Skip to main content
NCBI home page
Molecular and Cellular Biology logoLink to Molecular and Cellular Biology
. 1998 Jun;18(6):3580–3585. doi: 10.1128/mcb.18.6.3580

The Nucleic Acid Binding Activity of Bleomycin Hydrolase Is Involved in Bleomycin Detoxification

Wenjin Zheng 1, Stephen Albert Johnston 1,*
PMCID: PMC108939  PMID: 9584198

Abstract

Yeast bleomycin hydrolase, Gal6p, is a cysteine peptidase that detoxifies the anticancer drug bleomycin. Gal6p is a dual-function protein capable of both nucleic acid binding and peptide cleavage. We now demonstrate that Gal6p exhibits sequence-independent, high-affinity binding to single-stranded DNA, nicked double-stranded DNA, and RNA. A region of the protein that is involved in binding both RNA and DNA substrates is delineated. Immunolocalization reveals that the Gal6 protein is chiefly cytoplasmic and thus may be involved in binding cellular RNAs. Variant Gal6 proteins that fail to bind nucleic acid also exhibit reduced ability to protect cells from bleomycin toxicity, suggesting that the nucleic acid binding activity of Gal6p is important in bleomycin detoxification and may be involved in its normal biological functions.

Bleomycin hydrolase is a cysteine peptidase that can cleave the glycopeptide bleomycin, which has been used widely as an anticancer drug in chemotherapy. Bleomycin hydrolase is highly conserved in organisms ranging from bacteria to humans (3, 12). The yeast form of this enzyme has been purified, and its crystal structure has been solved (12). We designated the yeast form of bleomycin hydrolase Gal6, as its expression is regulated by galactose (12, 35). We first identified Gal6p by virtue of its DNA binding property. It binds single-stranded DNA with high affinity (10 nM) and double-stranded DNA with much lower affinity (1 μM) (33).

The crystal structure of Gal6p revealed that it is a hexamers, its six subunits forming a ring with a central channel of about 20-Å diameter at the entrance. The peptidase active sites are located deep inside this tunnel, sequestered from the exterior. Other nucleic acid-binding proteins with a ring structure, e.g., Escherichia coli DNA polymerase β subunit (16), eukaryotic proliferating cell nuclear antigen (17), and Bacillus subtilis Trp RNA-binding attenuation protein (2), have been identified, as have other proteases with ring structures, e.g., the proteasome and tricorn, which form a unique self-compartmentalizing protease family (18, 21). However, Gal6 is the only ring protein with both of these activities.

Although a role for Gal6p in bleomycin detoxification has been well documented (7, 13, 25, 33), there is no evidence to date that its nucleic acid binding activity plays a role in this detoxification. The fact that bleomycin itself cleaves nucleic acids (14, 28) suggests that the peptidase and nucleic acid binding activities of Gal6p may be functionally linked. On the other hand, the DNA binding activity of Gal6p could be an in vitro artifact and may not reflect a biologically relevant function.

We report here the characterization of the nucleic acid binding activity of Gal6p. It binds to single-stranded DNA as well as single-stranded RNA. Immunolocalization demonstrates that most of the Gal6 protein is located in the cytoplasm, indicating that it may function mainly as an RNA-binding protein. Mutation studies show that the β-hairpin structure at the opening of the channel in Gal6 protein plays an essential role in the nucleic acid binding activity but has no effect on the peptidase activity. Cells bearing this mutation are more sensitive to bleomycin toxicity, implying that this binding activity is functional in vivo.

MATERIALS AND METHODS

Strains and media.

E. coli TG1 was used for plasmid amplification and single-stranded DNA plasmid production. E. coli BL21(DE3) was used for protein production. Bacteria were grown in L broth plus the necessary antibiotics. The Saccharomyces cerevisiae strains used were W303 (MATα ura3 leu2 his3 trp1) and Sc377 (MATα ura3 leu2 his3 TRP1::Δgal6). Cells were grown in rich medium (YEP) or selective medium and supplied with appropriate carbon sources. Yeast cells were transformed by the LiCl method (11).

Plasmids.

Plasmids pUC118 (31), YEP352 (8), and pVTU101 (30) have been described previously. Plasmid pWZ1-3 is pUC118 containing the GAL6 gene and HIS3 gene fragments cloned into the BamHI site. It was the source of single-stranded DNA for site-directed mutagenesis. The same parental plasmid containing the gal6 DNA-binding mutation (gal6db) was designated pWZ1-3a. Two oligonucleotides, each containing a BamHI site, were used for PCR with the GAL6 or gal6db gene from pWZ1-3 and pWZ1-3a. The PCR fragments were digested by BamHI and cloned into the BamHI site in pVTU102 such that the GAL6 and gal6db genes are controlled by the ADH1 promoter in the plasmid. These two plasmids were named pWZ1-4 and pWZ1-4a, respectively. Plasmid pKM260 was used for protein production in E. coli. It contains a T7 promoter followed by a six-histidine (His6) tag. Downstream of the His6 tag is a TEV protease cleavage sequence followed by an NcoI site (24). Two primers were used for PCR with the GAL6 or mutated gal6 gene. The PCR fragments were digested by NcoI/BamHI and cloned in frame into the NcoI/BamHI site in pKM260 and were designated pWZ1-5 and pWZ1-5a. Plasmid pGAL1/10 was constructed by cloning the fragment that contains four Gal4p binding sites from GAL1 and GAL10 (bp 340 to 540, based on the numbering by Yocum et al. [34]) into the SalI site of pUC118. The same fragment was also used to test whether Gal6p can bind to a Gal4p binding site in a gel mobility shift assay.

Protein purification.

E. coli BL21 was transformed with pWZ1-5 and pWZ1-5a. The resulting strains were grown at 37°C overnight in 2-ml cultures, each containing 25 μg of ampicillin and chloramphenicol per ml. Each culture then was inoculated into 1 liter of L broth with the same antibiotics and grown to an optical density at 600 nm of 0.6 to 1.0. Isopropyl-β-d-thiogalactopyranoside was added to a final concentration of 200 μg/ml. Cells were grown for another 8 h and then harvested. Proteins were purified by using Ni-nitrilotriacetic acid resin according to the standard protocol provided by Qiagen. Each protein from Ni-nitrilotriacetic acid beads was extensively dialyzed against buffer A (25 mM Tris-HCl [pH 8.5], 10% glycerol, 50 mM KCl, 1 mM EDTA, 7 mM 2-mercaptoethanol) and applied to a Bio-Scale Q5 column (Bio-Rad) that had been equilibrated with buffer A. The column was washed with 3 to 4 column volumes of buffer A before the protein was eluted with a 0 to 0.5 M KCl gradient in buffer A. The peak fractions were pooled, dialyzed against buffer A, and concentrated in a Centricon30 (Ambion). The protein was stored in this buffer plus 50% glycerol at −20°C.

Site-directed mutagenesis.

pWZ1-3 was transformed into E. coli TG1, and the single-stranded DNA was prepared by the standard protocol. The GAL6 gene in pWZ1-3 was mutated by using the Amersham Sculptor in vitro mutagenesis system. To make gal6db, the oligonucleotide 5′-AGTGTGGATTGCCGCGTCTGCGTCTACGTATTCCC-3′ was used for changing each of Lys242, Lys244, and Lys245 to alanine. The mutant was identified and confirmed by sequencing.

Gel mobility shift assay.

The single-stranded DNA and RNA oligonucleotides UASL (5′-AGCTTAGCGGAAATTTGTGGTCCGAGC-3′) were end labeled with 32P by using T4 kinase. Unless specified, the assay for each sample contained 1 ng of labeled oligonucleotide, 1 μg of sheared salmon sperm DNA, and 30 ng of purified protein in buffer A. The gel shift procedure has been described elsewhere (3). For competitive gel mobility shift assays, different amounts of competitive DNA were added to the standard reaction mixtures. For nicked DNA preparation, 1 μg of either single-stranded or double-stranded DNA was digested with 2 ng of DNase I for different times. EDTA was added to inactivate the DNase I. The resulting nicked DNAs were subjected to competitive gel mobility shift assays. Five different oligonucleotides were used to test the DNA binding specificity of Gal6p.

Immunolocalization.

Yeast strain W303 was transformed by pPS118, the plasmid that expresses the Gal4(1-72)–LacZ fusion protein. This fusion protein contains the nuclear localization signal of Gal4p and thus serves as a positive control for a nuclear-localized protein (27). Antibody to Gal6 protein was affinity purified by using highly purified Gal6p. This antibody detected only a single band with the predicted size of Gal6p in crude yeast extracts and did not cross-react with any other proteins in the extracts (data not shown). Antibody against LacZ protein was obtained from Cappel. Cells were grown in medium lacking uracil and containing 2% galactose to mid-log phase. Immunolocalization was accomplished as described by Rose et al. (26).

Enzyme assay.

Two substrates were used to measure the peptidase activity of Gal6p and its variant. A synthetic substrate, Arg-AMC (l-arginine 7-amido-4-methylcoumarin; Bachem), was used to determine the enzyme activity as previously described (33). Zeocin (Invitrogen) is a copper-chelated phleomycin D1 (a bleomycin family member), which, like bleomycin, is isolated from Streptomyces verticillus. It was also used in assays as previously described (33) except that 5 mM Zeocin was used. At different times, the reaction was stopped by adding 80 μl of running buffer (100 mM NaH2PO4 [pH 2.3]). The product was then resolved by using a Beckman P/ACE System 2000 as described by McCormick (23). The amount of product was quantitated and plotted.

Bleomycin and cisplatin sensitivity assay.

S. cerevisiae Sc377 was transformed with pWZ1-6, pWZ1-6a, or pVTU102. Bleomycin sensitivity was determined essentially by the method of Lim et al. (19); cisplatin sensitivity was determined by the method of McA’Nulty and Lippard (22).

RESULTS

Characterization of Gal6p DNA binding activity.

Previous studies showed that Gal6p is a DNA-binding protein which can bind to single-stranded DNA with high affinity and to double-stranded DNA with lower affinity. The protein was originally identified by virtue of its binding to a 27-base nucleotide that contains the UAS binding site of the Gal4 regulatory protein in an in vitro binding assay (33). We first tested whether Gal6 protein can bind to the Gal4 binding sites in a more natural context, using a 250-bp DNA fragment that contains four natural Gal4 binding sites. As shown in Fig. 1A, purified Gal4-VP16 protein can bind to this fragment at a concentration of 100 nM. In contrast, the binding of Gal6 protein to this fragment cannot be detected even at a concentration of 10 μM. We also tested the ability of Gal6p to protect the Gal4 sites in this 250-bp fragment by footprinting and found no evidence of protection even at 166 μM (data not shown).

FIG. 1.

FIG. 1

Open in a new tab

Gal6p binds to single-stranded DNA. (A) A gel mobility shift assay shows that Gal6p does not bind to double-stranded GAL4 binding sites. The DNA used for each reaction was 20 ng of a 200-bp PCR fragment from the GAL1-GAL10 promoter which contains four Gal4p binding sites. Lane 1, free DNA; lane 2, DNA plus 100 nM purified GAL4-VP16 protein; lanes 3 to 5, DNA plus 100 nM, 1 μM, and 10 μM, respectively, Gal6p. (B) Gal6p binds to single-stranded oligonucleotides without apparent sequence specificity. A gel mobility shift assay shows that Gal6p can bind to various single-stranded oligonucleotides. For each reaction, 30 ng of purified recombinant Gal6p along with 1 ng of single-stranded oligonucleotide was used. Note that lane 3 is with the LexA oligonucleotide. Arrowheads 1 and 2 indicate the protein-DNA complex and free DNA, respectively. Sequences of the oligonucleotides used: lane 1, 5′-GGCAAACAACCAAGCTCTACCAGAGCT-3′; lane 2, 5′-CCTTTTTCTGTTTTATGAGCTATTT-3′; lane 3, 5′-TCGAGTACTGTATGTACATACAGTAC-3′ (LexA oligonucleotide); lane 4, 5′-CGGGATCCAGAGCTGCTGAAACTATTTA-3′; lane 5, 5′-AGCTTAGCGGAAATTTGTGGTCCGAGC-3′.

A previous study had shown that a single-stranded oligonucleotide with the LexA binding site (26 bp) had lower affinity for Gal6p than for a single-stranded oligonucleotide (27 bp) from a Gal4 binding site (33). To further explore whether Gal6p binds single-stranded DNA with selectivity, we tested five unrelated oligonucleotides, including the LexA oligonucleotide. We found that Gal6p binds equally well to four of the oligonucleotides. Only the LexA oligonucleotide had significantly lower affinity. We surmise that the LexA oligonucleotide self-anneals to form enough double-stranded DNA to decrease the apparent Gal6p binding (Fig. 1B). Therefore, we conclude that Gal6p does not bind to double-stranded Gal4p binding sites, nor does it bind single-stranded oligonucleotides with selectivity.

Gal6p binds to single-stranded DNA and RNA.

The structure of Gal6p reveals no obvious DNA binding motif (12). The six subunits form a tunnel with a diameter of 20 Å. Though the net charge of the protein is neutral, there are 60 lysines located in the inside wall of the tunnel. At each end of the tunnel, each subunit has a β-hairpin structure, with three lysines (Lys242, Lys244, and Lys245) in each. Thus, nine lysines form a positively charged ring at both ends of the tunnel (marked in black in Fig. 2A). If the protein binds to DNA through the tunnel, the charged residues at each end of the tunnel may act as DNA-contacting residues. We tested this by changing each of these lysines to alanine to produce gal6db. This variant is stable in vivo and has wild-type peptidase activity (see Fig. 5C). As shown in Fig. 2B, this variant is severely compromised in its ability to bind single-stranded DNA. We estimate that the mutant protein gal6db has approximately a 1,000-fold decrease in affinity for DNA, i.e., from 10−8 to 10−5 M (Fig. 2B) (33). This finding is consistent with our observation for crude extracts (unpublished data).

FIG. 2.

FIG. 2

Open in a new tab

Lys242, Lys244, and Lys245 are important for DNA and RNA binding. (A) The structure of Gal6p as seen looking down the central tunnel. Black indicates Lys242, Lys244, and Lys245 (three lysines in each turn), which are in the portal of the tunnel. The structure is derived from reference 12. (B) Changing Lys242, Lys244, and Lys245 to alanines disrupts the DNA binding activity of Gal6p. A gel mobility shift assay shows that the mutant protein does not bind to single-stranded DNA. In lanes 1 to 4 and 5 to 8, 30, 100, 300, and 1,000 ng of purified protein were used with 1 ng of 32P-labeled UASL; lane 9 is DNA alone. The small amounts of protein-DNA complexes present in the gal6db binding reactions (lanes 3 and 4) have higher mobility than that of wild type because gal6db protein has 18 fewer positive charges. (C) A gel mobility shift assay shows that Gal6p is an RNA-binding protein. The DNA-binding mutant does not bind to RNA. In lanes 1 to 4 and 5 to 8, 30, 100, 300, and 1,000 ng of purified protein were used; lane 9 is RNA alone. In each reaction, 1 ng of an RNA corresponding to the same sequence as the UASL DNA oligonucleotide was used. Arrowheads 1 and 2 indicate protein-nucleic acid complexes and free oligonucleotides, respectively.

FIG. 5.

FIG. 5

Open in a new tab

The nucleic acid binding activity of Gal6p is involved in bleomycin detoxification. (A) The nucleic acid-binding mutant (gal6db) strain is more sensitive to Zeocin than the wild-type GAL6 strain but less sensitive than the gal6 deletion strain. □, GAL6; ◊, gal6db; ▵, Δgal6. (B) The gal6db strain has the same sensitivity to cisplatin as the wild-type GAL6 strain and the deletion strain. ○, GAL6; ◊, gal6db; ▵, Δgal6. (C) A peptidase assay shows that extracts from the wild-type and nucleic acid-binding mutant strains have the same peptidase specific activity on the Arg-AMC substrate, and a Western blot shows that the wild-type and mutant strains express the same amount of protein. (D) Purified wild-type Gal6p (○) and gal6db (□) have the same enzyme activity toward Zeocin. For all assays, 5 mM Zeocin was incubated with 0.1 μM Gal6p or gal6dbp in a 20-μl reaction. The amount of conversion of Zeocin to products was determined by capillary electrophoresis.

To investigate whether Gal6p can also bind RNA, we used an RNA oligonucleotide with the same sequence as the DNA oligonucleotide used above (UASL) in a gel mobility shift assay. As shown in Fig. 2C, Gal6p also binds this RNA. We estimate the affinity of the wild-type protein for RNA is 10 nM, which is in the same range as that for single-stranded DNA. To characterize whether Gal6p uses the same motif for DNA and RNA binding, we tested the ability of the DNA-binding-defective protein to bind to RNA. This protein binds single-stranded DNA and RNA with comparable low affinities (Fig. 2C). These results indicate that Gal6 protein requires the same amino acids for both DNA and RNA binding and that these include the lysines at the portal of the tunnel.

Gal6p binds to the ends of nucleic acid.

The finding that the portal lysines are important for nucleic acid binding suggests two mechanisms for binding. One model is that Gal6p binds to a DNA fragment as the DNA end penetrates into the tunnel, with the positive charges in the β-hairpins playing an important role in stabilizing the protein-DNA complex (Fig. 3A, top). Alternatively, the protein may sit on the DNA through charge-charge interaction, and the three β-hairpins at the opening of the tunnel may act as a clamp to bind single-stranded DNA. In the latter case, the prediction follows that Gal6p may bind to single-stranded DNA other than through its ends (Fig. 3A, bottom).

FIG. 3.

FIG. 3

Open in a new tab

Gal6p binds to single-stranded DNA with ends and nicked DNA. (A) Two possible models of how Gal6p may bind to single-stranded DNA. Top, the protein binds through the ends of single-stranded DNA entering the tunnel; bottom, the three β-hairpins may bind DNA independent of the ends. (B) A single-stranded circular DNA does not compete with single-stranded oligonucleotide (oligo) for Gal6p binding. The amounts of Gal6p and DNA used are 30 ng (10 nM) and 1 ng (UASL), respectively. The amounts of circular single-stranded DNA (pUC118 construct, 3.5 kb) used in lanes 1 to 5 are 0, 1, 10, 100, and 1,000 ng. (C) A circular single-stranded DNA treated with DNase I can effectively compete with single-stranded oligonucleotide for Gal6p binding. One microgram of circular, single-stranded DNA was digested with 2 ng of DNase I for 0 (lane 1), 6 (lane 2), 12 (lane 3), 18 (lane 4), and 24 (lane 5) min. The resulting DNA samples were used in a competitive gel mobility shift assay as for panel B. (D) Circular double-stranded DNA (pUC118) does not compete with single-stranded oligonucleotide for Gal6p binding. The amounts of double-stranded DNA used for the assay are 0 (lane 1), 1 (lane 2), 10 (lane 3), 100 (lane 3), and 1,000 (lane 4) ng. (E) Double-stranded DNA treated with DNase I can compete effectively with single-stranded oligonucleotide for Gal6p binding. The double-stranded DNA was treated with 2 ng of DNase I for 0 (lane 1), 5 (lane 2), 10 (lane 3), 20 (lane 4), 30 (lane 5), and 40 (lane 6) min.

To help distinguish between these two models, we examined whether circular, single-stranded DNA is effectively bound by Gal6p. Different amounts of single-stranded, circular DNA were used to compete UASL (single-stranded DNA oligonucleotide) for Gal6p binding. As shown in Fig. 3B, circular, single-stranded DNA cannot compete with single-stranded UASL for Gal6p binding even at a 1,000-fold mass excess. This result suggests that circular, single-stranded DNA is not an efficient substrate for Gal6p binding. However, this DNA becomes a strong competitor for Gal6p binding after it is digested by DNase I. The binding activity is proportional to the concentration of ends introduced by DNase I (Fig. 3C). This finding indicates that the Gal6p protein binds to the ends of single-stranded DNA.

Although Gal6p binds single-stranded DNA with high affinity, single-stranded DNA ends are not abundant in living cells. One possible source of single-stranded DNA in the cell is nicked DNA. Since the peptide substrate of Gal6p, bleomycin, is a DNA-cleaving agent, it is possible that Gal6p binds to nicked templates that arise from bleomycin cleavage. We tested whether Gal6p binds to nicks in double-stranded DNA by treating pUC118 double-stranded DNA with DNase I to generate nicked substrates, and the ability of this nicked DNA to compete the single-stranded UASL was determined. As shown in Fig. 3D, untreated double-stranded DNA does not compete with the single-stranded oligonucleotide for Gal6p binding. However, the double-stranded DNA treated with DNase I is a good competitor (Fig. 3E). More extensive treatment with DNase I increased competition for Gal6p binding. This result suggests that Gal6p can bind to nicked double-stranded DNA.

Immunolocalization.

As shown above, Gal6p can bind to single-stranded DNA and RNA with high affinity in vitro. To help resolve which nucleic acid (if any) is the natural substrate in vivo, we immunolocalized Gal6p. In yeast, many proteolytic enzymes are located in vacuoles, functional counterparts of lysosomes in mammalian cells. If Gal6p is vacuolarly localized, it is likely that the nucleic acid binding activity of Gal6p is an artifact. On the other hand, nuclear localization would suggest DNA, and cytoplasmic localization would suggest RNA, as the substrate. Yeast cells with or without the GAL6 gene were grown to mid-log phase, and affinity purified anti-Gal6p antibody was used to determine the subcellular location of the Gal6 protein by immunolocalization. As shown in Fig. 4A and D, Gal6p is predominantly located in the cytoplasm and appears to be excluded from the nucleus. Though not apparent in Fig. 4, staining for Gal6p was absent in the vacuoles (data not shown). A nucleus-localized β-galactosidase protein was used as a positive control for nuclear localization (Fig. 4B and E). Only background staining is evident in a strain with GAL6 deleted (Fig. 4C and F). Treatment of the cells with bleomycin did not alter the localization of the Gal6 protein, nor did the localization change with cell cycle (data not shown). Although the possibility that some portion of Gal6p is located in the nucleus cannot be excluded, this result shows that a majority of Gal6 protein is in the cytoplasm, suggesting that RNA may be its predominant target if it does bind nucleic acid in vivo.

FIG. 4.

FIG. 4

Open in a new tab

Immunolocalization shows that most of the Gal6 protein is located in the cytoplasm. (A) Localization of Gal6p. Immunopurified antibody against Gal6p was used as described in Materials and Methods. Note the lack of staining in nuclei. (B) Nuclear localized β-galactosidase. Antibody to β-galactosidase was used, with cells expressing a GAL4–β-galactosidase fusion protein treated in the same manner as for panel A. (C) Δgal6 control. A strain with GAL6 deleted was treated as for panel A and probed with Gal6 antibody. (D to F) 4′,6-Diamidino-2-phenylindole staining of nuclei in the same cells of the corresponding panel.

The nucleic acid binding activity of Gal6p is involved in bleomycin detoxification.

To determine whether the observed nucleic acid binding activity of Gal6p is biologically active, we tested the bleomycin sensitivity of strains that express either wild-type Gal6 protein or the variant Gal6 protein that lacks nucleic acid binding activity. As shown in Fig. 5A, a strain that expresses the nucleic acid-binding-defective protein is more sensitive to bleomycin than a congenic strain that expresses wild-type protein. However, the binding-defective variant GAL6 strain displays slightly more bleomycin resistance than a strain with GAL6 deleted, presumably because the binding-defective Gal6p retains normal peptidase activity. These three strains have similar sensitivities to another DNA-damaging agent, cisplatin (Fig. 5B), implying that the effect of DNA binding is specifically related to bleomycin detoxification. Relative to this point, Fig. 5C (top) indicates that crude extracts of the two strains have essentially the same peptidase activity toward the synthetic substrate Arg-AMC and that the proteins are produced at the same levels in yeast (bottom). As shown in Fig. 5D, the purified wild-type and mutant proteins also have the same activity toward bleomycin (Zeocin), demonstrating that the alterations affecting nucleic acid binding did not decrease the intrinsic bleomycin detoxification activity in vitro. These data indicate that the nucleic acid binding activity is involved in, but not essential for, bleomycin detoxification and is probably specific for bleomycin. These data are also consistent with the observation that Gal6 protein can hydrolyze bleomycin in the absence of nucleic acid (33).

DISCUSSION

In this study, we have characterized the nucleic acid binding activity of Gal6p, showing that Gal6p can bind to RNA, single-stranded DNA, and nicked DNA. Immunolocalization of Gal6p demonstrates that Gal6 protein is predominantly located in the cytoplasm. Finally, the nucleic acid binding activity of Gal6p was shown to affect the ability of yeast to detoxify bleomycin. Gal6p was originally isolated as a DNA-binding protein and characterized as having strong single-stranded and weak double-stranded DNA affinity (33). We now show that Gal6p also binds to nicked double-stranded DNA and RNA with comparable affinities. We find that Gal6p can bind with equal affinity to a variety of oligonucleotides of random sequence, indicating it has no specific sequence requirements.

The structure of Gal6p is a hollow cylinder with a positively charged channel (12). This structure resembles those of other well-characterized DNA-binding proteins (for a review, see reference 15) and peptidases (for a review, see reference 18). The diameter of the channel at the entrance in Gal6p is about 20 Å, which is enough to accommodate single-stranded or double-stranded DNA, as shown from modeling (11a). Our results show that Gal6p binding requires the ends of single-stranded DNA, supporting the model that DNA ends may insert into the channel. The requirement for the lysines at the entrance of the channel suggests that charge-charge interactions are important for this binding. We have shown that Gal6p also binds RNA. Immunolocalization data reveal that most of the Gal6 protein is located in the cytoplasm, supporting the conclusion that RNA may be the predominant binding substrate of Gal6p in the cell. Since the DNA-binding mutant cannot bind to RNA, it is very likely that Gal6p binds DNA and RNA in the same way. The cytoplasmic localization of Gal6p may also reflect its possible role as a peptidase involved in the turnover of amino acids.

Based on the Gal6 protein structure, the channel is the only obvious access to the peptidase active sites of Gal6p (1). As the only known natural substrate of Gal6p is bleomycin, a nucleic acid-binding and -cleaving agent, it is possible that Gal6p binds to the nucleic acid and cleaves bleomycin bound to it. This proposal is supported by our data showing that the nucleic acid-binding mutant is more sensitive to bleomycin than wild-type cells. We have also observed that single-stranded DNA can stimulate the hydrolysis of a substrate, Arg-AMC, by Gal6p twofold (data not shown). However, the relatively small size of the substrate and the potential interactions between the positively charged arginine and DNA may influence the interpretation of this result. It has been shown that bleomycin can cleave double-stranded regions of tRNA (9), mRNA (4, 5), and rRNA (10). As these RNAs are abundant in the cytoplasm, it seems likely that RNA molecules in the cytoplasm would bind much of the bleomycin as it first enters the cell. Most of the detoxification of bleomycin may therefore take place in the cytoplasm, though the toxicity itself may arise primarily from damage to DNA. However, other RNA-damaging and -modifying agents such as ricin (6), aminoglycoside antibiotics (32), and onconase (20) are strong growth inhibitors. Thus, it is possible that the RNA cleaving activity of bleomycin plays a role in its cytotoxicity and Gal6p protects RNA.

We have provided evidence that RNA may be a physiological binding substrate for Gal6p. However, it is still possible that a small amount of Gal6p is present in the nucleus, where it is able to cleave bleomycin bound to DNA. The affinity of Gal6p for single-stranded DNA ends is 10 nM, and we estimate that there are 18,000 to 67,000 Gal6 monomers per cell, depending on growth conditions (unpublished data). Therefore, it is reasonable to expect that a small amount of Gal6p can efficiently bind nicks in DNA caused by bleomycin. Studies from other groups have shown that strains defective in double-stranded DNA break repair are more sensitive to bleomycin than wild-type cells whereas nuclear excision repair-defective cells have the same bleomycin sensitivity as wild-type cells (1). These data indicate that double-stranded DNA breaks caused by bleomycin are much more detrimental. It is possible that when bleomycin binds to DNA and cleaves the first strand, the single-stranded DNA ends are exposed and recruit bleomycin hydrolase to provide a high local concentration of detoxification activity to block the occurrence of the second break. In this way, bleomycin hydrolase could prevent lethal double-stranded DNA breaks.

We have shown that the mutations in GAL6 which inactivate nucleic acid binding also make yeast cells more sensitive to bleomycin but have no significant effect on its cisplatin sensitivity. While this correlation argues that the nucleic acid binding plays a role in bleomycin detoxification, it is also possible that this binding is merely correlative. For example, the positive charges of Gal6p may be important for binding some other targets such as acidic peptides. Arguing that the nucleic acid binding is not an artifact is the recent finding that the less positively charged rat bleomycin hydrolase also binds DNA (29).

By sequence comparison, bleomycin hydrolase is highly conserved from bacteria to mammals (3, 12). It would not be a surprise to find that all bleomycin hydrolases have the same mechanism to detoxify bleomycin. Supporting this idea, we have found that the bacterial homolog PepC, like the mammalian homolog (29), is a single-stranded DNA-binding protein (unpublished data). Further investigation of whether other mammalian Gal6-like proteins are nucleic acid-binding proteins and whether the nucleic acid binding activity is important for bleomycin detoxification may provide useful information for bleomycin chemotherapy. Also remaining to be determined are the normal cellular function of Gal6p and its nucleic acid binding activity. Regardless, these studies provide the first evidence that the nucleic acid binding activity of Gal6p functions in vivo.

ACKNOWLEDGMENTS

We thank Leemor Joshua-Tor and the Johnston lab for helpful discussions.

This work was supported by grants from NIH (CA67982) and the Council for Tobacco Research (4247R1) to S.A.J. and a molecular cardiology training fellowship to W.Z.

REFERENCES

  • 1.Abe H, Wada M, Kohno K, Kuwano M. Altered drug sensitivities to anticancer agents in radiation-sensitive DNA repair deficient yeast mutants. Anticancer Res. 1994;14:1807–1810. [PubMed] [Google Scholar]
  • 2.Antson A A, Otridge J, Brzozowski A M, Dodson E J, Dodson G G, Wilson K S, Smith T M, Yang M, Kurecki T, Gollnick P. The structure of trp RNA-binding attenuation protein. Nature. 1995;374:693–700. doi: 10.1038/374693a0. [DOI] [PubMed] [Google Scholar]
  • 3.Berti P J, Storer A C. Alignment/phylogeny of the papain superfamily of cysteine proteases. J Mol Biol. 1995;246:273–283. doi: 10.1006/jmbi.1994.0083. [DOI] [PubMed] [Google Scholar]
  • 4.Carter B J, de Vroom E, Long E C, van der Marel G A, van Boom J H, Hecht S M. Site-specific cleavage of RNA by Fe(II) · bleomycin. Proc Natl Acad Sci USA. 1990;87:9373–9377. doi: 10.1073/pnas.87.23.9373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Dix D J, Lin P N, McKenzie A R, Walden W E, Theil E C. The influence of the base-paired flanking region on structure and function of the ferritin mRNA iron regulatory element. J Mol Biol. 1993;231:230–240. doi: 10.1006/jmbi.1993.1278. [DOI] [PubMed] [Google Scholar]
  • 6.Endo Y, Mitsui K, Motizuki M, Tsurugi K. The mechanism of action of ricin and related toxic lectins on eukaryotic ribosomes. The site and the characteristics of the modification in 28 S ribosomal RNA caused by the toxins. J Biol Chem. 1987;262:5908–5912. [PubMed] [Google Scholar]
  • 7.Enenkel C, Wolf D H. BLH1 codes for a yeast thiol aminopeptidase, the equivalent of mammalian bleomycin hydrolase. J Biol Chem. 1993;268:7036–7043. [PubMed] [Google Scholar]
  • 8.Hill J E, Myers A M, Koerner T J, Tzagoloff A. Yeast/E. coli shuttle vectors with multiple unique restriction sites. Yeast. 1986;2:163–167. doi: 10.1002/yea.320020304. [DOI] [PubMed] [Google Scholar]
  • 9.Holmes C, Abraham A, Hecht S, Florentz C, Giege R. Fe-bleomycin as a probe of RNA conformation. Nucleic Acids Res. 1996;24:3399–3406. doi: 10.1093/nar/24.17.3399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Holmes C E, Carter B J, Hecht S M. Characterization of iron (II) · bleomycin-mediated RNA strand scission. Biochemistry. 1993;32:4293–4307. doi: 10.1021/bi00067a019. [DOI] [PubMed] [Google Scholar]
  • 11.Ito H, Fukuda Y, Murata K, Kimura A. Transformation of intact yeast cells treated with alkali cations. J Bacteriol. 1983;153:163–168. doi: 10.1128/jb.153.1.163-168.1983. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11a.Joshua-Tor, L. Personal communication.
  • 12.Joshua-Tor L, Xu H E, Johnston S A, Rees D C. Crystal structure of a conserved protease that binds DNA: the bleomycin hydrolase, Gal6. Science. 1995;269:945–950. doi: 10.1126/science.7638617. [DOI] [PubMed] [Google Scholar]
  • 13.Kambouris N G, Burke D J, Creutz C E. Cloning and characterization of a cysteine proteinase from Saccharomyces cerevisiae. J Biol Chem. 1992;267:21570–21576. [PubMed] [Google Scholar]
  • 14.Kane S A, Hecht S M. Polynucleotide recognition and degradation by bleomycin. Prog Nucleic Acid Res Mol Biol. 1994;49:313–352. doi: 10.1016/s0079-6603(08)60054-9. [DOI] [PubMed] [Google Scholar]
  • 15.Kelman Z, Finkelstein J, O’Donnell M. Protein structure. Why have six-fold symmetry? Curr Biol. 1995;5:1239–1242. doi: 10.1016/s0960-9822(95)00247-8. [DOI] [PubMed] [Google Scholar]
  • 16.Kong X P, Onrust R, O’Donnell M, Kuriyan J. Three-dimensional structure of the beta subunit of E. coli DNA polymerase III holoenzyme: a sliding DNA clamp. Cell. 1992;69:425–437. doi: 10.1016/0092-8674(92)90445-i. [DOI] [PubMed] [Google Scholar]
  • 17.Krishna T S, Kong X P, Gary S, Burgers P M, Kuriyan J. Crystal structure of the eukaryotic DNA polymerase processivity factor PCNA. Cell. 1994;79:1233–1243. doi: 10.1016/0092-8674(94)90014-0. [DOI] [PubMed] [Google Scholar]
  • 18.Larsen C N, Finley D. Protein translocation channels in the proteasome and other proteases. Cell. 1997;91:431–434. doi: 10.1016/s0092-8674(00)80427-4. [DOI] [PubMed] [Google Scholar]
  • 19.Lim S T, Jue C K, Moore C W, Lipke P N. Oxidative cell wall damage mediated by bleomycin-Fe(II) in Saccharomyces cerevisiae. J Bacteriol. 1995;177:3534–3539. doi: 10.1128/jb.177.12.3534-3539.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Lin J J, Newton D L, Mikulski S M, Kung H F, Youle R J, Rybak S M. Characterization of the mechanism of cellular and cell free protein synthesis inhibition by an anti-tumor ribonuclease. Biochem Biophys Res Commun. 1994;204:156–162. doi: 10.1006/bbrc.1994.2439. [DOI] [PubMed] [Google Scholar]
  • 21.Lupas A, Flanagan J M, Tamura T, Baumeister W. Self-compartmentalizing proteases. Trends Biochem Sci. 1997;22:399–404. doi: 10.1016/s0968-0004(97)01117-1. [DOI] [PubMed] [Google Scholar]
  • 22.McA’Nulty M M, Lippard S J. The HMG-domain protein Ixr1 blocks excision repair of cisplatin-DNA adducts in yeast. Mutat Res. 1996;362:75–86. doi: 10.1016/0921-8777(95)00037-2. [DOI] [PubMed] [Google Scholar]
  • 23.McCormick R M. Capillary zone electrophoretic separation of peptides and proteins using low pH buffers in modified silica capillaries. Anal Chem. 1988;60:2322–2328. doi: 10.1021/ac00172a003. [DOI] [PubMed] [Google Scholar]
  • 24.Melcher K A, Johnston S A. Gal4 interacts with TATA-binding protein and coactivators. Mol Cell Biol. 1995;15:2839–2848. doi: 10.1128/mcb.15.5.2839. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Pei Z, Calmels T P, Creutz C E, Sebti S M. Yeast cysteine proteinase gene ycp1 induces resistance to bleomycin in mammalian cells. Mol Pharmacol. 1995;48:676–681. [PubMed] [Google Scholar]
  • 26.Rose M D, Winston F, Hieter P. Methods in yeast genetics: a laboratory course manual. Cold Spring Harbor, N.Y: Cold Spring Harbor Laboratory Press; 1990. [Google Scholar]
  • 27.Silver P A, Chiang A, Sadler I. Mutations that alter both localization and production of a yeast nuclear protein. Genes Dev. 1988;2:707–717. doi: 10.1101/gad.2.6.707. [DOI] [PubMed] [Google Scholar]
  • 28.Stubbe J, Kozarich J W. Mechanisms of bleomycin-induced DNA degradation. Chem Rev. 1987;87:1107–1136. [Google Scholar]
  • 29.Takeda A, Masuda Y, Yamamoto T, Hirabayashi T, Nakamura Y, Nakaya K. Cloning and analysis of cDNA encoding rat bleomycin hydrolase, a DNA-binding cysteine protease. J Biochem (Tokyo) 1996;120:353–359. doi: 10.1093/oxfordjournals.jbchem.a021420. [DOI] [PubMed] [Google Scholar]
  • 30.Vernet T, Dignard D, Thomas D Y. A family of yeast expression vectors containing the phage f1 intergenic region. Gene. 1987;52:225–233. doi: 10.1016/0378-1119(87)90049-7. [DOI] [PubMed] [Google Scholar]
  • 31.Vieira J, Messing J. Production of single-stranded plasmid DNA. Methods Enzymol. 1987;153:3–11. doi: 10.1016/0076-6879(87)53044-0. [DOI] [PubMed] [Google Scholar]
  • 32.Wang Y, Killian J, Hamasaki K, Rando R R. RNA molecules that specifically and stoichiometrically bind aminoglycoside antibiotics with high affinities. Biochemistry. 1996;35:12338–12346. doi: 10.1021/bi960878w. [DOI] [PubMed] [Google Scholar]
  • 33.Xu H E, Johnston S A. Yeast bleomycin hydrolase is a DNA-binding cysteine protease. Identification, purification, biochemical characterization. J Biol Chem. 1994;269:21177–21183. [PubMed] [Google Scholar]
  • 34.Yocum R R, Hanley S, West R, Jr, Ptashne M. Use of lacZ fusions to delimit regulatory elements of the inducible divergent GAL1-GAL10 promoter in Saccharomyces cerevisiae. Mol Cell Biol. 1984;4:1985–1998. doi: 10.1128/mcb.4.10.1985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Zheng W, Xu, Eric H, Johnston S A. The cysteine-protease bleomycin hydrolase is a new component of the galactose regulon in yeast. J Biol Chem. 1997;272:30350–30355. doi: 10.1074/jbc.272.48.30350. [DOI] [PubMed] [Google Scholar]

Articles from Molecular and Cellular Biology are provided here courtesy of Taylor & Francis

RESOURCES