Biochirnica et Biophysica Acta, 1161 (1993) 235-243 235 © 1993 Elsevier Science Publishers B.V. All rights reserved 0167-4838/93/$06.00

BBAPRO 34403

Characterization of two major isoforms of juvenile hormone esterase from Trichoplusia ni (Lepidoptera)

Grace Jones a, Maria Manczak a, Mietek Wozniak b and R'ykandar Ko'rrati a

a School of Biological Sciences, University of Kentucky, lexington, KY (USA) and b Graduate Center for Toxicology, University of Kentucky, Lexington, KY (USA)

(Received 2 June 1992)

Key words: Juvenile hormone; Juvenile hormone esterase; Post-translational modification; Isoform; Peptide mapping; Glycosylation

The two major isoforms of juvenile hormone (JH) esterase isolated from Trichoplusia ni were fragmented by cyanogen bromide and trypsin digestion. The resulting CNBr or CNBr/trypsin fragments were characterized and compared biochemically by SDS-PAGE, isoelectric focusing, two-dimensional eleetrophoresis and HPLC. Similar and unique fragments were examined for sequence, antigenic determinants and carbohydrate moieties. The studies identified small regions of the proteins which possess either potentially different sequences or different post-translational modifications. The location of a glyeosylated asparagine residue was determined, as well as a region containing an epitope probably composed of a linear sequence of residues. An N-terminal region was identified that contained charge variation between the two isoforms and the sequence was obtained for the only unique CNBr/trypsin fragment detected from that region. These are the first data on mapping of regions of charge variation, epitope location and glycosylation sites for this enzyme from any insect species.

Introduction

Juvenile hormone (JH) plays an important role in insect development and metamorphosis. This hormone prevents ecdysteroid-induced expression of new genes, but allows reexpression of previously expressed genes [1]. JH esterase is a major biochemical component in the regulation of JH levels during the last larval sta- dium of Lepidoptera, as a sharp increase in activity of the enzyme permits metamorphosis to proceed [2,3]. A first peak occurs just before wandering and the second peak appears during the prepupal stage of most Lepi- doptera [4].

JH esterase from Trichoplusia ni has been charac- terized in general terms by several authors [5-8]. The enzyme possesses two major isoforms with isoelectric points (p I ) 5.5 and 5.3 (isoform JHE-A and JHE-B, respectively) and two minor isoforms with p l 5.1 and 5.6 [9]. Both major isoforms occur mainly as glyco- proteins in the fat body, the site of their biosynthesis [10]. These forms are secreted into the hemolymph,

Correspondence to: G. Jones, School of Biological Sciences, Univer- sity of Kentucky, Lexington, KY 40546, USA. Abbreviations: JH, juvenile hormone; Con A, concanavalin A; ALO, after lights on; IEF, isoeleetric focusing, TBS, tris-buffered saline; JHE, juvenile hormone esterase.

with the same relative proportion of activity as in the fat body [8].

The molecular basis of the different electrophoretic mobility of JH esterase isoforms from T. ni is still unknown. JH esterase isolated from a closely related species, Heliothis virescens, is weakly immunologically related to the major isoforms of JH esterase isolated from T. ni [11], and occurs as a mixture of two proteins, the minor form having a two-amino-acid residue exten- sion on the NH2-terminal end [12]. JH esterase from Heliothis zea possesses two isoforms with different p I values, one of which occurs as a glycoprotein [13]. However, no data are available on the biochemical properties which distinguish individual purified iso- forms of this enzyme in any species or on the role of primary structure/post translat ional modification in causing the existence of multiple isoforms. Knowledge of the basis of the occurrence of the isoforms would show whether they are due to two genes, which if the case would offer two promoters for comparison of conserved sequences. Alternatively, if post-transla- tional modification is the cause, elucidation of how a consistent percentage is converted to each isoform may reveal a previously unknown regulatory mechanism controlling isoform ratios.

In the present study, we have used peptide mapping in combination with the size and charge characteristics

236

to identify and sequence regions containing similar and different charges between the two major isoforms. In addition, we have mapped the locations of glycosyla- tion and epitope sites and assessed their contribution to charge variation between isoforms.

Materials and Methods

Insects. Larvae of T. ni were reared on a pinto bean diet at 28 + I°C, 14 h light/10 h dark cycle as de- scribed elsewhere [14]. Last instar, day-three larvae (L5D3) were collected between 12-13 h after lights on (ALO) and treated topically with 1 nmol of a JH analog, fenoxycarb [ethyl(2-phenoxy-phenoxy-ethyl) carbamate] in 1 /.~1 ethanol carrier. Such treatment induces a high level of the enzyme, prior to harvest for purification [2-4]. Larvae were taken 12 h after treat- ment.

Chemicals. Juvenile hormone III was obtained from Calbiochem-Behring (La Jolla, CA) while 10-3H - labeled JH III was purchased from New England Nu- clear (Boston, MA) with a specific activity of 12 Ci/mmol. A stock concentration of JH III was pre- pared as 5- 10 -4 M unlabeled and 2.10 -6 M labeled JH Ill.

Molecular size markers for protein SDS-polyacryl- amide gels and reagent for protein assay were pur- chased from Bio-Lab (Richmond, CA). Concanavalin A (Con A)-Sepharose 4B and all other reagents were obtained from Sigma (St. Louis, MO).

Purification of JH esterase. JH esterase was isolated from the hemolymph of last instar, day four larvae after JH analog treatment by the method described elsewhere [11], with some modifications. Con A-Sep- harose 4B column chromatography was used instead the Sephadex G-100 gel filtration step described previ- ously. For further purification, the glycosylated (Con A reactive) forms of JH esterase were taken.

Cyanogen bromide (CNBr) fragmentation. For preparing CNBr fragments for single or 2-dimensional electrophoretic analysis, 10 to 40 ~g of JH esterase (0.15-0.6 nmol) were dissolved in 100/.H of 70% triflu- oroacetic acid (TFA). 5 mg of CNBr was added and the solution was incubated for 6 h in N 2 atmosphere at room temperature. After fragmentation the sample was lyophilized, dissolved in 50 /zl of water and lyophilized again. Conditions were used so as to achieve partial hydrolysis for the purpose of mapping the posi- tions of met residues (see text below).

Gel electrophoresis. SDS-PAGE was performed ac- cording to Ref. 15, with a 10-20% acrylamide concen- tration gradient. Wide-range isoelectric focusing (IEF) was performed in 5% polyacrylamide gels containing 2% ampholine pH 3.5-10.0 or 1.6% ampholine pH 3.5-10.0 and 0.4% ampholine pH 2.5-5.0. After IEF, the gel was soaked in 5% HC10 4 for 2 h, then in 25%

isopropyl alcohol/10% acetic acid for 1 h. This proce- dure was repeated once and the gel was soaked in water for 1 h, then in equilibration buffer (5% 2-mer- captoethanol, 0.3% SDS, 0.37 M Tris-HC1 (pH 8.8)) for 2 h. The fragments were finally transferred to nitro- cellulose in 25 mM Tris, 192 mM glycine containing 20% methanol as described [17] or stained by silver according to Ref. 18. The p l of each fragment of protein was calculated by comparison with the p l of standard proteins. Two-dimensional (2-D) electrophor- esis was performed according to Ref. 16 using slab gel IEF (2% ampholine pH 4.0-6.5) as the first dimension and SDS-PAGE (10-20% gradient gel) as the second dimension.

Immunological methods. A white, female, New Zealand rabbit was immunized with 250-280/zg of JH esterase, isoform A (pI 5.5), after additional purifica- tion by semipreparative SDS-PAGE (10% gel). Ho- mogenized gel pieces containing JH esterase were in- jected together with complete Freund adjuvant. Three booster injections were given in incomplete Freund adjuvant and the rabbit was bled three weeks after the last injection. Immunoblotting was carried out accord- ing to the procedure of Ref. 19 with blocking solution containing 20 mM Tris-HC1 (pH 7.4), 150 mM NaC1, 20% horse serum and 5% BSA, and washing solution containing 20 mM Tris-HCl (pH 7.4), 150 mM NaC1, 0.5% milk powder, 0.5% Triton X-100, 0.2% SDS. The antigens were probed with either 1 : 1000 diluted rabbit antiserum generated against hea t /SDS-PAGE- denatured JHE-A [8] or by antiserum generated against the N-terminus of JHE-A, and the antigen-antibody complex was detected with [125I]goat-anti-rabbit IgG. Antiserum against the N-terminal sequence (C-L-S-L- K-A-A-D-A-P-L-A-H-I-D-S-G-L) was prepared by Im- munodynamics. In order to directly compare and map the fragments immunor~active with each antiserum, blots initially analyzed with one antiserum were then stripped of that antibody and then reprobed with the other primary antibody. Autoradiograms resulting from each analysis were then overlaid to permit determina- tion of fragments immunoreactive to only one or to both antisera. For the purpose of obtaining sequence analysis (instead of mapping of met residues, glycosyla- tion sites, etc.), CNBr fragments were digested to smaller typsin fragments that are more quantitatively recovered from the membrane, and which provide clearer sequence determinations. Tryptic digestion of CNBr fragments and sequencing of HPLC-purified CNBr/trypsin fragments was conducted as described by Stone et al. [20]. Sequencing of N-termini of the intact proteins on PVDF membranes was performed similar to that described elsewhere [21].

Glycoprotein staining. Glycoproteins or glycopep- tides were transferred to nitrocellulose after separation by SDS-PAGE or IEF and detected by probing with

peroxidase-labeled Con A. The blot was incubated with 2% BSA in TBS (20 mM Tris-HCl (pH 7.4), 150 mM NaC1, 1 mM CaCI2, 1 mM MgCI 2) for 1 h at room temperature, then 2 ~g /ml of Con A-peroxidase was added and incubated for 1 h, then the blot was washed 4 times with 0.1% BSA in TBS and developed with 4-chloronaphthol [22].

Protein concentration. The protein concentration was measured according to the Bradford method [23] with Bio-Rad protein assay reagent using BSA as a stan- dard.

Results

SDS-PAGE of CNBr-cleavage fragments The patterns of sizes of CNBr-cleavage fragments of

the two JHE isoforms resolved by SDS-PAGE and silver staining were very similar (Fig. 1, panel 1). In independent experiments, 12 major fragments common to both isoforms were reproducibly observed. Staining was performed under conditions that enhance develop-

237

ment of color of the fragments. Under such conditions, fragment b (Fig. 1) was distinctly orange, in contrast to the grey color of fragments a, c, d of similar molecular size (30-35 kDa).

Concanavalin A affinity was used to probe the loca- tion of N-linked mannose-type glycosylation sites. This probe identified fragments b, c, f and g as possessing carbohydrate (Fig. 1, panel 3).

The location of epitopes and order of fragments was determined with two different antisera. Probing of the CNBr fragments with an antiserum generated against the N-terminal sequence of the protein located three fragments of 13, 29 and 35 kDa for both isoforms (Fig. 1, panel 2). Very long exposures did not reveal binding of the antibodies to any fragment less than 13 kDa. Thus, it appears that the first three methionine residues from the N-terminus follow at positions 13, 28 and 35 kDa. A fragment (e z) was reproducibly observed for JHE-B that was not observed for JHE-A.

When the CNBr fragments separated by SDS-PAGE were probed with antiserum generated against dena-

1 2 3 ,NSET CA A B C8 C~..~ CA A

A B I I

BCe!3"C~ CA A B Ca

Frag. kDa

a 34.5 b 33.0 c 31.6 d 30.2 e 27.8 e 21.9 f l 20 .6 g 18.9 h 17.5 i 15.3 j 14.4 k 13.1 1 11.8

Fig. 1. Electrophoretic pattern of two major isoforms of JH esterase after CNBr fragmentation and their interaction with antibodies or Con A. Panel 1, fragments visualized after silver staining. C A and Cn, control samples of 10 ng of JHE-A or JI-IE-B treated only with TFA; A, B, lanes containing 20/.~g of JHE-A or JHE-B, respectively, after CNBr fragmentation. Band e z was detected among fragments for JHE-B, but not for JHE-A. Inset: understaining of region containing fragments a-d, showing the resolution of those four fragments. Panel 2, immunoblot of fragments using antibodies generated against heat/SDS-PAGE-denatured JHE-A (no asterisk), or using antibodies generated against the N-terminus of JHE-A (asterisk). The blot shown in panel 2 was first probed with the anti-JI-IE-A antibody, exposed to X-ray film, and then stripped of the signal and exposed to the anti-N-terminal antibody. X 1 and X 2 indicate low abundance fragments not detected by silver staining,

but detectable immunologically. Panel 3, reactivity of fragments b, c, f and g with Con A lectin.

238

tured isoform A, it was observed that the epitope(s) for this antiserum were located on fragments other than the 13, 29 and 35 kDa partial digest fragments that are the most N-terminal fragments (Fig. 1, panel 2). Epi- topes were observed on fragments f, g, i and j. Epitopes were also observed associated with fragments a -d , although the resolution on single-dimension SDS- PAGE made it difficult to assess whether all four fragments, a -d , in fact possess epitopes.

2-dimensional electrophoresis of CNBr fragments Resolution of which fragments, a -d , possess epi-

topes to the anti-JHE-A antiserum was accomplished by 2-D electrophoresis ( IEF /SDS-PAGE) . It was pos- sible to relate the four fragments a - d from single dimension SDS-PAGE to the corresponding spots on 2-D gels. For example, band b stained strongly orange following either single or two-dimensional electro- phoresis, permitting identification of spot 11 on the 2-D gels as the same fragment as band b on SDS-PAGE gels (Fig. 2). Analysis of the other fragments yielded the identification of bands a, c and d (Fig. 1) as corresponding to spots 10a, 12 and 12a on the 2-D gels (Fig. 2). All four spots yielded immunoblot signals following probing with the anti-JHE-A antiserum. Since JH esterase is 66 kDa, and the four fragments are each in the size range of 30-35 kDa, some of these frag- ments must overlap in sequence, and therefore in epitope(s), with other fragments.

Comparative charge on N-terminal fragments The possible location of moieties for charge varia-

tion on the N-terminal-most fragments was assessed by probing the CNBr fragments with antibodies following isoelectric focusing. Silver staining following electrofo- cusing showed a strong similarity between the two isoforms for the most abundant fragments (Fig. 3, panel 1). Such a strong similarity following partial protein cleavage shows that there are few locations of net charge difference between the two proteins and that most of the corresponding partial digest fragments possess indistinguishable isoelectric points. When these fragments were probed with the anti-JHE-A antibod- ies, there was a strong similarity in the identity of fragments possessing epitopes (Fig. 3). These results show that many of the fragments of the two forms possessing homologous epitopes are also fragments that exhibit no distinguishable difference in net charge. As shown above, these fragments are in the C-terminal half of the protein.

When these CNBr fragments of this same blot were reprobed with the anti-N-terminal antibody, a pair of fragments were detected for each of the two isoforms. However, the two fragments detected for JHE-B were each charged differently than the corresponding frag- ment from JHE-A (Fig. 3, panel 3). These data clearly

MW

pH 6. 5.5 5.0

42

31

21

14

MW

pH

6.0 5.5 5.0

42.7

31.0.

21.5

14.4

Fig. 2. Two-dimensional electrophoretic analysis of CNBr fragments of JHE-A. Upper panel, visualization of fragments with silver stain- ing; Lower panel, immunoblot of fragments with antibodies gener- ated against heat/SDS-PAGE-denatured JHE-A. Spots 10a, 11, 12 and 12a correspond to bands a-d on SDS-PAGE (Fig. 1). All four

fragments contain an epitope(s) toward the antibodies.

establish that at least one moiety conferring a differ- ence in charge between the two isoforms is located on these N-terminal fragments.

N-terminal sequence As a test of whether there is a primary sequence

difference conferring charge variation located at the N-terminus of the isoforms, both JHE-A and JHE-B were subjected to N-terminal sequencing. The se- querjces of the two isoforms were found to be the same (Fig. 4). The N-terminal sequence obtained for JHE-A was longer, and antibodies were prepared against a

synthetic peptide corresponding to this additional re- gion, as a probe of whether a difference in epitopes (structure) exists there between the two isoforms. The identical immunoreactivity of the two isoforms with these anti-N-terminal antibodies (Figs. 1 and 3) shows that the residues and/or posttranslational modifica- tions participating in this N-terminal epitope are iden- tical in the two isoforms.

Sequence analysis of unique CNBr / trypsin fragments As a further identification of similar or charge-varied

regions of the two isoforms, sequence analysis was performed on CNBr/tryptic fragments separated by

239

HPLC. Comparison of the HPLC separation profiles of these fragments for the two isoforms shows that there exists a high degree of similarity in sequence and charge of the corresponding forms of the two frag- ments (Fig. 5). However, for each isoform, several unique fragments were resolved. For example, for JHE-A, peaks 64, 84, and 92 were unique and not observed in the HPLC profile for JHE-B. Conversely, JHE-B peaks 79 and 85 were unique for JHE-B.

Under the premise that similarly eluting fragments from JHE-A and JHE-B have the same sequence and posttranslational modifications (if any), the fragments recovered from the corresponding peaks of the two

pH

1 2 3

A B A B A B

4

8

7

m

Fig. 3. Electrophoretic pattern of fragments derived from JH esterase isoforms analyzed on IEF gels containing a 5 : 1 mixture of ampholine pH 3.5-10 and ampholine pH 2.5-5.0 (panel 1). Panel 1, lanes marked A or B contain JHE-A or JHE-B, respectively, after CNBr fragmentation; Panel 2, interaction of fragments with antibodies generated against heat/SDS-PAGE-denatured JHE-A. Most of the corresponding pairs of immunoreactive fragments from the two isoforms possess indistinguishable net charges; Panel 3, interaction of fragments with antibodies generated against an oligopeptide corresponding to the N-terminus of JHE-A. The results show a distinct difference in charge of the fragments connected to the N-terminus, with the fragments from JHE-B being more acidic. The blot shown in panel 3 was first probed with the anti-JHE-A antibody, exposed to X-ray film and then stripped of the signal and exposed to the anti-terminal antibody. Solid bars on panel 3 denote that the

positions of the immunoreactive fragments are not the same as the positions of immunoreactive fragments detected in,panel 2.

240

Nterminal seuuences

JHE-B JHE-A HV JHE MS JHE

LPSLSADAEAPS LPSLSADAEAPSPLSLKAD-APLAHIDSGLL

WQETNSRSWAHLDSGIIRG RIPSTEEVVVRTESG

HPLC ~ sequences for JHE-A. JHE-B

JHE-B I00 GSQYQDIESPTAYQSK JHE-B 153a gVGXIEXLTYVFK JHE-B 153b JHE-A 79 GSQYQDIESPTAYQSK JHE-A l19a WGHIEDLTYVFK JHE-A ll9b HV JHE GMQYEDIVSPTIIRSK HV JHE GVGHIEDLTYVFK HV JHE

tXXAPVYqYQF VVGAPVYLYQFSYESPSSAIKQE TG-GAPLYLYRFAYEGQNSIIKKV

*JHE-A 84 VTTLR *JHE-B 85 KIDTAYYXGTK JHE-A 145 LIELPPEKL JHE-A 90 GVPYAKQPIGELR HV JHE VTTLR HV JHE TIERKYYNGTI HV JHE LIDLPAEKL }IV JHE GVPYAKQPVGELR

Fig. 4. Alignments of sequences obtained for JHE-A and JHE-B fragments with corresponding sequences of other example reported esterases. HV JHE, Heliothis virescens JH esterase [12] where such alignments could be deduced; MS JHE, Manduca sexta JH esterase. Lower case letters indicate uncertain designation of the identity of the residue. Upper case letter X indicates the residue was not identified by sequencing. Asterisk indicates fragments sequenced that were unique for JHE-A or JHE-B. Alignments were made by first searching the HV JHE sequence with the

fragment sequence from T. ni and further cross-checking the alignment by eye.

samples should have the same amino-acid sequence. Conversely, unique peaks for a given isoform would correspond to fragments with a sequence and/or charge that is not the same as that found on the

79 119 %

0~

B 1°°%1 153 ,4

_ _ _ ~ . . . . . . . . L_ I . . . . . . . J _ _ I 4 0 . 0 0 6 0 , 0 0 80~ 80 ~-

ELUTION TINIE-MIMJTES Fig. 5. HPLC elution profile of CNBr/trypsin fragments for JHE-A (upper) and JHE-B (lower). Numbers indicate peak number as detected in HPLC eluate. Arrows with asterisk point to peaks unique to the given isoform. Arrows without asterisks indicate peaks com- mon to both isoforms. Bars on either plot denote absences of a peak found in the corresponding elution point of the other plot. These data document the isolation by HPLC of discrete fragments contain-

ing unique sequence and/or post-translational modification.

homologous fragment of the other isoform. As a con- trol that this premise is correct, fragments for JHE-A (79, 119a, 119b) and JHE-B (100, 153a, 153b, respec- tively) which have indistinguishable HPLC elution properties between the two isoforms were sequenced. As shown in Fig. 4, the sequences of these correspond- ing pairs of fragments for the two isoforms were the same.

The sequences of unique fragments were analyzed for possible sites of posttranslational modification, as such modifications may be the reason why the frag- ments exhibited unique HPLC retention properties. The sequence of a fragment from an HPLC peak unique to JHE-A possesses a potential phosphorylation site of T-X-R (JHE-A peak 84; fragment sequence = VTTLR, Fig. 4). Another fragment unique to isoform A (JHE-A peak 64; LALSK) has no obvious sulfation or phosphorylation site, but it is so short that other parts of such motifs may be in a flanking sequence. A unique fragment for the other isoform (JHE-B 85) appears to possess carbohydrate on a classical N-X-T motif (Fig. 4). The residue position corresponding to the asparagine residue was the only residue in the sequence that provided no signal for any amino acid during sequencing. However, a closely related esterase form with a similar sequence in the corresponding fragment region ofHeliothis virescens (TIERKYYNGTI [12]) possesses an asparagine at that position. Further, glycosylated asparagine residues do not provide a sig- nal corresponding to any amino acid during sequencing under the conditions used in this study.

Discussion

The data obtained in the present study permit the first comparison of locations of met residues, glycosyla-

241

tion site, epitopes and regions of charge variation for identified and isolated isoforms of JH esterase.

Comparison of positions of methionine residues Results of immunoblotting of partial digest frag-

ments with the anti-N-terminal antibody enable map- ping of locations of the met residues most proximal to the N-terminus (Fig. 6). These positions are 13, 29 and 35 kDa from the N-terminus.

Positions of glycosylation Several pieces of evidence suggest the location of at

least one glycosylation site. Con-A lectin did not bind to any of the fragments determined to be within the 35 kDa N-terminal half of the protein, showing that the detected carbohydrate is in the C-terminal half of the protein. This conclusion is consistent with the identifi- cation of a probable N-linked glycosylation site on a fragment that maps on a consensus esterase sequence to a location 39 kDa from the N-terminus (Fig. 6).

Since partial fragments b and c (Fig. 1) both possess a glycosylated residue(s), each must extend into the C-terminal part of the protein beyond the location of a (the) glycosylation site. Further, since the first met residue from the N-terminus is at about 13 kDa, the 32-33 kDa fragments b and c must extend to at least 45-46 kDa into the protein. This result is also consis- tent with the placement of a glycosylated residue at a position 39 kDa from the N-terminus (corresponding to HPLC fragment JHE-B 85, Fig. 4). That is, both frag- ments b and c will contain the glycosylation site. Even if both fragments possess the C-terminal residue of the protein, each would then initiate on the N-terminal side of the glycosylation site (39 kDa from the N- terminus), and thus contain that glycosylation site.

Regions containing epitopes The patterns observed following immunoblotting

with antiserum prepared against total, denatured (by heat/SDS-PAGE) JHE-A permit identification of re- gion(s) containing the epitope(s). Since the antiserum was prepared against denatured JH esterase, the de- tected epitopes are likely to be composed of a linear

series of amino acids and any posttranslational modifi- cations occurring on those residues. No epitopes for this antiserum were detected within the N-terminal 35 kDa region.

Previous studies established that the anti-JHE anti- serum will cross-react with the minor percentage of each isoform that exists non-glycosylated. Thus, at least one epitope exists in the C-terminal half of each iso- form that is not located at the glycosylation site. Previ- ous studies have also shown that the molecular sizes of JHE glycosylated and unglycosylated forms are indis- tinguishable under SDS-PAGE conditions that re- solved as little as 0.5 kDa difference (or, several man- nose residues). Since glycosylation at a given site with mannose usually involves at least several sugar molecules, it is unlikely that JH esterase from T. ni is glycosylated at more than one such site, and thus the glycosylation site at 39 kDa from the N-terminus is likely the sole such glycosylated site.

Fragments f and g, which are 19-20 kDa, also contain both epitopes and glycosylation sites, neither site being in the N-terminal half of the protein. On the basis of these results and the above discussions, we propose that both of these fragments contain the glyco- sylated site at 39 kDa from the N-terminus, and that the epitope contained on these fragments is located in the interval between the center of the protein and 59 kDa from the N-terminus. It can be concluded that the glycosylation! site is not the (only) epitope on this fragment, since the antibodies generated against JHE-A cross react with the minor percentage of JH esterase molecules that are not glycosylated. Immunoblotting of the fragments after SDS-PAGE identified a 17.5 kDa fragment in both isoforms that contains no epitope or glycosylation site and that is not the N-terminal frag- ment. On the basis of the mapped locations of methio- nine residues and the glycosylation site, this fragment initiates on the C-terminal side of the glycosylation site (Fig. 6).

Regions contain charge variation The results of immunoblotting partial digest frag-

ments with the anti-N-terminal antibody, after electro-

region not containing epitope or glycosylation

I 1

V B ' ~ b ~ B.- ' t~

16kd 33kd 49kd 66kd

I I I I

region c o n t a i n i n g c h a r g e var ia t ion region containing e p i t o p e

Fig. 6. Map of the location of various functional regions of JH esterase from T. ni M, location of met residues in the N-terminal half of the protein; triangle, location of glycosylated asparagine residue; bars indicate the location of sequenced internal fragments (bars with asterisk denote regions with unique HPLC elution properties). Notations above and below the map demark that the regions do or do not contain an

epitope or glycosylation site. The fragment numbers refer to fragment peaks identified in Fig. 5.

242

focusing of the fragments, established that at least two of the 13, 28 and 35 kDa N-terminus-connected frag- ments possess charge variation between the two iso- forms.

Sequencing of the N-termini of the two isoforms shows they have identical sequences for the first 12 residues. The indistinguishable immunoreactivity of the two isoforms toward antibodies generated against the remaining residues out to position 30 of isoform A suggests that this region contains no differences in sequences that impart differences in charge. Therefore, the basis for charge variation must reside in the inter- val between positions 3 kDa to 28 kDa from the N-terminus (Fig. 6).

We have sequenced fragments from JHE-A that have unique HPLC retention properties from the cor- responding fragments in JHE-B. The only fragment thus identified that occurs in the interval 3 kDa to 28 kDa from the N-terminus is JHE-A 84 (Fig. 4), which maps to a position initiating about 19 kDa from the N-terminus (Fig. 6). The identification of a fragment mapped to the C-terminal half of the protein that differs in retention on HPLC (JHE-B 85) also demon- strates the existence of additional differences in either sequence or post-translational modification at that lo- cation. These locations are the prime candidates for attention in future studies to biochemically test the exact nature of charge variation between the two iso- forms.

Sequencing of fragments with identical HPLC reten- tion has also permitted mapping of regions of the two isoforms that have identical charge (Fig. 6). That these fragments have the same HPLC retention also shows that these regions in the two isoforms do not contain differences in post-translational modifications con- tributing to charge variation between isoforms.

Surface-located moieties Empirically, the most hydrophillic portion of a pro-

tein antigen has the highest likelihood of being used as an antigenic determinant and strongly hydrophillic re- gions also tend to be on the surface of proteins [24]. The fragment(s) with the region(s) of highest anti- genicity is also that possessing carbohydrate, and glyco- sylation usually occurs to residues at /3-turns at the surface of proteins [25]. Thus, the positions mapped to 39 kDa from the N-terminus and to a C-terminal region of the protein (Fig. 6) that possess carbohydrate and the strongest epitope are probably on the surface of the native protein.

Conclusions

The biochemical basis for the major isoforms of JH esterase found in the hemolymph and fat body tissue has thus far been elusive. Absent the knowledge of the

location and nature of these charge variations, their functional significance has remained experimentally unapproachable. The present study has provided the first identification of small, discrete, mapped locations that possess potentially either sequence differences and/or differences in posttranslational modifications confering variation in charge. We have also provided the first experimental evidence on the proteins them- selves as to the actual location of positions of glycosyla- tion and on the subregion of the isolated protein possessing an epitope(s). Using these data as tools, it will now be possible to extend the inferences drawn here on comparative structural organization of the two isoforms to questions on the functional significance of isoform charge variation and glycosylation state.

Acknowledgements

This work was supported, in part, by NIH grant DK 39197 and NSF DCB 9005184, and by a Biomedical Research Support Grant. We express our appreciation to Davy Jones for his input and suggestions during these studies.

References

1 Horodyski, F.M. and Riddiford, L.M. (1989) Develop. Biol. 132, 292-303.

2 Hammock, B.D. (1985) in Comprehensive Insect Physiology, Bio- chemistry, and Pharmacology (Kerkut, G.A. and Gilbert, L., eds.), pp. 431-472, Plenum Press, New York.

3 Jones, G. (1986) in Insect Neurochemistry and Neurophysiology (Borkovec, A.B. and Gelman, D.B., eds.), pp. 315-318, Humana Press, New Jersey.

4 Jones, D., Jones, G., Wing, K.D., Rudnicka, M. and Hammock, B.D. (1982) J. Comp. Physiol. 148, 1-10.

5 Wozniak, M., Jones, G., Hiremath, S. and Jones, D. (1987) Biochim. Biophys. Acta 926, 26-39.

6 Rudnicka, M. and Jones, D. (1987) Insect Biochem. 17, 373-382. 7 Hanzlik, T.N. and Hammock, B.D. (1987) J. Biol. Chem. 262,

13584-13591. 8 Wozniak, M. and Jones, G. (1990) Mol. Cell. Endocrinol. 70,

255-262. 9 Jones, D., Jones, G., Rudnicka, M., Click, A.J. and Sreekrishna,

S.C. (1986) Experientia 42, 45-47. 10 Jones, G., Jones, D. and Hiremath, S. (1987) Insect Biochem. 17,

897-904. 11 Wozniak, M. and Jones, D. (1987) Biochem. Biophys. Res. Com-

mun. 144, 1281-1286. 12 Hanzlik, T.N., Abdel-Aal, Y.A.I., Harshman, L.G. and Hammock

B.D. (1989) J. Biol. Chem. 264, 12419-12425. 13 Abdel-Aal, Y.A.I., Hanzlik, T.N., Hammock, B.D., Harshamn,

L.G. and Prestwich, G. (1988) Comp. Biochem. Physiol. 90B, 117-124.

14 Shorey, H.H. and Hale, R.L. (1965) J. Econ. Entomol. 58, 522- 524.

15 Laemmli, K. (1970) Nature 227, 680-685. 16 Knowles, R.W. (1987) in Histocompatability Testing (Dupont, B.,

ed.), pp. 1-45, Springer, New York. 17 Otey, C.A., Kalnoski, M.H. and Bulinski J.C. (1986) Anal.

Biochem. 157, 71-76. 18 Morrisey, J.H. (1981) Anal. Biochem. 117, 307-310.

19 Burnette, W.N. (1981) Anal. Biochem. 112, 195-203. 20 Stone, K., McNulty, D., LoPrest, Crawford, J., DeAngelis, R. and

Williams, K. (1992) in Techniques in Protein Chemistry (Angletti, R., ed.), pp. 23-34, Academic Press, New York.

21 Hunkapiller, M.W., Hewick, R.M., Dreyer, W.J. and Hood, L.E. (1983) Methods Enzymol. 91, 399-412.

243

22 Harlow, E. and Lane, D. (1988) Antibodies. A laboratory manual. Cold Spring Harbor Laboratory, pp. 594-595.

23 Bradford, M.M. (1976) Anal. Biochem. 72, 248-254. 24 Hopp, T.P. (1989) Methods Enzymol. 178. 25 Beeley, J.G. (1977) Biochem. Biophys. Res. Comm. 76, 1051-1055,