- Email: [email protected]
Copper(II)-based metal affinity chromatography for the isolation of the anticancer agent bleomycin from Streptomyces verticillus culture
Journal of Inorganic Biochemistry 115 (2012) 198–203
Contents lists available at SciVerse ScienceDirect
Journal of Inorganic Biochemistry journal homepage: www.elsevier.com/locate/jinorgbio
Copper(II)-based metal affinity chromatography for the isolation of the anticancer agent bleomycin from Streptomyces verticillus culture Jiesi Gu, Rachel Codd ⁎ School of Medical Sciences (Pharmacology) and Bosch Institute, University of Sydney, New South Wales 2006, Australia
a r t i c l e
i n f o
Article history: Received 24 November 2011 Received in revised form 27 January 2012 Accepted 31 January 2012 Available online 9 February 2012 Keywords: Metal affinity chromatography Bleomycin Copper complexes Green chemistry
a b s t r a c t The glycopeptide-based bleomycins are structurally complex natural products produced by Streptomyces verticillus used in combination therapy against testicular and other cancers. Bleomycin has a high affinity towards a range of transition metal ions with the 1:1 Fe(II) complex relevant to its mechanism of action in vivo and the 1:1 Cu(II) complex relevant to its production from culture. The affinity between Cu(II) and bleomycin was the underlying principle for using Cu(II)-based metal affinity chromatography in this work to selectively capture bleomycin from crude S. verticillus culture. A solution of standard bleomycin was retained at a binding capacity of 300 nmol mL− 1 on a 1-mL bed volume of Cu(II)-loaded iminodiacetate (IDA) resin at pH 9 via the formation of the heteroleptic immobilized complex [Cu(IDA)(bleomycin)]. Bleomycin was eluted from the resin at pH 5 as the metal-free ligand under conditions where pKa (IDA) b pHb pKa β-hydroxyhistidine amide (bleomycin). Bleomycin was captured on a Cu(II)-loaded IDA resin at pH 9 in 50% yield from bleomycin-containing S. verticillus culture that was pre-processed using XAD-2 resin to remove endogenously bound Cu(II). The approximate 25-fold purification of bleomycin from complex culture supernatant under aqueous conditions in a single step demonstrates the potential of Cu(II)-based metal affinity chromatography as a green chemistry platform for streamlined access to this high-value therapeutic agent. © 2012 Elsevier Inc. All rights reserved.
1. Introduction Bleomycin is the generic name for a family of metal-chelating glycopeptide compounds produced by Streptomyces verticillus with antibacterial and antitumor properties [1–7]. Bleomycin was first isolated as a mixture of 11 compounds (bleomycin A1–A6 and B1–B5) [1–3] with six additional analogs (demethyl A2, A2′-a, A2′-b, A2′-c, B1′ and B6) subsequently identified [8]. Bleomycin is an important chemotherapeutic agent for the treatment of several types of cancers with the combination of cisplatin, etoposide and bleomycin providing a 90% cure rate for testicular cancer [9]. Bleomycins feature diverse structural elements, including a metal binding region, a linker region, a bithiazole tail and a disaccharide motif [4–7]. Congeners are differentiated by the nature of the amine group substituted at the Cterminus. Clinical Blenoxane® (Bristol-Myers Squibb) is composed of about 60% bleomycin A2 with C-substituted aminopropyldimethylsulfonium; and 30% bleomycin B2 with C-substituted agmatine (Fig. 1). Other members of the bleomycin-type family include the phleomycins, tallysomycins and zorbamycin [10]. Phleomycin differs from bleomycin by a saturated thiazole motif (Fig. 1, boxed) [10,11]. The ability to coordinate metal ions is a key factor in the mechanism of action of bleomycin which has prompted many studies of
⁎ Corresponding author. Tel.: + 61 2 9351 6738; fax: + 61 2 9351 4717. E-mail address: [email protected] (R. Codd). 0162-0134/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.jinorgbio.2012.01.015
its coordination chemistry [12–21]. The reaction between bleomycin, Fe(II), oxygen and a one-electron reductant in vivo produces activated bleomycin, a low-spin ferric-peroxide species (O22 −-Fe(III)-bleomycin), which initiates DNA cleavage as the first step in its cytotoxic action [18,22,23]. The total synthesis of bleomycin [24] is not feasible for pharmaceutical-scale production and current access to this agent relies on bacterial fermentation. Conventional production of bleomycin involves culturing S. verticillus, followed by a multi-step purification process with several steps involving extraction using organic solvents [1,2]. A proportion of bleomycin produced by S. verticillus in culture is present as the 1:1 Cu(II) complex, which reflects the magnitude of the Cu(II)–bleomycin logK value (logK 12.6) [1,15]. The nature of the Cu(II)–bleomycin coordination complex has been revealed from X-ray crystallography [19,25] and spectroscopy [26] as a five-coordinate species with equatorial nitrogen donor atoms from imidazole, the deprotonated amide of β-hydroxyhistidine, pyrimidine, and the β-aminoalanine secondary amine; with axial donation from the β-aminoalanine primary amine. Given the affinity between Cu(II) and bleomycin, this work sought to examine the utility of Cu(II)-based immobilized metal affinity chromatography (IMAC) for capturing bleomycin from bacterial culture. This aqueous-compatible technique could reduce reliance on organic solvents and could reduce the number of chromatographic steps required for bleomycin purification. IMAC is used routinely for the purification of histidine-tagged (His-tag) recombinant proteins.
J. Gu, R. Codd / Journal of Inorganic Biochemistry 115 (2012) 198–203
199
Fig. 1. Structure of bleomycin A2 and bleomycin B2. The modified bithiazole motif of phleomycin D1 is boxed. Nitrogen donor atoms in the Cu(II)–bleomycin coordination sphere are shown in blue and italics.
The resin comprises a 1:1 complex between Ni(II) and immobilized iminodiacetate (IDA) to form resin-immobilized [Ni(IDA)(OH2)3]. The recombinant protein coordinates via the N- or C-terminal histidine tag to the three sites occupied by water ligands. After washing unbound components, the purified construct can be competitively eluted with imidazole or by lowering the pH value of the buffer to pHb pKa (histidine). In more recent times, Ni(II)-based IMAC has been used for the capture of low-molecular-weight non-proteinaceous ligands, including siderophores [27] and a related hydroxamic acid-based compound trichostatin A [28]. In this work, Cu(II)-based IMAC has been used to capture bleomycin from crude S. verticillus culture with high selectivity. The method describes a potential new green chemistry pathway towards this therapeutic agent.
from solutions in 96-well plates on a SpectraMax M5 plate reader. Analytical reverse-phase (RP)-HPLC-HPLC was conducted on an Agilent 1200 HPLC system using an Eclipse XDB-C18 column (particle size: 5 μm; column dimensions: 150 mm× 4.6 mm i.d.) with a linear gradient of 10–25% B over 20 min (A: 90% H2O; B: 10% ACN, both containing 0.1% trifluoroacetic acid), a flow rate of 0.5 mL min− 1 and an injection volume of 20 μL. Electrospray ionization mass spectrometry (ESI-MS) was carried out using a Finnigan LCQ mass spectrometer (San Jose, CA, USA) in positive ion mode. The mobile phase was methanol, with a flow rate = 0.20 mL min− 1; cone voltage was 25 V and the injection volume was 20 μL. Bacterial cultures were aerated on a Ratek OM6 orbital shaker. 2.3. Preparation and analysis of solutions of phleomycin or bleomycin
2. Experimental section 2.1. Materials 2.1.1. Chemistry All chemicals were used as received: NaCl (99%), K2HPO4 (>98%), ZnSO4∙7H2O; and phleomycin from S. verticillus (purity not specified) were from Sigma-Aldrich. NaH2PO4·2H2O (>98%) and CuSO4·5H2O (>98%) were from Ajax Finechem and Na2HPO4 (>99%) and CuCl2·2H2O (>99%) were from Merck. Na2EDTA·2H2O (99%) and 4-(2-hydroxyethyl) piperazine-1-ethanesulfonic acid (HEPES, 99%) were from Astral Scientific. Bleomycin complex (as sulfate salts, purity >99%) was from BioAustralis. 2.1.2. Biology The permanent stock of S. verticillus ATCC 15003 obtained from the American Type Culture Collection was kept frozen in DMSO at −80 °C. Corn starch, peptone, soybean flour and glucose were from Sigma-Aldrich. The following buffers were prepared: 25 mM HEPES, 0.5 M NaCl, pH 6, 7, 8 or 9 (binding conditions); or pH 5 (elution conditions); and 20 mM NaH2PO4/Na2HPO4, 0.5 M NaCl, pH 8. The following resins and columns were used: Ni(II) Sepharose™ 6 Fast Flow resin (GE Healthcare); pre-packed His GraviTrap affinity columns (1 mL, GE Healthcare); Micro Bio-Spin Chromatography Columns (Bio-Rad); and Amberlite XAD-2 (Supelco). 2.2. Instrumentation The pH measurements were made using a Microprocessor pH meter (Model: pH 211) with a 5-mm glass pH electrode (Model: HI 1330) from Hanna instruments. UV/visible spectra (from 450 to 850 nm) and single absorbance values (290 nm or 300 nm) were obtained
Solutions of phleomycin (10 mM, 5 mM, 2.5 mM, 1.25 mM, 0.625 mM, 0.313 mM, 0.156 mM) or bleomycin (10 mM, 5 mM, 2.5 mM, 1.25 mM, 0.625 mM, 0.313 mM, 0.156 mM, 0.078 mM, 0.039 mM, 0.020 mM) were prepared in HEPES buffer (25 mM HEPES, 0.5 M NaCl) at pH 6, 7, 8, or 9 (binding conditions); or at pH 5 (elution conditions). Aliquots (200 μL) of these solutions were transferred to a 96-well plate and the absorbance value at 300 nm (phleomycin) or 290 nm (bleomycin) was measured to generate standard curves. Selected solutions were used for IMAC experiments. Nine 2.5-μL aliquots of an aqueous 320 mM solution of CuCl2 were added to an 80-μL aliquot of phelomycin (10 mM, pH 6.0) in a 96-well plate and the electronic absorption spectrum of the solution was acquired from 450 to 850 nm after each addition. An aqueous sample of phleomycin as supplied was analyzed by positive ion ESI-MS. 2.4. Copper(II)-based immobilized metal affinity chromatography of phleomycin or bleomycin The first series of IMAC experiments was conducted using columns that were manually packed with a 250-μL bed volume of Cu(II)-charged IDA resin. First, the column was packed with Ni(II)charged IDA resin as supplied and the Ni(II) was removed using EDTA solution and exchanged for Cu(II) according to manufacturer's instructions. To remove unbound Cu(II), the column was preconditioned with 1 column volume (CV) of imidazole solution (0.25 M imidazole, 0.3 M NaCl, 50 mM NaH2PO4/Na2HPO4, pH 8), 5 CV of Milli-Q water and 5 CV of 20 mM NaH2PO4/Na2HPO4, 0.5 M NaCl, pH 8. Prior to sample addition, the column was equilibrated with 5 CV of binding buffer (pH 6, 7, 8 or 9). For each IMAC experiment, 1 CV of sample (250 μL) containing phelomycin or bleomycin was loaded onto the column. The column was washed with 20 CV of
200
J. Gu, R. Codd / Journal of Inorganic Biochemistry 115 (2012) 198–203
binding buffer and the bound components were then eluted with 6 CV of elution buffer (pH 5.0). Fractions of 250 μL were collected (27 fractions in total). The resin was stripped of Cu(II) and freshly recharged with Cu(II) between each experiment. Further experiments were conducted on a 1-mL His GraviTrap affinity column. Similar to above, the resin was stripped of Ni(II) and recharged with Cu(II). An aliquot of sample (250 μL) was loaded in each experiment. After loading, the column was washed with 4 CV of binding buffer and the bound components were then eluted with 4 CV of elution buffer. Fractions of 250 μL were collected (33 fractions in total). Phleomycin or bleomycin in the fractions was quantified by measuring the absorbance of a 200-μL aliquot of each fraction in a 96-well plate at 300 nm (phleomycin) or 290 nm (bleomycin). 2.5. Bacterial culture of S. verticillus S. verticillus was grown according to methods adapted from the literature [1,29]. First, the strain was cultured in 100 mL of seed medium consisting of (w/v) 1.0% glucose, 1.0% corn starch, 0.75% soybean meal, 0.75% peptone and 0.3% NaCl (medium pH adjusted to 6.5–7.0). Cells were inoculated by pipetting 1 mL of the permanent stock into the sterile media. The pre-culture was grown at 28 °C on an orbital shaker (220 rpm) for 48 h. The final culture was grown in medium containing (w/v) 2.5% corn starch, 0.5% glucose, 3.5% soybean meal, 0.1% K2HPO4, 0.05% ZnSO4∙7H2O and 0.01% CuSO4∙5H2O (medium pH adjusted to 6.5–7.0). The medium was distributed among four 250 mL conical flasks (100 mL of media in each). An aliquot (1 mL) of the pre-culture was inoculated into each flask. Growth was carried out on a rotary shaker (220 rpm) at 28 °C for 10 days. The culture suspension was pelleted by centrifugation (3900 rpm, 20 min) and the supernatant was retained. 2.6. XAD-2 chromatography of S. verticillus culture An aliquot of the culture supernatant (70 mL) was manually sorbed onto a column (30 cm ×2.5 cm, CV= 147 cm3) containing Amberlite XAD-2 resin that had been pre-equilibrated with 3 CV of Milli-Q water. The column was washed with Milli-Q water (2 CV), a 5% aqueous EDTA solution (3 CV), a 0.0025 N-sulfuric acid–methanol solution (1:1) (5 CV) and 100% methanol (4 CV). The column was re-equilibrated with Milli-Q water (2 CV). The flow rate was 5 mL min− 1. Forty fractions (20 mL each) were collected manually. Fractions 22–30 (the sulfuric acid–methanol eluate) which contained bleomycin were pooled and concentrated via rotary evaporation to an orange colored liquid (10.8 mL). An aliquot of this liquid (5 mL) was freeze-dried to yield a red-brown powder (19.5 mg). 2.7. Copper(II)-based immobilized metal affinity chromatography of S. verticillus culture The freeze-dried powder that was obtained following XAD-2 chromatography was re-dissolved in binding buffer (1 mL, pH 9.0). The solution was diluted two-fold and an aliquot (250 μL) was loaded onto a Cu(II)-charged 1-mL His GraviTrap column and subjected to the IMAC procedure as described in Section 2.4. 3. Results and discussion 3.1. General comments This work sought to explore the utility of Cu(II)-based metal affinity chromatography for the purification of the anticancer agent bleomycin. The 1:1 Cu(II):bleomycin coordination sphere has been established [19,25,26]. The most likely coordination mode between bleomycin and resin-immobilized [Cu(IDA)(OH2)3] would involve the three equatorially disposed nitrogen donor atoms from imidazole, the deprotonated amide of β-hydroxyhistidine and pyrimidine to
form the heteroleptic complex [Cu(IDA)(bleomycin)] (Fig. 2). The β-aminoalanine primary and secondary amine-based nitrogen donor atoms are depicted in Fig. 2 as unbound to Cu(II), although it cannot be discounted that these atoms could be involved in the coordination sphere of a distal [Cu(IDA)(OH2)3] unit. The heteroleptic complex [Cu(IDA)(bleomycin)] has a single negative charge about the first coordination shell arising from the deprotonated amide of β-hydroxyhistidine and a positive charge at the C-substituted aminopropyldimethylsulfonium (bleomycin A2) or agmatine (bleomycin B2) group to give an overall zwitterionic species. To control for any ion-exchange effects in the metal affinity chromatography, buffers used under binding conditions and elution conditions contained a high concentration of NaCl. At sufficiently high bleomycin concentrations, the ligand would compete against immobilized IDA for Cu(II) and effectively strip Cu(II) from the resin thereby preventing bleomycin binding. The balance between the target ligand binding to the immobilized metal or stripping the immobilized metal from the resin is prescribed by the respective logK values of complexation between Cu(II) and IDA (logK 16.1) or bleomycin (logK 12.6). The high extinction coefficient of phleomycin and bleomycin in the UV region (e.g., Cu(II)-phelomycin at 300 nm, ε≈0.8×104 M− 1 cm− 1) prompted the use of a UV-based detection system for establishing the binding profiles of these standards on a Cu(II)-loaded IDA matrix. An alternative detection method of using a Cu(II)-addition assay and measuring absorbance in the visible region of the spectrum (e.g., Cu(II)– phelomycin at 600 nm, ε≈110 M− 1 cm− 1) [30] was found to lack sensitivity for the low concentrations of phleomycin and bleomycin used. 3.2. Optimizing binding conditions of phleomycin or bleomycin Initial experiments were undertaken to determine the retention of standard phleomycin or bleomycin on a Cu(II)-loaded IDA resin and to establish optimal binding conditions with regard to pH value and loading capacity. Commercial phleomycin was supplied as a Cu(II)containing complex and without introducing a step to remove Cu(II), it was expected that the proportion of Cu(II)–phleomycin in the mixture would not be retained on the resin. Only the phleomycin present as free ligand would have potential for binding. The positive ion ESIMS for phleomycin gave signals from m/z 1488.53 to 1493.00 with an isotope pattern that matched well with the simulated isotope pattern for the Cu(II)-loaded phleomycin species [M− H+ + Cu2 +] + (Fig. 1S, Table 1S; Supplementary Material). Lower intensity signals at m/z 1438.73–1453.53 represented free phleomycin. The spectrum showed that a significant proportion (60–80%) of the commercial phleomycin sample was Cu(II)-loaded. Characterization of phleomycin by electronic absorption spectroscopy using Cu(II) titration was in agreement with the ESI-MS data and showed that about 60–70% of the commercial phleomycin was present as Cu(II)–phleomycin. When 625 nmol of phleomycin was adsorbed onto a 250 μL-bed volume of Cu(II)-loaded IDA resin at pH 6, the sample was quantitatively eluted from the resin with no binding evident (Fig. 3a). This indicated that neither Cu(II)-bound nor free phleomycin was interacting with the resin. A similar result was obtained at pH 7, although the broader tail of the peak shape indicated some interaction between the free ligand and the resin. From a 625 nmol loading at pH 8, about 150 nmol of phleomycin was retained on the resin, which was eluted upon washing the resin with the low-pH elution buffer. At pH 5, the integrity of the coordination between the immobilized Cu(II)-IDA moiety was maintained, yet that between Cu(II) and phleomycin involving the deprotonated β-hydroxyhistidine amide was disrupted [31] thereby releasing phleomycin from the immobilized heteroleptic complex [Cu(IDA)(phleomycin)] as free ligand. At pH 9, about 200 nmol of phleomycin was retained on the column and this value was selected as optimal for phleomycin or bleomycin binding. Higher pH values were not examined due to the precipitation of Cu(OH)2 on the resin.
J. Gu, R. Codd / Journal of Inorganic Biochemistry 115 (2012) 198–203
201
Fig. 2. Copper(II)-based metal affinity chromatography for the capture of bleomycin.
Due to the high concentration of Cu(II)–phleomycin in the commercial sample, all remaining optimization experiments were conducted with standard bleomycin, which was supplied as free ligand. When 2500 nmol of bleomycin was applied to a 250 μL-bed volume of Cu(II)-loaded IDA resin at pH 9, the sample was quantitatively eluted as the blue Cu(II)–bleomycin complex (Fig. 3b). At this concentration, the bleomycin was competing with the IDA for Cu(II) and was effectively stripping the Cu(II) from the resin, which was consistent with the blue color of the eluent and the white color of
the resin at the end of the experiment. Reducing the bleomycin loading to 78 or 39 nmol on freshly prepared Cu(II)-loaded IDA resin resulted in binding of about 9 (12%) or 5 (13%) nmol of bleomycin, respectively. These experiments showed that a larger bed volume of resin was required. At a bleomycin loading of 625 nmol on a 1-mL bed volume of Cu(II)-loaded IDA resin at pH 9, about 45% of the bleomycin eluted in the unbound fraction. At loadings of 39, 78 or 156 nmol, the retention of bleomycin on the resin was close to quantitative (Fig. 3c), with an average of about 9% appearing in the unbound fraction. These experiments
Fig. 3. Binding profiles determined from a UV-based assay of standard phleomycin (a) or bleomycin (b)–(c) on Cu(II)-loaded IDA resin: (a) phleomycin at constant loading (625 nmol) on 0.25 mL resin at variable pH; (b) constant pH (pH 9) on 0.25 mL resin with variable bleomycin loading; or (c) constant pH (pH 9) on 1 mL resin with variable bleomycin loading. Panel (d) shows the bleomycin binding profile at pH 9 of the S. verticillus culture supernatant on 1 mL Cu(II)-loaded IDA resin as detected using RP-HPLC measurements. The fractions analyzed by RP-HPLC displayed in Fig. 4 are marked with ticks on the upper axis. The binding-elution conditions in (a)–(d) are denoted with dotted lines.
202
J. Gu, R. Codd / Journal of Inorganic Biochemistry 115 (2012) 198–203
showed that the binding capacity of standard bleomycin towards Cu(II)loaded IDA was approximately 300 nmol mL− 1 resin. This capacity is comparable with the 40 mg mL− 1 capacity of a 150 kDa His-tag protein as stipulated in the manufacturer's specifications for Ni(II) Sepharose™ 6 Fast Flow resin. 3.3. Capture of bleomycin from S. verticillus culture Bleomycin is produced in S. verticillus culture as one of multiple secondary bacterial metabolites. The culture medium itself contains many components with some present in high concentrations (Section 2.5). As a result of competition from alternate Cu(II)-binding ligands, the binding profile of bleomycin-containing S. verticillus culture supernatant towards Cu(II)-loaded IDA resin would be expected to be different from that of standard bleomycin. This was evident from the IMAC experiment (Fig. 3d) which showed that 50% of bleomycin was retained on the resin with the remainder appearing in the fractions collected from the column when operated under binding conditions. There was a difference in the binding profiles of the congeners bleomycin A2 and bleomycin B2. The complexity of crude S. verticillus culture supernatant harvested at day 10 was evident from the RP-HPLC trace (Fig. 4a), which showed at least 50 peaks. The yield of bleomycin produced in culture is dependent upon the presence of Cu(II) in the medium [30], with some of the bleomycin produced as the 1:1 Cu(II)–bleomycin complex. In order to proceed to Cu(II)-based metal affinity chromatography, the culture was subject to purification on XAD-2 resin to remove Cu(II) from Cu(II)-coordinated bleomycin. Removal of the Cu(II) was achieved by loading the solution on the non-ionexchangeable macroreticular resin Amberlite XAD-2 which adsorbed the bleomycin. The Cu(II) ions were removed from the adsorbed Cu(II)–bleomycin complex upon washing the resin with EDTA. The resin-bound, Cu(II)-free bleomycin was eluted with an acidic methanol solution. The RP-HPLC trace from the XAD-2-treated eluent (Fig. 4b) showed a similar level of complexity to the untreated culture. The Cu(II)-free bleomycin-containing culture was processed on a 1-mL bed volume of Cu(II)-loaded IDA resin using conditions established for the bleomycin standard. In this instance, the UV-based detection method for pure bleomycin was not able to be used, due to the presence of multiple UV-active species. Each fraction volume collected under binding conditions (fractions 1–17) and under elution conditions (fractions 18–33) was analyzed by RP-HPLC. The first fraction collected following the elution of the void volume (fraction 5) showed the presence of multiple species which were not retained on the resin due to no or limited Cu(II) binding affinity (Fig. 4c). The RPHPLC trace from the solution of fraction 7 (Fig. 4d, black) showed two peaks at tR 10.73 min and tR 13.15 min, which increased in relative intensity compared to the trace from the same sample that was spiked with a sample of standard bleomycin (Fig. 4d, gray). This showed that the S. verticillus culture contained bleomycin and that based on previous RP-HPLC measurements [11], both bleomycin A2 (tR 10.73 min) and bleomycin B2 (tR 13.15 min) were present. Using the standard addition method and analysis of peak areas, the amount of bleomycin in the original S. verticillus culture was estimated to be approximately 14.3 mg L− 1, which is consistent with yields reported in the literature [32]. The presence of bleomycin in fractions collected under binding conditions showed that the binding capacity of the Cu(II)-loaded IDA resin towards bleomycin was reduced compared to the bleomycin standard due to the presence of competing ligands in the culture. This reduced capacity has been observed previously in the case of capture of siderophores and analogs from bacterial culture using Ni(II)-based metal affinity chromatography [27,28]. Compared to bleomycin B2 (tR 13.15 min), a higher relative concentration of bleomycin A2 (tR 10.73 min) was evident in the fractions collected under binding conditions (fraction 9, Fig. 4e), which suggested that the analogs had differential affinities towards the Cu(II)-loaded IDA resin. Although bleomycin analogs can be fractionated using cation-
Fig. 4. RP-HPLC traces of (a) crude; or (b) XAD-2-treated S. verticillus culture supernatant; or fractions collected from the XAD-2-treated S. verticillus culture supernatant during processing on Cu(II)-loaded IDA resin (refer Fig. 3d): (c)–(f) under binding conditions; and (g)–(j) under elution conditions. The trace from a sample of fraction 7 (F7) spiked with standard bleomycin is shown in (d) (gray). The trace from a two-day-old sample of fraction 23 (F23) (black) and the same sample spiked with standard bleomycin (gray) is shown in the inset of (g). The gradient in (a) was used in (b)–(j) and has been omitted for clarity.
exchange chromatography [2,8,33], the high concentration of NaCl (0.5 M) present in buffers used during the binding and elution steps suggested that the differential binding was due to metal-affinity rather than ion-exchange phenomena. The resin was washed under binding conditions until the concentration of unbound components was minimized (Fig. 4f) after which point, the column was operated under elution conditions. Two well-resolved peaks attributable to bleomycin A2 (tR 10.78 min) and bleomycin B2 (tR 13.19 min) were evident in the RP-HPLC trace from fractions collected under elution conditions (fraction 23, Fig. 4g). Bleomycin A2 and bleomycin B2 had modest stability in the elution buffer at pH 5, as shown from the reduced intensity of the peaks in an aged sample of the same fraction (Fig. 4g, inset, black) compared to a bleomycin-spiked sample of the aged sample (Fig. 4g, inset, gray). The pH value of the bleomycincontaining eluent would need to be readjusted from pH 5 to pH
J. Gu, R. Codd / Journal of Inorganic Biochemistry 115 (2012) 198–203
8 for increased stability. Subsequent fractions (Figs. 4h and i) showed a higher relative concentration of bleomycin B2 compared to bleomycin A2, which was consistent with the differential affinities towards the Cu(II)-loaded IDA resin under binding conditions (Fig. 4e). The differential binding profiles of bleomycin A2 and bleomycin B2 towards the Cu(II)-IDA resin was analyzed from the RP-HPLC traces from individual fractions, as shown in the overall binding profile of the culture (Fig. 3d). The elution of the majority of bleomycin was complete at 7.75 mL (fraction 31) (Fig. 4j). A comparison of the RP-HPLC trace from the XAD-2 eluent (Fig. 4b) and the components eluted in fraction 23 (Fig. 4g) showed a high level of purification of bleomycin from a complex mixture using single-step Cu(II)-based metal affinity chromatography. The level of purification can be estimated using measures of reduced peak numbers (with a common baseline threshold of 13) from 45 to 2, giving a purification factor of 22.5; or reduced cumulative peak areas of 930 to 34, giving a purification factor of 27.4. Based on these estimates, Cu(II)-based metal affinity chromatography gave an approximate 25-fold purification of bleomycin from S. verticillus XAD-2 eluent in a single step. 4. Conclusion Copper(II)-based metal affinity chromatography was used to purify the anticancer agent bleomycin about 25-fold directly from XAD-2 treated S. verticillus culture. The method worked via the formation of the resin-immobilized heteroleptic complex [Cu(IDA)(bleomycin)]. The value of the method lies in its simplicity and its aqueous compatibility. Streamlined access to bleomycin using Cu(II)-based metal affinity chromatography could reduce production costs and improve the sustainability of bleomycin pharmaceutics production. At present, chemotherapy with bleomycin can require up to 360 mg compound [34] per patient at a cost of approximately AUD$9000. The method could be useful in the discovery of new bleomycins. A new bleomycin derivative NC0604 has been identified from S. verticillus var. pingyangensis n.var [29] which was shown to be a more potent anticancer agent than the parent [35]. Other studies have used fermentation optimization [32] and bioengineering approaches [36] for improving yields of native bleomycin and generating new bleomycin analogs. Copper(II)-based metal affinity chromatography could be used to map bleomycin metabolomes in these systems. Acknowledgments Dr. Rachel Codd was a member of Professor Hans C. Freeman's research group in the early days of her career and acknowledges the pleasure of being immersed in the field of Cu(II) metalloproteins and more broadly in biological inorganic chemistry which she continues to practice. Dr. Saroja Rao is thanked for assistance with culturing S. verticillus. Financial support was provided from the University of Sydney (Bridging Grant, 2009).
203
Appendix A. Supplementary data Supplementary data to this article can be found online at doi:10. 1016/j.jinorgbio.2012.01.015. References [1] H. Umezawa, K. Maeda, T. Takeuchi, Y. Okami, J. Antibiot. 19 (1966) 200–209. [2] H. Umezawa, Y. Suhara, T. Takita, K. Maeda, J. Antibiot. 19 (1966) 210–215. [3] M. Ishizuka, H. Takayama, T. Takeuchi, H. Umezawa, J. Antibiot. 20 (1967) 15–24. [4] S.M. Hecht, J. Nat. Prod. 63 (2000) 158–168. [5] U. Galm, M.H. Hager, S.G. Van Lanen, J. Ju, J.S. Thorson, B. Shen, Chem. Rev. 105 (2005) 739–758. [6] J. Chen, J. Stubbe, Nat. Rev. Cancer 5 (2005) 102–112. [7] B. La Ferla, C. Airoldi, C. Zona, A. Orsato, F. Cardona, S. Merlo, E. Sironi, G. D'Orazio, F. Nicotra, Nat. Prod. Rep. 28 (2011) 630–648. [8] A. Fujii, T. Takita, K. Maeda, H. Umezawa, J. Antibiot. 26 (1973) 396–397. [9] L.H. Einhorn, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 4592–4595. [10] L. Wang, B.-S. Yun, N.P. George, E. Wendt-Pienkowski, U. Galm, T.-J. Oh, J.M. Coughlin, G.A. Zhang, M. Tao, B. Shen, J. Nat. Prod. 70 (2007) 402–406. [11] L. Du, C. Sánchez, M. Chen, D.J. Edwards, B. Shen, Chem. Biol. 7 (2000) 623–642. [12] G.M. Ehrenfeld, N. Murugesan, S.M. Hecht, Inorg. Chem. 23 (1984) 1496–1498. [13] J. Kuwahara, T. Suzuki, Y. Sugiura, Biochem. Biophys. Res. Commun. 129 (1985) 368–374. [14] A.M. Calafat, H. Won, L.G. Marzilli, J. Am. Chem. Soc. 119 (1997) 3656–3664. [15] Y. Sugiura, K. Ishizu, K. Miyoshi, J. Antibiot. 32 (1979) 453–461. [16] N.J. Oppenheimer, L.O. Rodriguez, S.M. Hecht, Proc. Natl. Acad. Sci. U. S. A. 76 (1979) 5616–5620. [17] J.C. Dabrowiak, J. Inorg. Biochem. 13 (1980) 317–337. [18] J.W. Sam, X.-J. Tang, J. Peisach, J. Am. Chem. Soc. 116 (1994) 5250–5256. [19] M. Sugiyama, T. Kumagai, M. Hayashida, M. Maruyama, J. Biol. Chem. 277 (2002) 2311–2320. [20] K.E. Loeb, J.M. Zaleski, C.D. Hess, S.M. Hecht, E.I. Solomon, J. Am. Chem. Soc. 120 (1998) 1249–1259. [21] K.D. Goodwin, M.A. Lewis, E.C. Long, M.M. Georgiadis, Proc. Natl. Acad. Sci. U. S. A. 105 (2008) 5052–5056. [22] R.M. Burger, J. Peisach, S.B. Horwitz, J. Biol. Chem. 256 (1981) 11636–11644. [23] T.E. Westre, K.E. Loeb, J.M. Zaleski, B. Hedman, K.O. Hodgson, E.I. Solomon, J. Am. Chem. Soc. 117 (1995) 1309–1313. [24] Y. Aoyagi, K. Katano, H. Suguna, J. Primeau, L.-H. Chang, S.M. Hecht, J. Am. Chem. Soc. 104 (1982) 5537–5538. [25] Y. Iitaka, H. Nakamura, T. Nakatani, Y. Muraoka, A. Fujii, T. Takita, H. Umezawa, J. Antibiot. 31 (1978) 1070–1072. [26] A. Decker, M.S. Chow, J.N. Kemsley, N. Lehnert, E.I. Solomon, J. Am. Chem. Soc. 128 (2006) 4719–4733. [27] N. Braich, R. Codd, Analyst 133 (2008) 877–880. [28] N. Ejje, E. Lacey, R. Codd, RSC Adv. 2 (2012) 333–337. [29] C.-Y. Chen, S. Si, Q. He, H. Xu, M.-Y. Lu, Y. Xie, Y. Wang, R. Chen, J. Antibiot. 61 (2008) 747–751. [30] T. Takita, Y. Muraoka, T. Nakatani, A. Fujii, Y. Iitaka, H. Umezawa, J. Antibiot. 31 (1978) 1073–1077. [31] H. Sigel, R.B. Martin, Chem. Rev. 82 (1982) 385–426. [32] N. Zhang, X. Zhu, D.J. Yang, J. Cai, M. Tao, L. Wang, Y. Duan, B. Shen, Z. Xu, Appl. Microbiol. Biotechnol. 86 (2010) 1345–1353. [33] H. Umezawa, T. Takita, Struct. Bond. 40 (1980) 73–99. [34] C. Bokenmeyer, J. Clin. Oncol. 26 (2008) 1783–1785. [35] W. Shi, C.-Y. Chen, X. Zhang, H. Xu, R. Chen, Q. He, Oncol. Rep. 24 (2010) 629–635. [36] M. Tao, L. Wang, E. Wendt-Pienkowski, N. Zhang, D.J. Yang, U. Galm, J.M. Coughlin, Z. Xu, B. Shen, Mol. Biosyst. 6 (2010) 349–356.
Contents lists available at SciVerse ScienceDirect
Journal of Inorganic Biochemistry journal homepage: www.elsevier.com/locate/jinorgbio
Copper(II)-based metal affinity chromatography for the isolation of the anticancer agent bleomycin from Streptomyces verticillus culture Jiesi Gu, Rachel Codd ⁎ School of Medical Sciences (Pharmacology) and Bosch Institute, University of Sydney, New South Wales 2006, Australia
a r t i c l e
i n f o
Article history: Received 24 November 2011 Received in revised form 27 January 2012 Accepted 31 January 2012 Available online 9 February 2012 Keywords: Metal affinity chromatography Bleomycin Copper complexes Green chemistry
a b s t r a c t The glycopeptide-based bleomycins are structurally complex natural products produced by Streptomyces verticillus used in combination therapy against testicular and other cancers. Bleomycin has a high affinity towards a range of transition metal ions with the 1:1 Fe(II) complex relevant to its mechanism of action in vivo and the 1:1 Cu(II) complex relevant to its production from culture. The affinity between Cu(II) and bleomycin was the underlying principle for using Cu(II)-based metal affinity chromatography in this work to selectively capture bleomycin from crude S. verticillus culture. A solution of standard bleomycin was retained at a binding capacity of 300 nmol mL− 1 on a 1-mL bed volume of Cu(II)-loaded iminodiacetate (IDA) resin at pH 9 via the formation of the heteroleptic immobilized complex [Cu(IDA)(bleomycin)]. Bleomycin was eluted from the resin at pH 5 as the metal-free ligand under conditions where pKa (IDA) b pHb pKa β-hydroxyhistidine amide (bleomycin). Bleomycin was captured on a Cu(II)-loaded IDA resin at pH 9 in 50% yield from bleomycin-containing S. verticillus culture that was pre-processed using XAD-2 resin to remove endogenously bound Cu(II). The approximate 25-fold purification of bleomycin from complex culture supernatant under aqueous conditions in a single step demonstrates the potential of Cu(II)-based metal affinity chromatography as a green chemistry platform for streamlined access to this high-value therapeutic agent. © 2012 Elsevier Inc. All rights reserved.
1. Introduction Bleomycin is the generic name for a family of metal-chelating glycopeptide compounds produced by Streptomyces verticillus with antibacterial and antitumor properties [1–7]. Bleomycin was first isolated as a mixture of 11 compounds (bleomycin A1–A6 and B1–B5) [1–3] with six additional analogs (demethyl A2, A2′-a, A2′-b, A2′-c, B1′ and B6) subsequently identified [8]. Bleomycin is an important chemotherapeutic agent for the treatment of several types of cancers with the combination of cisplatin, etoposide and bleomycin providing a 90% cure rate for testicular cancer [9]. Bleomycins feature diverse structural elements, including a metal binding region, a linker region, a bithiazole tail and a disaccharide motif [4–7]. Congeners are differentiated by the nature of the amine group substituted at the Cterminus. Clinical Blenoxane® (Bristol-Myers Squibb) is composed of about 60% bleomycin A2 with C-substituted aminopropyldimethylsulfonium; and 30% bleomycin B2 with C-substituted agmatine (Fig. 1). Other members of the bleomycin-type family include the phleomycins, tallysomycins and zorbamycin [10]. Phleomycin differs from bleomycin by a saturated thiazole motif (Fig. 1, boxed) [10,11]. The ability to coordinate metal ions is a key factor in the mechanism of action of bleomycin which has prompted many studies of
⁎ Corresponding author. Tel.: + 61 2 9351 6738; fax: + 61 2 9351 4717. E-mail address: [email protected] (R. Codd). 0162-0134/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.jinorgbio.2012.01.015
its coordination chemistry [12–21]. The reaction between bleomycin, Fe(II), oxygen and a one-electron reductant in vivo produces activated bleomycin, a low-spin ferric-peroxide species (O22 −-Fe(III)-bleomycin), which initiates DNA cleavage as the first step in its cytotoxic action [18,22,23]. The total synthesis of bleomycin [24] is not feasible for pharmaceutical-scale production and current access to this agent relies on bacterial fermentation. Conventional production of bleomycin involves culturing S. verticillus, followed by a multi-step purification process with several steps involving extraction using organic solvents [1,2]. A proportion of bleomycin produced by S. verticillus in culture is present as the 1:1 Cu(II) complex, which reflects the magnitude of the Cu(II)–bleomycin logK value (logK 12.6) [1,15]. The nature of the Cu(II)–bleomycin coordination complex has been revealed from X-ray crystallography [19,25] and spectroscopy [26] as a five-coordinate species with equatorial nitrogen donor atoms from imidazole, the deprotonated amide of β-hydroxyhistidine, pyrimidine, and the β-aminoalanine secondary amine; with axial donation from the β-aminoalanine primary amine. Given the affinity between Cu(II) and bleomycin, this work sought to examine the utility of Cu(II)-based immobilized metal affinity chromatography (IMAC) for capturing bleomycin from bacterial culture. This aqueous-compatible technique could reduce reliance on organic solvents and could reduce the number of chromatographic steps required for bleomycin purification. IMAC is used routinely for the purification of histidine-tagged (His-tag) recombinant proteins.
J. Gu, R. Codd / Journal of Inorganic Biochemistry 115 (2012) 198–203
199
Fig. 1. Structure of bleomycin A2 and bleomycin B2. The modified bithiazole motif of phleomycin D1 is boxed. Nitrogen donor atoms in the Cu(II)–bleomycin coordination sphere are shown in blue and italics.
The resin comprises a 1:1 complex between Ni(II) and immobilized iminodiacetate (IDA) to form resin-immobilized [Ni(IDA)(OH2)3]. The recombinant protein coordinates via the N- or C-terminal histidine tag to the three sites occupied by water ligands. After washing unbound components, the purified construct can be competitively eluted with imidazole or by lowering the pH value of the buffer to pHb pKa (histidine). In more recent times, Ni(II)-based IMAC has been used for the capture of low-molecular-weight non-proteinaceous ligands, including siderophores [27] and a related hydroxamic acid-based compound trichostatin A [28]. In this work, Cu(II)-based IMAC has been used to capture bleomycin from crude S. verticillus culture with high selectivity. The method describes a potential new green chemistry pathway towards this therapeutic agent.
from solutions in 96-well plates on a SpectraMax M5 plate reader. Analytical reverse-phase (RP)-HPLC-HPLC was conducted on an Agilent 1200 HPLC system using an Eclipse XDB-C18 column (particle size: 5 μm; column dimensions: 150 mm× 4.6 mm i.d.) with a linear gradient of 10–25% B over 20 min (A: 90% H2O; B: 10% ACN, both containing 0.1% trifluoroacetic acid), a flow rate of 0.5 mL min− 1 and an injection volume of 20 μL. Electrospray ionization mass spectrometry (ESI-MS) was carried out using a Finnigan LCQ mass spectrometer (San Jose, CA, USA) in positive ion mode. The mobile phase was methanol, with a flow rate = 0.20 mL min− 1; cone voltage was 25 V and the injection volume was 20 μL. Bacterial cultures were aerated on a Ratek OM6 orbital shaker. 2.3. Preparation and analysis of solutions of phleomycin or bleomycin
2. Experimental section 2.1. Materials 2.1.1. Chemistry All chemicals were used as received: NaCl (99%), K2HPO4 (>98%), ZnSO4∙7H2O; and phleomycin from S. verticillus (purity not specified) were from Sigma-Aldrich. NaH2PO4·2H2O (>98%) and CuSO4·5H2O (>98%) were from Ajax Finechem and Na2HPO4 (>99%) and CuCl2·2H2O (>99%) were from Merck. Na2EDTA·2H2O (99%) and 4-(2-hydroxyethyl) piperazine-1-ethanesulfonic acid (HEPES, 99%) were from Astral Scientific. Bleomycin complex (as sulfate salts, purity >99%) was from BioAustralis. 2.1.2. Biology The permanent stock of S. verticillus ATCC 15003 obtained from the American Type Culture Collection was kept frozen in DMSO at −80 °C. Corn starch, peptone, soybean flour and glucose were from Sigma-Aldrich. The following buffers were prepared: 25 mM HEPES, 0.5 M NaCl, pH 6, 7, 8 or 9 (binding conditions); or pH 5 (elution conditions); and 20 mM NaH2PO4/Na2HPO4, 0.5 M NaCl, pH 8. The following resins and columns were used: Ni(II) Sepharose™ 6 Fast Flow resin (GE Healthcare); pre-packed His GraviTrap affinity columns (1 mL, GE Healthcare); Micro Bio-Spin Chromatography Columns (Bio-Rad); and Amberlite XAD-2 (Supelco). 2.2. Instrumentation The pH measurements were made using a Microprocessor pH meter (Model: pH 211) with a 5-mm glass pH electrode (Model: HI 1330) from Hanna instruments. UV/visible spectra (from 450 to 850 nm) and single absorbance values (290 nm or 300 nm) were obtained
Solutions of phleomycin (10 mM, 5 mM, 2.5 mM, 1.25 mM, 0.625 mM, 0.313 mM, 0.156 mM) or bleomycin (10 mM, 5 mM, 2.5 mM, 1.25 mM, 0.625 mM, 0.313 mM, 0.156 mM, 0.078 mM, 0.039 mM, 0.020 mM) were prepared in HEPES buffer (25 mM HEPES, 0.5 M NaCl) at pH 6, 7, 8, or 9 (binding conditions); or at pH 5 (elution conditions). Aliquots (200 μL) of these solutions were transferred to a 96-well plate and the absorbance value at 300 nm (phleomycin) or 290 nm (bleomycin) was measured to generate standard curves. Selected solutions were used for IMAC experiments. Nine 2.5-μL aliquots of an aqueous 320 mM solution of CuCl2 were added to an 80-μL aliquot of phelomycin (10 mM, pH 6.0) in a 96-well plate and the electronic absorption spectrum of the solution was acquired from 450 to 850 nm after each addition. An aqueous sample of phleomycin as supplied was analyzed by positive ion ESI-MS. 2.4. Copper(II)-based immobilized metal affinity chromatography of phleomycin or bleomycin The first series of IMAC experiments was conducted using columns that were manually packed with a 250-μL bed volume of Cu(II)-charged IDA resin. First, the column was packed with Ni(II)charged IDA resin as supplied and the Ni(II) was removed using EDTA solution and exchanged for Cu(II) according to manufacturer's instructions. To remove unbound Cu(II), the column was preconditioned with 1 column volume (CV) of imidazole solution (0.25 M imidazole, 0.3 M NaCl, 50 mM NaH2PO4/Na2HPO4, pH 8), 5 CV of Milli-Q water and 5 CV of 20 mM NaH2PO4/Na2HPO4, 0.5 M NaCl, pH 8. Prior to sample addition, the column was equilibrated with 5 CV of binding buffer (pH 6, 7, 8 or 9). For each IMAC experiment, 1 CV of sample (250 μL) containing phelomycin or bleomycin was loaded onto the column. The column was washed with 20 CV of
200
J. Gu, R. Codd / Journal of Inorganic Biochemistry 115 (2012) 198–203
binding buffer and the bound components were then eluted with 6 CV of elution buffer (pH 5.0). Fractions of 250 μL were collected (27 fractions in total). The resin was stripped of Cu(II) and freshly recharged with Cu(II) between each experiment. Further experiments were conducted on a 1-mL His GraviTrap affinity column. Similar to above, the resin was stripped of Ni(II) and recharged with Cu(II). An aliquot of sample (250 μL) was loaded in each experiment. After loading, the column was washed with 4 CV of binding buffer and the bound components were then eluted with 4 CV of elution buffer. Fractions of 250 μL were collected (33 fractions in total). Phleomycin or bleomycin in the fractions was quantified by measuring the absorbance of a 200-μL aliquot of each fraction in a 96-well plate at 300 nm (phleomycin) or 290 nm (bleomycin). 2.5. Bacterial culture of S. verticillus S. verticillus was grown according to methods adapted from the literature [1,29]. First, the strain was cultured in 100 mL of seed medium consisting of (w/v) 1.0% glucose, 1.0% corn starch, 0.75% soybean meal, 0.75% peptone and 0.3% NaCl (medium pH adjusted to 6.5–7.0). Cells were inoculated by pipetting 1 mL of the permanent stock into the sterile media. The pre-culture was grown at 28 °C on an orbital shaker (220 rpm) for 48 h. The final culture was grown in medium containing (w/v) 2.5% corn starch, 0.5% glucose, 3.5% soybean meal, 0.1% K2HPO4, 0.05% ZnSO4∙7H2O and 0.01% CuSO4∙5H2O (medium pH adjusted to 6.5–7.0). The medium was distributed among four 250 mL conical flasks (100 mL of media in each). An aliquot (1 mL) of the pre-culture was inoculated into each flask. Growth was carried out on a rotary shaker (220 rpm) at 28 °C for 10 days. The culture suspension was pelleted by centrifugation (3900 rpm, 20 min) and the supernatant was retained. 2.6. XAD-2 chromatography of S. verticillus culture An aliquot of the culture supernatant (70 mL) was manually sorbed onto a column (30 cm ×2.5 cm, CV= 147 cm3) containing Amberlite XAD-2 resin that had been pre-equilibrated with 3 CV of Milli-Q water. The column was washed with Milli-Q water (2 CV), a 5% aqueous EDTA solution (3 CV), a 0.0025 N-sulfuric acid–methanol solution (1:1) (5 CV) and 100% methanol (4 CV). The column was re-equilibrated with Milli-Q water (2 CV). The flow rate was 5 mL min− 1. Forty fractions (20 mL each) were collected manually. Fractions 22–30 (the sulfuric acid–methanol eluate) which contained bleomycin were pooled and concentrated via rotary evaporation to an orange colored liquid (10.8 mL). An aliquot of this liquid (5 mL) was freeze-dried to yield a red-brown powder (19.5 mg). 2.7. Copper(II)-based immobilized metal affinity chromatography of S. verticillus culture The freeze-dried powder that was obtained following XAD-2 chromatography was re-dissolved in binding buffer (1 mL, pH 9.0). The solution was diluted two-fold and an aliquot (250 μL) was loaded onto a Cu(II)-charged 1-mL His GraviTrap column and subjected to the IMAC procedure as described in Section 2.4. 3. Results and discussion 3.1. General comments This work sought to explore the utility of Cu(II)-based metal affinity chromatography for the purification of the anticancer agent bleomycin. The 1:1 Cu(II):bleomycin coordination sphere has been established [19,25,26]. The most likely coordination mode between bleomycin and resin-immobilized [Cu(IDA)(OH2)3] would involve the three equatorially disposed nitrogen donor atoms from imidazole, the deprotonated amide of β-hydroxyhistidine and pyrimidine to
form the heteroleptic complex [Cu(IDA)(bleomycin)] (Fig. 2). The β-aminoalanine primary and secondary amine-based nitrogen donor atoms are depicted in Fig. 2 as unbound to Cu(II), although it cannot be discounted that these atoms could be involved in the coordination sphere of a distal [Cu(IDA)(OH2)3] unit. The heteroleptic complex [Cu(IDA)(bleomycin)] has a single negative charge about the first coordination shell arising from the deprotonated amide of β-hydroxyhistidine and a positive charge at the C-substituted aminopropyldimethylsulfonium (bleomycin A2) or agmatine (bleomycin B2) group to give an overall zwitterionic species. To control for any ion-exchange effects in the metal affinity chromatography, buffers used under binding conditions and elution conditions contained a high concentration of NaCl. At sufficiently high bleomycin concentrations, the ligand would compete against immobilized IDA for Cu(II) and effectively strip Cu(II) from the resin thereby preventing bleomycin binding. The balance between the target ligand binding to the immobilized metal or stripping the immobilized metal from the resin is prescribed by the respective logK values of complexation between Cu(II) and IDA (logK 16.1) or bleomycin (logK 12.6). The high extinction coefficient of phleomycin and bleomycin in the UV region (e.g., Cu(II)-phelomycin at 300 nm, ε≈0.8×104 M− 1 cm− 1) prompted the use of a UV-based detection system for establishing the binding profiles of these standards on a Cu(II)-loaded IDA matrix. An alternative detection method of using a Cu(II)-addition assay and measuring absorbance in the visible region of the spectrum (e.g., Cu(II)– phelomycin at 600 nm, ε≈110 M− 1 cm− 1) [30] was found to lack sensitivity for the low concentrations of phleomycin and bleomycin used. 3.2. Optimizing binding conditions of phleomycin or bleomycin Initial experiments were undertaken to determine the retention of standard phleomycin or bleomycin on a Cu(II)-loaded IDA resin and to establish optimal binding conditions with regard to pH value and loading capacity. Commercial phleomycin was supplied as a Cu(II)containing complex and without introducing a step to remove Cu(II), it was expected that the proportion of Cu(II)–phleomycin in the mixture would not be retained on the resin. Only the phleomycin present as free ligand would have potential for binding. The positive ion ESIMS for phleomycin gave signals from m/z 1488.53 to 1493.00 with an isotope pattern that matched well with the simulated isotope pattern for the Cu(II)-loaded phleomycin species [M− H+ + Cu2 +] + (Fig. 1S, Table 1S; Supplementary Material). Lower intensity signals at m/z 1438.73–1453.53 represented free phleomycin. The spectrum showed that a significant proportion (60–80%) of the commercial phleomycin sample was Cu(II)-loaded. Characterization of phleomycin by electronic absorption spectroscopy using Cu(II) titration was in agreement with the ESI-MS data and showed that about 60–70% of the commercial phleomycin was present as Cu(II)–phleomycin. When 625 nmol of phleomycin was adsorbed onto a 250 μL-bed volume of Cu(II)-loaded IDA resin at pH 6, the sample was quantitatively eluted from the resin with no binding evident (Fig. 3a). This indicated that neither Cu(II)-bound nor free phleomycin was interacting with the resin. A similar result was obtained at pH 7, although the broader tail of the peak shape indicated some interaction between the free ligand and the resin. From a 625 nmol loading at pH 8, about 150 nmol of phleomycin was retained on the resin, which was eluted upon washing the resin with the low-pH elution buffer. At pH 5, the integrity of the coordination between the immobilized Cu(II)-IDA moiety was maintained, yet that between Cu(II) and phleomycin involving the deprotonated β-hydroxyhistidine amide was disrupted [31] thereby releasing phleomycin from the immobilized heteroleptic complex [Cu(IDA)(phleomycin)] as free ligand. At pH 9, about 200 nmol of phleomycin was retained on the column and this value was selected as optimal for phleomycin or bleomycin binding. Higher pH values were not examined due to the precipitation of Cu(OH)2 on the resin.
J. Gu, R. Codd / Journal of Inorganic Biochemistry 115 (2012) 198–203
201
Fig. 2. Copper(II)-based metal affinity chromatography for the capture of bleomycin.
Due to the high concentration of Cu(II)–phleomycin in the commercial sample, all remaining optimization experiments were conducted with standard bleomycin, which was supplied as free ligand. When 2500 nmol of bleomycin was applied to a 250 μL-bed volume of Cu(II)-loaded IDA resin at pH 9, the sample was quantitatively eluted as the blue Cu(II)–bleomycin complex (Fig. 3b). At this concentration, the bleomycin was competing with the IDA for Cu(II) and was effectively stripping the Cu(II) from the resin, which was consistent with the blue color of the eluent and the white color of
the resin at the end of the experiment. Reducing the bleomycin loading to 78 or 39 nmol on freshly prepared Cu(II)-loaded IDA resin resulted in binding of about 9 (12%) or 5 (13%) nmol of bleomycin, respectively. These experiments showed that a larger bed volume of resin was required. At a bleomycin loading of 625 nmol on a 1-mL bed volume of Cu(II)-loaded IDA resin at pH 9, about 45% of the bleomycin eluted in the unbound fraction. At loadings of 39, 78 or 156 nmol, the retention of bleomycin on the resin was close to quantitative (Fig. 3c), with an average of about 9% appearing in the unbound fraction. These experiments
Fig. 3. Binding profiles determined from a UV-based assay of standard phleomycin (a) or bleomycin (b)–(c) on Cu(II)-loaded IDA resin: (a) phleomycin at constant loading (625 nmol) on 0.25 mL resin at variable pH; (b) constant pH (pH 9) on 0.25 mL resin with variable bleomycin loading; or (c) constant pH (pH 9) on 1 mL resin with variable bleomycin loading. Panel (d) shows the bleomycin binding profile at pH 9 of the S. verticillus culture supernatant on 1 mL Cu(II)-loaded IDA resin as detected using RP-HPLC measurements. The fractions analyzed by RP-HPLC displayed in Fig. 4 are marked with ticks on the upper axis. The binding-elution conditions in (a)–(d) are denoted with dotted lines.
202
J. Gu, R. Codd / Journal of Inorganic Biochemistry 115 (2012) 198–203
showed that the binding capacity of standard bleomycin towards Cu(II)loaded IDA was approximately 300 nmol mL− 1 resin. This capacity is comparable with the 40 mg mL− 1 capacity of a 150 kDa His-tag protein as stipulated in the manufacturer's specifications for Ni(II) Sepharose™ 6 Fast Flow resin. 3.3. Capture of bleomycin from S. verticillus culture Bleomycin is produced in S. verticillus culture as one of multiple secondary bacterial metabolites. The culture medium itself contains many components with some present in high concentrations (Section 2.5). As a result of competition from alternate Cu(II)-binding ligands, the binding profile of bleomycin-containing S. verticillus culture supernatant towards Cu(II)-loaded IDA resin would be expected to be different from that of standard bleomycin. This was evident from the IMAC experiment (Fig. 3d) which showed that 50% of bleomycin was retained on the resin with the remainder appearing in the fractions collected from the column when operated under binding conditions. There was a difference in the binding profiles of the congeners bleomycin A2 and bleomycin B2. The complexity of crude S. verticillus culture supernatant harvested at day 10 was evident from the RP-HPLC trace (Fig. 4a), which showed at least 50 peaks. The yield of bleomycin produced in culture is dependent upon the presence of Cu(II) in the medium [30], with some of the bleomycin produced as the 1:1 Cu(II)–bleomycin complex. In order to proceed to Cu(II)-based metal affinity chromatography, the culture was subject to purification on XAD-2 resin to remove Cu(II) from Cu(II)-coordinated bleomycin. Removal of the Cu(II) was achieved by loading the solution on the non-ionexchangeable macroreticular resin Amberlite XAD-2 which adsorbed the bleomycin. The Cu(II) ions were removed from the adsorbed Cu(II)–bleomycin complex upon washing the resin with EDTA. The resin-bound, Cu(II)-free bleomycin was eluted with an acidic methanol solution. The RP-HPLC trace from the XAD-2-treated eluent (Fig. 4b) showed a similar level of complexity to the untreated culture. The Cu(II)-free bleomycin-containing culture was processed on a 1-mL bed volume of Cu(II)-loaded IDA resin using conditions established for the bleomycin standard. In this instance, the UV-based detection method for pure bleomycin was not able to be used, due to the presence of multiple UV-active species. Each fraction volume collected under binding conditions (fractions 1–17) and under elution conditions (fractions 18–33) was analyzed by RP-HPLC. The first fraction collected following the elution of the void volume (fraction 5) showed the presence of multiple species which were not retained on the resin due to no or limited Cu(II) binding affinity (Fig. 4c). The RPHPLC trace from the solution of fraction 7 (Fig. 4d, black) showed two peaks at tR 10.73 min and tR 13.15 min, which increased in relative intensity compared to the trace from the same sample that was spiked with a sample of standard bleomycin (Fig. 4d, gray). This showed that the S. verticillus culture contained bleomycin and that based on previous RP-HPLC measurements [11], both bleomycin A2 (tR 10.73 min) and bleomycin B2 (tR 13.15 min) were present. Using the standard addition method and analysis of peak areas, the amount of bleomycin in the original S. verticillus culture was estimated to be approximately 14.3 mg L− 1, which is consistent with yields reported in the literature [32]. The presence of bleomycin in fractions collected under binding conditions showed that the binding capacity of the Cu(II)-loaded IDA resin towards bleomycin was reduced compared to the bleomycin standard due to the presence of competing ligands in the culture. This reduced capacity has been observed previously in the case of capture of siderophores and analogs from bacterial culture using Ni(II)-based metal affinity chromatography [27,28]. Compared to bleomycin B2 (tR 13.15 min), a higher relative concentration of bleomycin A2 (tR 10.73 min) was evident in the fractions collected under binding conditions (fraction 9, Fig. 4e), which suggested that the analogs had differential affinities towards the Cu(II)-loaded IDA resin. Although bleomycin analogs can be fractionated using cation-
Fig. 4. RP-HPLC traces of (a) crude; or (b) XAD-2-treated S. verticillus culture supernatant; or fractions collected from the XAD-2-treated S. verticillus culture supernatant during processing on Cu(II)-loaded IDA resin (refer Fig. 3d): (c)–(f) under binding conditions; and (g)–(j) under elution conditions. The trace from a sample of fraction 7 (F7) spiked with standard bleomycin is shown in (d) (gray). The trace from a two-day-old sample of fraction 23 (F23) (black) and the same sample spiked with standard bleomycin (gray) is shown in the inset of (g). The gradient in (a) was used in (b)–(j) and has been omitted for clarity.
exchange chromatography [2,8,33], the high concentration of NaCl (0.5 M) present in buffers used during the binding and elution steps suggested that the differential binding was due to metal-affinity rather than ion-exchange phenomena. The resin was washed under binding conditions until the concentration of unbound components was minimized (Fig. 4f) after which point, the column was operated under elution conditions. Two well-resolved peaks attributable to bleomycin A2 (tR 10.78 min) and bleomycin B2 (tR 13.19 min) were evident in the RP-HPLC trace from fractions collected under elution conditions (fraction 23, Fig. 4g). Bleomycin A2 and bleomycin B2 had modest stability in the elution buffer at pH 5, as shown from the reduced intensity of the peaks in an aged sample of the same fraction (Fig. 4g, inset, black) compared to a bleomycin-spiked sample of the aged sample (Fig. 4g, inset, gray). The pH value of the bleomycincontaining eluent would need to be readjusted from pH 5 to pH
J. Gu, R. Codd / Journal of Inorganic Biochemistry 115 (2012) 198–203
8 for increased stability. Subsequent fractions (Figs. 4h and i) showed a higher relative concentration of bleomycin B2 compared to bleomycin A2, which was consistent with the differential affinities towards the Cu(II)-loaded IDA resin under binding conditions (Fig. 4e). The differential binding profiles of bleomycin A2 and bleomycin B2 towards the Cu(II)-IDA resin was analyzed from the RP-HPLC traces from individual fractions, as shown in the overall binding profile of the culture (Fig. 3d). The elution of the majority of bleomycin was complete at 7.75 mL (fraction 31) (Fig. 4j). A comparison of the RP-HPLC trace from the XAD-2 eluent (Fig. 4b) and the components eluted in fraction 23 (Fig. 4g) showed a high level of purification of bleomycin from a complex mixture using single-step Cu(II)-based metal affinity chromatography. The level of purification can be estimated using measures of reduced peak numbers (with a common baseline threshold of 13) from 45 to 2, giving a purification factor of 22.5; or reduced cumulative peak areas of 930 to 34, giving a purification factor of 27.4. Based on these estimates, Cu(II)-based metal affinity chromatography gave an approximate 25-fold purification of bleomycin from S. verticillus XAD-2 eluent in a single step. 4. Conclusion Copper(II)-based metal affinity chromatography was used to purify the anticancer agent bleomycin about 25-fold directly from XAD-2 treated S. verticillus culture. The method worked via the formation of the resin-immobilized heteroleptic complex [Cu(IDA)(bleomycin)]. The value of the method lies in its simplicity and its aqueous compatibility. Streamlined access to bleomycin using Cu(II)-based metal affinity chromatography could reduce production costs and improve the sustainability of bleomycin pharmaceutics production. At present, chemotherapy with bleomycin can require up to 360 mg compound [34] per patient at a cost of approximately AUD$9000. The method could be useful in the discovery of new bleomycins. A new bleomycin derivative NC0604 has been identified from S. verticillus var. pingyangensis n.var [29] which was shown to be a more potent anticancer agent than the parent [35]. Other studies have used fermentation optimization [32] and bioengineering approaches [36] for improving yields of native bleomycin and generating new bleomycin analogs. Copper(II)-based metal affinity chromatography could be used to map bleomycin metabolomes in these systems. Acknowledgments Dr. Rachel Codd was a member of Professor Hans C. Freeman's research group in the early days of her career and acknowledges the pleasure of being immersed in the field of Cu(II) metalloproteins and more broadly in biological inorganic chemistry which she continues to practice. Dr. Saroja Rao is thanked for assistance with culturing S. verticillus. Financial support was provided from the University of Sydney (Bridging Grant, 2009).
203
Appendix A. Supplementary data Supplementary data to this article can be found online at doi:10. 1016/j.jinorgbio.2012.01.015. References [1] H. Umezawa, K. Maeda, T. Takeuchi, Y. Okami, J. Antibiot. 19 (1966) 200–209. [2] H. Umezawa, Y. Suhara, T. Takita, K. Maeda, J. Antibiot. 19 (1966) 210–215. [3] M. Ishizuka, H. Takayama, T. Takeuchi, H. Umezawa, J. Antibiot. 20 (1967) 15–24. [4] S.M. Hecht, J. Nat. Prod. 63 (2000) 158–168. [5] U. Galm, M.H. Hager, S.G. Van Lanen, J. Ju, J.S. Thorson, B. Shen, Chem. Rev. 105 (2005) 739–758. [6] J. Chen, J. Stubbe, Nat. Rev. Cancer 5 (2005) 102–112. [7] B. La Ferla, C. Airoldi, C. Zona, A. Orsato, F. Cardona, S. Merlo, E. Sironi, G. D'Orazio, F. Nicotra, Nat. Prod. Rep. 28 (2011) 630–648. [8] A. Fujii, T. Takita, K. Maeda, H. Umezawa, J. Antibiot. 26 (1973) 396–397. [9] L.H. Einhorn, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 4592–4595. [10] L. Wang, B.-S. Yun, N.P. George, E. Wendt-Pienkowski, U. Galm, T.-J. Oh, J.M. Coughlin, G.A. Zhang, M. Tao, B. Shen, J. Nat. Prod. 70 (2007) 402–406. [11] L. Du, C. Sánchez, M. Chen, D.J. Edwards, B. Shen, Chem. Biol. 7 (2000) 623–642. [12] G.M. Ehrenfeld, N. Murugesan, S.M. Hecht, Inorg. Chem. 23 (1984) 1496–1498. [13] J. Kuwahara, T. Suzuki, Y. Sugiura, Biochem. Biophys. Res. Commun. 129 (1985) 368–374. [14] A.M. Calafat, H. Won, L.G. Marzilli, J. Am. Chem. Soc. 119 (1997) 3656–3664. [15] Y. Sugiura, K. Ishizu, K. Miyoshi, J. Antibiot. 32 (1979) 453–461. [16] N.J. Oppenheimer, L.O. Rodriguez, S.M. Hecht, Proc. Natl. Acad. Sci. U. S. A. 76 (1979) 5616–5620. [17] J.C. Dabrowiak, J. Inorg. Biochem. 13 (1980) 317–337. [18] J.W. Sam, X.-J. Tang, J. Peisach, J. Am. Chem. Soc. 116 (1994) 5250–5256. [19] M. Sugiyama, T. Kumagai, M. Hayashida, M. Maruyama, J. Biol. Chem. 277 (2002) 2311–2320. [20] K.E. Loeb, J.M. Zaleski, C.D. Hess, S.M. Hecht, E.I. Solomon, J. Am. Chem. Soc. 120 (1998) 1249–1259. [21] K.D. Goodwin, M.A. Lewis, E.C. Long, M.M. Georgiadis, Proc. Natl. Acad. Sci. U. S. A. 105 (2008) 5052–5056. [22] R.M. Burger, J. Peisach, S.B. Horwitz, J. Biol. Chem. 256 (1981) 11636–11644. [23] T.E. Westre, K.E. Loeb, J.M. Zaleski, B. Hedman, K.O. Hodgson, E.I. Solomon, J. Am. Chem. Soc. 117 (1995) 1309–1313. [24] Y. Aoyagi, K. Katano, H. Suguna, J. Primeau, L.-H. Chang, S.M. Hecht, J. Am. Chem. Soc. 104 (1982) 5537–5538. [25] Y. Iitaka, H. Nakamura, T. Nakatani, Y. Muraoka, A. Fujii, T. Takita, H. Umezawa, J. Antibiot. 31 (1978) 1070–1072. [26] A. Decker, M.S. Chow, J.N. Kemsley, N. Lehnert, E.I. Solomon, J. Am. Chem. Soc. 128 (2006) 4719–4733. [27] N. Braich, R. Codd, Analyst 133 (2008) 877–880. [28] N. Ejje, E. Lacey, R. Codd, RSC Adv. 2 (2012) 333–337. [29] C.-Y. Chen, S. Si, Q. He, H. Xu, M.-Y. Lu, Y. Xie, Y. Wang, R. Chen, J. Antibiot. 61 (2008) 747–751. [30] T. Takita, Y. Muraoka, T. Nakatani, A. Fujii, Y. Iitaka, H. Umezawa, J. Antibiot. 31 (1978) 1073–1077. [31] H. Sigel, R.B. Martin, Chem. Rev. 82 (1982) 385–426. [32] N. Zhang, X. Zhu, D.J. Yang, J. Cai, M. Tao, L. Wang, Y. Duan, B. Shen, Z. Xu, Appl. Microbiol. Biotechnol. 86 (2010) 1345–1353. [33] H. Umezawa, T. Takita, Struct. Bond. 40 (1980) 73–99. [34] C. Bokenmeyer, J. Clin. Oncol. 26 (2008) 1783–1785. [35] W. Shi, C.-Y. Chen, X. Zhang, H. Xu, R. Chen, Q. He, Oncol. Rep. 24 (2010) 629–635. [36] M. Tao, L. Wang, E. Wendt-Pienkowski, N. Zhang, D.J. Yang, U. Galm, J.M. Coughlin, Z. Xu, B. Shen, Mol. Biosyst. 6 (2010) 349–356.