Tyrosinase Mutants Are Capable of Prodrug Activation in Transfected Nonmelanotic Cells¹

The concept of GDEPT³(1) is based on the conversion of a nontoxic prodrug by a product of foreign gene expression and is actively pursued in cancer gene therapy (2, 3, 4, 5, 6). Examples in which such paradigms have been successfully used include treatments of malignant mesothelioma(7), melanoma (8), disseminated liver metastasis from colon cancer (9), glioma(10), and sarcomas (11).

Tyrosinase has been investigated as a potential prodrug-converting enzyme for some time (12, 13) and has been used in targeted melanoma therapy (14). The existing obstacles limiting the efficacy of tyrosinase-mediated GDEPT include:(a) low expression of whole tyrosinase transgenes in nonmelanotic cancer cells; (b) low enzyme activity of full-length tyrosinase in transfected cells; and (c)relatively low toxicity of available prodrug metabolites (6, 15). The latter in particular has been a limiting factor in extending tyrosinase-mediated GDEPT beyond experimental melanotic melanoma treatment (16, 17).

Tyrosinase catalyzes two major reactions that occur primarily in lysosomes (or premelanosomes): (a) the hydroxylation of l-tyrosine with the formation of l-DOPA(3,4-dehydroxy-l-phenylalanine); and(b) the subsequent oxidation of DOPA with the formation of DOPA quinone (18). Further metabolic transformation of the DOPA quinone, either autocatalytic or catalyzed by TRPs (TRP-1, TRP-2), leads to melanin formation (19, 20). The quinone-generating activity of tyrosinase explains its cytotoxicity and antiproliferative activity. Tyrosinase-generated quinones may cause substantial cytotoxicity because of the ability to bind covalently with free SH groups and to alkylate DNA. The former effect interferes with normal cell metabolism by depleting cells of the reduced glutathione (21), and the latter leads to inhibition of cell proliferation (22).

The wild-type tyrosinases and TRPs are sorted to melanosomes because of the specific COOH-terminally located sorting (cytoplasmic) amino acid signals (23, 24, 25). We hypothesized that in the absence of the above sorting signals, tyrosinase would accumulate in various subcellular compartments and could be used to modulate toxicity to cells because tyrosinase-generated quinones will not be sequestered exclusively in lysosomes/premelanosomes. In addition, by introducing deletions in noncatalytic domains of tyrosinase, we expected to alter the processing and retention of precursor protein in the endoplasmic reticulum (17, 26) and, thereby, to increase expression of tyrosinase in model transfected cells. Therefore, we engineered several COOH-terminally truncated tyrosinase mutants and compared their expression efficiency, enzyme activity, and prodrug activation in several transformed (both nontumorigenic and tumorigenic) cell lines.

Our results suggest that tyrosinase mutants are indeed enzymatically active and can convert certain model prodrugs into cytotoxic agents with resultant cytotoxic and antiproliferative effects in several transfected cell lines, including gliosarcoma, adenocarcinoma, and fibrosarcoma cells.

The cDNA clone encoding for human tyrosinase (EC 1.14.18.1; ATCC) was used as a template for PCR in the presence of a forward primer encoding a COOH-terminal 5′-sequence (27). Four variants of the tyrosinase coding sequence were prepared (Fig. 1): (a) a full-length human tyrosinase clone(T) contained the entire 1.6-kb human tyrosinase full-length sequence; (b) a mutant lacking the COOH-terminal transmembrane and cytoplasmic domains (TΔTC)—a secretable tyrosinase mutant—carried a deletion of the 3′-terminus(1396–1571 bp) corresponding to the 457–512-AA COOH-terminal transmembrane-cytoplasmic domains (23); and (c)a mutant lacking the COOH-terminal cytoplasmic domain(TΔC)—the cell surface-anchored tyrosinase(28)—contained an A→T substitution that resulted in a replacement of a lysine codon at position 505 to a termination (TAG)codon downstream of the human tyrosinase transmembrane domain(29). The fourth construct—a tyrosinase-Lamp (TL)mutant—was constructed by fusing human tyrosinase truncated at Gln(506) position (six AAs downstream of transmembrane domain) to eight COOH-terminal AA (SHAGYQTI) of lysosome-associated membrane glycoprotein-1 (Lamp-1; Refs. 30, 31). All four of the constructs contained identical 5′-coding sequences and were cloned into the pcDNA3 eukaryotic expression vector (Invitrogen, Carlsbad, CA)under control of a CMV promoter/enhancer.

The 1.6-kb EcoRI fragment containing human tyrosinase cDNA sequence was isolated from the pmel 34 phagemid clone (ATCC, Rockville,MD). This fragment was amplified using 2.5 units of Pwo DNA polymerase, 30 pmol each primer, and 300 μmeach dNTP (DNA amplification kit; Boehringer-Mannheim, Indianapolis,IN). The primary structure of selected clones was confirmed by DNA sequencing. The 5′-sense (forward) primer used was:5′-ATCGAATTCGCCGCCACCATGGCCCTGGCTGTTTTGTACTG-3′. An EcoRI restriction site (underlined) was introduced upstream of the initiation codon (bold face) surrounded with a Kozak sequence (32). The antisense primer (reverse) was:5′-GCAGCCACTCCTCATGGAGAAAGAGGATTACCACAGCTTGTATC AGAGCCATTTATAA-3′. The stop codon is shown in boldface. The PCR product was digested with EcoRI restriction endonuclease and inserted into EcoRI-EcoRV sites of the pcDNA3 vector (Invitrogen, Carlsbad, CA).

The tyrosinase construct lacking the COOH-terminal transmembrane-cytoplasmic domain (TΔTC) was obtained by amplification of the same cDNA fragment using the antisense primer 5′GAACAAGCGAGTCGGATCCGGTCATGGCTC-3′ (BamHI restriction site is shown in boldface). The TΔTC sequence was inserted into the EcoRI-BamHI sites of the pBluescript II KS(+) vector (Stratagene, La Jolla, CA) to generate a TAG stop codon downstream from the truncated tyrosinase cDNA, and was then excised and inserted into the EcoRI-NotI sites of the pcDNA3 vector.

The tyrosinase mutant lacking only the COOH-terminal cytoplasmic domain(TΔC) was obtained by amplification of the full-length tyrosinase T using the reverse primer:5′-GCTTGCTGTGTCGTCACAAGAGATAGCAG-3′ (stop codon is in boldface). The PCR product was digested with EcoRI restriction endonuclease, and the insert was cloned into the EcoRI-EcoRV sites of the pcDNA3 vector as described above.

The tyrosinase fusion protein (TL, a fusion of tyrosinase with eight COOH-terminal cytoplasmic amino acids of Lamp-1) was obtained by PCR amplification of the full-length tyrosinase construct (T). This fusion was designed to improve tyrosinase sorting to lysosomes and to avoid retention of recombinant tyrosinase in Golgi or endoplasmic reticulum(33). The used reverse primer was:5′-GCTTGCTGTGTCGTCACAAGAGAAAGCAGAGTCACGCAGG CTACCAGACTATCTAGGC-3′ (stop codon in boldface). The EcoRI-digested PCR fragment was cloned into the EcoRI-EcoRV sites of the pcDNA3 vector. The integrity of cDNA was confirmed in all of the four clones by DNA sequencing.

Total RNA was extracted using the RNA STAT-60 according to the protocol provided by the manufacturer (Tel-Test ’B’, Friendswood, TX), with several modifications. Briefly, cells were lysed, and RNA was extracted twice. First-strand cDNA was synthesized as follows: 1 μg of total RNA was incubated in 1× first-strand buffer with 0.5 μg oligo(dT)15,0.5 mm dNTPs (Pharmacia Biotech Inc., Piscataway,NJ), and 200 units SUPERSCRIPT II RNase H-Reverse Transcriptase (Life Technologies, Inc., Gaithersburg, MD) for 90 min at 42°C. The reaction was terminated by heating at 95°C for 10 min. Prior to PCR amplification, the cDNA stock was diluted 7-fold. A 2-μl aliquot of the cDNA dilution was used in each experiment. The final volume of each reaction was 50 μl, containing 1× PCR buffer, 1.5 mmMgCl₂, 0.2 mm dNTPs, 0.4μ m each primer, and 2 units of Taq DNA polymerase (Life Technologies, Inc.). The primers for these cDNA fragments have been described previously (19). The chosen primers span at least one exon in the tyrosinase gene, i.e., the differentiation between potential genomic DNA contamination and generated cDNA was made possible. After amplification, aliquots of the PCR products were loaded onto 2% (w/v) agarose gels and were resolved by electrophoreses and photographed.

Human embryo kidney cells (293) and 9L gliosarcoma cells (Brain Tumor Research Center, University of California, San Francisco, CA; Ref.34) were grown in 10% FBS and DMEM, (Cellgro; Mediatech,Washington, DC). Human MCF-7 adenocarcinoma was grown in 10% FBS and MEM, (Cellgro, Mediatech, Washington, DC), and MCF10A adenoma was grown in Mammary Epithelial Cell Growth Medium 5% FBS,(Clonetics/BioWhittaker, Walkersville, MD). Human fibroblasts (NHDF-Ad)were grown in 10% FBS and Fibroblast Basal Medium(Clonetics/BioWhittaker, Walkersville, MD) and fibrosarcoma HT-1080 cells were cultured in 10% FBS RPMI medium (BioWhittaker,Walkersville, MD). The murine B16F1 melanotic melanoma cell line was grown in 10% FBS and DMEM/Ham’s F10.

Cells (2 × 10⁶/10-cm dish) were transfected with tyrosinase constructs using TransIT-100 reagent(PanVera Corporation, Madison, WI), as suggested by the manufacturer with modifications. Briefly, cells were plated 24 h before transfection. TransIT-100 (32.4 μl) was added dropwise to 810 μl of DMEM and was incubated for 5 min. Plasmid DNA (17 μg, 0.5 mg/ml) was added to the mixture and incubated for another 5 min. Complete medium was removed, and cells were washed with serum-free DMEM. Fresh serum-free DMEM was added, and cells were incubated with plasmid-TransIT-100 complex for 4 h. After incubation, complete culture medium was added to the cells. In a separate experiment, cells were transfected using different amounts of DNA (1–48 μg) to determine toxicity limits of tyrosinase expression.

The tyrosine hydroxylase catalytic activity was measured in protein extracts from 293HEK cells transfected with various tyrosinase constructs as described in Ref. 35 at 48 h after transfection. The assay was performed as described in Ref.36. Briefly, cells were lysed in PBS (pH 7.0), containing 1% Igepal. The lysates were supplemented with 2 μCi l-[³H] tyrosine (40–60 Ci/mmol;NEN-DuPont, Boston, MA). After 1 h of incubation at 37°C,supernatants were applied to microcolumns filled with the 50W-X8 resin(Bio-Rad, Hercules, CA) and the flow-through was mixed with charcoal(1.5 g in 20 ml of H₂O). A control sample was prepared in the absence of cell lysate. Tyrosine hydroxylase(tyrosinase) activity was expressed in nanomoles of tyrosine hydroxylated per minute, normalized per milligram of cell protein(37).

The melanin concentration was determined by measuring absorbance at 475 nm and was derived from a calibration curve using synthetic melanin(38).

293HEK cells (plated in 10-cm Petri dishes) were transfected with four tyrosinase expression constructs in triplicate. Forty-eight h after transfection, cells were harvested and washed with HBSS. The cells were then homogenized in a buffer containing 0.25 m sucrose, 25 mm KCl, 5 mm MgCl₂ and 20 mm Tris-HCl (pH 7.8). Postnuclear supernatant was obtained by centrifugation at 30,000 × gusing Beckman centrifuge (SW-41 rotor). The supernatant was separated in 30–10% OptiPrep gradient (Nicomed Pharma AS, Oslo, Norway),for 16 h at 30,000 × g (SW-41;Beckman). Gradients were divided into 0.5 ml fractions and DOPA-oxidase activity was determined (19). Peak fractions containing enzyme activity were selected, lysed by the addition of 50 mm Tris (pH 6.8), 0.1% SDS, 0.1% Igepal,supplemented with 1 mm PMSF and Complete inhibitors (Boehringer-Mannheim). Lysates were analyzed using 10%SDS-PAGE and transferred onto polyvinylidene difluoride membranes(Bio-Rad, Hercules CA), followed by blocking in 1% defatted milk(Bio-Rad) and 0.1% Tween 20 in PBS (pH 7.4). Washed membranes were then incubated with mouse monoclonal antibody raised against bacterial recombinant tyrosinase (NCL-Tyros; Novocastra Laboratories Ltd., United Kingdom; Ref. 39) diluted 1:500 with 1% defatted milk in PBS, followed by incubation with antimouse-peroxidase(Boehringer-Mannheim, Indianapolis, IN).

NAcSCAP was synthesized according to the method described in Ref.(40). The product was purified by crystallization from 25% isopropanol. HPP (Aldrich), or NAc4SCAP was dissolved in DMSO and added to culture medium at concentrations of 0–15 μm for HPP (41) or 0–2 mm for NAcSCAP(14). The concentration of DMSO in cell culture was kept below 0.5%.

To determine cell proliferation arrest mediated by the prodrug treatment of transfected and control mock transfected cells, cells were plated in 24-well plates at 2.5 × 10⁴ cells/well 24 h after transfection. Forty-eight h after transfection, cell culture media were supplemented with the prodrug (0–5 mm; overnight incubation for 16 h). By the end of the incubation time, media were supplemented with[³H]thymidine (0.4 μCi/well; NEN DuPont,Bellirica, MA) for 3 h to estimate cell proliferation with the subsequent cell harvesting and radioactivity measurement. Proliferation was estimated as a percentage of [³H]thymidine incorporation into the tyrosinase-transfected cells compared with mock-transfected control cells. Inhibition of proliferation in different cell lines was compared using IC₅₀values determined as drug concentration. Cytotoxicity of prodrugs was determined using CytoTox 96 Kit (Promega, Madison, WI). Percentage of dead cells was determined by measuring the release of lactate dehydrogenase as compared with the mock-transfected nontreated control cells.

Expression of tyrosinase mutants was studied in detail using transfected 293HEK cells as a model (see also Table 2 for other cell lines). To determine relative levels of mutant tyrosinase mRNA expression, total RNA isolated from transfected cells was subjected to RT-PCR analysis. PCR revealed the presence of cDNA as 1.6-, 1.55-,1.5-, and 1.4-kb fragments corresponding to T, TL, TΔC, and TΔTC tyrosinase mutants, respectively. When the amplified tyrosinase DNA band intensities were normalized to those of β-actin DNA,truncated variants TΔC and TΔTC showed a factor of 2, higher relative mRNA steady-state expression as compared with T and TL (Fig. 2). Western blot analysis of tyrosinase expression in transfected HEK cells was performed after density gradient centrifugation in fractions containing the highest total activity of DOPA oxidase (Fig. 3). The oxidase activity of the full-length tyrosinase was preferentially localized in microsomes and in dense lysosomes, whereas the truncated variant was completely excluded from lysosomes and was localized in the light vesicle fraction in addition to microsomes. The fractions that contained peaks of DOPA-oxidase activity also showed the presence of antityrosinase-reactive proteins (Fig. 4). An antityrosinase monoclonal antibody showed positive reaction with all of the tyrosinase mutants bearing COOH-terminal deletions (Fig. 4). The estimated protein M_rs of expression products were 72,000, 70,000, 65,000, and 62,000 corresponding to T, TL, TΔC and TΔTC mutants, respectively.

Specific activity of tyrosinase (35, 36, 37) was measured in 293HEK cells at 48 h after transfection. Cells transfected with T and TL cDNA contained specific activity that was factor of∼2.2–2.5 higher (P = 0.002) than in the control, mock-transfected cells. In cells transfected with TΔC and TΔTC cDNA, i.e., with truncated mutants, the enzyme activity was substantially higher by a factor of 3 and 4, respectively,than in control cells (P = 0.02; Fig. 5). Melanin biosynthesis was detectable in all of the cells transfected with tyrosinase cDNAs (Table 1). The intracellular melanin content ranged from 9.6 to17.7 μg/mg total cell protein (with control absorbance values subtracted from experimental ones) and was the lowest for TΔTC variant, presumably because of partial secretion of this mutant protein from cells. This was confirmed by measuring the melanin content of the cell culture supernatant. The highest extracellullar melanin content was detected for TΔTC mutant (13.6 μg/mg cell protein) and the lowest one (3.6μg/mg cell protein) for full-length tyrosinase (Table 1).

To determine whether transfection of 293HEK cells with tyrosinase would have any direct effects on proliferation, we performed thymidine incorporation assays for different amounts of DNA used for transfection. Transfection with increasing amounts of T and TL cDNAs resulted in a substantial decrease in thymidine incorporation (not shown). We observed a 20% reduced[³H]thymidine incorporation at the highest amount of cDNA used for transfection (24 μg/million cells). Expression of either TΔT or TΔTC mutants did not result in any significant decrease of thymidine incorporation in transfected cells.

In the next set of experiments, we tested antiproliferative (by measuring thymidine incorporation) and cytotoxic (by lactate dehydrogenase release) effects in 293HEK transfected cells using two different prodrugs: HPP and NAcSCAP. The IC₅₀for HPP had previously been determined as 11.7 μm in the work of Riley et al. (21). Thus, we varied HPP concentration in the range of 0–15μ m and performed standard[³H]thymidine incorporation tests. Expression of TL mutant was associated with the most pronounced prodrug-induced toxicity (Fig. 6). The results obtained with NAcSCAP as a prodrug (6, 14, 15) are summarized in Fig. 7. The 50% inhibition of [³H]thymidine incorporation occurred at 250 μm for T-transfected cells (full-length tyrosinase) and at 500μ m for cells expressing TL and TΔC mutants. Half-inhibition of cell proliferation for 293HEK cells expressing TΔTC mutant was observed at 2 mmonly (Fig. 7,B). The complete inhibition of cell proliferation was observed at 5 mm (Fig. 7 B) for all of the four tyrosinase mutants. Mock-transfected 293HEK cells showed about 20% inhibition of cell proliferation at this concentration. This number accounts for the presence of 0.5% DMSO(used as a solvent for NAcSCAP) as well as for spontaneous cell death at the level of 5–10%.

Cell cytotoxicity was determined as a fraction of prodrug-treated dead cells and was compared with that of control mock-transfected cells (Fig. 7 A). After 16 h incubation with NAcSCAP,the highest level of cell death (70% of the total cell population) was observed in the case of the full-length tyrosinase transfectants. When the treatment was extended to 48 h, a 100% cell death of transfected cells was observed at 2 mm NAcSCAP. In identical experimental conditions, 70% of control cells survived the treatment.

Experiments similar to those described above were performed using 9L rat gliosarcoma. The cytotoxicity data are summarized in Fig. 8,A, and antiproliferative activity is shown in Fig. 8,B. Results of these studies showed that the cytotoxicity effect reached 80–90% after 16-h incubation in the presence of NAcSCAP (Fig. 8 A). These numbers were comparable with the results obtained with the B16F1 murine melanotic cell line(92 ± 2%). As in the case of 293HEK cells, 48 h after prodrug treatment, transiently transfected 9L cells showed 100%cell death at 2 mm NAcSCAP. Seventy to 75% of mock-transfected cells survived.

Antiproliferative effects of tyrosinase-mediated prodrug conversion was also observed (Fig. 8 B); the complete arrest of cell proliferation was observed at lower than for 293HEK cells (2 mm).

In a series of separate experiments we determined inhibitory effects in a group of other cell lines (Table 2). These experiments were performed by treating cells (either mock-transfected or transfected with TΔC mutant with similar to wild-type tyrosinase antiproliferative and cytotoxic effects but with lower basal cytotoxicity) with NAcSCAP. These experiments were conducted: (a) to determine whether the approach would be applicable to other tumor lines; and (b) to determine differential effects between carcinoma and noncarcinoma cell pairs (e.g., adenoma versus adenocarcinoma and fibroblast versus fibrosarcoma). The results in Table 2 show that TΔC transfection followed by NAcSCAP treatment was cytotoxic and inhibitory to cells of the tested group. For MCF-7 human mammary adenocarcinoma cells, the IC₅₀ occurred at ∼1 mm of prodrug. For the tested HT-1080 fibrosarcoma line, the IC₅₀ was 0.15 mm, whereas that of the nontransfected melanoma was 0.5 mm (positive control). Interestingly,inhibitory concentrations were usually higher for nonmalignant cell counterparts.

The importance of tyrosinase expression in mammalian cells is in its intricate involvement in melanin biosynthesis. Tyrosinase catalyzes the synthesis of melanin precursors DOPA quinone and DOPA chrome. These can be toxic to cells because of their ability to deplete cells of reduced glutathione and other free thiols (21)and alkylate DNA (6) and to cause the arrest of DNA synthesis via the quinone-mediated mechanism of thymidylate synthase inhibition (22). Expression of tyrosinase in cells can potentially lead to such toxic effects because of the catalytic oxidation of tyrosine that enters cellular metabolism after being imported from the cell exterior (16). Moreover,tyrosinase-mediated conversion of nontoxic prodrugs into toxic metabolites has been proposed as a potential strategy of melanoma treatment (14, 41, 42, 43, 44), i.e., treatment of tumors that usually express high levels of tyrosinase. Furthermore,transfection of amelanotic melanomas with tyrosinase cDNA increases the toxicity of some tyrosinase substrates, e.g.,cysteaminophenol derivatives. These compounds usually do not exhibit toxicity in nonmelanotic cells (14, 41). Therefore, this suggests tyrosinase-mediated conversion of cysteaminophenols as a paradigm for gene therapy of other malignant tumors.

However, very slow maturation of tyrosinase and the concomitant endoplasmic retention of the nascent tyrosinase polypeptide has been described in amelanotic melanoma and potentially in other nonmelanotic cells (33). Therefore, we hypothesized that tyrosinase expression in transfected cells can be improved by engineering tyrosinase mutants. We assumed that: (a) COOH-terminally truncated mutants of tyrosinase will be expressed at higher levels than the wild-type enzyme; (b) these mutants may have a low basal level of cell toxicity but can still catalyze conversion of prodrugs[e.g., HPP and cysteaminophenol derivative (NAcSCAP)] into cytotoxic products.

To test these two assumptions, we first determined the cytotoxicity of CMV-promoter-driven tyrosinase expression in transfected 293HEK cells. As expected, expression of TΔC and TΔTC variants did not result in elevated rate of cell death. This may be explained by the fact that membrane-bound or secreted variants of tyrosinase are not present in prelysosomes or mature lysosomes and produce less toxic intermediates. Cell fractionation results supported this assumption,showing that DOPA-oxidase activity peaks in TΔC and TΔTC tyrosinases were colocalized with plasma membrane and/or microsomal fractions (Fig. 3). In contrast, the peaks of DOPA-oxidase activity in T and TL (both constructs showed identical distribution of DOPA-oxidase activity) were localized in microsomal fractions as well as in heavy fractions corresponding to lysosomes.

As shown by Western blot analysis, full-size tyrosinase as well as its deletion mutants and the fusion variant were expressed in transfected cells (Fig. 4). Antibodies reacted strongly with proteins isolated from cells transfected with a truncated TΔTC variant, which suggested higher expression than other tyrosinases. This result was in accordance with a higher steady-state mRNA level (Fig. 2) and correlated with the expression of specific tyrosinase activity (Fig. 5). It is well known that tyrosinase mRNA levels may not correlate with the melanin content and that pigment synthesis levels are determined by the posttranscriptional regulation of tyrosinase expression(45). Our results suggest either that tyrosinase activity may be down-regulated by melanin biosynthesis in the cell or that tyrosinase accumulates in non-lysosomal compartments (Table 1). We further demonstrated that expression of tyrosinase mutants bearing COOH-terminal deletions of the transmembrane domain and/or cytoplasmic sorting signal had a very limited effect on 293HEK cells viability,whereas the wild-type tyrosinase expression resulted in substantial inhibition. The wild-type tyrosinase (T) and the COOH-terminal Lamp-1 fusion protein (TL) showed an inhibition of cell proliferation at the level of 20–25% (at 25-μg expression vector transfected/10⁶ cells). The two other deletion tyrosinase mutants showed a basal inhibition comparable with control mock-transfected cells. This result may indicate that partially secreted (TΔTC) or cell-surface expressed (TΔC) tyrosinases(29) are localized in a compartment with lower local concentration of substrate (tyrosine) as opposed to tyrosinase variants sorted into a lysosomal (or premelanosomal) compartment, the site of tyrosine import (46) and melanin precursor biosynthesis. Furthermore, the specific activity of tyrosinase measured in cells transfected with the truncated tyrosinase cDNA (TΔC and TΔTC) was significantly higher than for lysosome-targeted variants (T and TL;Fig. 5). The lack of the melanosome/lysosome sorting signal (25, 47) in the two tyrosinase deletion variants TΔC and TΔTC resulted in lower cell melanization and in a significantly higher melanin concentration in the medium (2–3 fold; Table 1), presumably attributable to a more active secretion of the enzyme TΔTC. Thus, as anticipated, the deletion mutants possessed higher tyrosinase specific activity and had lower basal cell cytotoxicity.

Furthermore, we compared tyrosinase-mediated cytotoxic and antiproliferative effects in cells transfected with mutant and wild-type tyrosinases and treated with two model prodrugs: HPP and NAcSCAP (Figs. 6 and 7; Table 2). Both compounds are hydrophobic and insoluble in water and potentially should exert cytotoxicity on partition into the cell plasma membrane and subsequent intracellular transport (including lysosomes) as a result of constitutive retrograde membrane uptake by cells (48). To investigate differences in prodrug conversion by different tyrosinase mutants, a 16-h treatment experiment was performed initially. HPP inhibited 293HEK cell proliferation at very low concentrations—50% inhibition for T- and TL- transfectants at 2 μm—but was ineffective in the case of truncated tyrosinase transfections. Therefore,antiproliferative effects of HPP oxidation were induced almost exclusively by lysosome-sorted tyrosinase (Fig. 6). This effect can be attributed to specific lysosomal transport of HPP. With the further increase of prodrug concentration, no significant concentration-dependent inhibition was observed (data not shown). Similarly to HPP, cysteaminophenol-mediated inhibition of proliferation and cytotoxicity in mock-transfected 293HEK cells was very inefficient even at very high concentrations (>1 mm; Fig. 7). Another observed similarity was that the expression of tyrosinase mutant with deleted sorting/membrane attachment domain produced less NAcSCAP-mediated cytotoxicity and half-inhibited proliferation at the concentration 4- to 8-fold higher than in transfectants expressing other tyrosinases. Overall, partial or complete deletion of tyrosinase cytoplasmic/transmembrane domains appears to produce less efficient prodrug-converting mutants in transfected 293HEK cells and, at the same time, results in a lower basal cytotoxicity when compared with mock-transfected cells.

In contrast, cysteaminophenol treatment of 9L gliosarcoma cells resulted in a lack of significant difference in prodrug cytotoxicity mediated by various tyrosinase mutants (Fig. 8,A). The complete arrest of cell proliferation with any of tyrosinase mutants could be observed in the presence of 2 mmNAcSCAP, i.e., at the concentration that is 2.5 times lower than in the case of tyrosinase-transfected 293HEK cells. Interestingly,unlike the 293HEK cells, the nontransfected gliosarcoma cells showed much higher antiproliferative response to the prodrug(IC₅₀, 1.2 mm; complete inhibition at 2 mm; Fig. 8,B). At the same time, the level of cell death in treated mock-transfected 9L cells was only 10%. However, in tyrosinase-transfected 9L cells the antiproliferative effect of NAcSCAP was augmented by obvious cytotoxicity (>50% of total cells at 1 mm). Very similar results were observed with other tumor cells used in our study (Table 2).

In conclusion, our results suggest that tyrosinase gene expression can be used in gene therapy paradigm of prodrug activation not only for melanoma treatment but also for nonmelanotic cell treatment when transfected with tyrosinase/tyrosinase mutants. We have also shown that NAcSCAP does not exhibit cytotoxic effects in nontransfected tumor cells even after a prolonged treatment, but that the prodrug does interfere with DNA replication. Importantly, tyrosinase expression in transfected nonmelanotic tumors caused cell death attributable to prodrug conversion to a toxic product. All of the tyrosinase deletion mutants exhibited similar antiproliferation and cytotoxic effects in tumor cells and were as efficient as the wild-type tyrosinase. Finally,the observed cytotoxic effects were more prominent in more highly proliferating tumor cell lines (gliosarcoma, fibrosarcoma, and adenocarcinoma) as compared with the nontumorigenic counterparts. We anticipate that future research using more efficient vector systems will facilitate nonmelanotic tumor treatments based on tyrosinase/prodrug GDEPT system.

We are grateful to Dr. Xandra O. Breakefield for valuable suggestions and comments during the manuscript preparation and to Dr. Nikolay Sergeyev for technical expertise.