Most nucleotide biosynthesis in most cells occurs via the nearly ubiquitous de novo synthetic pathways, starting from amino acids and their derivatives (see Purine Ribonucleotide Metabolism and Pyrimidine Ribonucleotide Metabolism). However, most cells possess capabilities for taking up nucleosides and nucleobases and converting them to nucleotides. Because these processes involve reutilization of previously synthesized purine and pyrimidine rings, they are called salvage pathways. These pathways are much shorter and simpler than the de novo pathways. On the other hand, there is much more variability from organism to organism, and from tissue to tissue in the same organism, in salvage synthetic capabilities. The metabolic importance of salvage pathways has come to light, first, along with the realization of the serious consequences of hereditary deficiencies in certain salvage enzymes and, second, with the many ways in which salvage pathways are being exploited to create or enhance the effectiveness of chemotherapy against a variety of diseases.
1. Transport of Nucleosides and Nucleobases
Substrates for salvage pathways can come from intracellular nucleic acid degradation (see DNA Degradation In Vivo). For most cells, however, salvage precursors are derived from the extracellular environment. An extreme example is protozoan parasites, which derive all their nucleic acid precursors from salvage substrates in the blood of infected organisms. In fact, these parasites have evolved to the extent that they completely lack de novo pathways and are totally dependent on salvage (1, 2); development of specific salvage inhibitors now constitutes one of the most active research areas of biochemical parasitology.
Nucleosides are hydrophilic molecules and do not readily diffuse through membranes; to an extent, the same is true for nucleobases. Thus, most cells contain specific transporter systems, which take up these molecules by facilitated diffusion (3, 4). In recent years, however, concentrative transport systems have been described, especially for nucleosides. These systems depend on sodium and involve cotransport of the nucleoside with Na+. Most animal cells contain a broad-specificity nucleoside transporter that functions by facilitated diffusion, the active transport systems being limited primarily to specialized cells, for example, those that take up and use adenosine as a physiological regulator. The nucleobase transport systems are less well characterized, and the question of whether cells contain multiple base transporters or a single broad-specificity transporter has not yet been resolved.
2. Salvage Pathways
Nucleobases are anabolized by phosphoribosyltransferases , which use 5-phosphoribosyl-1-pyrophosphate (PRPP), Nucleosides can also be salvaged by a nucleoside phosphorylase followed by a phosphoribosyltransferase.
while nucleosides are anabolized primarily by kinases.
In principle, this route could be used for nucleotide synthesis, starting with a base, given that nucleoside phosphorylases are readily reversible. However, these enzymes generally act within cells in the direction of nucleoside degradation.
Salvage enzymes vary considerably in tissue distribution and specificity. Mammalian cells salvage purines primarily at the free base level through the action of two phosphoribosyltransferases: hypoxanthine guanine phosphoribosyltransferase (HGPRT), which converts hypoxanthine and guanine to inosinic acid (IMP) and GMP, respectively; and adenine phosphoribosyltransferase (APRT), which specifically converts adenine to AMP. HGPRT is the enzyme missing in cells of males suffering from Lesch-Nyhan syndrome (see Purine Ribonucleotide Metabolism). Pyrimidines, by contrast, are more likely to be salvaged at the nucleoside level. In the de novo pathway, orotic acid is converted to orotidylate by a phosphoribosyltransferase, but comparable pyrimidine salvage enzymes are of extremely limited distribution.
Protozoa that rely on nucleotide salvage pathways, particularly for purines, bear a distinctive set of enzymes— nucleoside hydrolases. These organisms take up purine nucleosides from the blood of infected animals, for example, and then hydrolyze the nucleoside to ribose plus the base; the base is then salvaged by a phosphoribosyltransferase (5). Because the nucleoside hydrolases are quite distinct from known proteins, they are being explored as targets for drugs that would specifically inhibit the growth of an infecting parasite by blocking this salvage route, which is essential for the parasite but not the host.
As discussed more thoroughly below, nucleoside and base analogues are widely used as antiviral, antimicrobial, and anticancer drugs. Because of the interest in developing or improving the effectiveness of drugs that act as DNA synthesis inhibitors, there has been particular emphasis on the deoxyribonucleoside kinases (6). Four such enzymes, with overlapping specificities, are found in human cells. Properties and subcellular distributions of these enzymes are summarized in Table 1. It is of particular interest that deoxycytidine kinase (dCK) is also the principal enzyme for utilization of purine deoxyribonucleosides (the Km values are much higher for the purines than for deoxycytidine; hence the name of the enzyme).
Table 1. Properties of the Human Deoxyribonucleoside Kinases
|
Enzyme Cycle Regulation |
Natural Substrates |
Intracellular Distribution |
Cell Cycle Regulation |
|
Deoxycytidine |
|||
|
kinase (dCK) |
dCyd, dAdo, dGuo |
Cytosol |
Constitutive |
|
Thymidine kinase 1 (TK1) |
dThd, dUrd |
Cytosol |
S-phase-specific |
|
Thymidine kinase 2 (TK2) |
dThd, dUrd, dCyd |
Mitochondria |
Constitutive |
|
Deoxyguanosine kinase (dGK) |
dGuo, dAdo, dIno |
Mitochondria- |
Constitutive |
A key to understanding some clinical disorders is knowledge of the tissue distribution of salvage enzymes. For example, the immunodeficiency state that results from adenosine deaminase deficiency can be related to this factor (see Purine Ribonucleotide Metabolism). Most or all tissues in an affected individual accumulate adenosine and deoxyadenosine as a result of the genetic block to purine catabolism. Nucleoside salvage enzymes are particularly active in the reticuloendothelial system, where apoptosis leads to degradation and reutilization of cell components, including nucleic acids. Thus, blood cells accumulate excessive amounts of ATP, and in particular, dATP, which interferes with white cell proliferation as part of the immune response, through its inhibition of ribonucleotide reductase.
A salvage enzyme of considerable interest is the family of thymidine kinases specified by herpes viruses. As exemplified by the herpes simplex thymidine kinase, these enzymes have two distinctive properties. First, they are bifunctional enzymes, with thymidylate kinase activity as well. Thus, the enzyme is evidently designed to catalyze two sequential reactions. Second, the nucleoside kinase activity has much broader substrate specificity than most cellular nucleoside kinases. As discussed below, antiviral chemotherapeutic strategies exploit this distinctive property.
3. Salvage Enzymes as Selectable Markers
Because nearly all cells possess de novo nucleotide synthetic capabilities, the salvage enzymes are usually not required for cell viability. In addition, a large number of nucleobase and nucleoside antimetabolites are available, most of them inhibitors of specific enzymes (see Nucleotides, Nucleosides, And Nucleobases). These factors create favorable conditions for the use of salvage enzymes as selectable genetic markers, ie, genetic characteristics that promote the selective survival or growth of desired cell types. For example, the mammalian enzyme encoding HGPRT is widely used as a selectable marker in genetic analysis. 6-Thioguanine is metabolized by HGPRT to give the thiol analogue of inosinic acid, which is toxic. Cells lacking HGPRT can be selected for because they grow in 6-thioguanine-containing medium, while other cells are killed. Hence, one can estimate mutation rates by culturing wild-type cells in the presence of thioguanine and enumerating the cells that grow. Another advantage of this gene for genetic analysis is that it is carried on the X-chromosome. Thus, in male-derived cell lines, only one mutation, rather than two independent events is required to give the resistant phenotype.
A comparable selectable marker is the gene encoding thymidine kinase. 5-Bromodeoxyuridine is anabolized similarly to thymidine;
However, incorporation of bromodeoxyuridine into DNA is a lethal event. Thus, thymidine kinase deficiency leads to a BrdUrd-resistant phenotype. Because most large DNA viruses encode a thymidine kinase, this is a useful system for manipulating viral genes, for example, the use of vaccinia virus as a vector in the generation of multivalent vaccines (7).
On the other hand, one can select for the presence of active salvage enzymes. The best example is the use of "HAT medium" in somatic cell genetic analysis and in preparing monoclonal antibodies. These techniques involve fusing cells of different origins and culturing in HAT medium to select for those cells that have undergone fusion. HAT is an acronym for the medium’s constituents,— hypoxanthine, aminopterin, and thymidine. Aminopterin inhibits dihydrofolate reductase, blocking the synthesis of tetrahydrofolate needed for de novo synthesis of purine nucleotides and thymidine nucleotides. Thus, cells can grow in HAT medium only if they express active thymidine kinase and HGPRT, for salvage synthesis of thymidine and purine nucleotides, respectively. In monoclonal antibody production, one of the cell lines to be fused lacks thymidine kinase, and the other lacks HGPRT. Thus, only cells resulting from a fusion event have functional copies of both enzymes and can grow.
A final example relates to the use of a selectable marker to force expression of a nonselected marker. For cloning into expression systems in mammalian cells, one often incorporates into the cloning vector the Escherichia coli xpt gene, which encodes a distinctive phosphoribosyltransferase that acts on xanthine and guanine. During and after transformation of cells with the recombinant DNA, the cells are cultured in the presence of mycophenolic acid, which blocks de novo guanine nucleotide synthesis by inhibiting IMP dehydrogenase, the enzyme that converts IMP to XMP (which would then be converted to GMP). Thus, the only cells that can grow are those that have taken up and expressed the xpt gene, which bypasses this metabolic block. Being carried on the same vector, the gene of interest is also cloned and/or expressed, even though its expression was not directly selected for.
4. Salvage Enzymes and Molecular Pharmacology
A large number of antiviral, antibacterial, antiparasitic, and anticancer drugs act by inhibiting or otherwise manipulating pathways in nucleotide metabolism and thereby interfering with nucleic acid synthesis. Several examples are cited in Nucleotides, nucleosides, and nucleobases. In most cases, the active species is a nucleotide. Because nucleotide molecules are charged, and because specific transport systems for nucleotides don’t exist, these molecules penetrate membranes poorly, if at all.
Therefore, in order to generate a therapeutically effective nucleotide intracellularly, its precursor must be administered extracellularly as a nucleobase or nucleoside analogue. Effective use of such drugs demands extensive understanding of salvage pathways in the target cell, subcellular distribution of the enzymes involved, degradative enzymes that might compete with the pathway leading to the desired nucleotide analogue, levels of the target enzyme, and cell-cycle regulation of all the enzymes involved (8).
As an example of the factors involved, consider the fluorinated pyrimidines, 5-fluorouracil (FUra) and 5-fluorodeoxyuridine (FdUrd), analogues used for four decades in treating various cancers. It was established in 1958 (9) that these drugs are converted in vivo to 5-fluorodeoxyuridine monophosphate (FdUMP), an analogue of deoxyuridine monophosphate, the substrate for thymidylate synthase (Fig. 1; see Deoxyribonucleotide Biosynthesis And Degradation), and that FdUMP is a potent inhibitor of thymidylate synthase and, hence of DNA replication. Figure 1 shows also the metabolic pathways that both activate these analogues and divert them from their desired endpoint (10). From the figure, one can see that coadministration with FdUrd of a thymidine phosphorylase inhibitor should increase the effectiveness of the analogue by blocking its catabolism. Note that there are multiple routes for activation of FUra; note also that FUra can enter pools of RNA precursors which, in principle, could limit its selectivity by diminishing the specificity of its effect against DNA synthesis. There is evidence, however, that, in some tumors, the effectiveness of FUra actually depends in part on its incorporation into RNA, where it stimulates translational miscoding.
Figure 1. Metabolism of fluorinated pyrimidines by salvage and other enzymes. dR-P is deoxyribose 5′-phosphate, dR-1-P deoxyribose-1-phosphate, and R-1-P ribose-1-phosphate. Enzymes are 1, thymidylate synthase; 2, dihydrofolate reductase; 3, serine transhydroxymethylase; 4, thymidine kinase; 5, thymidine phosphorylase; 6, uridine phosphorylase; 7, uridine kinase; 8, uridylate kinase; 9, nucleoside diphosphate kinase; 10, ribonucleoside diphosphate reductase; 11, RNA polymerase.
