Biochemistry 811

Lecture on purine metabolism

D. Nelson, last modified Nov. 20, 2002 8AM

Reading: Berg, Tymoczko and Stryer 5th edition, Chapter 25


In the next two lectures, we will be discussing purine and pyrimidine metabolism. In genome 
sequencing and annotation, it is necessary to sort the discovered genes into categories. 
One of these category lists is shown in your handout.  It has 13 divisions plus 
unknown as a 14th category.  Purine and pyrimidine metabolism is one of the 13 major divisions.
Purines and pyrimidines are the building blocks of nucleic acids, but they are also found 
in other important molecules such as ATP, GTP, cyclic nucleotides, nucleotide cofactors such as 
NAD+, NADP+, FAD, CoA, S-adenosylmethionine (sold in food supplement stores as SAM-e).
sugar nucleotides (UDP glucose) and choline nucleotides (CDP choline). Nucleotides can serve as 
allosteric effectors for enzymes including the ATP Binding Cassette transporters.  CFTR, the 
protein responsible for cystic fibrosis, is in this family.  Even bare nucleosides 
have important roles.  The induction of sleep may be mediated by uridine through a uridine 
receptor in the CNS (Uridine receptor: discovery and its involvement in sleep mechanism. Sleep. 
2001 24, 251-60)  There are four adenosine receptors characterized.  The adenosine A2A receptor 
is a new target for drug therapy of Parkinsons disease.  These drugs may have fewer debilitating 
side effects than L-DOPA. (The adenosine A2A receptor of the basal ganglia. J Physiol. 2001 
Apr 15;532(Pt 2):284).


Before we start, I would like to mention some things about purine metabolism maps in KEGG.

The Kyoto Encyclopedia of Genes and Genomes

The human purine metabolism map is in your handout. It is also possible to color code human disease genes and these maps are available from the pull down menu. This map is also in your handout. The human map has 96 shaded boxes representing human enzymes. These are not all different enzymes, since some of them occur in 6-8 different locations in the chart. There are actually only 46 different human enzymes shown on this map. Comparison with the OMIM (Online Mendelian Inheritance in Man) human disease map shows 32 enzymes (70%) are linked to human diseases.

A little history

Purines and pyrimidines are major constituents of nucleic acids. The biochemical history of nucleic acids begins in the 19th century. In 1869, the Swiss biochemist Friedrich Miescher treated pus from hospital bandages with pepsin and dilute hydrochloric acid. He then extracted the mixture with ether and recovered a sediment from the bottom of the aqueous phase. The sediment was white cell nuclei. He isolated an acidic compound from this fraction that was rich in phosphorus. The high phosphorus content was unusual since the only other biological molecule known at that time to have phosphorus was lecithin (phosphatidylcholine). Miescher called the new substance nuclein. He had a hard time getting it published, because Ernst Hoppe-Seyler, a very influential scientist of the day, refused to publish it until he could confirm the result himself, which took two years. This nuclein preparation was DNA complexed with protein. Miescher and others who continued to work on nuclein devised ways to remove the protein, and the term nucleic acid was coined in 1889.

The chemical composition of nucleic acids from many different sources was studied by hydrolysis and analysis of the resulting chemical products. The composition included pentose sugars, phosphoric acid and purine and pyrimidine bases. The chemical compounds purine and pyrimidine are shown in Fig 5.4, p. 119. pyrimidine is a single aromatic ring with two nitrogens. Purine is a double ring that is pyrimidine with an imidazole ring fused to it. The five bases of RNA and DNA are just modifications of these basic structures. One easy way to remember which compounds are purines is to remember that purine has two syllables and the purine bases have two rings. It is very common to refer to purine nucleotides or pyrimidine nucleotides, so you need to know which is which. There is another type of base used in nucleotides called pyridine. It is very similar to pyrimidine except it has only one nitrogen in the ring. Nicotinamide adenine dinucleotide NAD+ is a pyridine nucleotide. NAD is an enzyme cofactor and it is not used in nucleic acids.

The five bases found in these nucleic acids were adenine and guanine (both purines) and cytosine, thymine and uracil (pyrimidines) see Fig. 5.4, p. 119. When uracil was seen thymine was not and when thymine was seen uracil was not. Also, when thymine was seen the pentose was always a deoxy sugar, but when uracil was seen, the pentose was not missing any hydroxyl groups. I use pentose here , because the sugar was not identified as D-ribose until 1909. We now know these as RNA and DNA, but even as late as 1920, the RNA was only isolated from yeast and plants, while the DNA was found in thymus glands and other animal sources. This led to the mistaken assignment of the two types of nucleic acids as plant and animal nucleic acids. This was never accepted and soon shown to be wrong, when both types of nucleic acids were isolated from the same source.

The names of the five bases indicate their original source. Guanine was initially found in guano, that is bird and bat droppings, frequently mined for fertilizer because it is so rich in nitrogen. Guanine has five nitrogens per molecule, so it is a good vehicle for the excretion of waste nitrogen in a concentrated form. Uracil is found in the urine of all carnivorous mammals. The etymology in the dictionary showed ur(ine) + ac(etic) + il. Il means a substance related to, therefore uracil is a substance related to urine and acetic acid. It probably precipitates from urine when acetic acid is added. Thymine comes from thymus gland. Glandular tissue is rich in nucleic acids and it was also the source of adenine. Aden is a greek root meaning gland. Cytosine also was discovered in thymus.

Purine Metabolism

Purine metabolism is illustrated on the KEGG pathway maps at Purine Metabolism

Under file menu go to page setup and select landscape orientation at 50% to print.

For Homo sapiens see Human Purine Metabolism or just use the pull down menu.

Your handout has an overview map of the biosynthesis and interconversions of nucleotides. This map shows where the precursors come from and where some of the products are used. This is a handy reference to follow as we look in more detail at the individual pathways. More detail is given in the KEGG map.

This figure looks intimidating, but it is really not so complex as it seems at first. At the top of the figure we have the de novo pathway of purine biosynthesis. This is given in more detail on pp. 698-700.

As we look down in the middle region we see IMP (inosine monophosphate). This is the precursor of adenine and guanine monophosphate. Note that interconversions among the adenine nucleotides are on the left and the guanine nucleotides are on the right. The degradation route is straight down from xanthine monophosphate XMP to uric acid. Humans stop at urate though they have the next enzyme, it is defective. On the bottom right is a separate loop involved in diadenosyl tetraphosphate or diadenosyl triphosphate metabolism. These compounds may signal DNA synthesis and promote growth. Their breakdown is important to prevent cancer. The FHIT (fragile histidine triad) gene 3.6.1.29 is a tumor suppressor. The gene is large covering 500kb of DNA. This gene is often broken in head and neck cancers. The gene 3.6.1.17 AP4A hydrolase may also be a tumor suppressor.

The purine de novo pathway

Note that the first step is production of PRPP (phosphoribosyl pyrophosphate). This enzyme is phosphoribosyl pyrophosphate synthetase. There are actually two genes for this enzyme in humans PRPS1 and PRPS2. They are on opposite ends of the the X chromosome. This is the first step of the pathway and therefore it is regulated by feedback inhibition of ADP and GDP. There is a disease in humans caused by mutants in PRPS1 that do not react to feedback inhibition. The enzyme is then called superactive and it overproduces PRPP and its products like uric acid, leading to a form of gout. Most diseases are caused by lack of enzyme activity. This one is caused by overactivity. This first step in the pathway requires ATP converted to AMP, thus it consumes two ATP equivalents making it irreversible.

The second step in the pathway catalyzed by GPAT (glutamine PRPP amidotransferase) is the committed step. That is because the PRPP has additional uses in histidine and tryptophan biosynthesis and pyrimidine biosynthesis. Once the amino group of glutamine is added, the pathway must go forward to IMP. Therefore, this step is also inhibited by ADP and GDP, just like the first step. The glutamine nitrogen becomes nitrogen 9 of the purine ring, and the rest of the ring structure is built on the ribose molecule. The next step (3) is GAR synthetase (for glycinamide ribonucleotide synthetase). Note the three reactions (3 GARS, 4 GART and 6 PAIS) are all catalyzed by a single multifunctional enzyme. Step 3 requires ATP and adds glycine to the amino group that came from glutamine in the second step. Step four adds a formyl group from N10 formyl tetrahydrofolate (Fig. 23.13, P. 675). Step 5 is another ATP requiring step that donates an amine from glutamine to form the imine at carbon 4. Now the ring is closed by another ATP requiring step to form AIR. This is now the imidazole part of the purine ring.

Step 7 is addition of a carboxyl group to the growing ring. This enzyme has two components in E. coli, PurE and PurK. PurE catalyzes the reaction without ATP hydrolysis, but PurK improves the efficiency in vivo by ATP hydrolysis. In vertebrates, this reaction does not require ATP. This step also does not require biotin, the usual group involved in carboxylations. The product is called CAIR. In step 8, CAIR is converted to SAICAR by SAICAR synthase, by addition of aspartate to the carboxyl added in the previous step. In humans these two enzymes are linked in a single polypeptide chain. The gene for the second step in this pathway catalyzed by glutamine:PRPP amidotransferase (GPAT) is found adjacent to the AIR carboxylase(EC 4.1.1.21)/SAICAR synthase (EC 6.3.2.6) gene on chromosome 4. They are only 625 bases apart and they are transcribed in opposite directions. The following is a quote from OMIM "Although there are several examples for bidirectional transcription in higher eukaryotes, GPAT-AIRC was the first example for bidirectional transcription of tightly coupled genes that are not structurally related but are involved in the same pathway. This may be a eukaryotic equivalent of a prokaryotic operon."

In yeast, AIR carboxylase is the ADE2 gene, and SAICAR synthase is the ADE1 gene. Mutants in these genes result in pink colored yeast colonies because the AIR compound gets oxidized to a pink product if either of the next two genes are defective. Complementation of the defect gives white colonies. This has been used as a genetic tool in some selection screens in yeast.

Step 9 removes the carbon skeleton of aspartate, just leaving behind the amino group and splitting off fumarate. The product is AICAR. This picks up another formyl group from N10- formyl tetrahydrofolate to form FAICAR (step 10). This undergoes ring closure by IMP synthase to make inosine monophosphate (step 11). Steps 10 and 11 are coded in a single polypeptide called ATIC, this is another example of a bifunctional enzyme. IMP is the precursor to AMP and GMP.

N-10 formyl tetrahydrofolate is needed to make purines at steps 4 and 10 in the pathway. Inhibitors of tetrahydrofolate reductase can prevent these steps and block purine synthesis. Since bacteria and cancer cells are faster growing than human cells, methotrexate, methopterin and aminopterin are effective treatments for some cancers. These will harm faster growing tissues like the blood and the intestinal lining if given for too long a time.

The purine ring is built from many different sources. The nitrogens all come from amino acids. Two are from glutamine, one from glycine and one from aspartate. Three of the carbons are from single carbon donations from N10-formyl tetrahydrofolate and CO2. The largest fragment with two carbons and a nitrogen is from glycine.

To form AMP from IMP, the same biochemical trick we already saw in SAICAR synthase (step 8) is used again. Aspartate is added in a GTP dependent reaction, then fumarate is removed to leave the amino group behind. In fact, the enzyme that removes the fumarate is the same enzyme adenylosuccinate lyase (ADSL). Homozygosity for mutations in the ADSL gene results in a clinical disorder called succinylpurinemic autism. Quote from OMIM "In 3 children with severe psychomotor delay and autism, Jaeken and Van den Berghe (1984) found succinyl- adenosine and succinylaminoimidazole carboxamide ribotide in the body fluids. Normally these compounds are not found in blood and CSF but may be detected in trace amounts in urine. The compounds are dephosphorylated derivatives of the 2 substrates of adenylosuccinate lyase. Build up of these compounds due to pathway blockage is probably responsible for the neurological symptoms. This enzyme is involved in both de novo synthesis of purines and formation of adenosine monophosphate from inosine monophosphate. Enzyme deficiency may be the basic defect in a subgroup of children with genetically determined autism.

To form GMP from IMP, the pathway goes through an intermediate xanthine monophosphate. XMP is formed by an NAD+ dependent oxidation of the purine ring. IMP dehydrogenase IMPDH1 is positioned at the branch point in the synthesis of adenine and guanine nucleotides and is thus the rate-limiting enzyme in the de novo synthesis of guanine nucleotides. Inhibition of cellular IMP dehydrogenase activity results in an abrupt cessation of DNA synthesis and a cell-cycle block at the G1-S interface. There is another IMPDH2 gene on another chromosome. The two proteins are 84% identical. After this step the newly formed oxy group is converted to an amino group donated by glutamine. This step requires ATP that is converted to AMP, so it takes two ATP equivalents. Since IMP required 5 ATP (6 high energy phosphate bonds), it costs 8 ATP equivalents to make one GMP. AMP costs 7 ATP equivalents. For this reason it is best for the cell to recover these nucleotides from used mRNA and recycle them rather than making them from the de novo pathway.

Purine Nucleotide interconversions, making ADP, ATP, GDP and GTP

AMP is called adenylate and GMP is guanylate. The cell needs to make GTP and ATP from these monophosphates. This is done by adenylate kinase and guanylate kinase to make the ADP and GDP forms (also the dADP and dGDP forms). These reactions require ATP. There are three forms Of adenylate kinase AK1 is found in red cells and muscle. Its absence causes mild chronic hemolytic anemia. AK2 is found in the mitochondrial inter-membrane space and AK3 is found in the mitochondrial matrix. There are also three forms of guanylate kinase GUK1, GUK2 and GUK3. GUK1 functions in the recovery of cGMP and is, therefore, thought to regulate the supply of guanine nucleotides to signal transduction pathways. Brady et al. (1996) stated that the guanylate kinases are targets for cancer chemotherapy and are inhibited by the antitumor drug 6-thioguanine.

Once the diphosphates are formed a non-specific enzyme nucleoside diphosphate kinase converts them on to the triphosphate forms. This enzyme also requires ATP, but is it not base specific and it is not sugar specific. The enzyme will convert any ribose or deoxyribose nucleotide diphosphate to a triphosphate. Because of the large flux of ADP to ATP in oxidative phosphorylation, OXPHOS is the most common way to convert ADP to ATP. The nucleoside diphosphate (NDP) kinase is mainly needed for the non-adenine nucleotide interconversions.

Excerpt from OMIM "Human erythrocyte nucleoside-diphosphate kinase (NDP kinase) is a hexameric enzyme consisting of two kinds of polypeptide chains, A and B. There was 88% identity between the A and B sequences. Chain A was identical with human Nm23 protein, which has been reported as a potential suppressor protein in tumor metastasis and chain B was identical with Nm23-H2 protein.

Postel et al. (1993) reported evidence suggesting that the protein encoded by 1 of the 2 closely related NM23 genes (see NME2; 156491) may be a transcription factor. The gene that may be turned on is the MYC oncogene (190080). Although a dozen DNA binding proteins had been identified for the MYC gene, only one, called PuF was known to regulate MYC transcription in vitro. Cloning of PuF demonstrated perfect identity with the NM23 gene. Therefore the nucleoside-diphosphate kinase A subunit is PuF, a transcription factor that controls MYC oncogene expression."

In all there are four of these NDP kinase genes in humans they are also known as NME1, 2, 3 and 4 for protein expressed in Non-MEtastatic cells. Gene loss of NME1 is associated with metastasis, so this is a tumor suppressor gene and the others may be also. Here is an example of a protein with two unrelated functions.

Regulation of de novo purine biosynthesis

Fig. 25.16, p. 710 and Fig. 27.7 in your handout show the main features of regulation of this pathway. The end products the adenine and guanine nucleotides all inhibit the committed step two at specific A and G sites on the enzyme. ADP and GDP inhibit the production of PRPP in step one, but this is not complete, because PRPP is used in other pathways. PRPP stimulates the committed step to make purines. At the junction where IMP can go to either GMP or AMP, the first step in each branch is inhibited by the product. GMP inhibits IMP dehydrogenase and AMP inhibits adenylosuccinate synthetase. To keep production of these nucleotides balanced GTP is needed for AMP production and ATP is needed for GMP synthesis. Therefore, if there is too much ATP it stimulates the production of more GMP and too much GTP stimulates the production of more AMP

Purine degradation

Please refer to figure 25.17, p. 710. Feeding of labeled nucleic acids shows that most nucleic acids eaten are not absorbed and incorporated into cellular components. The main need for purine degradation comes from within cells, as they recycle RNA. To break down nucleotides, the monophosphate form is converted to a nucleoside (ribose and base only) by a nucleotidase The KEGG map shows a single enzyme (EC 3.1.3.5) working on all four purine nucleotides IMP, AMP, XMP and GMP. The nucleosides are then degraded to free bases and ribose 1 phosphate by purine nucleoside phosphorylase (PNP). Phosphate is added during release of the base. Adenosine is not degraded this way but it must first be converted to inosine by adenosine deaminase. However, the KEGG map shows one enzyme (EC 2.4.2.1) converting all four nucleosides including adenosine to their bases. This is an error in the KEGG map. The three bases hypoxanthine (from inosine), xanthine (from xanthosine) and guanine are all converted to uric acid. Hypoxanthine and guanine are first converted to xanthine by xanthine oxidase or guanine deaminase, then xanthine is further oxidized by xanthine oxidase again to make uric acid.

Adenosine deaminase deficiency can cause severe combined immunodeficiency or SCID. The enzyme can act on adenosine or deoxyadenosine, but if it is missing, deoxyadenosine accumulates and is converted to dATP. This compound inhibits DNA synthesis at ribonucleotide reductase by preventing the synthesis of the other dNTPs. Rapidly proliferating cells are affected. These include blood cells like the lymphocytes needed for immune competence. This disease has been treated by gene therapy.

Excerpt from OMIM Blaese et al. (1995) reported results of the first gene therapy clinical trial for ADA-deficient SCID. Ex vivo retroviral-mediated transfer of the ADA gene was performed on the T cells of 2 children. Patient 1, 4-year-old Ashanthi DeSilva, was begun on gene therapy on 14 September 1990 and received a total of 11 infusions. Patient 2, 9-year-old Cindy Cutshall, was enrolled in the protocol on 31 January 1991 and received a total of 12 infusions. The number of blood T lymphocytes normalized, as did many cellular and humoral immune responses. Gene treatment ended after 2 years, but integrated vector and ADA gene expression in T cells persisted. Blaese et al. (1995) concluded that although many components remained to be perfected, gene therapy was a safe and effective addition to the treatment for some patients with this form of SCID. (note T cells are very long lived cells and this therapy should be life long.)

A deficiency of purine nucleoside phosphorylase can also cause SCID because it is the step after adenosine deaminase. A block at purine nucleoside phosphorylase causes the same buildup of dATP with the same result.

In your handout Fig. 26.20 notice that AMP can be converted to IMP by AMP deaminase. This is a different enzyme than adenosine deaminase. The IMP can then be converted back to AMP by two enzymes from the de novo pathway, adenylosuccinate synthase and adenylosuccinate lyase. This creates a cycle that generates fumarate in the lyase step. (Fig. 27.11 p. 912 2nd ed, Fig 27.11 1st ed.). This purine nucleotide cycle is important in skeletal muscle as a way to replenish TCA cycle intermediates as fumarate.

There are three versions of AMP deaminase, AMPD1, 2 and 3. AMPD1 is the muscle form called myoadenylate deaminase. Excerpt from OMIM *102770 ADENOSINE MONOPHOSPHATE DEAMINASE 1; AMPD1 Myoadenylate deaminase (MADA; EC 3.5.4.6) catalyzes the deamination of AMP to IMP in skeletal muscle and plays an important role in the purine nucleotide cycle. Deficiency of the muscle-specific myoadenylate deaminase is apparently a common cause of exercise-induced myopathy and probably the most common cause of metabolic myopathy in the human. It is the experience of most large centers that 1 to 2% of all muscle biopsies submitted for pathologic examination are deficient in AMP deaminase enzyme activity. The chief complaint, often dating from childhood, is muscle weakness or cramping after exercise. Fatigue after exertion is prolonged.

Allopurinol is an inhibitor of xanthine oxidase

Xanthine oxidase is the final step in producing uric acid. The uric acid is not very soluble in water and it precipitates at high concentrations. If for any reason there is an over production of uric acid, gout develops from the deposition of uric acid crystals in the joints. Hypoxanthine (from inosine) and xanthine (from guanine) are more water soluble and do not cause gout like symptoms. Therefore, inhibition of xanthine oxidase is a good strategy to treat gout. The inhibitor allopurinol is a hypoxanthine analog (p. 710). Only the location of one nitrogen is different. It inhibits xanthine oxidase and causes the concentration of xanthine and hypoxanthine to rise, while preventing or limiting the formation of uric acid. Xanthinuria type I is caused by a defect in xanthine oxidase. This disorder, which was first described by Dent and Philpot (1954), is characterized by excretion of very large amounts of xanthine in the urine and a tendency to form xanthine stones. Uric acid is strikingly diminished in serum and urine. (thus the basis of allopurinol).

The purine salvage pathway

The purine bases created by degradation of RNA and intermediates of purine synthesis were costly for the cell to make, so there are pathways to recover these bases in the form of nucleotides. Two phosphoribosyl transferases, one for adenine (APRT) and one for guanine or hypoxanthine (HGPRT) can join the bases to PRPP with the liberation of PPi. This is cleaved to form Pi thus driving the reaction (see p. 698) Loss of the HGPRT enzyme results in Lesch-Nyan syndrome (pp. 710-711). In this disease, the rate of purine synthesis is increased about 200 times and uric acid levels rise to cause gout. There are also severe mental abberations, since patients self mutilate by biting their lips and fingers off. The basis for the higher rate of purine synthesis is increased levels of PRPP, which is a substrate for the HGPRT enzyme. Since the PRPP cannot be reduced by this pathway, its level rises causing the activation of purine synthesis. It is not known what causes the mental disorder. It cannot be just increased levels of uric acid, since this also occurs in gout, but without mental problems.

Below are three related entries from OMIM related to Lesch-Nyhan Syndrome LNS 308950 LESCH-NYHAN PHENOTYPE WITH NORMAL HGPRT A male patient was described with self-mutilation, mental retardation, choreoathetosis, spasticity and hyperuricemia, identical to the clinical picture of HGPRT deficiency, although HGPRT and purine salvage were normal. This suggests that LNS can be caused by some other defect rather than a loss of HGPRT enzymatic activity. This patient's HGPRT gene was not checked for mutations. Later work showed a case like this one (or maybe the same case) was caused by an increase in the Km for PRPP. When assayed in the lab the PRPP concentration was much higher than in the cell so a reduction in HGPRT activity was not noticed. When PRPP was low as in the cell the HGPRT activity was also low.

300322 LESCH-NYHAN SYNDROME; LNS Virtually complete deficiency of HPRT residual activity (less than 1.5%) is associated with the Lesch-Nyhan syndrome, whereas partial deficiency (at least 8%) is associated with the Kelley-Seegmiller syndrome (300323). LNS is characterized by abnormal metabolic and neurologic manifestations. In contrast, Kelley-Seegmiller syndrome is usually associated only with the clinical manifestations of excessive purine production.

308000 HYPOXANTHINE GUANINE PHOSPHORIBOSYLTRANSFERASE 1; HPRT1 (also called HGPRT) A single point mutant has been described that has less than 0.1% of normal HPRT activity.

[.0049 HPRT DEFICIENCY, PARTIAL [HPRT, GLY16SER] HPRT URANGAN In a patient with partial HPRT deficiency (enzyme activity less than 0.1%; 300323), Sculley et al. (1991) identified a G-to-A mutation at nucleotide 145 resulting in a substitution of serine for glycine-16.]

This patient has gout but not LNS. Patients with absence of the HPRT protein have LNS. This suggests that the LNS symptoms may be independent of the HPRT enzyme activity. The protein may have more than one function, with the 2nd function being responsible for the LNS. The 2nd function does not have to be an enzymatic activity, but could be a regulatory interaction with another protein. Loss of this interaction could then lead to LNS. There are single point mutations like F74L that do cause LNS. Patients with gout from other defects do not get the LNS symptoms, so the elevation in purines alone cannot account for this. This is reminiscent of the NDP kinase also serving as a transcription factor that affects metastasis.

There is also salvage of adenine by APRT. Absence of this enzyme does not cause gout, probably because adenosine deaminase converts adenosine to inosine so the flux through this pathway is lower than for the HGPRT salvage of hypoxanthine and guanine.

These salvage pathways are also useful for activation of prodrugs to active forms. 6-mercaptopurine is converted to a nucleotide by the purine salvage pathways. This nucleotide analog is then a competitive inhibitor of IMP in the synthesis of AMP and GMP. 6-Mercaptopurine (6-MP) can be inactivated by S-methylation, which is catalyzed by thiopurine methyltransferase. A small percentage of acute lymphoblastic leukemia patients are known to have low activity of this enzyme because they have one or two copies of an inactive allele. It is now possible to genotype patients before treatment to avoid overdosing them with 6-MP which can cause liver toxicity. Much of this progress has been made at St. Judes by Dr. Bill Evans group.

What happens to uric acid?

Uric acid is the end of purine metabolism in birds, reptiles, insects and primates. It is excreted. This is not the end in other mammals and in molluscs. In these organisms, the purine ring is opened by urate oxidase to form allantoin (your handout Fig 27.14). Humans have the urate oxidase gene, but it has a stop codon in exon 5 so the protein is not made. Urate, is a powerful antioxidant that directly scavenges reactive oxygen species and chelates iron. Ames et al. (1981) proposed that one of the mechanisms for the lengthening life span and the decrease in age-specific cancer rates during primate evolution may be the high plasma levels of urate that are 10-fold greater than those in lower mammals.

Bony fish continue the ring breakage by forming allantoic acid. This is further broken down by cartilaginous fish (sharks etc.) and amphibians to make glyoxylate and two molecules of urea. Crustaceans can break down urea to CO2 and NH3 by the enzyme urease. Land animals cannot tolerate ammonia since it is toxic, but when you live in the water it can be continuously excreted and diluted so it is not harmful as a waste product.