Bioconjugate Chem. 1999, 10, 693−700
693
Characterization of Lysozyme-Estrone Glucuronide Conjugates. The Effect of the Coupling Reagent on the Substitution Level and Sites of Acylation C. Mark Smales, Christopher H. Moore, and Leonard F. Blackwell* Institute of Fundamental Sciences, Chemistry, Massey University, Private Bag 11 222, Palmerston North, New Zealand. Received January 27, 1999
Estrone glucuronide conjugates of hen egg white lysozyme were prepared by the mixed anhydride and active ester coupling procedures. Both methods gave good yields of conjugates, but the active ester procedure gave a more diverse range of products, making it less suitable for preparing conjugates for homogeneous enzyme immunoassay. Conjugation of lysozyme with estrone glucuronide by the mixed anhydride procedure gave one major derivative exclusively acylated at lysine residue 33 whereas conjugation by the active ester method gave six derivatives which were acylated at one or more of lysine residues 33, 97, and 116. None of the lysine residues 1, 13, and 96, or the N-terminal R-amino group, were acylated in any of the conjugates isolated. The correlation of the conjugate structures with the protein environments of the amino groups in the crystal structure of lysozyme suggested that the sites of acylation were determined not only by the chemical nature of the acylating reagent but also by the surface accessibility and nucleophilicity of the individual lysine residues.
INTRODUCTION
The Ovarian Monitor home test for measuring the levels of the metabolites of ovarian estradiol (estrone glucuronide, or E1G) and progesterone (pregnanediol glucuronide, or PdG) in urine (1-5) allows the daily monitoring of a woman’s menstrual cycle and accurate identification of her fertile period. The basis of the home test is an homogeneous enzyme immunoassay, which has been used successfully for the determination of a variety of analytes in biological fluids (6, 7). The attraction of the homogeneous enzyme immunoassay format lies in the fact that the enzyme activity of hapten conjugates of some enzymes (hen egg white lysozyme in this case) is extensively inhibited by anti-hapten antibodies in the immune complex. As a result, a simple kinetic measurement serves to differentiate between bound and free enzyme label (8) without the need for time-consuming and technically demanding physical separation procedures. Hen egg white lysozyme is one of a limited number of enzymes which have been employed successfully in homogeneous enzyme immunoassays. It possesses an apparent specific activity with its bacterial substrate Micrococcus lysodeikticus which renders it ideal for the measurement of menstrual cycle levels of urinary E1G and PdG (9) without excessive dilution of the samples. Lysozyme possesses six lysine residues and a free R-NH2 terminal amino group, all of which can be acylated by a variety of acylating reagents (10) to give stable conjugates which retain lytic activity providing that the total number of lysine residues acylated is less than four (11). However, the lytic activity of lysozyme is adversely affected whenever any of the amino groups are modified by moieties that result in the loss of positive charge from the surface of the enzyme (12). It has been proposed that * To whom correspondence should be addressd. Phone: 646-350 4783. Fax: 64-6-350 5682. E-mail: L.F.Blackwell@ massey.ac.nz.
this loss in activity reflects an alteration in the electrostatic interaction between the positively charged enzyme and the negatively charged M. lysodeikticus cell wall (13). Hence, in developing procedures for conjugation of lysozyme with steroid glucuronides, it is desirable to use conditions which maximize the concentration of conjugate(s) with low degrees of steroid substitution. The estrone glucuronide-lysozyme conjugate mixtures utilized in the Ovarian Monitor assay system have been prepared with a mixed anhydride acylating reagent (1), but currently they are uncharacterized. Identification of the amino groups in a protein which are modified by various acylating reagents can be carried out by analysis of proteolytic digest patterns (14), mass spectrometric peptide mapping (15), and immunological studies (16). It has been suggested on the basis of these previous studies that the acylation position and orientation of haptens in enzyme conjugates is an important factor in determining the recognition and strength of binding of the hapten by antihapten antibodies (17). However, all of these studies have been undertaken on conjugates produced using a large excess of hapten to enzyme in the coupling procedure, therefore producing protein adducts with high levels of hapten substitution. The present study was undertaken to characterize the preferred acylation positions in the E1G-lysozyme conjugates and hence to evaluate the effect of the acylation reagent on the acylation positions. The conjugates were characterized using proteolytic digestion and electrospray mass spectrometry to identify (1) the positions of acylation in the steroid glucuronide lysozyme conjugates and (2) the effect of the coupling procedure on the conjugate products and sites of acylation as a prerequisite to an understanding of the factors which are important in determining the mechanism of inhibition by anti-E1G antibodies.
10.1021/bc9900441 CCC: $18.00 © 1999 American Chemical Society Published on Web 05/29/1999
694 Bioconjugate Chem., Vol. 10, No. 4, 1999 EXPERIMENTAL PROCEDURES
Reagents. The following reagents were obtained from the sources indicated: 17-oxoestra-1,3,5-3-β-yl-D-glucopyranosiduronic acid (estrone glucuronide or E1G[H]) was synthesized essentially according to Conrow and Bernstein (18) as described previously (19). Commercial hen egg white lysozyme (Grade 1, three times recrystallized and lyophilized, Sigma, St. Louis, MO) was further purified before use as described previously (19). Antisera were raised in sheep against thyroglobulin-estrone glucuronide conjugates. Trypsin (L-1-tosylamino-2-phenylethyl chloromethyl ketone treated) and M. lysodeikticus were obtained from Sigma. All other chemicals were analytical or reagent grade. Preparation Of Estrone Glucuronide (E1G[H])Lysozyme Conjugates. Conjugation of lysozyme with E1G[H] was achieved by both the mixed anhydride procedure of Erlanger et al. (20) and the N-hydroxysuccinimide/dicyclohexylcarbodiimide coupling method (21) at a 1.6:1 molar ratio of steroid glucuronide to lysozyme as described previously (19). The mixed anhydride reagent was prepared from E1G[H] (7.2 mg, 16.1 µmol) in dimethylformamide (116 µL) in a 0.5 mL glass reaction vial. The vial was placed in an aluminum block preequilibrated at 10 °C in a refrigerator and allowed to equilibrate for 3 h. Tri-n-butylamine (4.5 µL, 18.9 µmol), preequilibrated at 10 °C, was then added and the E1G[H] solution left for 5 min at 10 °C before the aluminum block was cooled to 0 °C. Once the aluminum block had cooled to 0 °C, it was removed from the freezer and placed in a cold room at 4 °C and left to equilibrate for 15 min. Isobutyl chloroformate (2.5 µL, 19.2 µmol) preequilibrated at 4 °C was then added and the resulting E1G[H] solution left to stand for 30 min. The mixed anhydride reagent was then stabilized by placing the aluminum block at the bottom of a refrigerated centrifuge preequilibrated at approximately -15 °C. Purified lysozyme (144 mg, 10.1 µmol) was dissolved in Milli-Q water (3.7 mL) with stirring and the pH adjusted to 8.00 with 0.5 M NaOH at room temperature. Ice was packed around the lysozyme reaction vessel, which was then placed in the cold room at 4 °C and left to equilibrate for 45 min with stirring while the pH was monitored. The temperature setting of the pH meter was set at 0 °C and stabilized at pH 8.54 after the equilibration period. After the solution stood at -15 °C for 1 h, the mixed anhydride reagent was added dropwise, with a dry 10 µL pipet at 4 °C, to the stirred lysozyme solution over 15 min. Associated with each addition of mixed anhydride reagent was a drop in the pH of the lysozyme solution which was readjusted back to approximately 8.54 with 0.5 M NaOH. The lysozyme solution was left with stirring for 2 h upon the final addition of mixed anhydride reagent (while 0.5 M NaOH was added as required to readjust the pH) after which time the reaction mixture, which appeared opaque, had stabilized at pH 8.39 and was stored at -10 °C until purified by cation-exchange chromatography. In a typical active ester experiment, the reagent was prepared from E1G[H] (7.6 mg, 17.0 µmol) in dimethylformamide (52 µL). N-Hydroxysuccinimide (22.5 mg, 195 µmol) and dicyclohexylcarbodiimide (29.9 mg, 145 µmol) were each separately dissolved in dimethylformamide (100 µL). From these stock solutions, N-hydroxysuccinimide (20 µL, 39 µmol) was added to the E1G[H] solution followed by the addition of dicyclohexylcarbodiimide (20 µL, 29 µmol). After a stirred solution of purified lysozyme (152 mg, 10.7 µmol), dissolved in 1% aqueous sodium hydrogen carbonate (6 mL, pH 9.10) at 0 °C in an ice
Smales et al.
bath, stood at room temperature for 1 h, the active ester reagent was added dropwise. After standing overnight with stirring at 4 °C, the reaction mixture, which appeared opaque, was dialyzed against Milli-Q water (3 × 2 L) and then stored at -10 °C until purified by cation-exchange chromatography. Purification of Estrone Glucuronide-Lysozyme Conjugates. The success of the conjugation experiments and purity of the derivatives was determined by analytical chromatography on a Mono-S cation-exchange column in 7 M urea buffers (19). Estrone glucuronide-lysozyme conjugates were purified from the conjugation reaction mixture by cation-exchange chromatography, also in 7 M urea buffers, on an S-Sepharose (fast flow) column (50 × 1.6 cm i.d.) connected to a Pharmacia Fast Protein Liquid Chromatography system. To the reaction mixture were added sufficient solid urea to give a 7 M solution and a calculated amount of solid sodium dihydrogen phosphate dihydrate to give a 50 mM phosphate solution. After the solution had been adjusted to pH 6.0 with 1 M NaOH, it was loaded, with a peristaltic pump, onto the column preequilibrated with 7 M urea/50 mM sodium dihydrogen phosphate dihydrate buffer (pH 6.0). The column was eluted with a linear salt gradient from 0 to 0.24 M NaCl in 792 min at a flow rate of 1.5 mL/min. The absorbance of the eluent fractions (6 mL) was read at 280 nm and the enzymic activities were determined from the rate of lysis of M. lysodeikticus using an Ovarian Monitor (1). Fractions from each peak were combined, dialyzed extensively against Milli-Q water, and freezedried before rechromatography on the same column. This procedure gave chromatographically homogeneous estrone glucuronide-lysozyme conjugates when analyzed by Mono-S cation-exchange chromatography (19). The purified conjugates were extensively dialyzed against Milli-Q water, freeze-dried, and then stored at -10 °C until required for digestion experiments. Enzyme Assays for Lytic Activity of Chromatography Column Fractions. The assays for lytic activity of the various column fractions were performed in a plastic 1 mL cuvette (Adindas Plastics, Melbourne, Australia) designed specifically for the Ovarian Monitor (1). The appropriate volume from each fraction was added to a cuvette, 0.04 M tris-maleate buffer (pH 7.0) was added to make the volume up to 340 µL, and the solution was equilibrated at 40 °C. M. lysodeikticus (10 µL of a 7.5 mg/mL suspension in 75 mM tris-maleate buffer, pH 7.0) was then added with vortex mixing and the initial transmission noted. After 20 min, the final transmission was noted, and the lytic activity, expressed as the change in transmission (∆T) per 20 min, was calculated. The volume of each fraction to be used in the assay was calculated from the concentration of the peak tube as determined by the absorbance at 280 nm. The dilution of this peak tube required to give an apparent change in transmission of 350 units in 20 min was determined. The same dilution was then used for all of the other fractions to measure their relative lytic activities. The inhibition of the lytic activity of each eluent fraction from the ionexchange columns by anti-E1G antibodies was measured using the Ovarian Monitor assay by adding a slight excess of the antiserum to the assay mixture and comparing it with the appropriate control. Reduction and Alkylation of Estrone Glucuronide-Lysozyme Conjugates with Iodoacetic Acid. Conjugates were reduced and alkylated with iodoacetic acid essentially as described by Rutherfurd et al. (22, 23). Typically, conjugate (5 mg) was dissolved in a mixture of 8 M urea and 0.25 M tris-HCl buffer (pH 8.75, 300
Technical Notes
µL) containing 1 mM EDTA prior to the addition of dithiothreitol (0.84 mg, 5.4 µmol). The resulting solution was flushed with argon gas, and the reduction allowed to proceed at 37 °C for 3 h prior to the addition of iodoacetic acid (1.57 mg, 8.4 µmol). Alkylation was allowed to proceed in the dark at room temperature for 45 min. Excess β-mercaptoethanol was then added and the alkylated protein was dialyzed against 8 M urea (3 × 200 mL). Tryptic Digestion. Tryptic digestion of the alkylated estrone glucuronide-lysozyme conjugates and unmodified lysozyme was carried out in 2 M urea solutions. After dialysis against 8 M urea, the alkylated protein solutions were diluted with 1% ammonium bicarbonate buffer (pH 8.50) to give 2 M urea solutions allowing for the volume of the trypsin solution. L-1-Tosylamino-2-phenylethyl chloromethyl ketone treated trypsin (2 mg) was dissolved in 330 µL of 1% ammonium bicarbonate (pH 8.50) and then the appropriate volume was added to the digest solution so that an enzyme-to-substrate ratio of 1.5:100 (w/w) was achieved. Digestion was allowed to proceed for 3.5 h at 37 °C after which the reaction mixture was stored at -10 °C until analysis by HPLC. Separation of Trypsin Fragments. The separation of the tryptic fragments was carried out on a Vydak C18 reversed-phase column (4.6 × 250 mm) using a Spectra Physics SP8800 high-performance liquid chromatography (HPLC) system. Aliquots from the completed tryptic digests were loaded onto the column preequilibrated with deionized, distilled water (containing 0.1% TFA) and then eluted with a linear gradient from 0 to 40% acetonitrile (containing 0.08% TFA) in 80 min at a flow rate of 1 mL/ min. Detection of the tryptic peptides was carried out at 220 nm with a Spectra Physics SP8490 detector and they were collected manually into 1 mL Eppendorf tubes. Peptides were concentrated on a Savant Speed Vac Concentrator to a volume of approximately 15 µL and then stored at -10 °C until analysis. Analysis of Tryptic Peptides. Tryptic peptides were analyzed by both peptide sequencing and electrospray mass spectrometry. Peptide sequencing was undertaken on an Applied Biosystems Edman degrading gas phase 476A Protein Sequencer. Mass spectral analyses were obtained on a VG Platform II Electrospray Mass Spectrometer. Immunochemical Analysis of Tryptic peptides. Peptides were screened for the presence of estrone glucuronide using the Ovarian Monitor (1). An assay mixture was prepared containing amounts of estrone glucuronide-lysozyme conjugate and E1G antibody such that a partially inhibited ∆T value of approximately 150 units in 40 min was obtained. To the assay mixture was added 5 µL of the concentrated peptide, and the change in transmission over 40 min was recorded. The difference between the ∆T value in the presence and absence of the peptide in the partially inhibited assay mixture was calculated as the immune reactivity. RESULTS
Preparation and Purification of Estrone Glucuronide-Lysozyme Conjugates. Acylation of hen egg white lysozyme with estrone glucuronide by the mixed anhydride and N-hydroxysuccinimide/dicyclohexylcarbodiimide coupling procedures resulted in two distinctly different reaction mixtures (Figure 1). The mixed anhydride procedure produced two major chromatographically distinct peaks (Figure 1a) when passed though an SSepharose fast-flow column, the largest being unreacted
Bioconjugate Chem., Vol. 10, No. 4, 1999 695
Figure 1. E1G-lysozyme reaction mixtures in 7 M urea on a S-Sepharose (fast flow) column. Mixed anhydride method (a), and active ester method (b). The profiles show A280 (9) and lytic activity profiles in the absence (b ) ∆T) and presence (2 ) ∆T) of excess E1G antibody. L ) lysozyme and E1-E6 are E1Glysozyme conjugates. For conditions and details see text.
lysozyme (L). The material which eluted earlier consisted almost entirely of conjugate family E3. E refers to the nomenclature of Smales et al. (19) where E1 is the estrone glucuronide lysozyme peak which elutes next to lysozyme and E6 is the peak which elutes furthest from lysozyme (the first peak eluted). On the other hand, when lysozyme was conjugated by the N-hydroxysuccinimide/ dicyclohexylcarbodiimide coupling procedure, seven chromatographically distinct peaks were obtained (Figure 1b). Again the largest peak was unreacted lysozyme and of the six conjugate peaks (E1-E6) which eluted earlier, E1 and E3 together comprised over half of the total conjugate yield. These six conjugate peaks were characteristic of acylation of lysozyme by steroid glucuronides at low concentrations of reagent, appearing in varying amounts in all conjugations. Even at a molar ratio of 3:1 for the active ester reagent the same six peaks were found but the amounts of E4, E5, and E6 were increased relative to E1, E2, and E3. All of the conjugate families (E1-E6) and unreacted lysozyme from both conjugation procedures had lytic activities which paralleled their A280 profiles (Figure 1), and the conjugates were inhibited by 70-80% when excess E1G antibody was present in the assay mixture as reported previously (19). The unreacted lysozyme (designated L) was not affected by the anti-E1G antiserum as expected, and this was shown by the congruence
696 Bioconjugate Chem., Vol. 10, No. 4, 1999
of the lytic activity in the presence and absence of the antiserum. Conjugate families E4, E5, and E6 had much lower specific activities relative to lysozyme (60.8, 57.6, and 42.3%, respectively) than the E1, E2, and E3 conjugate families (95.1, 88.2, and 94%, respectively) which had specific activities similar to that of native lysozyme. It was necessary to rechromatograph each conjugate family by passage twice through the SSepharose (fast flow) column to obtain chromatographically homogeneous derivatives. Characterization of the E1G-Lysozyme Conjugate Families. The identities of the lysine residues acylated by estrone glucuronide in the chromatographically homogeneous conjugates were determined by tryptic digestion of the reduced, S-carboxymethylated derivatives in 2 M urea solutions at pH 8.5 and 37 °C. The resulting tryptic peptides were analyzed by reversed-phase chromatography (Figures 2, 3, and 4), peptide sequencing, and electrospray mass spectroscopy (Tables 1 and 2). The tryptic peptides from the native lysozyme digest are shown in Figure 2a and are numbered in their order of elution from the reversed-phase column. Twenty major peaks were observed in the digest of native lysozyme (Figure 2a) of which peaks 2 (T1 + 2, K1), 7-10 (T3 and T3 + 4, K13), 15 (T6, K33), 16-18 (T13 and T12 + 13, K97), and 20 (T11, K96) contained lysine residues as shown by amino acid sequencing. T refers to the nomenclature of Canfield (24) for the tryptic peptides of hen egg white lysozyme. The small lysine dipeptide T15 (K116) was not observed. The potential cleavage sites between T1 and T2 and between T17 and T18 (under peak 3) could not be hydrolyzed as was previously reported by Noda et al. (25) and Okazaki et al. (26) for S-carboxymethylated lysozyme derivatives in the absence of 2 M urea. The potential cleavage sites between T3 and T4 and T12 and T13 were only partially hydrolyzed. Peptide T13 appeared as two peaks in the present work, and Yamada et al. (27) have suggested that this is due to the presence of a monomer-dimer mixture which presumably persists also in 2 M urea. In the tryptic maps for each conjugate family, certain lysine-containing peptides disappeared or diminished in height relative to the tryptic map of lysozyme. For example, it was found that peaks 6, 15, and 16-18, which contained the tryptic peptides T7 (6), T6 (15) and T13 + (T12 + 13) (16-18), respectively, were reduced in height in one or more of the conjugate digest profiles (see Figures 2-4 for examples). Associated with these changes, specific new peaks appeared later in the HPLC gradient than the native peptides (after peak 20 in Figure 2a). The amino acid sequence and mass of these later eluting lysine acylated peptides (c-m) were matched with the sequences and masses of the tryptic peptides from hen egg white lysozyme (28) (Table 2). By comparing the experimental data with the expected data, the acylated peptide, and thus the acylated lysine residue, could be assigned (Tables 1 and 2). Acylation with E1G could be readily determined in this way, and it was shown that conjugates resulting from both the active ester and mixed anhydride procedures, which eluted in the same place on the Mono S cation-exchange column (e.g., active ester E3 and mixed anhydride E3) were indeed equivalent (Table 1). For several of the acylated peptides, deamidation had occurred from glutamine to glutamic acid or asparagine to aspartic acid (as shown by amino acid sequencing, Table 2), while in other cases chymotryptic cleavage had occurred (Table 2). In all cases, the calculated and experimental mass spectral data were in good agreement and unambiguously identified the assigned
Smales et al.
Figure 2. Reversed-phase HPLC separation of tryptic digests of unmodified lysozyme (a), and active ester E1G-lysozyme conjugates E1 (b), and E2 (c). The numbers refer to unmodified tryptic peptides and the letters to acylated tryptic peptides. For details see text.
peptide as confirmed by amino acid sequencing (Table 2). However, the acylated peptides did not exhibit a lysine peak where expected from the known tryptic peptide sequences during sequence analysis. Instead, a new peak was observed which eluted from the sequencer HPLC column after all of the other amino acid peaks. This peak was always present when sequencing the modified peptides and was presumably due to an estrone glucuronide acylated lysine residue. Since the new peak eluted last from the sequencer HPLC column, it did not interfere with detection of any of the other amino acids in the sequence. No evidence was found for acylation occurring via ester linkages as has been observed previously in conjugation experiments (14). From the tryptic maps, it is evident that the Nterminal R-amino group and lysine residues 1, 13, and
Technical Notes
Bioconjugate Chem., Vol. 10, No. 4, 1999 697
Figure 3. Reversed-phase HPLC separation of tryptic digests of active ester E1G-lysozyme conjugate E3 (a), and mixed anhydride E1G-lysozyme conjugate E3 (b). The numbers refer to unmodified tryptic peptides and the letters to acylated tryptic peptides. For details see text.
96 were unmodified in all of the conjugates since the peptide peaks T1 + 2 (residues 1-5, peak 2, Figure 2a), T3 + (T3 + 4) (residues 6-14, peaks 7-11, Figure 2a), and T11 (residues 74-96, peak 20 Figure 2a) remained unchanged in height and area relative to the tryptic map for the lysozyme control. Thus, acylation of lysozyme with estrone glucuronide was specific for the -amino groups of lysine residues at positions 33, 97, and 116 when using the active ester method and almost solely for the -amino group of lysine 33 when the mixed anhydride reagent was used (Table 1). Active ester conjugates E1 and E2 were found to be acylated at the same positions, lysines 97 and 116 (Table 1). However, E1 consisted of two monoacylated conjugates, one where acylation has occurred at lysine 97 and the other where acylation has occurred at lysine 116 (Table 1). Partial acylation at lysine residue 97 can be clearly seen as peptide peaks 16-18 had diminished in height by approximately 60% relative to the native digest (Figure 2b). On the other hand, conjugate E2 appears to be a disubstituted conjugate with both lysine residues 97 and 116 acylated on each lysozyme molecule. Peptide peaks 16-18 had almost disappeared in the tryptic digest of conjugate E2, indicating that residue 97 was almost completely acylated. Both active ester conjugate E3 and mixed anhydride conjugate E3 were acylated at lysine residue 33 (Table 1). Active ester conjugates E4 and E5 were disubstituted conjugates acylated at lysine residues 33 and 116 (E4) and 33 and 97 (E5), respectively (Table 1). The most highly substituted conjugate (active ester E6) was trisubstituted at lysine residues 33, 97, and 116 (Table 1). Enzyme inhibition assays carried out in the presence of the native lysozyme tryptic peptides (peaks 1-20, Figure 2a) did not significantly increase the lytic activity
Figure 4. Reversed-phase HPLC separation of tryptic digests of active ester E1G-lysozyme conjugates E4 (a), E5 (b), and E6 (6). The numbers refer to unmodified tryptic peptides and the letters to acylated tryptic peptides. For details see text.
of the intact E1G-lysozyme mixed anhydride E3 conjugate in the inhibition assay (Figure 5a). The same result was obtained for peaks 1-20 obtained from the tryptic digest of E1 (Figure 5b) which therefore did not contain E1G after conjugation. However, inhibition assays carried out in the presence of the active ester E1 tryptic peptides f, i, and j increased the enzymic rates of lysis to approximately 350 units/40 min. This rate was the same as obtained for the control assay in the absence of the E1G antibody, thus the immune reactivities for these three peptides were high (Figure 5b), confirming the presence of at least one estrone glucuronide moiety. DISCUSSION
Since the active ester methodology is easier to perform than the mixed anhydride method, it is used more often in protein conjugation reactions, especially in the hands of nonorganic chemists. However, close inspection of the
698 Bioconjugate Chem., Vol. 10, No. 4, 1999
Smales et al.
Table 1. Summary of Acylation Positions of Lysozyme-Estrone Glucuronide Conjugates conjugate peak
E1G per lysozyme moleculea
AE E1
1
AE E2
1-1.5
AE E3 AE E4
1 2
AE E5
2
AE E6
3
MA E3
1
acylated peptide(s)
corresponding peptide peaksb
location of acylation position
T12 + 13 T15 + 16 T12 + 13 T15 + 16 T6 + 7 T6 + 7 T15 + 16 T6 + 7 T12 + 13 T6 + 7 T12 + 13 T15 + 16 T6 + 7
d, i, j c, f d, i, j c, f k, m, o k, m, o c, f k, k′, o i, j k, k′, m d, i, j c, f k, m, o
Lys 97 Lys 116 Lys 97 Lys 116 Lys 33 Lys 33 Lys 116 Lys 33 Lys 97 Lys 33 Lys 97 Lys 116 Lys 33
a
Determined from proteolytic studies and titration of lysine groups (19). b See Figures 2, 3, and 4.
present data showed that there were significant differences between the two methods in terms of the conjugates they produced, and hence, careful consideration needs to be given to the choice of reagent depending on the aims of the study. The active ester reagent was more reactive and, hence, less selective than the mixed anhydride reagent as it gave a more diverse range of conjugate products at the same ratio of E1G to lysozyme (Figure 1). Hence, a greater overall yield of conjugate was obtained than with the mixed anhydride reagent, and the reaction was less time consuming. However, the complexity of the product mixtures made purification of the different conjugates more difficult. For immunoassays, the extra complexity of the less selective active ester reaction was undesirable since the relatively large amounts of the more highly substituted conjugates produced, E4, E5, and E6, had lower specific activities. A greater concentration of these conjugates was required, therefore, in the assays for a given rate of lysis than for the lower substituted conjugates. E4, E5, and E6 also required more antiserum to inhibit their lytic activities due to their higher substitution levels with E1G, hence, less sensitive standard curves were obtained, making them less suitable for homogeneous enzyme immunoassay. The mixed anhydride reagent is thus preferable to the active ester reagent as a source of conjugate for homogeneous enzyme immunoassay as it gives an almost pure mono E1G conjugate in good yield and with a high specific activity. The difference between the two reagents is possibly due to the difference in their intrinsic reactivities. In support of this suggestion, the active ester reaction was almost instantaneous, as shown by the similarity of the FPLC traces for the reaction mixtures after 2, 5, 60, and 180 min and overnight when analyzed by chromatography on a Mono-S column in 7 M urea (19). On the other hand, the more selective mixed anhydride reagent reacted much more slowly, taking several hours as evidenced by the slow drop in pH as the reagent reacted with lysozyme. Under these conditions the estrone glucuronide moiety is more likely to react with the most reactive lysine residue having sufficient lifetime in solution to select among various possibilities. The identification of the acylation positions in the conjugates obtained with both acylating reagents was relatively straightforward using tryptic digestion experiments. However, it was necessary to use 2 M urea to maintain the solubility of the hydrophobic reduced and S-carboxymethylated acylated conjugates and peptides.
Under the 2 M urea conditions used for the digestion experiments, trypsin was active (29) and highly reproducible tryptic digests were obtained (Figures 2, 3, and 4). After calibration of the HPLC peaks by sequence data and mass analysis, visual inspection of the traces allowed ready identification of the major sites of acylation from a combination of the loss in amplitude of the lysinecontaining tryptic peptides in the native digest and the appearance of new characteristic acylated peptides in the conjugate digest (peaks i, j, etc.). As the coupling reactions were carried out under limiting conditions of reagent, the fact that conjugate E3 (lysine 33) was the major product of the mixed anhydride reaction indicates that lysine 33 is the most reactive residue at pH 8.5. The elution position of the mixed anhydride minor conjugate peak (Figure 1a) suggests that it is E5 with both lysines 33 and 97 acylated (Table 1) which in turn indicates that lysine 97 is the next in order of reactivity. This reactivity order suggested from the mixed anhydride reaction also holds for the active ester reaction. Conjugate E3 was still one of the major peaks, and of the two disubstituted conjugate fractions (E4 and E5 of Table 1), E5 was present in the greatest amount. Hence, the reactivity order suggested by the combined data is K33 > K97 > K116. Previous work involving the acylation and modification of the amino groups of hen egg white lysozyme has also shown a differential selectivity in the reactivity of the lysine residues (14-15). For example, when Suckau et al. (15) aminoacetylated lysozyme with acetic anhydride, the reactivities of the amino groups showed an order, K97 ) K33 > K1 (-NH2 and R-NH2) > K13 ) K116 > K96. The accessibility of the lysine side-chain -amino groups in hen egg white lysozyme as determined from their degree of extension in the crystal structure decreases in a similar order, K97 > K33 . K1 (-NH2) > K13 > K116 > K96 > R-NH2 (15). Nuclear magnetic resonance studies have shown that the solution environments of the amino groups in lysozyme are similar to those in the crystalline protein (30), hence the difference in the reactivities of the lysine residues can be related to the different protein environments as revealed by crystallography. Thus, the reactivity data of Suckau et al. (15) shows that lysine availability or exposure is a major factor in determining chemical reactivity during acylation reactions. The preferential acylation of lysozyme with estrone glucuronide at lysines 33 and 97 in the present work is consistent with the reactivity preference shown in the data of Suckau et al. (15) as lysines 33 and 97 are by far the most accessible of the seven having surface accessibilities of 40.1 and 47.4%, respectively. However, the preference for lysine 33 exhibited in the present work shows that other factors are involved. The pK value calculated for lysine 33 (9.7) is less than for lysine 97 (10.5) (31) which would result in an almost 10-fold higher concentration of the lysine 33 free base at pH 8.5 (relative to lysine 97) and, hence, could explain the greater reactivity. However, according to the results of Suckau et al. (15), acylation of lysine 116 with estrone glucuronide should not have occurred until after the acylation or partial acylation of lysine 1 (R and -NH2) and lysine 13. The acylation of residue 116 with estrone glucuronide in preference to lysine 1 (R and -NH2), lysine 13, or lysine 96 may be a result therefore of a relatively high local concentration of the E1G reagent in the vicinity of lysine 116. This could be achieved if E1G has an affinity for one or more of the sugar-binding sites in the active-site cleft since lysine 116 is situated just above this. The present results show that the lysozyme molecule
Technical Notes
Bioconjugate Chem., Vol. 10, No. 4, 1999 699
Table 2. Sequence and Mass Spectral Data of E1G Acylated Tryptic Peptides peptide peak c d f i j k k′ m o
sequence data CaKbGTDVQAWc KbIVSDGNGMNAWc CaKbGTDVQAWIR KbIVSDGNGMNAWVAWR KbIVSDGNGMDdAWVAWR cSLGNWVCaAAKbFESNFNTQ ATNR cSLGNWVCaAAKbFESNFNTEe ATNR GYSLGNWVCaAAKbFESNFNTQ ATNR GYSLGNWVCaAAKbFESNFNTEe ATNR
calculated (m/z)
experimental (m/z)
acylation location
1492.7 1718.9 1762.7 2231.9 2232.9 2943.2
1493.0 1719.3 1762.5 2231.5 2233.5 2943.9
Lys 116 Lys 97 Lys 116 Lys 97 Lys 97 Lys 33
2944.2
2944.7
Lys 33
3164.4
3163.6
Lys 33
3165.2
3166.8
Lys 33
a S-carboxymethylated Cys. b Estrone glucuronide acylated lysine residue. c Chymotryptic cleavage at this site. interchange. e Q f E amino acid interchange.
d
N f D amino acid
the length and rigidity of the link between the hapten and the enzyme and the binding kinetics are all important determinants of the degree of inhibition. A further understanding of the relative importance of these factors based on the structure and reactivity of the characterized conjugates E1-E6 is being carried out and may help in the rational design of other inhibitable enzyme conjugate systems. ACKNOWLEDGMENT
We wish to thank the following people for their contributions: Associate Professor David Harding and Mr. Dick Poll for use of their chromatographic equipment, discussions, and suggestions; Dr. Brian Anderson for his help with computer graphics; and Dr. Keith Henderson for raising antibodies. LITERATURE CITED
Figure 5. Immune reactivities of the tryptic peptides from unmodified lysozyme (A) and active ester E1G-lysozyme conjugate E1 (B) as measured using an Ovarian Monitor. The high immune reactivity of peptides f, i, and j (B) confirmed the presence of E1G.
is ideally constituted for homogeneous enzyme immunoassays for steroid glucuronides, since the most reactive lysine residues with respect to steroid glucuronide acylation are located near the active site. Steroid glucuronide conjugates of lysozyme, with low substitution levels and which are highly inhibited (>90%), can therefore be obtained using stoichiometric ratios of acylating reagent. Binding of the anti-hapten antibody to haptens attached at any of lysines 33, 97, or 116 results in a strong inhibitory effect on the lytic activity of the Ovarian Monitor conjugates, undoubtedly due to steric blocking of access of the M. lysodeikticus substrate to the active site. A combination of factors such as orientation of the hapten with respect to the active site (acylation position), the size of the substrate, the size of the enzyme itself,
(1) Brown, J. B., Blackwell, L. F., Holmes, J., and Smyth, K. (1989) New Assays for Identifying the Fertile Phase. Int. J. Gynecol. Obstet. Suppl. 1, 111-122. (2) Brown, J. B., Blackwell, L. F., Cox, R. I., Holmes, J. M., and Smith, M. A. (1988) Chemical and Homogeneous Enzyme Immunoassay Methods for the Measurement of Estrogens and Pregnanediol and their Glucuronides in Urine. Prog. Biol. Clin. Res. 285, 119-138. (3) Brown, J. B., Blackwell, L. F., Billings, J. J., Conway, B., Cox, R. I., Garrett, G., Holmes, J. M., and Smith, M. A. (1987) Natural Family Planning. Am. J. Obstet. Gynecol. 157, 10821089. (4) Thornton, S. J., Pepperell, R. J., and Brown, J. B. (1990) Home Monitoring of Gonadotrophin Ovulation Induction using the Ovarian Monitor. Fertility and Sterility 54, 10761082. (5) Brown, J. B., Barker, G., and Holmes, J. (1991) Use of the Home Ovarian Monitor in Pregnancy Avoidance. Am. J. Obstet. Gynecol. 165, 2008-2011. (6) Rubenstein, K. E., Schneider, R. S., and Ullman, E. F. (1972) “Homogeneous” Enzyme Immunoassay. A new Immunological Technique. Biochim. Biophys. Res. Commun. 47, 846-851. (7) Dhar, T. K., Samanta, A. K., and Ali, E. (1988) Homogeneous Enzyme Immunoassay of Estradiol using Estradiol-3-OCarboxymethyl Ether as Hapten. Steroids 51, 519-526. (8) Schneider, R. S., Lindquist, P., Wong, E. T., Rubenstein, K. E., and Ullman, E. F. (1973) Homogeneous Enzyme Immunoassay for Opiates in Urine. Clin. Chem. 19, 821-825. (9) Adlercreutz, H., Brown, J., Collins, W., Goebelsman, U., Kellie, A., Campbell, H., Spieler, J., and Braissand, G. (1982) The Measurement of Urinary Steroid Glucuronides as Indices of the Fertile Period in Women. World health organisation, Task Force on Methods for the Determination of the Fertile Period, Special Programme of Research, Development and Research Training in Human Reproduction. J. Steroid Biochem. 17, 695-702.
700 Bioconjugate Chem., Vol. 10, No. 4, 1999 (10) Habeeb, A. F. S. A., and Atassi, M. Z. (1970) Enzymic and Immunochemical Properties of Lysozyme. Evaluation of Several Amino Group Reversible Blocking Reagents. Biochemistry 9, 4939-4944. (11) Habeeb, A. F. S. A., and Atassi, M. Z. (1971) Enzymic and Immunochemical Properties of Lysozyme. V. Derivatives Modified at Lysine Residues by Guanidination, Acetylation, Succinylation, or Maleylation. Immunochemistry 8, 10471059. (12) Patthy, L., and Smith, E. L. (1975) Identification of Functional Arginine Residues in Ribonuclease A and Lysozyme. J. Biol. Chem. 250, 565-569. (13) Imoto, T., Johnson, L. N., North, A. C. T., Phillips, D. C., and Ruplet, J. A. (1972) Vertebrate Lysozymes. In The Enzymes (P. D. Boyer, Ed.) 3rd ed., Vol. VII, pp 781, Academic Press, New York. (14) Lee, C., Atassi, M. Z., and Habeeb, A. F. S. A. (1975) Enzyme and Immunochemical Properties of Lysozyme. Biochim. Biophys. Acta 400, 423-432. (15) Suckau, D., Mak, M., and Przybylski, M. (1992) Protein Surface Topology-Probing by Selective Chemical Modification and Mass Spectrometric Peptide Mapping. Proc. Natl. Acad. Sci. U.S.A. 89, 5630-5634. (16) Burnens, A., Demotz, S., Corradin, G., Binz, H., and Bosshard, H. R. (1987) Epitope Mapping by Chemical Modification of Free and Antibody-Bound Protein Antigen. Science 235, 780-783. (17) Gani, M., Coley, J., Piron, J., Humphreys, A. S., Arevalo, J., Wilson, I. A., and Taussig, M. J. (1994) Monoclonal Antibodies Against Progesterone. Effect of Steroid-Carrier Coupling Postion on Antibody Specificity. J. Steroid Biochem. Mol. Biol. 48, 277-282. (18) Conrow, R. B., and Bernstein, S. (1971) Steroid conjugates. VI. An Improved Koenigs-Knorr Synthesis of Aryl Glucuronides using Cadmium Carbonate, a New and Effective Catalyst. J. Org. Chem. 36, 863-870. (19) Smales, C. M., Cooke, D., and Blackwell, L. F. (1994) Use of Ion-Exchange and Hydrophobic-Interaction Chromatography for the Rapid Purification of Lysozyme-Estrone Glucuronide Conjugates. J. Chromatogr. B 662, 3-14. (20) Erlanger, B. F., Borek, F., Beiser, S. M., and Lieberman, S. (1959) Steroid-Protein Conjugates II: Preparation and Characterisation of Conjugates of Bovine Serum Albumin and Progesterone Deoxycorticosterone and Estrone. J. Biol. Chem. 234, 1090-1094.
Smales et al. (21) Rajkowski, K. M., and Cittanova, N. (1981) The Efficiency of Different Coupling Procedures for the Linkage of Oestriol16R-Glucuronide, Oestrone-3-Glucuronide and Pregnanediol3R-Glucuronide to Four Different Enzymes. J. Steroid Biochem. 14, 861-866. (22) Rutherfurd, K. J., Chen, S., and Shively, J. E. (1991) Isolation and Amino Acid Sequence Analysis of Bovine Adrenal 3β-Hydroxysteroid Dehydrogenase/Steroid Isomerase. Biochemistry 30, 8108-8116. (23) Rutherfurd, K. J., Chen, S., and Shively, J. E. (1991) Affinity Labeling of Bovine Adrenal 3β-Hydroxysteroid Dehydrogenase/Steroid Isomerase by 5′-[p-(fluorosulfonyl)benzoyl] Adenosine. Biochemistry 30, 8116-8123. (24) Canfield, R. E. (1963) Peptides Derived from Tryptic Digestion of Egg White Lysozyme. J. Biol. Chem. 238, 26912697. (25) Noda, Y., Fujiwara, K., Yamamoto, K., Fukuno, T., and Segawa, S. (1994) Specificity of Trypsin Digestion and Conformational Flexibility at Different Sites of Unfolded Lysozyme. Biopolymers 34, 217-226. (26) Okazaki, K., Imoto, T., Yamada, H., Kuroki, R., and Fujita, K. (1982) Specificity of Trypsin Cleavage of Aspartic Acid 101 Derivatives of Lysozyme by Trypsin. J. Biol. Chem. 257, 12559-12562. (27) Yamada, H., Ueda, T., Kuroki, R., Fukumura, T., Yasukochi, T., Hirabayashi, T., Fujita, K., and Imoto, T. (1985) Isolation and Characterisation of 101-β-Lysozyme that Possesses the β-Aspartyl Sequence at Aspartic Acid-101. Biochemistry 24, 7953-7959. (28) Canfield, R. E. (1963) The Amino Acid Sequence of Egg White Lysozyme. J. Biol. Chem. 238, 2698-2707. (29) Winter, G. (1986) Manual Sequence StrategysA Personal View. In Practical Protein Chemistry-A Handbook (A. Darbre, Ed.) pp 349, J. Wiley & Sons Ltd, New York. (30) Gerken, T. A., Jentoft, J. E., Jentoft, N., and Dearborn, D. G. (1982) Intramolecular Interactions of Amino Groups in 13C Reductively Methylated Hen Egg-White Lysozyme. J. Biol. Chem. 257, 2894-2900. (31) Spassov, V. Z., Karshikov, A. D., and Atanisov, B. P. (1989) Electrostatic Interactions in Proteins. A Theoretical Analysis of Lysozyme Ionization. Biochim. Biophys. Acta 999, 1-6.
BC9900441
