Absence of 3′-Terminal Residues from Transfer Ribonucleic Acid of Dormant Spores of Bacillus megaterium

Essentially all (>97%) of the transfer ribonucleic acid (tRNA) in log-phase and sporulating cells of Bacillus megaterium contains a complete 3'-cytidyl-cytidyl-adenosine terminus. However, about one-third of the tRNA in the dormant spore lacks the 3'-terminal adenosine 5'-monophosphate (AMP) residue, and some of the adjacent cytosine monophosphate residues are also missing. Examination of specific tRNAs indicated that those specific for isoleucine, leucine, and methionine are missing 30 to 40% of their terminal residue, whereas tRNAs specific for tyrosine lack 88% of the 3'-terminal AMP. Defective spore tRNA is not degraded during germination, but the missing residues are added back in the first minutes of the process. The enzyme catalyzing the addition reaction, tRNA nucleotidyltransferase, is present in the dormant spore at a level similar to that found in the vegetative cell.     

Reactions at the 3' terminus of transfer ribonucleic acid. IX. tRNA nucleotidyltransferase activity in bacteriophage T4-infected Escherichia coli

There was no detectable increase in tRNA nucleotidyltransferase activity upon infection of $\text{Escherichia coli}$ A19 with bacteriophage T4. Three mutant strains which contained low levels of tRNA nucleotidyltransferase activity also showed no increase in activity after infection. tRNA nucleotidyltransferase was purified from both uninfected and T4-infected cells and examined for possible modification. It was found that enzyme purified from both types of cells eluted from DEAE cellulose at the same specific conductivity. In addition, the molecular weight of tRNA nucleotidyltransferase purified from both uninfected and T4-infected cells was approximately 45,000 daltons as determined by chromatography on Sephadex G-100. These results suggest that T4-infection does not lead to synthesis of a new virus-specific tRNA nucleotidyltransferase nor does it cause modification of the host enzyme.     

tRNA nucleotidyltransferase is not essential for Escherichia coli viability.

The role of tRNA nucleotidyltransferase in Escherichia coli has been uncertain because all tRNA genes studied in this organism already encode the -C-C-A sequence. Examination of a cca mutant, originally thought to contain 1-2% enzyme activity, indicated that it actually produces an inactive fragment of 40 kd compared to 47 kd for the wild-type enzyme due to a nonsense mutation in its cca gene. To confirm that the residual activity in extracts of this strain is due to another enzyme, and that tRNA nucleotidyltransferase is non-essential, we have interrupted the cca gene in vitro, and transferred this mutant gene to a variety of strains. In all cases mutant strains are viable, although as much as 15% of the tRNA population contains defective 3' termini, and no tRNA nucleotidyltransferase is detectable. Mutant strains grow slowly, but can be restored to more normal growth by a relA mutation or by a decrease in RNase T activity. In the latter case the amount of defective tRNA decreases dramatically. These findings indicate that tRNA nucleotidyltransferase is not essential for E. coli viability, and therefore, that all essential tRNA genes in this organism encode the -C-C-A sequence.     

Reversible inactivation of tRNA nucleotidyltransferase from baker's yeast by tRNAPhe containing iodoacetamide-alkylated 2-thiocytidine in normal and additional positions.

2-Thiocytidine 5'-triphosphate, s2CTP, is able to replace CTP as a substrate for tRNA nucleotidyltransferase. s2CMP can be incorporated into both cytidine sites of the C-C-A terminus common to all tRNAs, and in the absence of ATP into at least two additional positions. This was shown by alkylation of the 2-thiocytidine residues with iodo[14C]acetamide, total nucleoside analysis, microgel electrophoresis and analysis of RNase T1 fragments of these tRNAs. The incorporation of the 3'-terminal AMP is not influenced by the additional s2CMP residues at pH 9.0. However, at pH 7.6 the additional s2CMP residues are hydrolysed and AMP can be incorporated into the normal position. Two different tRNAs with terminal 2-thiocytidine alkylated by iodoacetamide inhibit tRNA nucleotidyltransferase. This inhibition is significantly slower if an elongated species is used compared to a tRNA with alkylated 2-thiocytidine in the normal position 75. The addition of 2-mercaptoethanol reactivates the enzyme and leads to a cytidine containing tRNA. This reaction identifies the attacking nucleophile of the enzyme as cysteine residue, which is probably identical to a cysteine residue found in a similar experiment reported previously. The mechanism of the enzymatic and chemical reactions is discussed.     

Effects of crp mutations on adenosine 3'',5''-monophosphate metabolism in Salmonella typhimurium.

Wild-type Salmonella typhimurium could not grow with exogenous cyclic adenosine 3',5'-monophosphate (AMP) as the sole source of phosphate, but mutants capable of cyclic AMP utilization could be isolated provided the parental strain contained a functional cyclic AMP phosphodiesterase.All cyclic AMP-utilizing mutants had the growth and fermentation properties of cyclic AMP receptor protein (crp) mutants, and some lacked cyclic AMP binding activity in vitro. The genetic defect in each such mutant was due to a single point mutation, which was co-transducible with cysG. crp mutants isolated by alternative procedures also exhibited the capacity to utilize cyclic AMP. crp mutants synthesized cyclic AMP at increased rates and contained enhanced cellular cyclic AMP levels relative to the parental strains, regardless of whether or not cyclic AMP phosphodiesterase was active. Moreover, adenylate cyclase activity in vivo was less sensitive to regulation by glucose, possibly because the enzyme II complexes of the phosphotransferase system, responsible for glucose transport and phosphorylation, could not be induced to maximal levels. This possibility was strengthened by the observation that enzyme II activity (measured both in vitro by sugar phosphorylation and in vivo by sugar transport and chemotaxis) was inducible in the parental strain but not in crp mutants. The results suggest that the cyclic AMP receptor protein regulates cyclic AMP metabolism as well as catabolic enzyme synthesis.     

Separation of functionally distinct regions of a macromolecular substrate. Stimulation of tRNA nucleotidyltransferase by a nonreacting fragment of tRNA.

Rabbit liver tRNA nucleotidyltransferase catalyzes the incorporation of AMP and CMP into the model acceptor substrate, cytidine. The apparent Km for cytidine in this reaction is about 80 to 90 mM which is more than 10(4) greater than the Km values for the natural substrates, tRNA lacking the terminal AMP (tRNA-C-C) and tRNA lacking the terminal pCpA (tRNA-C). The Vmax values for the model reaction are only 5% and 2% of those for the reaction with the natural tRNA substrates. Addition of the tRNA fragments, tRNA lacking the terminal XpCpCpA sequence (tRNA-(X - 1)p) and tRNA lacking the terminal CpCpA (tRNA-Xp), greatly stimulates the rate of nucleotide incorporation into cytidine. In the case of CMP incorporation into cytidine, tRNA-Xp stimulates the reaction about 60-fold, to a rate similar to that of the normal reaction with tRNA-C. The tRNA fragment has no effect on the apparent Km of either cytidine or CTP, but only alters the Vmax of the reaction. Stimulation of the model reactions is maximal with tRNA fragments of specific chain lengths. These results provide direct evidence that the nonreacting regions of a substrate molecule play an important role in the catalytic efficiency of an enzyme.     

Schizosaccharomyces pombe contains separate CC- and A-adding tRNA nucleotidyltransferases

A specific cytidine-cytidine-adenosine (CCA) sequence is required at the 3′-terminus of all functional tRNAs. This sequence is added during tRNA maturation or repair by tRNA nucleotidyltransferase enzymes. While most eukaryotes have a single enzyme responsible for CCA addition, some bacteria have separate CC- and A-adding activities. The fungus, Schizosaccharomyces pombe, has two genes (cca1 and cca2) that are thought, based on predicted amino acid sequences, to encode tRNA nucleotidyltransferases. Here, we show that both genes together are required to complement a Saccharomyces cerevisiae strain bearing a null mutation in the single gene encoding its tRNA nucleotidyltransferase. Using enzyme assays we show further that the purified S. pombe cca1 gene product specifically adds two cytidine residues to a tRNA substrate lacking this sequence while the cca2 gene product specifically adds the terminal adenosine residue thereby completing the CCA sequence. These data indicate that S. pombe represents the first eukaryote known to have separate CC- and A-adding activities for tRNA maturation and repair. In addition, we propose that a novel structural change in a tRNA nucleotidyltransferase is responsible for defining a CC-adding enzyme.     

Mutagenesis of the folC gene encoding folylpolyglutamate synthetase-dihydrofolate synthetase in Escherichia coli

The folC gene of Escherichia coli, cloned in a pUC19 vector, was mutagenized by progressive deletions from both the 5' and the 3' ends and by TAB linker insertion. A number of 5'-deleted genes, which had the initiator ATG codon removed, produced a truncated gene product, in reduced amounts, from a secondary initiation site. The most likely position of this site at a GTG codon located 35 codons downstream of the normal start site. This product could complement the folC mutation in E. coli strain SF4 as well as a strain deleted in the folC gene. The specific activity of extracts of the mutant enzyme are 4-16% that of the wild type enzyme for the folylpolyglutamate synthetase activity and 6-19% for the dihydrofolate synthetase activity. The relative amount of protein expressed by the mutant, compared to the wild type, in maxicells was comparable to the relative specific activity, suggesting that the kcat of the mutant enzyme is similar to that of the wild type. Mutants with up to 14 amino acids deleted from the carboxy terminal could still complement the folC deletion mutant. Seven out of ten linker insertions dispersed through the coding region of the gene showed complementation of the folC mutation in strain SF4 but none of these insertion mutants were able to complement the strain containing a deleted folC gene. None of the carboxy terminal or linker insertion mutants had a specific activity greater than 0.5% that of the wild type enzyme. The dihydrofolate synthetase and folylpolyglutamate synthetase activities behaved similarly in all mutants, both retaining a large fraction of the wild type activity in the amino terminal deletions and both being very low in the carboxy terminal deletions and linker insertion mutants. These studies are consistent with a single catalytic site for the two activities catalyzed by this enzyme.     

Thymidine 5′-Monophosphate-Requiring Mutants of Saccharomyces cerevisiae Are Deficient in Thymidylate Synthetase

Thymidylate synthetase activity was measured in crude extracts of the yeast Saccharomyces cerevisiae by a sensitive radiochemical assay. Spontaneous non-conditional mutants auxotrophic for thymidine 5'-monophosphate (tmp1) lacked detectable thymidylate synthetase activity in cell-free extracts. In contrast, the parent strains (tup1, -2, or -4), which were permeable to thymidine 5'-monophosphate, contained levels of activity similar to those found in wild-type cells. Specific activity of thymidylate synthetase in crude extracts of normal cells or of cells carrying tup mutations was essentially unaffected by the ploidy or mating type of the cells, by the medium used for growth, by the respiratory capacity of the cells, by concentrations of exogenous thymidine 5'-monophosphate as high as 50 mug/ml, or by subsequent removal of thymidine 5'-monophosphate from the medium. Extracts of a strain bearing the temperature-sensitive cell division cycle mutation cdc21 lacked detectable thymidylate synthetase activity under all conditions tested. Its parent and another mutant (cdc8), which arrests with the same terminal phenotype under restrictive conditions, had normal levels of the enzyme. Cells of a temperature-sensitive thymidine 5'-monophosphate auxotroph arrested with a morphology identical to the cdc21 strain at the nonpermissive temperature and contained demonstrably thermolabile thymidylate synthetase activity. Tetrad analysis and the properties of revertants showed that the thymidylate synthetase defects were a consequence of the same mutation causing, in the auxotrophs, a requirement for thymidine 5'-monophosphate and, in the conditional mutants, temperature sensitivity. Complementation tests indicated that tmp1 and cdc21 are the same locus. These results identify tmp1 as the structural gene for yeast thymidylate synthetase.     

Isolation and properties of tRNA nucleotidyltransferase from wheat embryos.

From wheat embryos, tRNA nucleotidyltransferase (EC 2.7.7.25) was isolated. By chromatography on Sepharose 6B, DEAE-cellulose and affinity chromatography on tRNA-hydrazyl-Sepharose 4B, 7000-fold purification of the enzyme was achieved. The enzyme required for its activity Mg2+ or Mn2+ ion. ATP inhibited incorporation of CMP from CTP into lupin tRNA, and CTP acted as a competitive inhibitor of AMP incorporation from ATP. The regulatory role of ATP in incorporation of terminal CMP into tRNA is discussed. The incorporation of terminal CMP into tRNA deprived of terminal CCA or CA, was also studied.     

Mutants of Chinese hamster cells deficient in thymidylate synthetase

Stable mutants of Chinese hamster V79 cells deficient in thymidylate synthetase (TS; E.C. 2.1.1.45) have been selected from cultures grown in medium supplemented with folinic acid, aminopterin, and thymidine (FAT). After chemical mutagenesis, the frequency of colonies resistant to the "FAT" medium increased more than 100-fold over the spontaneous frequency. The optimal expression time of the mutant phenotype was 5-7 days after mutagen treatment. The recovery of FAT-resistant colonies in the selective medium was not affected by the presence of wild-type cells at a density below 9,000 cells per cm2. All 21 mutants tested exhibited thymidine auxotrophy; neither folinic acid nor deoxyuridine could support mutant cell growth. There was no detectable TS activity in all 11 mutants so far examined and only about 50% of wild-type activity in three prototrophic revertants, as measured by whole-cell and cell-free enzyme assays. The apparent Michaelis-Menten constant (Km) for deoxyuridine-5'-monophosphate and inhibition constant (Ki) for 5-fluoro-deoxyuridine-5'-monophosphate, measured by whole-cell enzyme assay, appear to be similar for the wild-type and revertant cell lines. Using 5-fluoro-[6-3H]-2'-deoxyuridine 5'-monophosphate as active site titrant, the relative amounts of TS in crude cell extract from the parental, revertant, and mutant cells were shown to exist in a 1:0.5:0 ratio. Furthermore, the enzymes from two revertants were more heat labile than that of V79 cells. These properties, taken together, suggest that the FAT-resistant, thymidine auxotrophic phenotype may be the result of a structural gene mutation at the TS locus. The availability of such a mutant facilitates studies on thymidylate stress in relation to DNA metabolism, cell growth, and mutagenesis.     

Defects of two temperature-sensitive lysyl-transfer ribonucleic acid synthetase mutants of Bacillus subtilis

Two temperature-sensitive mutants (lysS1 and lysS2) of the lysyl-transfer ribonucleic acid synthetase (l-lysine:tRNA ligase [adenosine 5'-monophosphate], EC 6.1.1.6) of Bacillus subtilis have been isolated. Although protein synthesis is inhibited in both mutants at the restrictive temperature (42 to 45 C), the mutants remain viable in a minimal medium. In comparison with the wild-type lysyl-tRNA synthetase, the l-lysine-dependent exchange of [(32)P]pyrophosphate with adenosine 5'-triphosphate (ATP) for both mutant enzymes is decreased. The lysS1 enzyme is completely defective in the ATP-dependent attachment of l-lysine to tRNA, whereas the lysS2 enzyme has 3- to 10-fold reduced levels of this activity. Temperature-resistant transformants have wild-type enzyme levels, whereas partial revertants to temperature resistance have varied levels of enzyme activity. The attachment and exchange activities of the lysS2 enzyme are more heat labile in vitro than the wild-type enzyme, as is the attachment activity of a partial revertant of the lysS1 mutant. The lysS1 and the lysS2 lysyl-tRNA synthetases have higher apparent K(m) values for lysine and ATP, in both the activation and the attachment reactions. The lysS2 enzyme has a V(max) for tRNA(lys) one-third that of the wild-type enzyme. Molecular weights of approximately 150,000 for the wild-type and lysS2 enzymes and approximately 76,000 for the lysS1 enzyme were estimated from sedimentation positions in sucrose density gradients assayed by the ATP-pyrophosphate exchange activity. We propose that the two mutations (lysS1 and lysS2) directly affect the sites for exchange activity, but indirectly alter attachment activity as a consequence of defective subunit association.     

High-level overexpression, rapid purification, and properties of Escherichia coli tRNA nucleotidyltransferase

The cloned Escherichia coli cca gene, described in the accompanying paper (Cudny, H., Lupski, J. R., Godson, G. N., and Deutscher, M. P. (1986) J. Biol. Chem. 261, 6444-6449), has been used to construct strains that overproduce tRNA nucleotidyltransferase, the enzyme that synthesizes the CCA terminus of tRNA. Strain UT481 (pEC4), which contains a 1.9-kilobase cca gene insert in plasmid pUC8, overproduces the enzyme by about 100-150-fold, probably under the control of the cca gene promoter. A second strain, containing a plasmid with a 1.5-kilobase insert, overproduces tRNA nucleotidyltransferase by about 650-fold, to a level of about 3-4% of the soluble cell protein. In this case, overexpression was dependent on the lac promoter of the plasmid. A rapid, two-step procedure was developed to purify large amounts of the enzyme from strain UT481 (pEC4) that was about 40% pure, free of ribonucleases, and suitable for use as a reagent for modification of tRNA molecules. Preparation of milligram quantities of homogeneous tRNA nucleotidyltransferase was accomplished by two further chromatographic steps. The structural and catalytic properties of this purified enzyme were similar to those from partially purified preparations previously described. The availability of large amounts of pure tRNA nucleotidyltransferase will not permit a variety of structural and functional studies of the enzyme that previously were not possible.     

Stereochemistry of internucleotidic bond formation by tRNA nucleotidyltransferase from baker's yeast.

Isomer A of adenosine 5'-O-(1-thiotriphosphate) (ATP alpha S) is a substrate for tRNA nucleotidyltransferase from baker's yeast, whereas isomer B is a competitive inhibitor. The tRNA resulting from this reaction has a phosphorothioate instead of a phosphate diester linkage at the last internucleotidic linkage between cytidine and adenosine. On limited digestion of this tRNA with RNase A, one can isolate cytidine 2',3'-cyclic phosphorothioate which can be deaminated to uridine 2',3'-cyclic phosphorothioate. It can be shown that this compound is the endo isomer and that, therefore, the phosphorothioate diester bond in the tRNA must have had the R configuration. This result indicates that no racemization during the condensation of ATP alpha S, isomer A, onto the tRNA had occurred. Whether inversion or retention of configuration had taken place awaits elucidation of the absolute configuration of isomer A of ATP alpha S.     

Enzymatic incorporation of ATP and CTP analogues into the 3' end of tRNA

Structural analogues of adenosine 5'-triphosphate and cytidine 5'-triphosphate were investigated as substrates for ATP(CTP):tRNA nucleotidyl transferase. Eight out of 26 ATP analogues and six out of nine CTP analogues were incorporated into the 3' terminus of tRNA. In general, for the recognition of the substrates the modification of the cytidine is less critical than is the modification of adenosine. An isosteric substitution on the ribose residue is possible in both CTP and ATP. The free hydroxyls of these triphosphates can be replaced by an amino group or hydrogen atom without loss of substrate properties. Modifications of positions 1, 2, 6, and 8 on the adenine ring of ATP are not allowed whereas modification on positions 2, 4 and 5 on the cytosine ring of CTP are tolerated by the enzyme. No differences can be observed in the substrate properties of ATP(CTP):tRNA nucleotidyl transferase isolated from different sources. Methods for preparation of tRNA species, which are shortened at their 3' end by one or more nucleotides, and analytical procedures for characterisation of these modified tRNAs are described.     

Salmonella typhimurium Mutants Defective in Cytidine Monophosphate Kinase (cmk)

Mutants of Salmonella typhimurium defective in cytidine 5'-monophosphate (CMP) kinase (cmk) have been isolated. The mutants also lack the ability to phosphorylate 2'-deoxyCMP, indicating that one enzyme is responsible for the phosphorylation of both CMP and deoxyCMP to the corresponding diphosphates. In glucose minimal medium the mutants grow at the same rate as the parental strain; however, they excrete large quantities of pyrimidines into the growth medium. Cytidine but not deoxycytidine has been identified among the excreted products. The mutant phenotype suggests that the physiological role of CMP kinase is that of rephosphorylating CMP arising from the breakdown of messenger ribonucleic acid. This proposed role of CMP kinase is supported by the fact that a cmk(-) mutant is much more sensitive to any partial impairment of cytidine 5'-triphosphate synthetase than is the cmk(+) parent strain. The gene cmk has been located on the Salmonella chromosome at 38.5 min. No markers which can be cotransduced with cmk by phage P22 have been found.     

Isolation and characterization of mutants at the APRT locus in the L-5178Y TK+/TK- mouse lymphoma cell line

2,6-Diaminopurine(DAP)-resistant mutants have been isolated from mouse lymphoma 5178Y TK+/TK- heterozygotes. In the presence of 50 microM DAP, two colony types were isolated. Small colonies contained 50% wild-type adenine phosphoribosyl transferase (APRT) activity (partial mutants), whereas large colonies have undetectable levels of APRT (aprt- mutants). aprt- mutants could be isolated following mutagenesis with ICR-191 or EMS from the partial mutants. Southern blot analysis of EcoRI digested wild-type DNA using a 3.1 kb mouse aprt genomic probe indicated sequence polymorphism at one or both EcoRI sites flanking the allele. Southern blot analysis of one of the partial mutants and one ICR-induced aprt- mutant (single step) indicated that both strains were hemizygous at the APRT locus. Such stable hemizygous strains would be useful in short-term mutagen tests.     

A mutation in Escherichia coli tRNA nucleotidyltransferase that affects only AMP incorporation is in a sequence often associated with nucleotide-binding proteins

Escherichia coli strain 5C15 contains a mutation in the cca gene that decreases AMP incorporation by tRNA nucleotidyltransferase while leaving CMP incorporation unaffected. Earlier studies of the purified mutant enzyme suggested that the mutation was localized to the AMP-incorporating site. In order to analyze this mutation in more detail, the cca gene from strain 5C15 was cloned into plasmid pUC8. Analysis of tRNA nucleotidyltransferase activity in extracts of a strain transformed with this plasmid demonstrated an elevated level of CMP incorporation, but low AMP incorporation, as expected from the properties of the original mutant. Sequence analysis of the mutant cca gene revealed only a single G to A point mutation leading to a glycine to aspartic acid substitution at position 70 of the peptide chain. The amino acid change was localized to one of two Gly-X-Gly-X-X-Gly sequences present in the protein. This sequence has been identified previously near the nucleotide-binding domain of various proteins, but it has not been noted in enzymes that incorporate nucleotide residues. However, other sequences often associated with ATP-binding domains are not found in tRNA nucleotidyltransferase. The implications of these findings for our understanding of nucleotide-binding domains are discussed.     

Nuclear-encoded mitochondrial tRNAs of Trypanosoma brucei have a modified cytidine in the anticodon loop.

The mitochondrial genome of Trypanosoma brucei does not appear to encode any tRNA genes. Isolated organellar tRNAs hybridize to nuclear DNA, suggesting that they are synthesized in the nucleus and subsequently imported into the mitochondrion. Most imported tRNAs have cytosolic counterparts, showing identical mobility on two-dimensional polyacrylamide gels. We have compared three nuclear-encoded mitochondrial tRNAs (tRNA(Lys), tRNA(Leu), tRNA(Tyr)) with their cytosolic isoforms by direct enzymatic sequence analysis. Our findings indicate that the primary sequences of the mitochondrial and the corresponding cytosolic tRNAs are identical. However, we have identified a mitochondrion-specific nucleotide modification of each tRNA which is localized to a conserved cytidine residue at the penultimate position 5' of the anticodon. The modification present in mature mitochondrial tRNA(Tyr) was not found in a mutant tRNA(Tyr) defective in splicing in either cytosolic or mitochondrial fractions. The mutant tRNA(Tyr) has been expressed in transformed cells and its import into mitochondria has been demonstrated, suggesting that the modified cytidine residue is not required for import and therefore may be involved in adapting imported tRNAs to specific requirements of the mitochondrial translation machinery.     

Spontaneous splicing mutations at the dihydrofolate reductase locus in Chinese hamster ovary cells.

We isolated and characterized three spontaneous mutants of Chinese hamster ovary cells that were deficient in dihydrofolate reductase activity. All three mutants contained no detectable enzyme activity and produced dihydrofolate reductase mRNA species that were shorter than those of the wild type by about 120 bases. Six exons are normally represented in this mRNA; exon 5 was missing in all three mutant mRNAs. Nuclease S1 analysis of the three mutants indicated that during the processing of the mutant RNA, exon 4 was spliced to exon 6. The three mutant genes were cloned, and the regions around exons 4 and 5 were sequenced. In one mutant, the GT dinucleotide at the 5' end of intron 5 had changed to CT. In a second mutant, the first base in exon 5 had changed from G to T. In a revertant of this mutant, this base was further mutated to A, a return to a purine. Approximately 25% of the mRNA molecules in the revertant were spliced correctly to produce an enzyme with one presumed amino acid change. In the third mutant, the AG at the 3' end of intron 4 had changed to AA. A mutation that partially reversed the mutant phenotype had changed the dinucleotide at the 5' end of intron 4 from GT to AT. The splicing pattern in this revertant was consistent with the use of cryptic donor and acceptor splice sites close to the original sites to produce an mRNA with three base changes and a protein with two amino acid changes. These mutations argue against a scanning model for the selection of splice site pairs and suggest that only a single splice site need be inactivated to bring about efficient exon skipping (a regulatory mechanism for some genes). The fact that all three mutants analyzed exhibited exon 5 splicing mutations indicates that these splice sites are hot spots for spontaneous mutation.