527.full

Glycobiology vol. 6 no. 5 pp. 527-536, 1996

Purification, biochemistry and molecular cloning of an insect glycosylasparaginasefrom Spodoptera frugiperda

Yuan Liu, Graham S.Dunn1 and Nathan N.Aronson, Jr.

Department of Biochemistry and Molecular Biology, College of Medicine,University of South Alabama, Mobile, AL 36688, USA, and 'School ofBiology and Biochemistry, University of Bath, Bath, UK2To whom correspondence should be addressed

Glycosylasparaginase (EC 3.5.1.26) from Sf9 cells (Spodop-tera frugiperda) was purified to homogeneity with a specificactivity of 2.1 unit/mg. The enzyme is composed of twonon-identical a/p subunits joined by strong non-covalentforces and has one glycosylation site located in the a sub-unit. Molecular masses of the subunits were determined tobe 28 kDa and 17 kDa by SDS-PAGE. Native enzyme ex-isted in quaternary structures of either heterodimer (<*P)or heterotetramer (a2P2)- These forms exhibited differentionic characteristics during DE52 anion exchange chroma-tography, and their molecular masses were determined tobe 47 kDa and 101 kDa by gel filtration. The enzyme wasthermostable, requiring 65-70°C to be denatured, and ithad a broad pH optimum from 4-10.5 with a pKa around6. SDS easily inactivated the enzyme. The Km of glyco-sylasparaginase for its normal substrate GlcNAc-Asn was0.88 mM. The enzyme also exhibited asparaginase activitywith a Km of 3.0 mM for asparagine. N-terminal aminoacids of the denatured subunits were sequenced and degen-erate primers were designed for cloning its cDNA usingPCR and 5' and 3' RACE. Glycosylasparaginase cDNAsfrom bovine and rat were also cloned using similar strate-gies, and primary structures of glycosylasparaginases fromsix species (human, bovine, rat, mouse, Sf9 cells and Fla-vobacterium) have been compared and related to a recentcrystal structure of the human enzyme.

Key words: glycosylasparaginase/glycoprotein degradation/lysosomal hydrolase/insect

Introduction

Glycosylasparaginase [glycoasparaginase, N4-(P-/V-acetyl-D-glucosaminyl)-L-asparaginase, EC 3.5.1.26] is a lysosomal hy-drolase involved in the stepwise hydrolysis of Asn-linked gly-coproteins to monosaccharides and amino acids (Aronson andKuranda, 1989). This amidase cleaves the P-amide on the as-paragine moiety of the N-glycosidic linkage of glycoproteins(Aronson and Kuranda, 1989; Mononen et al, 1993). Glyco-sylasparaginase was first purified to homogeneity from rat liverby Tollersrud and Aronson (1989). The rat enzyme has a mo-lecular mass of 49 kDa and is composed of two non-identicalsubunits of 24 kDa and 20 kDa, which are referred to as the aand P subunits. These subunits are glycosylated and joined bystrong noncovalent forces. The enzyme is heat stable up to60-70°C and resistant to SDS binding and denaturation. Gly-

cosylasparaginases have since been purified completely or par-tially from human, pig, cow, mouse, and Flavobacterium, andthey share many characteristics of the rat enzyme (Kaartinen etal, 1991; Tollersrud and Aronson, 1992; Tarentino and Plum-mer, 1993). Although eukaryotic glycosylasparaginases are ly-sosomal hydrolases, they have their greatest activity near neu-tral, or even basic pH.

Three complete cDNAs encoding human, mouse and Fla-vobacterium glycosylasparaginases have been cloned (Fisher etal, 1990; Tarentino et ai, 1994; Tenhunen et al, 1995). Thetwo subunits of these glycosylasparaginases are encoded by asingle gene. During translation of the human enzyme, the na-scent polypeptide, which is a pre-pro-protein, is translocatedinto the endoplasmic reticulum where there is cotranslationalcleavage of the signal peptide. Like other lysosomal enzymes,glycosylasparaginase is transported through the ER and theGolgi to lysosomes. In the ER the polypeptide is glycosylated,and an early processing step occurs which cleaves the pro-enzyme into the two subunits at position Aspl82-Thrl83 (cor-responds to Asp205-Thr206 in the previous studies) (Ikonen etal, 1993). This cleavage step produces a catalytically activeglycosylasparaginase (Fisher et al., 1993; Ikonen et al., 1993).Recent studies using the bacterial enzyme show that the cleav-age between Aspl51-Thrl52 (equivalent to human D182-Tl 83) is an autoproteolytic process (Guan et al., 1996), and theside chain nucleophile group on the N-terminal Thr of the psubunit participates in this activation step (Fisher et al., 1993;Guan et al., 1996). A previous study with an active-site inhibi-tor of the enzyme, 5-diazo-4-oxo-L-norvaline (DONV)(Tarentino and Maley, 1969), demonstrated that the same N-terminal Thrl83 of the mature human enzyme p subunit is alsoinvolved in substrate hydrolysis (Kaartinen et al., 1991). Mostrecently, glycosylasparaginase has been shown to belong to anew class of amidohydrolase termed 'N-terminal nucleophileamidases' (Brannigan et ai, 1995). In this set of enzymes anN-terminal Thr, Ser or Cys created by autoproteolysis serves asthe nucleophile in their catalytic reaction, and the exposeda-amino group on the same residue acts as the base which aidsdeprotonation and activation of the nucleophilic side-chain. Arecent crystal structure study of human glycosylasparaginaseshowed Thrl83 is located in the narrow center of tightly as-sociated a/p subunits and along with five other amino acidsinteracts with the aspartyl moiety of the substrate (Oinonen etal, 1995). Other studies have revealed aspects of the primarystructure of glycosylasparaginase with respect to its process-ing, maturation and reaction mechanism. A medically impor-tant example is the recessively inherited genetic disease AGU(aspartylglucosaminuria) which is caused by deficiency of gly-cosylasparaginase (reviewed in Mononen et al, 1993). Enrich-ment of the disorder occurs in Finland due to a founder with aCMOS substitution (Fisher and Aronson, 1991; Ikonen et al,1991; Mononen et al, 1991). This mutation causes defectivedisulfide bond formation in the a subunit, which in turn causes

© Oxford University Press 527

at Universidade E

stadual Paulista JÃºlio de M

esquita Filho on Novem

ber 14, 2013http://glycob.oxfordjournals.org/

Dow

nloaded from

http://glycob.oxfordjournals.org/


Y.Liu, G.S.Dunn and N.N.Aronson, Jr.

3 2 11.5

Marker 5

28.7 KD -

20.5 KD

- a subunit\

- B subunit0.00 0.0

c)

0 20 40 60 80 100

Fraction No.2.4

112 kDa-

53.2 kDa-

34.9 kDa28.7 kDa20.5 kDa

2.2-

2.0"

1.8

1.8-

1.4

3-Amylase (200 kDa)

Alcohol dehydrogenase (150 kDa)

SO glycosylasparaginase (101 kDa)

(high salt)

BSA (66 kDa)

Sf9 glycosylasparaginase (47 kDa)(low salt)

Carbonic anhydrase (29 kDa)

4

Ve/Vo

Fig. 1. Purification of glycosylasparaginase from Sf9 cells. A 5 liter culture of cells was pelleted, resuspended in pH 7.5 phosphate buffer, and homogenizedby sonication. Separation procedures are described in Materials and methods, (a) SDS-PAGE of glycosylasparaginase samples under denaturing conditions:Lanes 1-4, Coomassie Blue staining. Lane 5, silver staining. Lane 1, Con-A eluant; lane 2, low-salt DE52 eluant; lane 3, high-salt DE-52 eluant; lane 4,G-150-120 gel-filtration eluant; lane 5, CM52 eluant. (b) Two glycosylasparaginase activity peaks separated by DE-52 chromatography: Protein, absorption at280 nm, (•); Glycosylasparaginase activity, ( • ) . (c) Phastgel electrophoresis and immunoblot of Sf9 enzyme eluted from DE52 by low- and high-salt buffers:Holoenzyme and separated subunits were detected by antibodies against rat glycosylasparaginse a and (3 subunits. Lane 1, high-salt eluant denatured by 2%SDS, 1 M 2-mercaptoethanol and boiling; lane 2, low-salt eluant denatured by 2% SDS, 1 M 2-mercaptoethanol and boiling; lane 3, high-salt eluant,non-denaturing conditions; lane 4, low-salt eluant, non-denaturing conditions, (d) Molecular weight of Sf9 cell glycosylasparaginase based on gel filtration.

improper polypeptide folding and processing (Riikonen el ai,1994; McCormack et ai, 1995).

Although many new aspects of glycosylasparaginase struc-ture have been learned, more details of the mechanism bywhich the enzyme is activated and processed inside of cells aswell as its specificity towards substrates need to be determined.Our approach has been to collect and analyze primary struc-tural data for glycosylasparaginases from a number of speciesin order to learn which amino acids are conserved and aretherefore likely to be of functional importance. This study wastherefore designed from an evolutionary point of view to com-pare the glycosylasparaginase primary structure from severaldifferent species. Conserved amino acids become the basis inrelation to the recent human enzyme crystal structure for di-

recting mutagenesis investigations to further characterize thestructural biology and biochemistry of glycosylasparaginase. Anew glycosylasparaginase from the insect cell Spodoptera fru-giperda (Sf9) has been purified, its cDNA has been cloned, andits biochemical properties have been examined. Comparison ofthe insect glycosylasparaginase to the enzymes from other spe-cies (human, bovine rat, mouse and Flavobacterium) is alsoreported.

ResultsPurification of glycosylasparaginase from Sf9 cellsGlycosylasparaginase from Sf9 cells was purified approxi-mately 500-fold to homogeneity with a specific activity of 2.1

528

at Universidade E




Dow

nloaded from



Sf9 glycosylasparaginase

units/mg after concanavalin-A Sepharose affinity chromatog-raphy, DE52 anion exchange chromatography, G-150-120Sephadex molecular sieving chromatography and CM52 cationexchange chromatography (Table I, Figure 1). This enzyme isrichly expressed in this insect cell line, about 4-5-fold over thatin rat liver. In the Con-A step, about 15% of the enzyme ranthrough the column. Since this behavior was not due to columnoverloading, some oligosaccharide chains of Sf9 glycosylaspa-raginase may lack sufficient mannose structure for binding toCon-A. DE52 anion-exchange chromatography separated twoactivity peaks (Figure lb). Both high-salt and low-salt enzymefractions gave the same band pattern during SDS-PAGE underreducing conditions (Figure lc, lanes 1, 2), but the high-saltfraction had a single slow mobility band on native gel electro-phoresis compared to smaller proteins for the low-salt fraction(Figure lc, lanes 3, 4). The glycosylasparaginase fractionseluted by high- and low-salt buffer were separately concen-trated and chromatographed on G-150-120 Sephadex. Thelow-salt form eluted at a molecular weight of 47 kDa and thehigh-salt form eluted equivalent to a molecular weight of about101 kDa (Figure Id). Therefore, Sf9 glycosylasparaginase ap-pears to exhibit different quarternary structures, either a het-erodimer a(3 or a heterotetramer a232-

N-Terminal amino acid sequencing of Sf9glycosylasparaginase

Concentrated enzyme (10-20 |xg) was separated by denaturingSDS-PAGE into a and fi subunits (Figure la, lane 5) whichwere confirmed by immunoblotting with cross-reactive anti-bodies made against rat a and p subunits. These subunit bandswere transferred to a PVDF membrane and N-terminal aminoacids were sequenced. The amino acid sequence from bothsubunits and their comparison with the N-terminal subunits ofhuman and rat enzymes are shown in Figure 2. N-terminalsequences of the (3 subunit are more conserved among thesespecies, especially the first five identical amino acids TIGMV.Residue W11 of the a subunits, like the N-terminal T of (3subunits, is conserved, and both these amino acids are reportedto be essential for enzyme activity (Fisher etai, 1993; Oinonenet al, 1995; Riikonen et ai, 1995; Guan et ai, 1996).

SDS sensitivity

Native rat glycosylasparaginase, which resists SDS denatur-ation, has strong noncovalent interaction between its two sub-units. The rat enzyme remains completely active even in thepresence of 5% SDS at room temperature and neutral pH (Tol-lersrud and Aronson, 1989). The two subunits can be dissoci-

N-terminal amino acid sequence of glycosylasparaginase a-subunits

Human S S P L P L V V N T W P F K N A T E A A w R

Rat S N P L P L V V N T W P F K N A T E A A W W

Sf9 cells E K N I P I V I T T W S F T N A S Q K A I E

Consensus - - - - P - V - - T W - F - N A - - - A - -

N-terminal amino acid sequence of glycosylasparaginase p-subunit

Human T I G M V V I H K T G H I A A

Rat T I G M V V I H K T G H T A A

Sf9 cells T I G M V A V D S K G D V A A

Consensus T I G M V - - - - - - - - A A

Fig. 2. Experimental N-terminal sequences of Sf9 glycosylasparaginase aand p subunits (comparison to human and rat glycosylasparaginasesubunits).

ated in the presence of SDS only when the temperature is high(80°C) or the pH is below 5.5. The human enzyme is similarlyresistant to SDS (Tollersrud and Aronson, 1992). In contrast,Sf9 glycosylasparaginase irreversibly lost all activity in thepresence of 0.05% SDS at room temperature due to dissocia-tion of the enzyme into its subunits (Figure 3, inset).

Temperature stability

Glycosylasparaginase from all known species including hu-man, rat, pig, bovine and mouse enzymes has a high thermo-stability (Tollersrud and Aronson, 1992). Sf9 enzyme began toirreversibly inactivate at 60°C, and was completely inactivatedat 75°C (Figure 4). Native SDS-PAGE immunoblots showedthe slow migration of the native form of the enzyme shifted toits two denatured subunits when the temperature reached 60-65°C (Figure 4, inset). Enzyme denatured by the higher tem-perature precipitated in the gel loading well, which explains thelower amounts of immunoreactive forms in those samples.

pH dependence of activity and stability

The pH-activity profile of Sf9 glycosylasparaginase is shownin Figure 5a. Similar to glycosylasparaginases from other spe-cies (Tollersrud and Aronson, 1992), the Sf9 enzyme had abroad pH optimum between 5 and high basic pH at least up to10, where maximum enzyme activity was preserved even after

Table I. Purification of glycosylasparaginase from Sf9 cells

Enzyme sample

Cell lysateConcanavalin ADE-52 cellulose

Low-salt peakHigh-salt peak

G-150-120 gel filtrationLow MW peakHigh MW peak

CM-52 cellulose(High MW peak)

Total protein(mg)

26622

0.621.43

0.140.23

0.05

Total units(units)

2.331.21

0.110.38

0.0710.26

0.15

Specific activity(U/mg)

0.00440.055

0.180.26

0.531.15

2.1

Purification(fold)

112.5

40.660.0

119261

477

Yield(%)

10052.1

4.716.2

3.011.2

6.4

529

at Universidade E




Dow

nloaded from




0.2-

inoom

QO

0 . 1 -

SDS%

Native enzyme

Denatured subunits

g - CM wq 9 o o •-; ino o o o o o

f i t

o.o 4-0.0 0.1 0.2 0.3 0.4 0.5

SDS concentration (%)

0.6

Fig. 3. Effect of SDS on Sf9 glycosylasparaginase. Purified Sf9glycosylasparaginase (0.4 mU) was incubated with different concentrationsof SDS (0% to 0.5%) for I h at room temperature in a total volume of 3 uXOne half the sample was assayed for glycosylasparaginase activity ( • ) atpH 7.0 and 37°C for 1 h, and the other half was subjected to nativeSDS-PAGE (Phastgel) followed by immunoblotting with antibodies againstrat a and pj subunits (inset).

2 h at 37°C. The enzyme reaction appears to require a residuewith a pKa around pH 4. The pH-stability of the enzyme wasstudied by pretreating enzyme samples at different pH for 1 h atroom temperature prior to assay at pH 7.5 (Figure 5b). Sf9 cellglycosylasparaginase was stable over a broad range from pH4-10. The enzyme remained active and a single major highmolecular weight band representing the native enzyme mi-

00inQO

50 60 70 80

Temperature (°C)

90

Fig. 4. Temperature stability of Sf9 glycosylasparaginase. Purified Sf9glycosylasparaginase (0.25 mU) was pre-incubated at different temperaturesfor 15 min. One half of the sample was used to assay enzyme activity ( • ) .and the other half was subjected to native SDS-PAGE (Phastgel) anddetected by immunoblotting (inset).

grated on a Phastgel (Figure 5b, inset). Another faster-migrating band occurred and is similar to ones seen during pHor heat denaturation studies of rat glycosylasparaginase (Tol-lersrud and Aronson, 1989). The Sf9 enzyme was unstablebelow pH 3.5 and is totally dissociated into its subunits atpH 3.

Kinetics of Sf9 glycosylasparaginase and asparaginaseactivities

Hydrolysis of GlcNAc-Asn by the Sf9 enzyme displayed Mi-chaelis-Menten kinetics. The apparent A"m of Sf9 glycosylaspa-raginase for GlcNAc-Asn was 0.88 mM with a Vmax of 1.13nmol/min under the assay conditions. The Sf9 glycosylaspa-raginase also exhibited asparaginase activity. Hydrolysis of thenon-glycosylated amino acid displayed a Michaelis-Menten re-lationship with respect to the asparagine concentration. The Km

of the enzyme for asparagine was 3.0 mM and the Vmax was0.79 nmol/min.

Cloning of Sf9 glycosylasparaginase cDNA

The strategy for amplification and cloning of Sf9 glycosylaspa-raginase cDNA is depicted in Figure 6 (described in Materialsand methods). The complete cDNA and resulting deducedamino acid sequence of Sf9 glycosylasparaginase (Figures 7and 8) were obtained by combining the DNA sequences froma 5'-fragment BL1/NA89 and the 3'-RACE fragment amplifiedby primers NA65/NA2 (Figure 6). The calculated molecularmass of the a subunit is 20.8 kDa and the (3 subunit is 14.1kDa. The calculated pi of the enzyme is 6.92 using the GCGpackage of programs (University of Wisconsin). There is onlyone potential glycosylation site (N15 on the a subunit), whilehuman, bovine, rat and mouse glycosylasparaginase have thissite plus a second one on their (3 subunit. This pattern agreeswith immunoblotting experiments (Figure lc), which showedelectrophoresed denatured Sf9 enzyme had two immunoreac-tive a subunit bands, but only one for the (3 subunit. N-glycanase treatment also converted the two a-bands into one(unpublished data).

Cloning of rat and bovine liver glycosylasparaginase cDNAs

Partial rat liver cDNA sequence located at the middle region ofthe cDNA has been published previously (Fisher et ai, 1990).The known sequence was confirmed by RT-PCR, and 5'- and3'-RACE were used to clone the unknown portions of ratcDNA. Complete rat liver cDNA sequence and its deducedamino acid sequence are shown in Figures 7 and 8. CompletecDNA and amino acid sequences for bovine liver glyco-sylasparaginase were also obtained by similar methods (Fig-ures 7 and 8).

Sequence comparison of glycosylasparaginases from human,bovine, rat, mouse, insect (Sf9 cells) and Flavobacterium

Amino acid sequences from human, bovine, rat, mouse, Sf9cells (insect) and Flavobacterium were aligned by the 'pretty'program in the GCG computing package (Figure 8). Theiroverall sequences are highly conserved and are encoded bysingle genes. Identities compared to the human sequence are:83%, rat; 82%, mouse; 80%, bovine; 57%, insect; and 50%,Flavobacterium. There are 100 consensus residues amongmore than 300 amino acids in the glycosylasparaginases ofthese six species. T183 was conserved and is known to be anucleophile group both in the zymogen processing mechanism

530

at Universidade E




Dow

nloaded from




in

mQO

10

PH PH

Fig. 5. Effect of pH on activity and stability of Sf9 glycosylasparaginase. (a) Activity at different pH: Purified enzyme (0.5 mU) was assayed forglycosylasparaginase activity for 2 h in 20 mM sodium citrate, pH 3-5.5; or 20 mM sodium phosphate, pH 6-8; or 20 mM Tris-HCI, pH 8.5-10. All bufferscontained 0.15 M NaCl. (b) Stability of Sf9 glycosylasparaginase at different pH: Purified enzyme (1 mil) was mixed with 2 |xl buffer of varied pH (see a)and incubated at room temperature for 1 h. The pH of one half the sample was adjusted to 7.5 with 0.5 M sodium phosphate, and enzyme activity wasassayed by incubation at 37°C for 2 h. The other half of the sample was subjected to native SDS-PAGE (Phastgel) and analysed by immunoblotting (inset).

and in enzyme catalytic function (Fisher et at, 1993; Guan etai, 1996). In addition to T183, four other amino acids, R211,D214, T234 and G235, which were found to directly bind toaspartate in the human glycosylasparaginase crystal (Oinonenet ai, 1995), are conserved within these species. T201, anotherconserved residue in the (3 subunit, is functionally important informing a hydrogen bond with T183 (Oinonen et ai, 1995).Wl 1, which was suggested to be involved in glycosylasparagi-nase activity from a mutagenesis study (Riikonen et ai, 1995),

appears to be localized near the substrate binding pocket, andpossibly is involved in binding the carbohydrate portion ofsubstrate (Oinonen et ai, 1995). This tryptophan is conservedin all six species. A cluster of conserved amino acids includesC-terminal HD of the a subunit and TIGM on the N-terminusof the p subunit. Mutagenesis of these residues, such as alter-ation of the H, D (unpublished data) and T (Fisher et ai, 1993;Guan et ai, 1996), have suggested their importance in themechanism of zymogen processing and maturation as well as

Sf9 cell mRNA aaaaaaa

BL1 NA90

reverse transcription

NA65

polyT-adaptor (NA3)

First strand cDNA-* NAl-3

-* NA89

PCR1 (BL1/NA1-3); (NA90/NA2)PCR2 (BL1/NA89); (NA65/NA2)

adaptor (NA2)

ITA ligation into pBlueT7 vector and DNA sequencing

Fig. 6. Sf9 glycosylasparaginase cDNA amplification by RT-PCR (see Materials and methods).

531

at Universidade E




Dow

nloaded from




GAGAAGAACATGCCGATCGTAATAACGACATGGTCGTTTACAAATGCTTCACAAAAAGCC

TGGGAAGTTCTAAAGGATGAAGGCAAAGCTTTAGATGCTGTGGAACAAGGCGGTATAGTA

TGTGAAAACGAGCAATGTGATCGGACAGTAGGTTATGSAGGAAGTCCTGATGAAGATGGG

GAGACGACACTAGATGCTTTTATCATGGATGGGAGTACAATGAACGTTGGTGCGGTAGCA

GCACTCCGTAGGATAAAAAGTGCAATATCAGTTGCCCGACATGTAATGGAGTACACAACA

CATTCATTCCTAGCTGGAGAACTGGCTACGAAGTTTGCTGTTGAAATGGGATTCAAGGAG

GAGTrTCTTTCAACTGATGAATCCAGGGAATTATGGTCAAAATGGCGTTTTGAAAAACAA

TGCCAACCGAATTTTCGAAAGAATGTGAAACCAGATCCACGTAAACATTGTGGGCCTTAT

CATAAAAAGAGGAATTTTGTGGATTACATACATCCAGAAGTGTTTGTAGTTGATCAATAC

AATCACGACACGATTGGGATGGTTGCTGTGSACAGCAAGGGAGATGTAGCAGCT&STACC

TCCACTAATGGCGCTAAGTTCAAAATACCTGGAAGAGTAGGCGACTCTCCCATTCCCGGT

GCAGGGGCTTACGCTGACAATACAGTAGGAGGAGCAGCGGCTACAGGTAACGSCGACACC

ATGATGAGATTTTTACCTAGTTTCTTAGCGGTAGAGGAAATGCGTCGAGGAGCGTCACCG

GCGAATGCCGCGAAAACTGCAATCAAAAGAATCTCAGCACATCATCCTGATTTCATGGGT

GCTGTAATAGCTTTATCTAAGAATGGTCAATACGGGGCCGCGTGTAACGGTATCGAAACG

TTCCCATTCGTCGTACAAGATAAAACTCGAAAAACATTTGAAGTCGTTACTATTAAATGT

TGAtttttgtatatttcgtattatacctaaaaatctgtaaa*t»aatccaattttgaatt

BttcgtggctctggcggcggtctgggcgcgccagtAXGGCGTGGAAGCCGGGTCTTCTCCT

GCCGTTGCTGCTGTTACTGCTCGGCCCAGCCCCGGCGCGCTGCTTCGGCCAGCTGCCCTT

GGTCCTCAACACTTGGCCTTTCAGGAATGCAACCGTGGCAGCGTGGAAGACGTTGGCGGC

CGGAGATTCCGCCCTGGACGCGGTTGAGAGTGGCTGCGCGACCTGCGAGCAGCAGCAGTG

TGACGGCAGCGTGGGCTTCGGGGGCAGCCCGGACGAGTCGGGGGAGACCACGCTGGACGC

CATGATCATGGACGGCACTACCATGAACGTGGGAGCAGTCGGAGACCTTAGACGAATTAA

AAATGCCATTGGGCTGGCGCGCAAAGTCCTGGAACACACGACACACACGCTGTTGGCAGG

AGAGGCAGCCACTAAGTTTGCCGAAAGCATGGGCTTTATCAATGAGGATTTATCCACCAA

CGTTTCTCAGGCTCTTCATTCAGACTGGCGTGCTCGAAATTGCCAACCAAATTACTGGAA

AAATGTTATACCAGATTCTTCAAAATACTGTGGACCCTACAAACCACCTACTGTCTTAAA

ACGAGATGGTATCACCTACGAAGATACAGCACAGAGAATCGGTCATGATACTATTGGCAT

GGTCGTCATCCATAAGACAGGAAATATTGCTGCTGGTACATCTACAAATGGTATAAAATT

CAAAATACCTGGCCGAATAGGAGACTCACCGATCCCTGGGTCTGGGGCCTACGCTGACGA

CATGGTGGGAGCCGCGGCAGCCACGGGCGACGGGGACATCCTGATGCGCTTCGTCCCAAG

CTATCAAGCTGTAGAATATATGAGAAGAGGACCACAGCTTGTGAAAAAGTGATCTCAAGA

ATCCAGAAGTATTTTCCCAAATTCTTTGGGGCTGTGATATGTGCTAACGTGACCGGAAGT

TATGGTGCTGCTTGCAATAAACTTTCAACATTTACTCAGTTTCCATTCATGGTTTATAAT

CCTCTAAAAAGTGCGCCAACTGAGGAAAAAGTAGACTGCATCTAAtccagcttttctgtc

ageatctatatttaaagaagaaacagaagctgaaaaggctactcactgtcattgagtaat

gtgcctcatgattctaaactctgtaaaataagcacaacaataaattttatgctgaaacag

acgctgctcggcgggtctgcttctagctilSGCGCGGAAGTGGAATCTCCCTTTTCTTCT

CCTGCCGCTCGTTCTGGGTATACCCCTGGTGCGGGGCTCCAACCCTCTACCCCTGGTCGT

CAACACTTGGCCTTTTAAGAATGCCACCGAAGCAGCGTGGTGGACATTAGTGTCTGGAGG

TTCTGCCCTGGATGCGGTGGAGAAGGGCTGTGCTATGTGTGAAAAGGAGCAATGTGGTGG

TACTGTAGGCTTTGGAGGAAGTCCTGATGAAGTTGGGGAAACCACCCTGGATGCCATGAT

CATGGATGGCACTGCTATGGATGTGGGAGCAGTGGGAGGCCTTAGAAGAATTAAAAATGC

TATTGGAGTGGCTCGGAAAGTCTTGGAGCATACCACGCACACGCTTTTAGTGGGGGACTC

AGCCACCAAGTTTGCTGTAAGTATGGGGTTTACCAGTGAGGACTTATCTACCAATACCTC

AAGAGCTCTTCATTCAGATTGGCTTTCTCGAAACTGCCAGCCAAATTATTGGAGAAACGT

TATACCAGATCCTTCAAAATACTGTGGACCCTACAAACCACCTGATTTCTTAGAGCAGAA

TAACCGTGCCCACAAAGAAGTGGATATCCACAGTCATGATACTATTGGCATGGTTGTAAT

CCATAAGACGGGACATACTGCTGCTGGCACATCTACAAATGGTTTAAAATTCAAAATACC

TGGTCGTGTAGGAGACTCACCAATCCCTGGAGCCTATGCTGATGACATGGCTGGAGCAGC

TGCGGCCACTGGAGATGGTGACACACTCCTGAGATTTCTGCCAAGCTACCAAGCTGTAGA

ATATATGAGAGGAGGAGATGATCCAGCCAGAGCTTGCCAAAAAGTGATTTCAAGAATTCA

GAAATATTATCCAAAATTCTTTGGAGCTGTTATATGTGCCAATGTGACAGGAAGTTATGG

TGCTGCTTGCAACAGACTTCCAACATTTACTCAATTTAGTTTCATGGTTTATAACTCTCT

ACACAATCAGGCGATTGAAGAAAAAGTAGACTGCATGTAAtctctcttgtctgtcaacat

atgtatttaaagaggaaagacgtgacggcgggaaggttgctcactgacatggagtgatcc

tcaattatttcaagttcttgtgtaaattaagtgcaaaaataaatatgctgcagtgttgtt

tctgtatct

60

120

180

240

300

360

420

480

540

600

660

720

780

840

900

960

1020

60

120

180

240

300

360

420

480

540

600

660

720

780

840

900

960

1020

1080

1140

1200

60

120

180

240

300

360

420

480

540

600

660

720

780

840

900

960

1020

108O

1140

1200

Fig. 7. Glycosylasparaginase cDNA sequences from (A) Sf9 cells; (B)bovine liver; and (C) rat liver.

in the mechanism of enzyme activity. Disulfide bonds contrib-ute significantly to proper intra-subunit conformation and fold-ing of human glycosylasparaginase. Human enzyme containsfour pairs of disulfide bonds: C41-C46 and C140-C156 of the

a subunit, C263-C283 and C294-C322 of the p subunit (Oi-nonen et ai, 1995). All these cysteines are conserved in bo-vine, rat and mouse, suggesting a similar structure of disulfidebridges in these three species. Sf9 cell enzyme has three pairsof these cysteines, but lacks both C263 and C283. The Flavo-bacterium enzyme has the most different amino acid sequencepattern, suggesting a different mechanism might be involved inits folding. It contains two pairs of disulfide bonds, one pair isformed with the two cysteines in its a subunit, another pair isin the p subunit (Tarentino et ai, 1994). The bacterial enzymelacks about 30 amino acids at the C-terminus of its a subunitin comparison with other species, however the C-terminal HDis preserved. It is known that the mature human enzyme asubunit loses 10-20 amino acids from its C-terminus in thelysosomes (Ikonen et ai, 1993; Oinonen et ai, 1995). Thismodification does not affect enzyme activity.

Discussion

A new glycosylasparaginase from the insect Spodoptera fru-giperda has been purified and cloned. Purification includedCon-A, DE52, G-150-120 and CM52 chromatography (TableI, Figure 1). The purified amidase had a specific activity of 2.1unit/mg. N-terminal amino acid sequences of both a and (3subunits were obtained chemically (Figure 2), and use of maxi-mal degenerate primers based on these data allowed RT-PCRamplification of the cDNA for this insect glycosylasparaginase(Figure 6) and deduction of its amino acid sequence (Figure 8).The biochemical properties of Sf9 glycosylasparaginase aresimilar to glycosylasparaginases from other species. The en-zyme has two forms with molecular weight of 105 kDa (a2b2)and 52 kDa (ab) (Figure Id). It is composed of two subunitswhich are encoded by a single gene, and they are joined to-gether by strong noncovalent forces which are not easily bro-ken by a wide range of temperature (Figure 4) and pH (Figure5b). The enzyme is stable up to 60°C and then rapidly dena-tures between 65°C and 70°C (Figure 4). The enzyme has abroad pH optimum, maintaining more than 85% of its activityfrom pH 3.5 to 10 (Figure 5a). Within this pH range at roomtemperature the enzyme acts as a stable holoenzyme withoutdissociation into subunits. In contrast to glycosylasparaginasefrom other species (Tollersrud and Aronson, 1992), SDS couldeasily dissociate the two subunits of the Sf9 enzyme (Figure 3).

The two different conformations of Sf9 glycosylasparagi-nase, a3 and a2p2 (Figure lb-d), differ from the heterodimerconformation of enzyme from most other species. The humanenzyme was reported to be a heterotetramer which forms crys-tals that are associated as an a@Pa structure (Oinonen et ai,1995). However, a heterodimeric conformation of human gly-cosylasparaginase appears to exist naturally (Tollersrud et ai,1994). This controversy is explained by the two conformationsof Sf9 enzyme. The tetramer may be a more stable conforma-tion, however a(3 dimer is the core unit of the enzyme whichis structurally sufficient for enzyme activity. The ionic char-acteristics of the two different Sf9 glycosylasparaginase con-formations are different as indicated by their separation onDE52 anion exchange resin (Figure lb). The 47 kDa form (aP)was eluted by a lower salt concentration than the 101 kDa form(a2p2). An interesting structural feature of Sf9 glycosylaspa-raginase is that it has only one potential glycosylation sitewhich is located on its a subunit (Figure 8). Both human,bovine, rat and mouse enzymes contain one glycosylation siteon each of their subunits. During Con-A affinity chromatog-

532

at Universidade E




Dow

nloaded from




L°LHumanGAG

BovineGAGRatGAG

MouseGAGsf9GAG

FlavoGAGConsensus

SSPLPLWNTFGQLPLVLNTSNPLPLWNTSSPLPLWNTEKNIPIVITTTTNKPIVLST

P-V—T

WPFK.NATEAWPFR.NATVAWPFK.UAiEAWPFK.NATEAWSFT.NASOKWNFGLHANVEW.-F A

AWRALASGGSAWKTLAAGGSAWWTLVSGGSAWWTLLSGGSAWEVLKDEGKAWKVLSKGGKA W — L G-

ALDAVESGCAALDAVESGCAALDAVEKGCAALDAVENGCAALDAVEQGGIALDAVEKGVRALDAVE-G—

MCEREQCDGSTCEQQQCDGSMCEKEQCGGTVCEKEQCDGTVCENEQCDRTLVEDDPTERS— E

59VGFGGSPDELVGFGGSPDESVGFGGSPDEVVGFGGSPDEGVGYGGSPDEDVGYGGRPDRDVG-GG-PD—

HumanGAGBovineGAG

RatGAGMouseGAG

sf9GAGFlavoGAG

Consensus

******GETTLDAMIM

* *****DGTTMDVGAV

GETTLDAMIM DGTTMNVGAVGETTLDAMIMGETTLDAMIMGETTLDAFIMGRVTLDACIMG—TLDA-IM

DGTAMDVGAVDGTAMDVGAVDGSTMNVGAVDEN.YNIGSVD G-V

GDLRRIKNAIGDLRRIKNAIGGLRRIKNAIGGLRRIKNAIAALRRIKSAIACMEHIKNPI

IK--I

GVARKVLEHTGLARKVLEHTGVARKVLEHTGVARRVLEHTSVARHVMEYTSVARAVMEKT—AR-V-E-T

THTLLVGESATHTLLAGEAATHTLLVGDSATHTLLVGDSATHSFLAGELAPHVMLVGDGA-H—L-G—A

USTTFAQSMGFITKFAESMGFITKFAVSMGFTTKFAESMGFTTKFAVEMGFKLEFALSQGFK— FA GF-

HumanGAGBovineGAG

RatGAGMouseGAG

sf9GAGFlavoGAG

Consensus

HumanGAGBovineGAG

RatGAGMouseGAG

sf9GAGFlavoGAG

Consensus

HumanGAGBovineGAG

RatGAGMouseGAG

sf9GAGFlavoGAG

Consensus

NEDLSTSA.SQ ALHSDW.LA.R NCQPNYWRNV IPDPSKYCGP YKPPGILKQDNEDLSTNVSQ ALHSDW.RAR NCQPNYWKNV IPDSSKYCGP YKPPTVLKRDSEDLSTNTSR ALHSDW.LSR NCQPNYWRNV IPDPSKYCGP YKPPDFLEQNNEDLSTKTSR DLHSDW LSR NCQPNYWRNV IPDPSKYCGP YKPSGFLKQSEESLSTDESR ELWSKWRFEK QCQPNFRKNV KPDPRKHCGP YHKKRNFVDY

KENLLTAESE KEWKEWLKTS QYKP-E-L-T—S- W P

178IPIHKETEDDGITYEDTAQRNRAHKEV.DIISPHKEEVDIIHPEVFWDQ

IVNI

PRGHDTIGMWIGHDTIGMWHSHDTIGMWHSHDTIGMWYNHDTIGMVAENHDTIGMIA—HDXIGM—

A

IHKTGHIAAGIHKTGNIAAGIHKTGHTAAGIHKTGHTAAGVDSKGDVAAGLDAQGNLSGA

G

TSTNGIKFKITSTNGIKFKITSTNGLKFKITSTNGIKFKITSTNGAKFKICTTSGMAYKM—I-G K-

HGRVGDSPIPPGRIGDSPIPPGRVGDSPIPPGRVGDSPIPPGRVGDSPIPHGRVGDSPII-G£-G£SPI-

GAGAYADDTAGSGAYADDMVGAGAYADDMAGAGAYADDTAGAGAYADNTVGAGLFVDNEIG-G D

ILMRFLPSYQILMRFVPSYQTLLRFLPSYQTLLRFLPSYQTMMRFLPSFLEVIRTVGTHL

R

AVEYMRRGEDAVEYMRRGENAVEYMRGGDDAVEYMRGGDDAVEEMRRGASWELMNQGRT-VE-M—G—

/\s\f\S\A/\/\/\/\s

PTIACQKVIS RI QPTTACEKVIS RI QPARACQKVIS RI QPAIACQKVIL RI QPANAAKTAIK RI SPQQACKEAVE RIVKIVNRRGP—A RI

* * * * * * * AA

GAAAATGNGDGAAAATGDGDGAAAATGDGDGAAAATGDGDGGAAATGNGDGAATATGHGEG-A-ATG-G-

KHFPEFFGAVKYFPKFFGAVKYYPKFFGAVKYYPNFFGAVAHHPDFMGAVKNLKDIQVGF

ICANVTGSYGICANVTGSYGICANVTGSYGICASVNGSYGIALSKNGQYGIALNKKGEYGI G-YG

HumanGAGBovineGAG

RatGAGMouseGAG

sf9GAGFlavoGAG

Consensus

AACNKLSTFTAACNKLSTFTAACNRLPTFTAACNKLPTFTAACNGIET..AYCIQ DA-C

********QFSFMVYNSEQFPFMVYNPLQFSFMVYNSLQFSFMVSNSL. FPFWQDKTGFNFAVHDQK-F-F-V

****** I 3 2 3

KNQPTEEKVD CIKSAPTEEKVD CIHNQAIEEKVD CMHNEPTEKKVD CIRKTFEWTIK C.GNRLETPGFA LK

Fig. 8. Species comparison of glycosylasparaginase deduced amino acid sequences. Glycosylasparaginases from human, bovine, rat, mouse, Sf9 cells andFlavobacterium were aligned using the GCG program 'pretty'. The beginning of the a and P subunit is marked above the amino acids by arrows. Enzymeactivation cleavage site between D182 and T183 is indicated with an arrowhead. Disulfide bonds of human enzyme connect corresponding Cys residues(bold) according to the crystal structure [15]. Consensus amino acids located in the active site of human glycosylasparaginase [15] are underlined.N-glycosylation sequenons are underlined. Amino acids are numbered according to human glycosylasparaginase amino acid sequence with its mature asubunit N-terminal residue being 1. Amino acids that form a-helices (AAAA) and (3-sheets (****) are marked above the appropriate residues [15].

533

at Universidade E




Dow

nloaded from




raphy, there is a small amount of insect enzyme that runsthrough the column, probably due to lack of glycosylation.This small amount of enzyme is active, and therefore the oli-gosaccharide is not essential to maintain activity. This suggeststhe sugar portion probably does not seriously affect conforma-tion or stability of the enzyme once it is folded. Recent muta-genesis studies of human glycosylasparaginase glycosylationsites showed that glycosylation of both subunits of humanglycosylasparaginase is not essential for proper enzyme pro-cessing and activation. The glycosylation site on the p subunitprobably is responsible for lysosomal targeting of the enzymeand does appear to positively influence folding and stability(Park and Aronson, 1996).

Bovine and rat glycosylasparaginse cDNAs were also cloned(Figure 7). To date, complete primary structures of six speciesfrom this enzyme family are known (Figure 8). The overallsequences of glycosylasparaginase are highly conserved. En-zymes from human, bovine, rat and mouse have higher iden-tities between each other than enzymes from bacterium andinsect which reflects the evolutionary relationship of these spe-cies. About 30% of the amino acids are conserved among allthese species. Functional importance of some amino acids havebeen suggested by recent mutagenesis and crystal structurestudies (Fisher et al., 1993; Guan et al., 1996; Oinonen et al.,1995; Riikonen et al., 1995). Six conserved amino acids whichare all in the P subunit are suggested to be important in theenzyme active site. It also will be important to understand theresidues at the junction of the subunits, such as HI 81 and D182at the C-terminus of the a subunit, since they constitute theactivation cleavage site with Thrl83. The conserved aminoacids close to the N-terminal of the P subunit, such as1841GM(V/l)(V/A)]88, may be structurally important to forma fj sheet with the terminal Thr at the active site (Oinonen etal., 1995). Functions of conserved amino acid clusters in theN-terminal portion of the a subunit, such as 30ALDAVE35,50VG(F/Y)GGSPD57, and 6JTLDA-IMD70, are unknown.The N-terminal half of the a subunit is structurally close to theenzyme active site (Oinonen et al., 1995), suggesting it may beinvolved in the reaction mechanism. In the P subunit, twoconserved amino acid clusters, 2/0GR-GDSPI-G-G22/ and229G-AAATG-G257, may be close to the active site becauseR211, D214, T234 and G235 interact with substrate in thecrystal structure (Oinonen et al, 1995). The amino acid se-quence of Flavobacterium glycosylasparaginase varied themost from the enzymes of other species, especially the C-terminus of its a subunit. It has been known that the C-terminus of human enzyme a subunit is somewhat like a 'proregion' as it is cleaved after the enzyme has been transportedinto lysosomes. It is interesting to note the C-terminal part ofthe mature a subunit projects away from the core ot|3 hetero-dimer in the human enzyme crystal structure (Oinonen et al.,1995). This likely indicates that after the precursor zymogenwas translocated into the endoplasmic reticulum, correct fold-ing of this region appropriately oriented the tertiary structure toenable auto-cleavage at Thr 183, but it remained protrudingfrom the a subunit such that it was proteolytically cleaved inlysosomes. Although this region is far away from the enzymeactive site, its structural importance is suggested by the FinnishAGU allele. Thus, Cysl40 is located in this region and itsmutation to Ser causes the precursor zymogen to fold defec-tively due to inability to form the C140-C156 disulfide bond.Possibly this protruding region has a function during enzymetransport and targeting or a role in stabilising the holoenzyme

before it enters lysosomes. Expression of the bacterial enzymein a mammalian system (COS1 cells) yielded significant gly-cosylasparaginase activity comparable to the expression ofwild type human enzyme (unpublished data). However, expres-sion of a bacterial a and human p chimeric construct did notresult in glycosylasparaginase activity (unpublished), althoughthe HDTIGM activation cleavage site was preserved. Thesestudies suggest that folding of the bacterial enzyme a and Psubunits as well as their interaction and processing to form anactive enzyme may differ mechanistically from the humancounterpart.

Materials and methodsEnzyme assays

Glycosylasparaginase activity was assayed by measuring the release of N-acetylglucosamine from the substrate N4-((3-A'-acetylglucosaminyl)-L-asparagine (GlcNAc-Asn) (Bachem, Torrance, CA, U.S.A.) (Tollersrud andAronson, 1989). Reactions with 2.5 mM substrate in 20 |j.l of 20 mM sodiumphosphate buffer, pH 7.5, were incubated with enzyme for appropriate times at37°C and stopped by boiling for 3 min after adding 50 u.1 of 250 mM sodiumborate buffer, pH 8.8. Released N-acetylglucosamine was assayed by theMorgan-Elson reaction. One unit of enzyme liberates 1 u.mol of N-acetylglucosamine per minute. Asparaginase activity was assayed at 37°C bya coupled enzymatic procedure with glutamic/oxalacetic transaminase (GOT)and malate dehydrogenase (MD) to measure release of aspartic acid fromasparagine (Tarentino and Maley, 1969): i) aspartate + a-ketoglutarate - GOT—> oxalacetale + glutamate; ii) oxalacetate +NADH - MD —> malate + NAD+.The overall reaction was measured by the decrease in NADH absorption at 340nm. The assay was in 600 (xl of 20 mM sodium phosphate, pH 7.5, containing0.1 mM a-ketoglutarate (Sigma), 0.1 mg NADH (Sigma), and 7.5 units eachof GOT (Sigma) and MD (Sigma). L-Asparagine (1 mM) (Sigma) was used assubstrate, and a unit of enzyme liberates 1 u.mol of aspartic acid per minute.

Purification of glycosylasparaginase from Sf9 cells

Sf9 cells were spin-cultured by National Cell Culture Center (Cellex Biosci-ences, Inc., Minneapolis, MN, USA) in 10 liters of HyQ CCM-3 medium(Hyclone Lab. Inc., Logan, UT, USA) to a final cell density of l.75xlO6/mlwith a viability of 99%. Cells were collected at 4°C by centrifugation at 2000rpm for 10 min. The cell pellets were washed once with PBS and then resus-pended in 50 ml of 50 mM sodium phosphate buffer, pH 7.5, containing 0.15M NaCl. Cells were homogenized by sonication followed by centrifugation at15,000 rpm for 30 min at 4°C. The cell lysate supernate was subjected tochromatography in the cold on concanavalin A Sepharose (Sigma) equilibratedwith 50 mM sodium phosphate buffer, pH 7.5, containing 0.15 M NaCl.Non-binding proteins were washed from the column using 300 ml buffer, andbound glycosylated proteins were eluted by 0.2M methyl a-mannoside(Sigma). Eluted glycosylasparaginase was concentrated to about 20 ml in anAmicon ultrafiltration cell fitted with a YM-10 membrane (Amicon Corp.,Danvers, MA, USA). The sodium phosphate buffer was changed to 0.02 MTris-HCI, pH 8.5, by three repeated ultrafiltrations. The concentrated enzymesample was applied to a DE52-cellulose (Whatman Laboratory Division,Maidstone, England) anion-exchange column that had been pre-equilibratedwith 0.02 M Tris-HCI, pH 8.5. The column was eluted at 0.5 ml/min with a400 ml linear NaCl gradient (0-0.4 M in the same Tris buffer). Four mlfractions were collected, and those that contained glycosylasparaginase werecombined, concentrated to 3 ml and subjected to gel filtration on SephadexG-150-120 (Sigma) equilibrated with 0.02 M sodium phosphate buffer, pH5.5. The column was run at 0.25 ml/min and 2.5 ml fractions were collected.Fractions containing glycosylasparaginase activity were combined and con-centrated to 3 ml with a YM-10 membrane. This sample was subjected toCM52-cellulose (Whatman) cation-exchange chromatography on a columnequilibrated with 0.02 M sodium phosphate buffer, pH 5.5. Glycosylasparagi-nase was eluted by 100 ml of the same buffer containing a linear NaCl gradientfrom 0 to 0.4 M. The column was run at 0.25 ml/min and 2 ml fractions werecollected. Glycosylasparaginase fractions were combined and concentrated toI ml by Centricon 10 ultrafiltration (Amicon Corp.).

N-Terminal peptide sequence analysis

Purified Sf9 glycosylasparaginase (10-20 u,g) was heat denatured in the pres-ence of 2% SDS/I M 2-mercaptoethanol and applied to a 15% SDS-PAGE gel

534

at Universidade E




Dow

nloaded from




according to Laemmli. These conditions dissociated glycosylasparaginase intoits two subunits. After electrophoresis, the peptides were electroblotted at 400mA for 1 h with cooling to a pre-wetted PVDF membrane (Immobilon, Mil-lipore Corp,. Bedford, MA, USA) using protein transfer buffer containing0.3% Tris-base, 1.5% glycine and 20% methanol. The PVDF membrane wasstained with 2% Coomassie blue for lh and destained for lh at room tempera-ture with 10% (v/v) acetic acid/60% (v/v) methanol. followed by furtherdestaining in 10% (v/v) acetic acid/90% (v/v) methanol for 2-5 min until themembrane became almost white in color. The membrane was finally rinsedwith distilled water and air-dried. Peptides corresponding to the glycosylaspa-raginase a and B subunits were recognized by comparison with immunode-tected bands using antibodies against rat glycosylasparaginase a and B sub-units (Tollersrud and Aronson, 1992). The dye-stained subunits were excisedand the N-terminus of each peptide was sequenced using an automated Edmandegradation method. For these analyses we thank A.L. Tarentino (WadsworthCenter for Laboratories and Research. New York State Department of Health.Albany, NY, USA).

Subunil electrophoresis and immunoblotting

Methods were modified from those of Tollersrud and Aronson (1992). Nativegel electrophoresis was carried out on an 8-25% Phastgel (Pharmacia. Piscat-away, NJ, USA) using Sf9 enzyme samples that had been mixed with 2:1 (v/v)3x running buffer (50 mM Tris-HCl, pH 6.8, with 0.05% bromophenol blueand 10% glycerol). Denaturing gel electrophoresis was done on either an8-25% Phastgel or 15% standard SDS-PAGE gel. Enzyme samples weremixed with 2:1 (v/v) 3x running buffer supplemented with 6% SDS and 3 mM2-mercaptoethanol. After electrophoresis, the Phastgel was blotted to a pre-wetted PVDF membrane by diffusion at 70°C for 1 h. Regular gels wereelectroblotted to a pre-treated PVDF membrane in transfer buffer at 400 mAfor lh with cooling. Nonspecific binding was blocked for 30 min with 1.5%(w/v) BSA in TBST (10 mM Tris-HCl, pH 8.0; 150 mM NaCl; 0.05% Tween20). Membranes were then incubated for 30 min at room temperature withantiserum against rat glycosylasparaginase diluted 1:1000 in the same buffer.After three washes, the immunocomplex was detected by a secondary anti-body/alkaline phosphatase conjugate kit (Promega Corp., Madison. WI, USA).

cDNA cloning using PCR methods (Figure 6 and Table II)

SJ9 glycosylasparaginase cDNA. Maximal degenerate oligonucleotides for Sf9cell glycosylasparaginase were synthesized (GIBCO BRL) corresponding tothe a subunit amino-terminus sequence, EKNIPIV, and that of the B subunit,TIGMVA. The 5'-upstream sense primer BL1 was used with two nested 3'-downstream antisense primers NA1-3 and NA89, which are separated by sixbases. RT-PCR amplification was initially done to sequence the a subunit(Figure 6). Sf9 cells were cultured in TNM-FH medium (GIBCO BRL) supple-mented with 10% FBS. Total RNA was isolated from the cells, and first strandcDNA was synthesized by reverse transcription using M-MLV reverse tran-scriptase (GIBCO BRL). PCR1 was used to amplify a 5'-cDNA fragmentusing primers BL1/NA1-3: 4 min denaturation at 94°C for the first cycle,followed by 30 repeated cycles of 50 s of denaturation at 94°C, 50 s of primerannealing at 42°C, and 50 s of DNA amplification at 72°C. A 5 min extensionat 72°C was used at the end of this PCR program. After PCR 1, PCR2 was doneusing the nested primer NA89 with primer BL1. PCR2 amplification programwas the same as PCR1 except the annealing temperature was increased to50°C. PCR amplified DNA products were electrophoresed on a 2% agarose geland the DNA bands were gel isolated and purified using QIAEX DNA ab-sorption resin (QIAGEN). Purified DNA fragments were cloned into pBlueT7

Table II. Primers used in amplification of Sf9, rat and bovineglycosylasparaginase cDNAs

BLI 5'-GARAARAAYATICCIATHGT-3' (sense)NA1-3 5'-GCIACCATICCDATNGT-3' (antisense)NA89 5'-ACCATGCCAATAGTRTCRTG-3' (antisense)NA3 5'-GAGAATTCGTCGACAAGCT,7-3' (antisense)NA2 5'-GAGAATTCGTCGACAAGCTT-3' (antisense)NA90 5'-CGAAGTTTGCTGTTGAAATGGG-3' (sense)NA65 5'-CATGATACTATTGGCATG-3' (sense)NA1 5'-ACTATTGGCATGGTTGTA-3' (sense)NA11 5'-GCTGCTGGTACATCTACAAATG-3' (sense)NA80 5'-AAAGCCAATCTGAATGAAGAGC-3' (antisense)NA81 5'-GGTAGATAAGTCCTCACTGGT-3' (antisense)PolyC 5'-TTCTAGAATTCGGATC12-3' (sense)

vector (Novagen), a TA cloning vector designed for PCR products amplifiedby Taq polymerase. To avoid artifacts of Taq polymerase, four transformationclones were picked, and plasmid DNA prepared from each was used fordouble-stranded DNA sequencing. 3'-RACE was used to amplify the B subunitcDNA sequence (Figure 6). First strand cDNA was reverse transcribed fromSf9 cell total RNA with NA3, a polyT-adaptor primer (Table I). PCR1 wasdone with the 5' upstream sense primer NA90 and the antisense adaptor primerNA2. NA90 was synthesized according to the prior-determined cDNA se-quence of the Sf9 a subunit. PCR2 was done using the same antisense primerNA2 with nested sense primer NA65, which is located at the 5' end of the Bsubunit, 215 bp downstream from NA90.

Rat and bovine glycosylasparaginase cDNAs. Total RNA was isolated fromrat liver and first strand cDNA was synthesized by a reverse transcriptionreaction using NA3. PCR and 3'-RACE strategies and methods were the sameas for amplification of Sf9 glycosylasparaginase cDNA except different senseprimers NA1 and NA11 were used. 5' RACE was used to determine the cDNAsequence coding the N-terminus of the rat a subunit. After reverse transcrip-tion of total rat liver RNA, the cDNA was cleaned using Wizard DNA puri-fication resin (Promega) to desalt and to eliminate small nucleotides such asprimers. A 3'-terminal deoxynucleotidyl tailing reaction (dGn) was carried outin 30 |xl containing 100 mM cacodylate buffer, pH 6.8, 1 mM CoCl2, 0.1 mMdithiothreitol. A concentration of 2 mM deoxyguanine and 30 units terminaldeoxynucleotidyl transferase (Promega) were added to the reaction, which wasincubated at 37°C for 30 min. A 5 u,l reaction sample was then used directlyas template for a 25 (xl PCR reaction. Two rounds of PCR amplification weredone by using 3'-downstream nested gene specific antisense primers, NA80and NA81 (Table II), with a 5'-upstream polyC-linker sense primer. The samestrategies were used to clone the cDNA for bovine liver glycosylasparaginase.

AcknowledgmentsThis work was supported by Public Health Service Grant DK-33314 from theNational Institute of Diabetes and Kidney Diseases.

AbbreviationsPAGE, polyacrylamide-gel electrophoresis; PBS, phosphate-buffered saline;PVDF, polyvinylidene difluoride; DONV, 5-diazo-4-oxo-norvaline; RACE,rapid amplification of cDNA ends; GOT, glutamic/oxalacetic transaminase;MD, malate dehydrogenase.

References

Aronson,N.N.,Jr. and Kuranda.M.J. (1989) Lysosomal degradation of Asn-linked glycoproteins. FASEB J., 3, 2615-2622.

Brannigan,J.A., Dodgson.G., Duggleby.H.J., Moody.P.C.E., Smith.J.L., Tom-chick, DR. and Murzin,A.G. (1995) A protein catalytic framework with anN-terminal nucleophile is capable of self-activation. Nature, 378, 416—419.

Fisher.K.J., Tollersrud.O.K. and Aronson,N.N.,Jr. (1990) Cloning and se-quence analysis of a cDNA for human glycosylasparaginase: a single geneencodes the subunits of this lysosomal amidase. FEBS Lett., 269, 440-444.

Fisher,K.J. and Aronson,N.N.,Jr. (1991) Characterization of the mutation re-sponsible for aspartylglucosaminuria in three Finnish patients. J. Biol.Chem., 266, 12105-12113.

Fisher.K.J., Klein,M., Park,H., Vettese,M.B. and Aronson,N.N.,Jr. (1993)Post-translational processing and Thr-206 are required for glycosylasparagi-nase activity. FEBS Lett., 323, 271-275.

Guan.C, Cui,T., Rao,V., Liao,W., BennerJ., Lin,C.-L. and Combs,D. (1996)Activation of glycosylasparaginase: formation of active N-terminal threo-nine by intramolecular autoproteolysis. J. Biol. Chem., 271, 1732-1737.

Ikonen.E., Baumann.M., Gron,K., Syvanen.A.C, Enomaa,N., Halila.R., Au-la,P. and Peltonen,L. (1991) Aspartylglucosaminuria: cDNA encoding hu-man aspartylglucosaminidase and the missense mutation causing the dis-ease. EM BO J., 10, 51-58.

Ikonen.E., Julkunen,!., Tollersrud.O.K., Kakkinen,N. and Peltonen.L. (1993)Lysosomal aspartylglucosaminidase is processed to the active subunit com-plex in the endoplasmic reticulum. EMBO J., 12, 295-302.

Kaartinen.V., Williams,J.C, Tomoch.J., Yates,J.R.,III, Hood,L.E. andMononen,I. (1991) Glycoasparaginase from human leukocytes: inactivationand covalent modification with diazo-oxonorvaline. J. Biol. Chem., 266,5860-5869.

McCormack,A.L., MononenJ., Kaartinen,V. and Yates,J.R.,HI (1995) Local-ization of the disulfide bond involved in post-translational processing of

535

at Universidade E




Dow

nloaded from




glycosylasparaginase and disrupted by a mutation in the Finnish-type as-partylglucosaminuria. J. Biol. Chem., 270, 3212—3215.

MononenJ., Heisterkamp,N., Kaartinen,V., Williams.J.C, Yates,J.R.,III,Griffen,P.R., Hood.L.E. and Groffen,J. (1991) Aspartylglucosaminuria inthe Finnish population: identification of two point mutations in the heavychain of glycoasparaginase. Proc. Nail. Acad. Sci. USA, 88, 2941-2945.

Mononen.I., Fisher.K.J., Kaartinen.V. and Aronson,N.N.,Jr. (1993) Aspartyl-glycosaminuria: protein chemistry and molecular biology of the most com-mon lysosomal storage disorder of glycoprotein degradation. FASEB J., 7,1247-1256.

Oinonen.C, Tikkanen.R., Rouvinen.J. and Peltonen.L. (1995) Three-dimensional structure of human lysosomal aspartylglucosaminidase. NatureStruc. Biol., 12, 1102-1108.

Park.H. and Aronson,N.N.,Jr. (1996) Glycosylation and phosphorylation oflysosomal glycoasparaginase. Arch. Biochem. Biophys., 328, 73-77.

Riikonen.A., Ikonen.E., Sormunen.R., Lehto,V.-P., Peltonen,L. and Jalanko,A.(1994) Dissection of the molecular consequences of a double mutation caus-ing a human lysosomal disease. DNA Cell Biol., 13, 257-264.

Riikonen.A., Tikkanen.R., Jalanko,A. and Peltonen.L. (1995) Immediate in-teraction between the nascent subunits and two conserved amino acids Trp34

and Thr206 are needed for the catalytic activity of aspartylglucosaminidase.J. Biol. Chem., 270, 4903^(907.

Tarentino.A.L. and Maley,F. (1969) The purification and properties of a 0-as-partyl N-acetylglucosylamine amidohydrolase from hen oviduct. Arch. Bio-chem. Biophys., 130, 295-303.

Tarentino.A.L. and Plummer,T.H.Jr. (1993) The first demonstration of a pro-caryotic glycosylasparaginase. Biochem. Biophys. Res. Comm., 197, 179-186.

Tarentino.A.L., Quinones.G., Hauer,C.R., Changchien,L.-M. and Plummer-,T.H.,Jr. (1994) Molecular cloning and sequence analysis of Flavobacteriummeningosepticum glycosylasparaginase: a single gene encodes the a and (3subunits. Arch. Biochem. Biophys., 316, 399^06.

Tenhunen.K., Laan.M., Manninen,T., Palotie,A., Peltonen.L. and Jalanko.A.(1995) Molecular cloning, chromosomal assignment, and expression of themouse aspartylglucosaminidase gene. Genomics, 30, 244—250.

Tollersrud.O.K. and Aronson,N.N.,Jr. (1989) Purification and characterizationof rat liver glycosylasparaginase. Biochem. J., 260, 101-108.

Tollersrud.O.K. and Aronson,N.N.,Jr. (1992) Comparison of liver glyco-sylasparaginases from six vertebrates. Biochem. J., 282, 891—897.

Tollersrud.O.K., Heiskanen.T. and Peltonen.L. (1994) Human leucocyte gly-cosylasparaginase is an a/p-heterodimer of 19kDa ot-subunit and 17 and 18kDa (i-subunit. Biochem. J., 300, 541-544.

Received on February 14, 1996; revised on April 3, 1996; accepted on April15, 1996.

536

at Universidade E




Dow

nloaded from



527.full

Documents

lowsalt eluant

highsalt eluant

lowsalt de52 eluant

boiling lane

gelfiltration eluant

kdasf9 glycosylasparaginase

kdaso glycosylasparaginase

glycosylasparaginase