Ionic Charge, Hydrophobicity and Tryptophan ... - Bioscience Reports

1 downloads 0 Views 91KB Size Report
A high-affinity folate binding protein was isolated and purified from cow's milk by a combi- nation of cation exchange chromatography and methotrexate affinity ...
Bioscience Reports, Vol. 21, No. 3, June 2001 ( 2001)

Ionic Charge, Hydrophobicity and Tryptophan Fluorescence of the Folate Binding Protein Isolated from Cow’s Milk Jan Holm,1,4 Steen Ingemann Hansen,2 and Mimi Høier-Madsen3 Receiûed January 26, 2001 A high-affinity folate binding protein was isolated and purified from cow’s milk by a combination of cation exchange chromatography and methotrexate affinity chromatography. Chromatofocusing studies revealed that the protein possessed isoelectric points in the pHinterval 8–7. Polymers of the protein prevailing at pH values close to the isoelectric points seemed to be more hydrophobic than monomers present at pH 5.0 as evidenced by hydrophobic interaction chromatography and turbidity (absorbance at 340 nm) in aqueous buffer solutions (pH 5–8). Ligand binding seemed to induce a conformation change that decreased the hydrophobicity of the protein. In addition, Ligand binding quenched the tryptophan fluorescence of folate binding protein suggesting that tryptophan is present at the binding site and兾or ligand binding induces a conformation change that affects tryptophan environment in the protein. There was a noticeable discordance between the ability of individual folate analogues to compete with folate for binding and the quenching effect. KEY WORDS: Isolation of cow’s milk folate binding protein; ligand-induced changes in tryptophan fluorescence; hydrophobicity of folate binding protein.

INTRODUCTION The high-affinity folate binding protein (FBP) exists as a membrane-bound folate receptor (FR) as well as a soluble FBP in milk and other body fluids [1]. A gene located on chromosome 11q13 expresses at least three isoforms of FR兾FBP in humans [1]. The N-terminal amino acid sequence of one of these, FRalpha , is homologous to that of human milk FBP [2]. The presence of a high-affinity FBP was first demonstrated in cow’s milk [3]. A large scale purification of FBP from cow’s whey powder by means of a combination of cation exchange chromatography and methotrexate affinity chromatography made it feasible to characterize this protein extensively with regard to primary and secondary structure, physicochemical nature and ligand binding characteristics [4, 8]. The complete amino acid sequence of bovine milk FBP showed a high degree of homology with that of human milk FBP [9] supporting the view that bovine FBP could serve as, a prototype model in studies of the interaction between FBP兾FR and its ligand, folate. 1

Medicinsk Center, Centralsygehuset Roenne, Bornholm, DK-3700, Denmark. Department of Clinical Chemistry, Central Hospital Hillerød, DK-3400, Denmark. 3 Laboratory for Autoimmune Serology, State Serum Institute, Copenhagen, DK-2300, Denmark. 4 To whom all correspondence should be addressed. E-mail: [email protected]. Fax: 5691-1204. 305 2

0144-8463兾01兾0600-0305$19.50兾0  2001 Plenum Publishing Corporation

306

The present study has been devoted to a further characterization of purified bovine milk FBP with regard to ionic charge properties and ligand-induced changes in hydrophobicity and tryptophan fluorescence spectrum. MATERIALS AND METHODS The radiochemicals [14C] folate (pteroylglutamate) with a specific activity of 52.4 Ci兾mol and [3H]-folate with a specific activity of 29–43 Ci兾mmol were obtained from Amersham International Ltd., Amersham, UK. The following unlabeled analogues were used: folate, 5-formyltetrahydrofolate and 4-amino-benzoic acid (Sigma), pteroic acid (Aldrich) and methotrexate (Lederle) purified as previously described [10]. FBP was prepared and purified from cow’s whey powder by a combination of cation exchange chromatography on CM-Sepharose CL-6B and affinity chromatography on a methotrexate-AH-Sepharose 4B column [4]. Polyclonal antibodies against FBP raised in rabbits were employed in a previously described ELISA for quantitation of bovine milk FBP [11]. Equilibrium dialysis of FBP solutions predialyzed against 0.2 M acetate buffer, pH 3.5 at 4°C to remove endogenous folate was performed as described previously [6] for periods of 20 hr in 0.17 M Tris-HCl buffer (37°C, pH 7.4) with FBP in the internal (1000 µl) and radioligand in the external solution (200 ml). Due to a large volume of external solution, the concentration of radioligand was kept constant during the entire dialysis experiments. Triton X-100 at a concentration of 1 g兾l was added to both sides of the dialysis membrane [6]. Radioactivity was measured as previously described [6]. When necessary, unlabeled analogs were added to the external solutions together with [3H]-folate. Chromatofocusing Chromatofocusing experiments were performed in the pH interval 9–6 according to the instructions by the manufacturer (Pharmacia). A column (40 cmB2 cm2) packed with a PBE 94 gel (Pharmacia) was equilibrated overnight with the start buffer, ethanolamine HCl (Baker), pH 9.4 (0.025 M) and eluted (800 ml) with Polybuffer 96-acetate, pH 6.0 (Pharmacia). Flow rate was 30 ml兾hr and fractions of 10 ml were collected. The column was regenerated with 200 ml NaCl (1 M) for 7 hr and then 150 ml 0.025 M ethanolamine HCl (Baker) for 5 hr. Samples (1 ml volumes) of FBP (157–230 nM) were preincubated for 3 hr, at 25°C with 250 nM [3H]-folate in the start buffer (pH 9.4). A pH-meter (PH M72, Radiometer, Copenhagen) was used for measurement of pH. The concentration of FBP was determined both by radioactivity measurement and ELISA-technique. Hydrophobic Interaction Chromatography of Folate Binding Protein Hydrophobic interaction chromatography was performed on a column (2 cm2B40 cm) of Octyl-Sepharose CL-4B gel (Pharmacia). A flow rate of 10 ml兾 hr, 9–13 ml fractions and sample volumes of 5 ml were used. A solution of FBP

Bovine Milk Folate Binding Protein

307

(maximum folate binding, 356 nM) was preexposed to 2000 nM [14C]-folate for 3 hr at 25°C. The column was equilibrated and eluted (16 hr, 4°C) with 0.17 M Tris-HCl buffer, pH 7.4; then the column was eluted (8 hr) with 0.17 M Tris-HCl buffer, pH 7.4, containing Triton X-100 (10 g兾l). Triton X-100 was measured at 285 nm. Effluent fractions were assayed for FBP by radioactivity measurements and absorbance (OD) at 492兾620 nm in ELISA (50-fold diluted samples). Samples (5 ml volumes) of the effluent (fraction 5 and 26) from the Octyl-Sepharose CL-4B gel column were applied to a column (5.3 cm2B94 cm) of Ultrogel  (IBF). Triton X100 (1 g兾l) was added to the elution buffer, 0.17 M Tris-HCl of pH 7.4 (5°C). Eluant from the gel chromatographic column was assayed for FBP by radioactivity measurements and absorbance (OD) at 492兾620 nm in ELISA. Turbidity of FBP Absorbance measurements were performed on a Cobas FARA instrument (Roche Ltd.) at 340 nm with increasing concentrations of ligand-free or ligandbound FBP at pH 5.0 and 8.0. Samples of FBP incubated without or with equimolar concentrations of ligand in 0.2 M acetate buffer, pH 5.0 were added to equal volumes (80 µl) of either the same buffer (final pH 5.0) or 0.17 M Tris-HCl buffer, pH 9.0 (final pH 8.0) prior to absorbance measurements at 37°C (5 min). Fluorescence Spectroscopy Tryptophan emission spectra (excitation 290 nm) were recorded with a Perkin– Elmer luminescence spectrometer model LS-50 as described in a previous report [8]. Also, a few spectra were recorded with a Perkin–Elmer 203 fluorescence spectrophotometer. RESULTS Chromatofocusing of FBP Chromatofocusing revealed two distinct peaks of radioligand-bound FBP in the pH-interval 8-7 (Fig. 1). The profile of immunoreactive FBP determined by ELISA technique contained two coinciding peaks as well as an additional peak at pH 7.0. A sample blank (2 nM[3H] folate) applied to the column eluted after termination of the pH-gradient (data not shown). Hydrophobic Interaction Chromatography of FBP Hydrophobic interaction chromatography of FBP showed one large peak of radioligand-bound and immunoreactive FBP in the front effluent (Fig. 2) which eluted at the position (220 ml) of monomeric (30 kDa) FBP (estimated concentration 60 nM) on gel filtration together with a large peak of excess [14C] folate (Fig. 2, upper insert). A small fraction of immunoreactive FBP eluted after addition of Triton X100 (Fig. 2). This fraction contained two peaks of radioligand-bound and immunoreactive FBP (estimated concentration 0.6 nM) on gel filtration, one at 220 ml (30 kDa)

308

Fig. 1. Chromatofocusing of bovine milk in the pH-interval 9–6 on a column (40 cmB2 cm2) of PBE 94 gel equilibrated with ethanolamine HCl buffer, pH 9.4 and eluted with Polybuffer 96acetate, pH 6.0. Bovine milk folate binding protein at a concentration of 157–230 nM in a volume of 1000 µl was preexposed (3 hr, 25°C, pH 9.4) to 250 nM [3H]-folate. Right ordinate, pH gradient monitored (!). Left ordinate, concentration of FBP nM, determined by radioligand binding (䊊) and ELISA (䉮).

and one at 180 ml, the position of dimeric (60 kDa) FBP, (Fig. 2, lower insert). The third peak represented [14C] folate (Fig. 2, lower insert).

Turbidity Measurements Measurement of the turbidity (absorbance at 340 nm) of FBP solutions were performed to investigate aqueous solubility at different pH values with increasing concentrations of ligand-bound and ligand-free FBP. Preliminary experiments showed that the incubation temperature had no influence on the results (data not shown). Fig. 3 shows that ligand-free FBP even at high concentrations has a high water solubility (no turbidity) at pH 5. At pH 8.0 there is a pronounced decrease in solubility with increasing concentrations of ligand-free FBP (increasing turbidity). Contrastingly, ligand-bound (folate-bound) FBP has a high water solubility, both at pH 5.0 and 8 regardless of the FBP concentration. The slight and linear increase in the turbidity (absorbance at 340 nm) seen with increasing concentration of ligandbound FBP is not due to any absorbance effect of increasing concentrations of folate (data not shown). Other ligands, e.g., 5-formyltetrahydrofolate and methotrexate exerted similar effects on the solubility of FBP (data not shown).

Bovine Milk Folate Binding Protein

309

Fig. 2. Hydrophobic interaction chromatography of bovine milk FBP on a column (2 cm2B40 cm) of Octyl-Sepharose CL-4B gel . Bovine milk FBP at a concentration of 356 nM in a sample volume of 5 ml was preexposed to 2000 nM [14C] folate, (3 hr, 25°C, pH 7.4) prior to column application. After sample application the column was equilibrated (16 hr, 4°C) with Tris-HCl buffer, pH 7.4, and then eluted with Tris-HCl buffer containing 10 g兾l Triton X-100 for 8 hr. Effluent fractions were assayed for radioactivity (left ordinate, open squares), and immunoreactivity (50-fold dilution) in ELISA (right ordinate, absorbance (OD) at 492兾620 nm, open circles). The Triton X-100 concentration was estimated from absorbance measurements at 285 nm (right ordinate, dotted line). Ultrogel AcA 44 chromatography (for details, see text) of fractions eluted after hydrophobic interaction chromatography: Front effluent (upper insert), and fractions eluted after the addition of Triton X100 (lower insert). Left ordinate, cpm (●). Note different scales (log) of cpm. Right ordinate, absorbance (OD) at 492兾620 nm, (䊊).

Inhibition by Folate Analogues Experiments performed to study the effect of folate analogs on binding of [3H]folate (0.1 nM) to FBP (0.5 nM) are shown in Fig. 4. As can be seen, 5-formyltetrahydrofolate (folinic acid) inhibited folate binding with a molar inhibition ratio at 50% inhibition of 10. Other folate derivatives or components of pteroylglutamate, e.g., pteroic acid and purified methotrexate acted as weaker inhibitors with molar inhibition ratios at 50% of 1000 and H1000, respectively. No inhibition was seen in the presence of 4-amino benzoate. Effect of Folate Analogs on FBP Tryptophan Fluorescence Figure 5 shows the effect of increasing concentrations of 5-formyltetrahydrofolate and pteroate on the tryptophan emission spectrum of 0.5 µM FBP. Both analogs, in particular pteroate quenched tryptophan fluorescence. The extent of the quench at 340 nm as a function of the analog concentration resulted in titration curves

310

Fig. 3. Turbidity (absorbance at 340 nm) as a function of the concentration of FBP at pH 5.0 and 8.0 in the absence or presence of ligand (equal concentrations of FBP and ligand). FBP at pH 5.0 with (䊊) or without (●) folate. FBP at pH 8.0 with (䊐) or without (■) folate.

consisting of two straight lines with an intersection point corresponding to the stoichiometry of folate binding. The observed stoichiometry 0.6 mol analog兾mol FBP is very close to that obtained from equilibrium dialysis experiments using radiolabeled folate [8]. A few additional experiments performed with FBP at a concentration of 0.15 µM in the presence of 10-fold higher concentrations of folate兾folate analogues showed the following values for maximum quench of tryptophan fluorescence relative to that of folate: methotrexate, 0.9, pteroate, 1.0 and 5-formyltetrahydrofolate, 0.6. Other experiments showed that the ability of methotrexate to quench FBP tryptophan fluorescence was considerably lower at pH 3.5 than at pH 7.4 (data not shown). DISCUSSION The chromatofocusing profile of bovine milk FBP exhibits isoelectric points in the interval pH 7–8 (Fig. 1) confirming the cationic nature of the protein at nearneutral pH [4]. This is consistent with the behavior of bovine milk FBP on anionexchange chromatography at near-neutral pH [13] where like other cationic FBPs in human milk, ovarian carcinoma tissue, granulocytes and serum elutes in the front effluent [14–17]. There seemed to be some discrepancy between concentrations of immunoreactive and radioligand-bound FBP with regard to the third peak close to pH 7.0 in the chromatofocusing profile (Fig. 1). We cannot offer any explanation to

Bovine Milk Folate Binding Protein

311

Fig. 4. Effect of 5-formyltetrahydrofolate (●), pteroate (&), 4-aminobenzoate (■) and purified methotrexate (▼) on [3H] folate binding to 0.5 nM bovine milk FBP. Equilibrium dialysis experiments with 0.1 nM [3H] folate in the external solution together with unlabeled folate analogue. The curves were fitted to the following equation [12]: fG(1CF*K)兾1CF*KCI*M ): f is the fraction of maximum bound folate. F is the free folate concentration (constantG0.1 nM), K is the apparent association constant for folate binding, I is the apparent association constant for analogue binding, and M is the analogue concentration.

this finding but by analogy to human saliva immunoreactive non-functional FBP could represent unprocessed immunoreactive precursor or FBP with ligand-binding sites occupied by endogenous folate resistant to acidic dialysis at pH 3.5 [18–19]. The behavior of bovine milk FBP in hydrophobic interaction chromatography (Fig. 2) showed that a very small fraction (