< Back to previous page
Project
The effect of Mycoplasma enzymes on the antiviral and anticancer activity of nucleoside analogues
With more than 3 million new cases and 1.7 million deaths each year, cancer is the most important cause of death in Europe after cardiovasculardiseases (numbers according to the World Health Organization). Strategies to manage cancer include surgery, radiation therapy, immunotherapy, hormonal therapy, chemotherapy and others. Currently, chemotherapeutic intervention of cancer is largely based on inhibition of tumor cell proliferation by the administration of nucleoside-based drugs (nucleoside analogues; NAs). After intracellular activation (often one or more phosphorylation steps), NAs may act as antimetabolites, interfering with the tumor cell nucleo(s)(t)ide metabolism and DNA/RNA synthesis. Since many cells in adults are quiescent and therefore not in the process of replication, NAs display a certain level of selectivity towards tumor cells.
The microenvironment of a tumor encompasses cancer cells and resident or infiltrating non-tumor cells. In addition, several studies havereported a close association of different prokaryotes with tumor tissuein cancer patients, which may be attributed to an increased nutrient availability at the tumor site, bacterial chemotaxis or defective defense mechanisms in cancer cells. A high preferential colonization of tumor tissue with different mycoplasma species (e.g. Mycoplasma hyorhinis</>), compared with healthy or non-malignant diseased tissue, has been repeatedly reported. Mycoplasmas are characterized by the lack of a cellwall and a strongly reduced genome and are considered to be the smallest self-replicating organisms. Some mycoplasmas (e.g. Mycoplasma pneumoniae</> and Mycoplasma genitalium</>) are clearly associated with disease. However, most species cause asymptomatic infections andare considered to be part of the residential flora of the healthy humanbody. Mycoplasma-related pathogenesis is therefore mostly observed in immunocompromised individuals. The limited number of genes in mycoplasmasis reflected in their nucleo(s)(t)ide metabolism. Most mycoplasma species share the inability to synthesize purine and pyrimidine nucleotides de novo</> and therefore rely on the tissue of their host for preformed nucleic acid precursors which are efficiently internalized and metabolized using several transport mechanisms and salvage enzymes, respectively. Therefore we hypothesized that a mycoplasma infection may significantly affect the availability and metabolism of nucleo(s)(t)ides and nucleo(s)(t)ide analogues in the host tumor cells.
Thecytostatic and antiviral activity of nucleoside-based drugs was indeed dramatically altered in mycoplasma-infected tumor cell cultures due to the expression of catabolic mycoplasma-encoded enzymes. Taking into account the preferential mycoplasma colonization of tumor tissue and the occurrence of secondary mycoplasma infections in immunocompromised individuals (e.g. AIDS patients), chemotherapy of cancer/viral infections with NAs may be suboptimal. In this work, we aimed at a careful investigation of the influence of a mycoplasma infection on the biological potential ofnucleo(s)(t)ide-derived drugs.
Mycoplasma-encoded nucleoside phosphorylases (NPs) were found to be crucial in the altered activityof purine- and pyrimidine-based NAs. Depending on the nature of the nucleobase that was released from the drug by NP-catalyzed phosphorolysis, an increased or a decreased biological activity was observed in mycoplasma-infected tumor cell cultures. The cytostatic activity of the drugs could be restored by the co-administration of mycoplasma-targeting antibiotics or a specific NP inhibitor. We therefore cloned the thymidine phosphorylase (TP) </>and purine nucleoside phosphorylase (PNP) </>genes annotated in the genome of M. hyorhinis</>and expressed them as recombinant proteins in Escherichia coli</>.
M. hyorhinis TP</> was found to encode a nucleoside phosphorylase that, in contrast to human TP, does not discriminatebetween pyrimidine-based 2-deoxyribosides and ribosides. The kinetic parameters of this enzyme were determined in detail and surprisingly revealed that uridine was the preferred natural substrate followed by thymidine. These results therefore indicate that, even though the mycoplasma-encoded enzyme shows high sequence similarity with both human and Escherichia coli </>TP, it should be re-annotated as a pyrimidine nucleoside phosphorylase (PyNP-Hyor) rather than as a TP. The expressionof similar enzymes has been reported for a few other prokaryotes (e.g. Geobacillus stearthermophilus</>) and the parasite Giardia lamblia</>. PyNP-Hyor also efficiently catalyzed the phosphorolysis of therapeutic 5-halogenated pyrimidine nucleosides (e.g. floxuridine) totheir less active nucleobases, explaining the decreased antiviral/cytostatic activity of such compounds in mycoplasma-infected cell cultures.
Also, efficient phosphorolysis of clinically approved purine-based NAs (e.g. cladribine and fludarabine) was observed in M. hyorhinis</>-infected tumor cells resulting in an altered biological activity of the compounds. Depending on the released nucleobase, mycoplasma infection resulted in a decreased cytostatic activity (as observed for cladribine due to the release of the poorly cytostatic 2-chloroadenine base) or increased cytostatic activity (as observed for fludarabinedue to the release of the highly cytostatic 2-fluoroadenine base). The co-administration of potent PNP inhibitors efficiently restored the activity of the drugs. The recombinantly expressed PNP of M. hyorhinis</> (PNP-Hyor) was kinetically characterized and found to catalyze thephosphorolysis of all natural purine nucleosides [except for xanthosine; including (2-deoxy)adenosine], and several of their therapeutic derivatives. Since (2-deoxy)adenosine and its analogues are generally not accepted as a substrate for human PNP, the presence of mycoplasmas ata tumor site may significantly contribute to a selective phosphorolysisof such drugs. Elimination of mycoplasmas or suppression of their catabolizing enzymes may therefore potentially increase the therapeutic indexof purine-based NAs such as cladribine. Alternatively, a rational choice for drugs that are activated upon mycoplasma-mediated phosphorolysis (e.g. fludarabine) may increase therapeutic selectivity.
Next, a decreased cytostatic activity of gemcitabine was observed in M. hyorhinis</>-infected cell cultures. Strong evidence was provided for the expression of a mycoplasma-encoded (2-deoxy)cytidine deaminase catalyzing the deamination (and therefore inactivation) of gemcitabine insuch cultures. In addition, we also observed a PyNP-related inactivation of gemcitabine: its cytostatic activity could be efficiently restored by selective PyNP inhibition and was not compromised in cell cultures infected with a PyNP-deficient mycoplasma strain. These findings were surprising to us since gemcitabine proved not to be a direct substrate for NP-catalyzed phosphorolysis. Several observations suggest that mycoplasma PyNP activity may alter the host cell pyrimidine nucleo(s)(t)ide metabolism (i.e. by decreasing intracellular dTTP andincreasing dCTP levels) and therefore may indirectly affect the activity of cellular 2-deoxycytidine kinase (dCK; the enzyme responsible for the initial activation of gemcitabine). However, it is currently unclear whether dNTP levels are sufficiently different in the mycoplasma-infected tumor cell cultures to have a significant inhibitory effect on dCK. Alternatively, the formation of a tight multi-enzyme complex of catabolic mycoplasma pyrimidine nucleoside salvage enzymes [i.e. PyNP and (2-deoxy)cytidine deaminase] may also explain our findings. However, this possibility is still very hypothetic and no solid evidence for its occurrence has yet been provided.
Due to the compromising role that mycoplasma PyNP may have in the cytostatic activity of thymidine analogues, we explored the possibility to develop a prodrug of floxuridine that is not susceptible to NP-catalyzed inactivation. In collaboration with prof. C. McGuigan (Cardiff University, Cardiff, UK), the phosphoramidate prodrug (ProTide) technology was applied to design novel derivatives of floxuridine-5-monophosphate, the active metabolite of floxuridine. We found the naphtyl/benzyl-L-alaninyl phosphate motif (present in prodrug NUC-3073) instrumental to confer resistance to enzymatic breakdown by human and prokaryotic NPs. Therefore, and in contrast to floxuridine, NUC-3073 retained its activity in mycoplasma-infected tumor cells remarkably well. Furthermore, this drug successfully delivered the active (phosphorylated) metabolite of floxuridine (5-fluoro-2-deoxyuridine-5-monophosphate) into the intact tumor cells, independently of the presence of thymidine kinase (TK) activity. This compound may therefore represent a more resilient alternative for floxuridine and may also prove useful in the treatment of tumors with (acquired or inherent) TK-deficiency. NUC-3073 is planned to enter phase I clinical trials for the treatmentof colorectal cancer in 2014.
In a next phase of our research we aimed at the design of a universal inhibitor that concomitantlytargets both pyrimidine and purine NPs. Most reported NP inhibitors arenucleoside derivatives with modifications in the nucleobase and therefore show high specific inhibitory activity against one type of NP. Instead, we aimed at the development of derivatives of (2-deoxy)ribose-1-phosphate, the common product of all NP-catalyzed reactions to find a broad-acting NP inhibitor. We found that 3,5-dichlorobenzoyl-substituted 2-deoxy-D-ribose-1-phosphate (Cf2891) inhibited a variety of both mammalian and prokaryotic pyrimidine and purine NPs. Kinetic studies revealed that this compound competes with inorganic phosphate for binding to NPs and, depending on the nature of the inhibited enzyme, a competitive or non-competitive inhibition with regard to the nucleoside binding site was observed. Cf2891 may therefore serve as a lead molecule for the development of potent broad-substrate NP inhibitors to simultaneously protect pyrimidine- and purine-based NAs from catabolism when administered in combination therapies.
The study of mycoplasma-infected tumor cell cultures has allowed us to explore the relationship between mycoplasma infections and the (cytostatic/antiviral) efficiency of therapeutic NAs. Our results suggest that a mycoplasma infection at the tumor site (or a secondary mycoplasma infection in immunocompromised patients) may alter the therapeutic index of such drugs. In vivo</> studies in mice confirmed that the antitumor activity of gemcitabine and floxuridine is indeed compromised when treating mycoplasma-infected mammary tumors compared with uninfected control tumors. Due to experimental constraints in co-culturing prokaryotes (different from mycoplasmas) along with mammalian cells, it is currently unclear whether our results are also of relevance for other bacterial infections. We demonstrated that a rational choice of (pro)drugs or a combination therapy of NAs with (i) mycoplasma-targeting antibiotics or (ii) specific inhibitors of catabolic mycoplasma-encoded enzymes, may optimize the efficiency of chemotherapeutics in cellculture. Our findings justify that this principle should be further explored and validated in the clinical setting.
The microenvironment of a tumor encompasses cancer cells and resident or infiltrating non-tumor cells. In addition, several studies havereported a close association of different prokaryotes with tumor tissuein cancer patients, which may be attributed to an increased nutrient availability at the tumor site, bacterial chemotaxis or defective defense mechanisms in cancer cells. A high preferential colonization of tumor tissue with different mycoplasma species (e.g. Mycoplasma hyorhinis</>), compared with healthy or non-malignant diseased tissue, has been repeatedly reported. Mycoplasmas are characterized by the lack of a cellwall and a strongly reduced genome and are considered to be the smallest self-replicating organisms. Some mycoplasmas (e.g. Mycoplasma pneumoniae</> and Mycoplasma genitalium</>) are clearly associated with disease. However, most species cause asymptomatic infections andare considered to be part of the residential flora of the healthy humanbody. Mycoplasma-related pathogenesis is therefore mostly observed in immunocompromised individuals. The limited number of genes in mycoplasmasis reflected in their nucleo(s)(t)ide metabolism. Most mycoplasma species share the inability to synthesize purine and pyrimidine nucleotides de novo</> and therefore rely on the tissue of their host for preformed nucleic acid precursors which are efficiently internalized and metabolized using several transport mechanisms and salvage enzymes, respectively. Therefore we hypothesized that a mycoplasma infection may significantly affect the availability and metabolism of nucleo(s)(t)ides and nucleo(s)(t)ide analogues in the host tumor cells.
Thecytostatic and antiviral activity of nucleoside-based drugs was indeed dramatically altered in mycoplasma-infected tumor cell cultures due to the expression of catabolic mycoplasma-encoded enzymes. Taking into account the preferential mycoplasma colonization of tumor tissue and the occurrence of secondary mycoplasma infections in immunocompromised individuals (e.g. AIDS patients), chemotherapy of cancer/viral infections with NAs may be suboptimal. In this work, we aimed at a careful investigation of the influence of a mycoplasma infection on the biological potential ofnucleo(s)(t)ide-derived drugs.
Mycoplasma-encoded nucleoside phosphorylases (NPs) were found to be crucial in the altered activityof purine- and pyrimidine-based NAs. Depending on the nature of the nucleobase that was released from the drug by NP-catalyzed phosphorolysis, an increased or a decreased biological activity was observed in mycoplasma-infected tumor cell cultures. The cytostatic activity of the drugs could be restored by the co-administration of mycoplasma-targeting antibiotics or a specific NP inhibitor. We therefore cloned the thymidine phosphorylase (TP) </>and purine nucleoside phosphorylase (PNP) </>genes annotated in the genome of M. hyorhinis</>and expressed them as recombinant proteins in Escherichia coli</>.
M. hyorhinis TP</> was found to encode a nucleoside phosphorylase that, in contrast to human TP, does not discriminatebetween pyrimidine-based 2-deoxyribosides and ribosides. The kinetic parameters of this enzyme were determined in detail and surprisingly revealed that uridine was the preferred natural substrate followed by thymidine. These results therefore indicate that, even though the mycoplasma-encoded enzyme shows high sequence similarity with both human and Escherichia coli </>TP, it should be re-annotated as a pyrimidine nucleoside phosphorylase (PyNP-Hyor) rather than as a TP. The expressionof similar enzymes has been reported for a few other prokaryotes (e.g. Geobacillus stearthermophilus</>) and the parasite Giardia lamblia</>. PyNP-Hyor also efficiently catalyzed the phosphorolysis of therapeutic 5-halogenated pyrimidine nucleosides (e.g. floxuridine) totheir less active nucleobases, explaining the decreased antiviral/cytostatic activity of such compounds in mycoplasma-infected cell cultures.
Also, efficient phosphorolysis of clinically approved purine-based NAs (e.g. cladribine and fludarabine) was observed in M. hyorhinis</>-infected tumor cells resulting in an altered biological activity of the compounds. Depending on the released nucleobase, mycoplasma infection resulted in a decreased cytostatic activity (as observed for cladribine due to the release of the poorly cytostatic 2-chloroadenine base) or increased cytostatic activity (as observed for fludarabinedue to the release of the highly cytostatic 2-fluoroadenine base). The co-administration of potent PNP inhibitors efficiently restored the activity of the drugs. The recombinantly expressed PNP of M. hyorhinis</> (PNP-Hyor) was kinetically characterized and found to catalyze thephosphorolysis of all natural purine nucleosides [except for xanthosine; including (2-deoxy)adenosine], and several of their therapeutic derivatives. Since (2-deoxy)adenosine and its analogues are generally not accepted as a substrate for human PNP, the presence of mycoplasmas ata tumor site may significantly contribute to a selective phosphorolysisof such drugs. Elimination of mycoplasmas or suppression of their catabolizing enzymes may therefore potentially increase the therapeutic indexof purine-based NAs such as cladribine. Alternatively, a rational choice for drugs that are activated upon mycoplasma-mediated phosphorolysis (e.g. fludarabine) may increase therapeutic selectivity.
Next, a decreased cytostatic activity of gemcitabine was observed in M. hyorhinis</>-infected cell cultures. Strong evidence was provided for the expression of a mycoplasma-encoded (2-deoxy)cytidine deaminase catalyzing the deamination (and therefore inactivation) of gemcitabine insuch cultures. In addition, we also observed a PyNP-related inactivation of gemcitabine: its cytostatic activity could be efficiently restored by selective PyNP inhibition and was not compromised in cell cultures infected with a PyNP-deficient mycoplasma strain. These findings were surprising to us since gemcitabine proved not to be a direct substrate for NP-catalyzed phosphorolysis. Several observations suggest that mycoplasma PyNP activity may alter the host cell pyrimidine nucleo(s)(t)ide metabolism (i.e. by decreasing intracellular dTTP andincreasing dCTP levels) and therefore may indirectly affect the activity of cellular 2-deoxycytidine kinase (dCK; the enzyme responsible for the initial activation of gemcitabine). However, it is currently unclear whether dNTP levels are sufficiently different in the mycoplasma-infected tumor cell cultures to have a significant inhibitory effect on dCK. Alternatively, the formation of a tight multi-enzyme complex of catabolic mycoplasma pyrimidine nucleoside salvage enzymes [i.e. PyNP and (2-deoxy)cytidine deaminase] may also explain our findings. However, this possibility is still very hypothetic and no solid evidence for its occurrence has yet been provided.
Due to the compromising role that mycoplasma PyNP may have in the cytostatic activity of thymidine analogues, we explored the possibility to develop a prodrug of floxuridine that is not susceptible to NP-catalyzed inactivation. In collaboration with prof. C. McGuigan (Cardiff University, Cardiff, UK), the phosphoramidate prodrug (ProTide) technology was applied to design novel derivatives of floxuridine-5-monophosphate, the active metabolite of floxuridine. We found the naphtyl/benzyl-L-alaninyl phosphate motif (present in prodrug NUC-3073) instrumental to confer resistance to enzymatic breakdown by human and prokaryotic NPs. Therefore, and in contrast to floxuridine, NUC-3073 retained its activity in mycoplasma-infected tumor cells remarkably well. Furthermore, this drug successfully delivered the active (phosphorylated) metabolite of floxuridine (5-fluoro-2-deoxyuridine-5-monophosphate) into the intact tumor cells, independently of the presence of thymidine kinase (TK) activity. This compound may therefore represent a more resilient alternative for floxuridine and may also prove useful in the treatment of tumors with (acquired or inherent) TK-deficiency. NUC-3073 is planned to enter phase I clinical trials for the treatmentof colorectal cancer in 2014.
In a next phase of our research we aimed at the design of a universal inhibitor that concomitantlytargets both pyrimidine and purine NPs. Most reported NP inhibitors arenucleoside derivatives with modifications in the nucleobase and therefore show high specific inhibitory activity against one type of NP. Instead, we aimed at the development of derivatives of (2-deoxy)ribose-1-phosphate, the common product of all NP-catalyzed reactions to find a broad-acting NP inhibitor. We found that 3,5-dichlorobenzoyl-substituted 2-deoxy-D-ribose-1-phosphate (Cf2891) inhibited a variety of both mammalian and prokaryotic pyrimidine and purine NPs. Kinetic studies revealed that this compound competes with inorganic phosphate for binding to NPs and, depending on the nature of the inhibited enzyme, a competitive or non-competitive inhibition with regard to the nucleoside binding site was observed. Cf2891 may therefore serve as a lead molecule for the development of potent broad-substrate NP inhibitors to simultaneously protect pyrimidine- and purine-based NAs from catabolism when administered in combination therapies.
The study of mycoplasma-infected tumor cell cultures has allowed us to explore the relationship between mycoplasma infections and the (cytostatic/antiviral) efficiency of therapeutic NAs. Our results suggest that a mycoplasma infection at the tumor site (or a secondary mycoplasma infection in immunocompromised patients) may alter the therapeutic index of such drugs. In vivo</> studies in mice confirmed that the antitumor activity of gemcitabine and floxuridine is indeed compromised when treating mycoplasma-infected mammary tumors compared with uninfected control tumors. Due to experimental constraints in co-culturing prokaryotes (different from mycoplasmas) along with mammalian cells, it is currently unclear whether our results are also of relevance for other bacterial infections. We demonstrated that a rational choice of (pro)drugs or a combination therapy of NAs with (i) mycoplasma-targeting antibiotics or (ii) specific inhibitors of catabolic mycoplasma-encoded enzymes, may optimize the efficiency of chemotherapeutics in cellculture. Our findings justify that this principle should be further explored and validated in the clinical setting.
Date:1 Jun 2008 → 2 Dec 2013
Keywords:Cancer
Disciplines:Microbiology, Systems biology, Laboratory medicine, Palliative care and end-of-life care, Regenerative medicine, Other basic sciences, Other health sciences, Nursing, Other paramedical sciences, Other translational sciences, Other medical and health sciences
Project type:PhD project