garth L Fletcher

Ocean Sciences Centre, Memorial University of Newfoundland, St John’s, Newfoundland, Canada

Teleost fish that inhabit icy seawater synthesize

anti-freeze proteins⁄ polypeptides (AFPs) or antifreeze

glyco-proteins (AFGPs) for protection against freezing

Diverse species from numerous taxonomic groups

pro-duce AFPs that are grouped into four distinct classes

(types I, II, III and IV) based on their primary and

secondary structural characteristics [1–3] Regardless of

protein structure, all fish AFPs lower the solution

freezing point noncolligatively by binding to certain

surfaces of ice crystals, modifying their structure and

inhibiting further growth The difference between the

lowered freezing point and unaltered melting point is

termed thermal hysteresis and is used as a measure of

antifreeze activity [1,3,4]

Of the four classes of AFPs described thus far, the simplest is type I AFP found in right-eye flounders (Pleuronectes) and a few sculpin species (e.g Myoxo-cephalus) These polypeptides have high alanine con-tent (> 60 mol%), have an amphipathic a-helical secondary structure, and are usually quite small (3.3– 4.5 kDa) [2,5] Until the past decade, it was generally accepted that the synthesis of AFPs was confined solely to liver tissue (termed liver type) for secretion into blood for extracellular freeze protection How-ever, more recently, a novel subclass of type I AFPs was isolated and characterized from the skin of winter flounder Pseudopleuronectes americanus (for-merly Pleuronectes americanus) [6] These AFPs, which

Keywords

antifreeze; cDNA; protein expression;

snailfish; type I

Correspondence

R P Evans, Department of Biochemistry,

University of Alberta, Edmonton, Alberta

T6G 2H7, Canada

Fax: +1 780 492 0886

Tel: +1 780 492 3481

E-mail: robert.evans@ualberta.ca

(Received 13 June 2005, revised 8 August

2005, accepted 22 August 2005)

doi:10.1111/j.1742-4658.2005.04929.x

Type I antifreeze proteins (AFPs) are usually small, Ala-rich a-helical poly-peptides found in right-eyed flounders and certain species of sculpin These proteins are divided into two distinct subclasses, liver type and skin type, which are encoded by separate gene families Blood plasma from Atlantic (Liparis atlanticus) and dusky (Liparis gibbus) snailfish contain type I AFPs that are significantly larger than all previously described type I AFPs In this study, full-length cDNA clones that encode snailfish type I AFPs expressed in skin tissues were generated using a combination of library screening and PCR-based methods The skin clones, which lack both signal and pro-sequences, produce proteins that are identical to circulating plasma AFPs Although all fish examined consistently express antifreeze mRNA in skin tissue, there is extreme individual variation in liver expression – an unusual phenomenon that has never been reported previously Further-more, genomic Southern blot analysis revealed that snailfish AFPs are products of multigene families that consist of up to 10 gene copies per genome The 113-residue snailfish AFPs do not contain any obvious amino acid repeats or continuous hydrophobic face which typify the structure of most other type I AFPs These structural differences might have implica-tions for their ice-crystal binding properties These results are the first to demonstrate a dual liver⁄ skin role of identical type I AFP expression which may represent an evolutionary intermediate prior to divergence into distinct gene families

Abbreviations

AFGPs, antifreeze glycoproteins; AFPs, antifreeze proteins ⁄ polypeptides; IBM, ice-binding motif; ORF, open reading frame;

UTR, untranslated region.

are encoded by a separate subset of genes, were

desig-nated as skin-type AFPs They are synthesized as

mature polypeptides that lack both signal and

pro-sequences, which suggests that they remain

intracellu-lar [6] Recent publications of skin-type AFP isolation

from shorthorn (Myoxocephalus scorpius) and

long-horn (M octodecemspinosus) sculpins indicate that the

production of AFP in peripheral epithelial tissues may

be a common trait in many fish species [7,8] The

char-acterization of known skin-type AFPs and the presence

of antifreeze activity in skin tissues of other species has

led to the hypothesis that skin-type AFPs are

wide-spread ancestors of liver-type (plasma) AFPs [6,9]

Atlantic snailfish (Liparis atlanticus) and dusky

snailfish (L gibbus) belong to a large family

(Cyclo-pteridae) of benthic and pelagic marine fishes that

inhabit northern regions of the Atlantic Ocean

Snail-fish are closely related to sculpins, which belong to a

different family of the same order Scorpaeniformes

[10] Both species spawn during the winter months in

ice-laden inshore coastal regions around

Newfound-land, which makes them prime candidates for

produc-tion of AFPs Type I AFPs were previously isolated

and characterized from the blood plasma of both

Atlantic and dusky snailfish which are the largest

des-cribed to date (> 9.3 kDa) [11] We have also shown

recently that Atlantic snailfish skin tissues contain

type I AFPs that have identical molecular mass and

very similar amino acid content to their plasma

counter-parts [12]

Further analysis of the snailfish AFPs would be

helpful in determining the structure⁄ function

charac-teristics of these unusually large type I AFPs and to

clarify the relationship between skin and plasma

pro-teins Pursuant to this, an Atlantic snailfish skin

cDNA library was screened using a shorthorn sculpin

skin-type AFP clone as a probe Full-length cDNA

sequences of both Atlantic and dusky snailfish skin

type I AFPs were generated using a combination of

library clones with RT-PCR and RACE techniques

Results from this study show that skin AFPs from

both species are nearly identical to each other and

their skin transcripts produce proteins that are

identi-cal to their corresponding plasma proteins

Results

cDNA library screening and cloning of snailfish

skin AFP

A cDNA library was constructed to investigate the

presence of type I AFP mRNA in skin tissues Two

independent clones were identified from the library

screen using the open reading frame (ORF) of a shorthorn sculpin skin cDNA as a probe The

 260 bp clones (clone-c1 and clone-c2) contained identical sequences, apart from a small difference in the length of poly(A) tail and a few nucleotides at their 5¢ ends However, the clones appeared to be truncated versions of complete type I AFP messages As indica-ted by the underline in Fig 1, one reading frame gave

an Ala-rich 26 amino acid peptide, that lacks an obligatory in-frame start codon This sequence infor-mation was then used in 5¢-RACE reactions to ascer-tain the remainder of the skin AFP cDNA sequence RNA ligase-mediated RACE was used to clone the remaining 5¢ portion of the snailfish AFP cDNA The full L atlanticus skin cDNA is 568 bp and contains a complete 342 bp ORF (Fig 1) The ORF encodes an Ala-rich protein of 113 residues and was designated as Las-AFP (L atlanticus skin AFP) The putative start and stop codons are underlined as well as three pos-sible polyadenylation signal sequences [13]

The Las-AFP sequence was utilized to design appro-priate RT-PCR and 3¢)5¢-RACE primers for de novo cloning of AFP sequence from dusky snailfish skin RNA The 3¢-RACE procedure (primers indicated in Fig 2) produced a single band that was  450 bp, whereas 5¢-RACE gave a 370 bp product The overlap-ping sequences were combined into a 587 bp clone which contained a 342 bp ORF that encodes a 113 residue, Ala-rich, protein designated as Lgs-AFP (L gibbus skin AFP) The putative start and stop codons are underlined along with the standard poly-adenylation signal sequence

The AFP cDNAs cloned from the skin tissues of Atlantic and dusky snailfish have strikingly similar nuc-leotide sequences that encode nearly identical proteins apart from a few minor differences However, there is a

19 bp insertion in the Lgs-AFP 3¢-untranslated region (UTR) just before the poly(A) tail region Snailfish AFP cDNA sequences are similar to skin-type AFPs from winter flounder and sculpins in that they do not appear to contain a signal sequence or pro-sequences [6–8]

Surprisingly, the amino acid composition of proteins purified from Atlantic snailfish skin tissue is quite sim-ilar to the AFP predicted from the cDNA sequence (Table 1) Any differences may be attributed to varia-tions in analytical procedures and the fact that mix-tures of AFPs were analyzed Most importantly, the predicted molecular mass and N-terminal sequence for Las-AFP is identical to the isolated plasma proteins [11] Dusky snailfish also express the same type I AFP

in skin tissue that is circulating in their blood (Table 1)

Analysis of snailfish AFP genes

Total RNA from Atlantic snailfish tissues were probed

with a section from the 3¢-UTR of Las-AFP to

evalu-ate the distribution of snailfish AFP mRNA (Fig 3A)

A specific band was visible after a short exposure in skin tissue as well as a faint signal from gill is detect-able with longer exposures No other tissues gave detectable signals on this northern blot Similar results were observed in another fish, except that there was a

Fig 1 Nucleotide sequence and primary

translation product of L atlanticus skin AFP

cDNA The ORF is capitalized, whereas the

5¢- and 3¢-UTRs are in lower case letters.

The putative start and stop codons are

underlined in bold as are three possible

polyadenylation signal sequences The

sequence obtained from the initial las-c1

and las-c2 cDNA clones are underlined.

RT-PCR or RACE primer sequences are

shown above (5¢ fi 3¢) or below (3¢ fi 5¢)

the nucleotide sequence GenBank

Accession Number AY455862.

Fig 2 Nucleotide sequence and primary

translation product of L gibbus skin AFP

cDNA The ORF is capitalized, whereas the

5¢- and 3¢-UTRs are in lower case letters.

The putative start and stop codons and the

polyadenylation signal are underlined.

RT-PCR or RACE primer sequences are

shown above (5¢ fi 3¢) or below (3¢ fi 5¢)

the nucleotide sequence GenBank

Accession Number AY455863.

definite detectable signal in liver tissue RNA (Fig 3B).

RT-PCR performed using the same RNA (with ORF

primer set) gave positive bands for skin, gill, blood

cells, kidney, spleen and liver (Fig 3C) and experi-ments using 3¢-UTR primers gave the same expression pattern (data not shown)

Additional northern blot experiments with 3¢-UTR DNA probes gave very intense autoradiographic sig-nals in skin tissues from four Atlantic snailfish and a dusky snailfish (Fig 4A), these were confirmed by RT-PCR analysis (Fig 4B) Results of a northern blot using liver RNA from eight individual Atlantic snail-fish and a dusky snailsnail-fish showed that three of the Atlantic snailfish samples, but not the dusky snailfish, gave positive signals of varying intensities (Fig 4C) RT-PCR analysis confirmed this result with the excep-tion of one liver sample (Fig 4D) A recently pub-lished report of northern blots probed with shorthorn sculpin skin ORF cDNA, indicated that snailfish liver and skin tissues express type I AFP mRNA [11a]

To analyze snailfish genes, a Southern blot was probed with the same DNA probe applied to the pre-vious northern blots At least nine individual frag-ments can be distinguished when Atlantic snailfish DNA is digested with HindIII (Fig 5; lane 3), whereas

up to 10 are visible for dusky snailfish and the same restriction enzyme (Fig 5; lane 7) Results indicate that snailfish AFPs are expressed via a multigene fam-ily but the exact number gene copies cannot be deter-mined precisely here

Discussion

Using a combination of cDNA library screening and 5¢-RACE, a complete cDNA corresponding to type I AFP was cloned from Atlantic snailfish skin tissue and

Table 1 Amino acid composition (mol%) and molecular mass of snailfish type I AFPs.

Amino

acid

LaP-AFP (protein)

LaS-AFP (protein)

Las-AFP (cDNA)

LgP-AFP1 (protein)

LgP-AFP2 (protein)

Lgs-AFP (cDNA)

a

Based on ESI-MS analysis of HPLC peaks [11,12].bPredicted from cDNA sequence excluding Met LaP-AFP, L atlanticus plasma AFP; LaS-AFP, L atlanticus skin AFP; LgP-AFP, L gibbus plasma AFP; LgS-AFP, L gibbus skin AFP.

A

B

C

Fig 3 Tissue distribution of Atlantic snailfish skin AFP mRNA.

(A, B) Northern blots of samples from two individual fish with lanes

labeled with RNA tissue source Each lane contains 5 lg total RNA

and blots were probed with a 175 bp fragment of the 3¢-UTR

sequence of snailfish type I AFP cDNA (C) A typical result of

RT-PCR analysis Numbers correspond with the tissue labels from

the northern blots above c1 is a water only control; c2 is no RT

control The lower panel shows RT-PCR products generated from

b-actin primers used as a loading control.

subsequently in closely related dusky snailfish The

nucleotide and protein sequences are almost identical,

clearly suggesting that these AFPs shared a common

ancestral gene prior to snailfish species divergence

This differs from taxonomically related shorthorn and

longhorn sculpin skin AFPs which produce quite con-trasting proteins, whereas the UTRs of mRNA are nearly identical [8]

Based on the cDNA sequence, both snailfish species express 113 residue type I AFPs that are the largest described to date The predicted proteins lack signal or pro-sequences, which indicates that the mature poly-peptides remain intracellular This would imply that their location and function is analogous to the pre-sumptive intracellular skin AFPs of winter flounder [6] and sculpins [7,8] However, the molecular mass of snailfish skin proteins predicted from cDNA and their N-terminal sequence are identical to the results deter-mined for their purified plasma AFPs [11,12] Further-more, northern blots indicate that snailfish AFP mRNA has consistently significant expression only in skin tissue Taken together, the evidence indicates that the circulating plasma AFPs and skin localized AFPs are identical proteins that are normally expressed by the same skin-specific gene

These results represent the first definitive report of fish that synthesize identical AFPs for protection in two different physiological locations The assumption has been that skin-type AFPs are expressed via a different subset of genes from the liver multigene fam-ily [6–8] The evidence from snailfish contradicts the

A

B

C

D

Fig 4 Distribution of type I AFP mRNA in skin and liver tissues

from Atlantic and dusky snailfish (A) Northern blot analysis of skin

tissue RNA from four individual Atlantic snailfish and one dusky

snailfish Each lane contains 5 lg total RNA and blots were probed

with a 175 bp fragment of the 3¢-UTR sequence of snailfish type I

AFP cDNA (B) The corresponding RT-PCR results from identical

tissue samples Numbers correspond with the tissue labels from

the Northern blots c1 is a water only control; c2 is no RT control.

The lower panel shows RT-PCR products generated from b-actin

primers used as a loading control (C) Northern blot analysis of liver

tissue RNA from eight individual Atlantic snailfish and one dusky

snailfish Blots were probed as indicated above (D) The

corres-ponding RT-PCR results from the same tissue samples as

des-cribed above.

Fig 5 Southern blot analysis of Atlantic and dusky snailfish AFP genes Ten micrograms of genomic DNA were digested with the indicated restriction enzymes and run in each gel lane The blot was probed with the identical snailfish 3¢-UTR DNA fragment used

in the northern blots.

original hypothesis that separate sets of genes code for

unique AFP isoforms to provide extracellular and

intracellular antifreeze protection Although the exact

subcellular location has not yet been unequivocally

established for skin-type AFPs, evidence from winter

flounder indicates that skin AFPs are present in gill

cell cytoplasm as well as in contact with the plasma

membrane outside epithelial cells [14]

Clearly, snailfish AFPs produced in epithelial cells

are secreted into blood to provide extracellular

protec-tion but it is not clear whether some protein remains

inside these cells It is uncertain exactly how snailfish

AFPs are secreted from the cells that expresses them if

they do not contain the requisite signal sequences

There have been recent reports of mature type I AFPs

being exported from cells in winter flounder epidermis

despite the absence of a secretion signal or

pro-sequence [14,15] Furthermore, alternative pathways

for protein export that circumvent the usual

endo-plasmic reticulum–Golgi complex have been described

previously [16,17]

The northern blot experiments exhibited unexpected

variation in AFP expression patterns among individual

fish Whereas skin tissues consistently produced high

levels of AFP mRNA, expression in liver ranged from

undetectable to high levels This extreme individual

variation in mRNA expression has not been reported

previously for any species producing antifreeze

How-ever, studies have shown geographic-dependent

popu-lation differences in antifreeze gene copy number

[18,19] In fact, individual fish from one population of

Newfoundland ocean pout had demonstrable

differ-ences in antifreeze gene copies that indicate the

malle-ability of antifreeze genes within a given fish genome

[18] Furthermore, there is a report in the literature of

large variations in gene expression patterns in

trans-genic rainbow trout [20] It would be informative to

determine if the diverse nature of the snailfish

multi-gene family or if regulatory control regions within

snailfish AFP gene(s) are responsible for the variation

in observed tissue-specific gene expression

The physiological significance of the variegation in

snailfish mRNA expression is not clear because all fish

examined had significant levels of protein in blood and

skin during the winter It is possible that different

phy-siological or environmental cues initiate expression in

each tissue separately Previous studies have shown

that type I AFP expression in liver is seasonally

adjus-ted from low in summer to high in winter based on

environmental cues [1,3] Moreover, skin AFP

expres-sion is uniformly high in winter flounder but has an

annual variation in shorthorn sculpin It seems likely

that skin AFP expression provides the primary source

of AFP production in snailfish and the liver is an ancillary site of expression for contributing supple-mentary protection Snailfish may rely more on skin (and its AFP content) to provide the primary barrier

to ice crystal propagation

The primary structure of snailfish AFPs is unlike most other known type I AFPs Although they are extremely a-helical proteins – determined experiment-ally by CD spectrometry – they possess only moderate thermal hysteresis activity compared with other type I AFPs [11] Helical net and helical wheel representa-tions (Fig 6) indicate that Las-AFP contain none of the ice-binding motifs (IBM) that were originally sug-gested as important for ice binding [21–23] Recently, amino acid substitution experiments have determined that it is the conserved Ala-rich hydrophobic surface which is most important for ice-binding in type I AFPs [5,24–26] Las-AFP contain no full-length hydrophobic surface is free from interfering polar residue side chains Furthermore, snailfish AFPs do not contain

Fig 6 Schematic representations of Atlantic snailfish AFP secon-dary structure (A) Helical net, (B) helical wheel diagrams were constructed by the EMBOSS package located on the Canadian Bioinformatics Resource web page Hydrophilic residues DENQST are marked with diamonds Positively charged residues HKR are marked with octagons Aliphatic residues ILVM are marked with squares.

the requisite hydrogen bonding amino acids necessary

to create the elaborate terminal cap structures found in

most type I AFPs [21] The lack of complete

hydro-phobic face and terminal caps might be responsible for

the low activity of these AFPs It should be noted,

however, that the predicted structure of snailfish AFP

may not exactly correspond with structural data

provi-ded by experimental methods It is possible that the

protein contains kinks or bends in the backbone

around the helix-breaking proline residues

Based on protein primary structure, most type I

AFPs cluster into three distinct groups, depending

on the nature of their highly conserved N-terminal

sequences (Fig 7) Two of the groups contain the

clas-sic 11 residue (ThrX10) repeat sequence, whereas the

third group contains no such repeat structure

Although all polypeptides that fit in the three groups

are small ( 3.3 to 4.5 kDa), the unusually large

snail-fish and shorthorn sculpin skin AFPs are outliers that

do not conform to either of the categories Similarly,

the novel hyperactive winter flounder type I AFP,

which is unusually large (15 kDa) and without obvious

amino acid repeats, would not fit into either major

group [27] Interestingly, there seems to be no

connec-tion between the AFP structural groups and

phylo-genic classification or tissue source of the proteins

With the discovery of snailfish skin proteins, it is

apparent that type I AFPs can be divided into distinct

structural subclasses based on size and the absence of

amino acid repeat structure This subclass could have

unique evolutionary origins and a distinctive

mechan-ism for ice-binding separate from the three groups

mentioned above Perhaps the fundamental property

of a type I AFP, as represented by snailfish AFPs, is

an Ala-rich protein with a-helical secondary structure that is capable of ice binding

Experimental procedures

Tissue sample collection Twelve Atlantic snailfish (L atlanticus) were collected by divers near Logy Bay, Newfoundland, in winter 2000 Two specimens of dusky snailfish (L gibbus) were collected from Placentia Bay, Newfoundland during winter 1999 Tissues were removed from anesthetized fish, immediately frozen in liquid nitrogen and stored at)70
C

Skin library construction and screening Total RNA from Atlantic snailfish skin tissue was isolated using TRIzol reagent (Invitrogen Canada Inc, Burlington,

ON, Canada) and poly(A)+mRNA was isolated from total RNA using an Oligotex mRNA Kit (Qiagen Inc, Mississ-auga, ON, Canada) A skin cDNA library was constructed,

as described by the manufacturer, using Lambda ZAP II library and ZAP-cDNA Synthesis Kit and Gigapack Gold III packaging extracts (Stratagene, La Jolla, CA, USA) The primary skin cDNA library contained around

5· 105

clones Normally,  50 000 plaques were grown on

15 cm NZYCM plates for primary screening; 9 cm plates were used in secondary and tertiary screens

Hybond-N nylon membranes (Amersham Biosciences, Piscataway, NJ, USA) were prepared and screened

Fig 7 Classification of known type I AFP

sequences based on primary structural

char-acteristics Amino acid sequence alignments

of the groups created by CLUSTALX analysis.

Columns of identical amino acids are shown

with black backgrounds, whereas those

with a majority of identical amino acids are

shaded gray Abbreviations used: SS-3,

shorthorn sculpin plasma AFP 3; AS-3,

Arc-tic sculpin plasma AFP 3; GS-5, grubby

scul-pin plasma AFP 5; lss-AFP, longhorn sculscul-pin

plasma AFP; wfs-AFP2, winter flounder skin

AFP 2; wfl-HPLC6, winter flounder liver

HPLC-6; AP-AFP, American plaice plasma

AFP; wfl-AFP9, winter flounder liver AFP9;

YT-AFP, yellowtail flounder AFP; sssAFP-2,

shorthorn sculpin skin AFP 2; Las-AFP,

L atlanticus AFP.

according to the manufacturer Briefly, membranes were

hybridized at 42
C overnight in the following buffer: 5·

NaCl⁄ Pi, 5· Denhardt’s, 0.5% SDS, 50% formamide and

100 lgÆmL)1 calf thymus DNA Probe was labeled with

[32P]dCTP using an All-in-One Random-Primed Labeling

Mix (Sigma-Aldrich, Oakville, ON, Canada) and purified

prior to use with ProbeQuant G-50 Micro Columns

(Amer-sham Biosciences) The final wash was performed in 1.0·

NaCl⁄ Pi, and 0.1% SDS, at 52
C for 20 min A 300 bp

DNA fragment corresponding to the ORF of shorthorn

sculpin skin (s3–2) clone [7] was used as a probe to screen

 2.0 · 105 clones of the primary cDNA library Positive

plaques were first isolated and then pBluescript

phage-mids, to be used for sequencing inserts, were produced

using an in vitro excision protocol (Stratagene)

Northern blot analysis

Total RNA from various tissues of Atlantic and dusky

snailfish were isolated using TRIzol reagent (Invitrogen

Canada Inc) as described by the manufacturer

Five-micro-gram aliquots of total RNA were separated on 1%

for-maldehyde gels and analyzed by a nonradioactive northern

blotting procedure using positively charged nylon

mem-branes (Roche Diagnostics Canada, Laval, QC, Canada)

RNA was transferred to membranes using a VacuGene XL

Vacuum Blotting System (Amersham Biosciences) and

cross-linked with UV light The membrane was hybridized

at 50
C overnight in DIG Easy Hyb buffer (Roche

Diag-nostics) Probe was labeled with DIG-11–dUTP using a

DIG-High Prime DNA Labeling kit or in some cases with

a PCR DIG Probe Synthesis Kit (Roche Diagnostics) with

chemiluminescent signal detection using CDP-Star The

final wash was performed in 0.1· NaCl ⁄ Pi, and 0.1% SDS,

at 50
C for 2 · 15 min A 175 bp DNA fragment

corres-ponding to the 3¢-UTR of the skin clone was used as a

probe

Southern blot analysis

Genomic DNA was isolated from liver of Atlantic and

dusky snailfish using a Wizard Genomic DNA Purification

Kit (Promega, Madison, WI, USA) Aliquots of RNAse

A-treated genomic DNA were digested with various

restric-tion endonucleases (Invitrogen) Five-microgram aliquots of

the digestion products were separated in a 0.8% agarose gel,

transferred to positively charged nylon membranes using a

VacuGene XL Vacuum Blotting System (Amersham

Bio-sciences) and cross-linked with UV light A

chemilumines-cent-based nonradioactive method was used to detect

sequences on the membrane Briefly, the membrane was

hybridized at 42
C overnight in DIG Easy Hyb buffer

(Roche Diagnostics) Probe was labeled with DIG-11–dUTP

using a PCR DIG Probe Synthesis Kit (Roche Diagnostics)

with chemiluminescent signal detection using CDP-Star

The final wash was performed in 0.5· NaCl ⁄ Pi, and 0.1% SDS, at 65
C for 2 · 15 min A 175 bp DNA fragment cor-responding to the 3¢ UTR of the skin clone was used as a probe

RACE procedure Both 5¢- and 3¢-RACE reactions were performed using the RNA ligase-mediated GeneRacerTM Kit, as described by the manufacturer (Invitrogen Canada Inc) One microgram

of DNase-treated total RNA combined with Thermo-scriptTM reverse transcriptase (Invitrogen Canada Inc) was used to generate adapter-linked first strand cDNA for 1 h

in a 50
C reaction The first-strand cDNA was combined with the appropriate primers and touchdown PCR amplifi-cation was performed using DyNAzyme EXTTM DNA polymerase (Finnzymes, Oy, Finland) in an Eppendorf Mastercycler The touchdown cycling conditions consisted

of an initial 95
C denaturing step (2 min), followed by 10 cycles of 94
C (15 s), 72
C decreased to 60
C (15 s),

72
C (60 s) and 25 more cycles of 94
C (15 s), 60
C (15 s), and 72
C (60 s) In order to obtain a product in most reactions, dimethylsulfoxide was added at 10% (v⁄ v) RACE reaction products were separated on 1% agarose gels and then purified using spin columns provided in the kit GeneRacerTM Kit or by CONCERTTM Gel Extraction System (Invitrogen Canada Inc) A TOPO TA Cloning kit was used to clone the purified RACE products for sequencing into a pCR4-TOPO cloning vector (Invitrogen Canada Inc) At least three independent clones were isola-ted and the purified plasmids sequenced

RT-PCR analysis One microgram of DNase-treated total RNA from each of the specified tissues was combined with 70 pmol of an anchored poly(T) primer ThermoscriptTMreverse transcrip-tase (Invitrogen Canada Inc) was then used to generate first-strand cDNA in a 1 h reaction at 50
C, as described

by the manufacturer Normally, 1⁄ 10th of the RT reaction was combined with the appropriate primers and touchdown PCR amplification was performed using DyNAzyme EXTTM DNA polymerase in an Eppendorf Mastercycler The touchdown cycling conditions consisted of an initial

95
C denaturing step (2 min), followed by 10 cycles of

94
C (15 s), 72
C decreased to 60
C (15 s), 72
C (60 s) and 25 more cycles of 94
C (15 s), 60
C (15 s), and 72
C (60 s) RT-PCR products were separated on 1% agarose gels and visualized using ethidium bromide

DNA sequencing Sequencing was performed on the pBluescript phagemids

or pCR4-TOPO plasmids using T3 and T7 primers at the

DNA sequencing facility in The Centre for Applied

Genom-ics (Hospital for Sick Children, Toronto, ON, Canada)

Bioinformatics programs

Homologous nucleotide and protein sequences were

searched through blast searches on the NCBI web server

The NCBI orf finder was utilized to identify putative

open reading frames in the nucleotide sequences Helical

net and helical wheel diagrams were constructed using

emboss package located on the Canadian Bioinformatics

Resource web page (all located at http://www.ncbi.nlm.nih

gov/) Swiss PDB software used to generate a

three-dimen-sional model of Las-AFP clustalx and treeview (1.6.1)

software were used to create an unrooted neighbor-joining

tree

Acknowledgements

We thank M King and Dr M Shears at the OSC for

technical assistance and the OSC divers for sample

col-lection We also thank Dr Ming Kao for help with

antifreeze activity measurements This study was

sup-ported by a grant from NSERC

References

1 Fletcher GL, Goddard SV, Davies PL, Gong Z, Ewart

KV & Hew CL (1998) New insights into fish antifreeze

proteins: physiological significance and molecular

regu-lation In Cold Ocean Physiology (Po¨rtner HO & Playle

RC, eds), pp 240–265 Cambridge University Press,

New York

2 Ewart KV, Lin Q & Hew CL (1999) Structure,

func-tion and evolufunc-tion of antifreeze proteins Cell Mol Life

Sci 55, 271–283

3 Fletcher GL, Hew CL & Davies PL (2001) Antifreeze

proteins of teleost fishes Annu Rev Physiol 63, 359–390

4 Hew CL & Yang DS (1992) Protein interaction with

ice Eur J Biochem 203, 33–42

5 Harding MM, Ward LG & Haymet AD (1999) Type I

‘antifreeze’ proteins Structure–activity studies and

mechanisms of ice growth inhibition Eur J Biochem

264, 653–665

6 Gong Z, Ewart KV, Hu Z, Fletcher GL & Hew CL

(1996) Skin antifreeze protein genes of the winter

flounder, Pleuronectes americanus, encode distinct and

active polypeptides without the secretory signal and

pro-sequences J Biol Chem 271, 4106–4112

7 Low WK, Miao M, Ewart KV, Yang DS, Fletcher GL

& Hew CL (1998) Skin-type antifreeze protein from

the shorthorn sculpin, Myoxocephalus scorpius

Expres-sion and characterization of a Mr9700 recombinant

protein J Biol Chem 273, 23098–23103

8 Low WK, Lin Q, Stathakis C, Miao M, Fletcher GL

& Hew CL (2001) Isolation and characterization of skin-type, type I antifreeze polypeptides from the long-horn sculpin, Myoxocephalus octodecemspinosus J Biol Chem 276, 11582–11589

9 Low WK, Lin Q, Ewart KV, Fletcher GL & Hew CL (2002) The skin-type antifreeze polypeptides: a new class of type I AFPs In Fish Antifreeze Proteins (Ewart

KV & Hew C L, eds), pp 161–186 World Scientific Publishing, Singapore

10 Scott WB & Scott MG (1988) Atlantic Fishes of Canada University of Toronto Press, Toronto, ON

11 Evans RP & Fletcher GL (2001) Isolation and charac-terization of type I antifreeze proteins from Atlantic snailfish (Liparis atlanticus) and dusky snailfish (Liparis gibbus) Biochim Biophys Acta 1547, 235–244

11a Evans RP & Fletcher GL (2005) Type I antifreeze pro-teins: Possible origins from chorion and keratin genes

in Atlantic snailfish J Mol Evol doi: org/10.1007/ s00239-004-0067-y

12 Evans RP & Fletcher GL (2004) Isolation and purifica-tion of antifreeze proteins from skin tissues of snailfish, cunner and sea raven Biochim Biophys Acta 1700, 209–217

13 Graber JH, Cantor CR, Mohr SC & Smith TF (1999)

In silicodetection of control signals: mRNA-3¢-end-processing sequences in diverse species Proc Natl Acad Sci USA 96, 14055–14060

14 Murray HM, Hew CL, Kao KR & Fletcher GL (2002) Localization of cells from the winter flounder gill expressing a skin type antifreeze protein gene Can J Zool 80, 110–119

15 Murray HM, Hew CL & Fletcher GL (2003) Spatial expression patterns of skin-type antifreeze protein

in winter flounder (Pseudopleuronectes americanus) epidermis following metamorphosis J Morph 257, 78–86

16 Mignatti P, Morimoto T & Rifkin DB (1992) Basic fibroblast growth factor, a protein devoid of secretory signal sequence, is released by cells via a pathway inde-pendent of the endoplasmic reticulum–Golgi complex

J Cell Physiol 151, 81–93

17 Menon RP & Hughes RC (1999) Determinants in the N-terminal domains of galectin-3 for secretion by a novel pathway circumventing the endoplasmic reticulum–Golgi complex Eur J Biochem 264, 569–576

18 Hew CL, Wang NC, Joshi S, Fletcher GL, Scott GK, Hayes PH, Buettner B & Davies PL (1988) Multiple genes provide the basis for antifreeze protein diversity and dosage in the ocean pout, Macrozoarces ameri-canus J Biol Chem 263, 12049–12055

19 Hayes PH, Davies PL & Fletcher GL (1991) Popula-tion differences in antifreeze gene copy number and arrangement in winter flounder Genome 34, 174–177

20 Uzbekova S, Amoros C, Cauty C, Mambrini M,

Perrot E, Hew CL, Chourrout D & Prunet P (2003)

Analysis of cell-specificity and variegation of transgene

expression driven by salmon prolactin promoter in

stable lines of transgenic rainbow trout Transgenic Res

12, 213–227

21 Sicheri F & Yang DS (1995) Ice-binding structure and

mechanism of an antifreeze protein from winter

floun-der Nature 375, 427–431

22 Lin Q, Ewart KV, Yang DS & Hew CL (1999)

Stu-dies of a putative ice-binding motif in winter

floun-der skin-type anti-freeze polypeptide FEBS Lett 453,

331–334

23 Zhang W & Laursen RA (1998) Structure–function

relationships in a type I antifreeze polypeptide The

role of threonine methyl and hydroxyl groups in

anti-freeze activity J Biol Chem 273, 34806–34812

24 Baardsnes J, Kondejewski LH, Hodges RS, Chao H, Kay C & Davies PL (1999) New ice-binding face for type I antifreeze protein FEBS Lett 463, 87–91

25 Baardsnes J, Jelokhani-Niaraki M, Kondejewski LH, Kuiper MJ, Kay CM, Hodges RS & Davies PL (2001) Antifreeze protein from shorthorn sculpin: identification

of the ice-binding surface Protein Sci 10, 2566–2576

26 Fairley K, Westman BJ, Pham LH, Haymet ADJ, Harding MM & Mackay JP (2002) Type I shorthorn sculpin antifreeze protein Recombinant synthesis, solution conformation, and ice growth inhibition studies J Biol Chem 277, 24073–24080

27 Marshall CB, Fletcher GL & Davies PL (2004) Hyper-active antifreeze protein in a fish Nature 429, 153

28 Evans RP & Fletcher GL (2005) Type I antifreeze proteins: possible origins from chorion and keratin genes in atlantic snailfish J Mol Evol in press