Practical aspects of including functional endpoints in developmental toxicity studies. Case study: immune function in HuCD4 transgenic mice exposed to anti-CD4 MAb in utero.

Keliximab is a human-cynomolgus monkey chimeric (Primatized) monoclonal antibody with specificity for human and chimpanzee CD4. As the preclinical safety assessment of biopharmaceuticals requires evaluation in pharmacologically responsive species, comprehensive toxicology studies, including reproductive toxicity, of this antibody were conducted in a human CD4 transgenic mouse model. The reproductive toxicology studies included a pre- and postnatal development study that incorporated immunotoxicological evaluation of offspring (F1) mice. The potential effects of exposure to treating maternal mice (F0) with keliximab during pregnancy and lactation on offspring viability, physical growth, neurobehavioral development, reproductive function, lymphoid tissue morphological structure, lymphocyte subsets and host resistance to Candida albicans infection were assessed. The results showed no impairment of these functions. The use of F1 transgenic mice in study with keliximab provides an example of a novel practical approach to assess developmental immunotoxicity within a study of pre- and postnatal development designed in accordance with ICH Guidelines.

Preclinical development of keliximab, a Primatized anti-CD4 monoclonal antibody, in human CD4 transgenic mice: characterization of the model and safety studies.

The preclinical safety assessment of biopharmaceuticals necessitates that studies be conducted in species in which the products are pharmacologically active. Monoclonal antibodies are a promising class of biopharmaceuticals for many disease indications; however, by design, these agents tend to have limited species cross-reactivity and tend to only be active in primates. Keliximab is a human-cynomolgus monkey chimeric (Primatized) monoclonal antibody with specificity for human and chimpanzee CD4. In order to conduct a comprehensive preclinical safety assessment of this antibody to support chronic treatment of rheumatoid arthritis in patients, a human CD4 transgenic mouse was used for chronic and reproductive toxicity studies and for genotoxic studies. In addition, immunotoxicity studies were conducted in these mice with Candida albicans, Pneumocystis carinii and B16 melanoma cells to assess the effects of keliximab on host resistance to infection and immunosurveillance to neoplasia. The results of these studies found keliximab to be well tolerated with the only effects observed being related to its pharmacologic activity on CD4+ T lymphocytes. The use of transgenic mice expressing human proteins provides a useful alternative to studies in chimpanzees with biopharmaceutical agents having limited species cross-reactivity.

Benzofuranyl ureas with potent cardiovascular teratogenicity in rats.

Studies of embryo-fetal development in rats were conducted with two 5-lipoxygenase inhibitors. SB-202235 (1,000 mg/kg/day) or SB-210661 (50, 100, or 500 mg/kg/day) was administered orally by gavage to female rats on days 6-17 postcoitus (pc) or days 7-16 pc. SB-202235 (1,000 mg/kg/day) and SB-210661 (100 mg/kg/day) reduced maternal body weight gain for the treatment period by 16% and 21%, respectively, relative to controls. SB-202235 (1,000 mg/kg/day) or SB-210661 (50 or 100 mg/kg/day), did not affect numbers of resorptions, dead or live fetuses/litter, but 500 mg/kg/day of SB-210661 caused 100% embryo lethality. SB-202235 (1,000 mg/kg/day) and SB-210661 (50 and 100 mg/kg/day) reduced fetal body weight by 15-30% and produced extensive cardiovascular malformations, as well as diaphragmatic hernias. SB-210661 also caused thymic abnormalities and cryptorchidism. Cardiovascular defects included abnormalities in aorticopulmonary septation, the aortic arch, pulmonary trunk, and ventricular septal defects are discussed relative to comparable human syndromes of cardiovascular malformation.

Quantitative mechanistically based dose-response modeling with endocrine-active compounds.

A wide range of toxicity test methods is used or is being developed for assessing the impact of endocrine-active compounds (EACs) on human health. Interpretation of these data and their quantitative use in human and ecologic risk assessment will be enhanced by the availability of mechanistically based dose-response (MBDR) models to assist low-dose, interspecies, and (italic)in vitro(/italic) to (italic)in vivo(/italic) extrapolations. A quantitative dose-response modeling work group examined the state of the art for developing MBDR models for EACs and the near-term needs to develop, validate, and apply these models for risk assessments. Major aspects of this report relate to current status of these models, the objectives/goals in MBDR model development for EACs, low-dose extrapolation issues, regulatory inertia impeding acceptance of these approaches, and resource/data needs to accelerate model development and model acceptance by the research and the regulatory community.

Developmental toxicity and toxicokinetics of two endothelin receptor antagonists in rats and rabbits.

Embryo-fetal development studies with toxicokinetic evaluations were conducted in rats and rabbits after oral or intravenous administration of two endothelin receptor antagonists. In the rat studies, females were administered SB-217242 (0.01-300 mg/kg/day) orally or SB-209670 (0.01-50 mg/kg/day) intravenously from days 6-17 postcoitus (pc). External and visceral fetal examinations were performed at necropsy on day 21 pc. Maternal body weight and food consumption were decreased only at 300 mg/kg/day SB-217242. Embryolethality was seen at 300 mg/kg/day SB-217242. Decreased fetal body weight occurred at 300 mg/kg/day SB-217242 and 50 mg/kg/day SB-209670. Dose-dependent increases in the mean percentage of fetuses per litter with malformations were seen at > or = 50 mg/kg/day SB-217242 and > or = 10 mg/kg/day SB-209670. Craniofacial, great vessel, heart, and thyroid were the predominant malformations. In the rabbit studies, females were administered SB-217242 (0.01-50 mg/kg/day) orally or SB-209670 (0.01-25 mg/kg/day) intravenously from days 6-20 pc. There was no drug-related effect on maternal body weight or food consumption. Embryolethality was observed at 50 mg/kg/day of SB-217242. Dose-related increases in the mean percentage of fetuses per litter with malformations were seen at > or = 10 mg/kg/day SB-217242 and > or = 10 mg/kg/day SB-209670. The malformations were similar to those observed in the rat studies, except that craniofacial development was not altered by SB-209670. The malformations observed are consistent with the pattern of endothelin-1 gene expression described in mouse embryonic pharyngeal arches and heart, and with the craniofacial and cardiovascular malformations observed in endothelin-1-deficient mice. Given the known role for endothelins in development, and concordant malformations in rats and rabbits observed in this study, teratogenicity is likely to be a class effect of endothelin receptor antagonists.

An evaluation of the novel selective estrogen receptor modulator, idoxifene, for effects on reproduction in rats and rabbits.

Idoxifene, a tissue-specific selective estrogen receptor modulator, was evaluated in male and female rats and female rabbits after oral administration for effects on fertility and/or embryo-fetal development. In all studies, adult toxicity was evident at doses >/=0.03 mg/kg/day in rats and >/=0.1 mg/kg/day in rabbits as evidenced by decreased body weight and/or food consumption. In the male fertility study, rats were treated with 0.003, 0.3, or 3.0 mg/kg/day for 64 to 68 days. Doses >/=0.3 mg/kg/day decreased seminal vesicle and prostate weights and impaired posttesticular sperm development, resulting in decreased epididymal sperm count and weight, but did not affect male fertility. In the female fertility study, rats were treated for 2 weeks prior to mating until insemination with 0.003, 0.03, or 3.0 mg/kg/day. Disrupted estrous cycles, impaired fertility, increased preimplantation loss, and increased vaginal fluid at necropsy were evident at >/=0.03 mg/kg/day. In the early embryonic development study, pregnant female rats were treated from days 0 to 6 postcoitus (pc) with 0.003, 0.03, or 3.0 mg/kg/day idoxifene. Partial or complete preimplantation loss was seen at 0.03 and 3.0 mg/kg/day, respectively. In the embryo-fetal development study, pregnant rats were treated from days 6 to 17 pc with 0.003, 0.03, or 3.0 mg/kg/day. At 3.0 mg/kg/day there was maternal lethality, excess vaginal fluid, embryo-fetal death, generalized fetal edema, and developmental delays. Excess vaginal fluid but no fetal effects were seen at 0.03 mg/kg/day. There were no treatment-related effects at 0.003 mg/kg/day in any rat reproduction study performed. In the rabbit embryo-fetal development study, pregnant New Zealand White rabbits were treated from days 6 to 20 pc with 0.01, 0.1, or 1.0 mg/kg/day idoxifene. At 1.0 mg/kg/day there was maternal lethality, vaginal or uterine bleeding, abortion/premature deliveries, and embryolethality. Vaginal or uterine bleeding was seen at 0.1 mg/kg/day. No treatment-related effects were observed at 0.01 mg/kg/day. Although systemic toxicity was evident in all the studies, the effects of idoxifene on rat and rabbit reproduction were considered to be due to the pharmacological activity of the compound.

Effect of prazosin on sperm transport in male rats.

Our objective was to investigate ejaculation and transport of sperm in the reproductive tract of male rats treated with an alpha-adrenergic receptor antagonist. Males were dosed (s.c.) with vehicle or 1.4 mg/kg prazosin. Sperm recovered in utero and ex vivo from the vas deferens and cauda epididymis were evaluated. Mating behavior and sperm motility were unaffected by prazosin. Prazosin treated males ejaculated fewer sperm (12.58 +/- 8.12 vs. 110.5 +/- 29.15 million), and the distal vas deferens contained fewer sperm (2.72 +/- 0.84 vs. 24.42 +/- 3.25 million) relative to controls. Prazosin-treated males had more sperm in the cauda epididymis relative to controls indicating inhibition of sperm transport to the vas deferens. These data demonstrate that inhibition of sperm transport from the cauda epididymis to the distal vas deferens is related to low ejaculate sperm counts in prazosin treated rats.

Spontaneous and induced alterations in the cardiac membranous ventricular septum of fetal, weanling, and adult rats.

Alterations of the cardiac membranous ventricular septum were studied using macrodissection, scanning electron and light microscopy of fetal, weanling, and adult Sprague-Dawley rats. Membranous ventricular septal defects (VSDs) were observed in 2.0% of fetuses on day 21 postcoitus (pc) but not in weanling or adult rats. The most common observation was a nonpatent depression in the membranous septum with an incidence of 38.1, 10.5, 4.3% for fetuses on days 17, 19, or 21 pc, respectively, 11.8% for weanlings, and 9.1% for adults. VSDs were characterized by a split in the endocardial cushion cells in the interventricular component of the membranous septum. Nonpatent depressions were characterized by a split in the endocardial cushion cells in the atrioventricular component of the septum, and they persisted postnatally as a blind-ended diverticulum directed above the tricuspid valve. The cardiovascular teratogens, trimethadione and trypan blue, produced in fetuses nonpatent depressions and VSDs morphologically similar to untreated fetuses. Maternal diet restriction (25% of controls) lowered fetal (day 21 pc) body weight by 47% but did not affect the incidence of ventricular septal alterations, suggesting that intrauterine growth retardation is not necessarily associated with alterations in the development of the ventricular septum. We conclude that neither VSDs nor nonpatent depressions in Sprague-Dawley rats affect postnatal survival and that VSDs close spontaneously during neonatal life.

Isolation of rat sperm from the vas deferens for sperm motion analysis.

A technique for isolation of viable spermatozoa from the rat vas deferens is presented. A single vas deferens from a mature Sprague-Dawley rat contains 35.4 +/- 3.3 million spermatozoa, and a 0.5-cm segment from the distal end of the vas will spontaneously expel 8.96 +/- 1.39 million sperm relatively free of cellular debris when placed into a balanced salt solution. Treatment of rats (n = 5) with alpha-chlorohydrin (100 mg/kg, single dose) significantly reduced, relative to controls (n = 5, water 10 mL/kg), sperm motility by 72%, swimming velocity by 44%, and head displacement motion by 50%, but did not affect the number of sperm in the vas deferens nor sperm recovery. In summary, the isolation of rat sperm from the vas deferens can be achieved with minimum handling in a manner that is not confounded by sperm toxicity.

Behavioral evaluations in developmental toxicity testing: MARTA survey results.

During 1991, the Middle Atlantic Reproduction and Teratology Association (MARTA) conducted a survey of laboratories performing behavioral evaluations as part of GLP developmental toxicity studies. This survey was conducted to determine the extent to which an "industry standard" had evolved for behavioral test batteries. The most commonly used developmental parameters were eye opening, pinna unfolding, and sexual maturation (physical landmarks); surface righting, pupil constriction, and nonautomated acoustic startle (reflexive landmarks). Locomotor activity was used by 80% of the laboratories. The majority (76%) of laboratories conducted at least one learning and/or retention evaluation; nearly one-fourth of the laboratories routinely performed two. The most common learning tests were watermaze (primarily a simple two-choice discrimination task) and passive avoidance. Automated startle paradigms (habituation, prepulse modification, and/or startle elicitation) were evaluated by 28% of the laboratories. This survey showed a remarkable similarity in methodology across laboratories and a progressive increase in the number of GLP studies that included behavioral assessments. The results indicate that behavioral tests have become a common component of developmental toxicity assessments of pharmaceuticals.

Toxicity of cadmium in the perfused human placenta.

Cadmium, a placental toxicant in rodents, was studied in the in vitro isolated dually perfused human placental lobule for periods of up to 12 hr to determine if cadmium can also be toxic in the human placenta. Placental lobules were perfused with a modified M199 medium containing 0, 10, 20, or 100 nmol of cadmium chloride/ml added only initially to the maternal perfusate. Every 4 hr, the perfusates in both the maternal and the fetal circuits were replaced with fresh perfusate containing no Cd. Measurements during perfusion were oxygen consumption, net fetal oxygen transfer, fetal pressure, fetal volume loss, glucose utilization, lactate production, human chorionic gonadotropin (hCG), and zinc transfer. Postperfusion, morphology, and tissue slice studies were performed to evaluate cellular metabolic function and uptake of an amino acid (alpha-[14C]-aminoisobutyric acid). In all cadmium experiments, there were no significant alterations in oxygen consumption, lactate production, glucose utilization, or amino acid uptake compared with controls; however, there were dose-related changes in the synthesis and release of the protein hormone, hCG, beginning within 4 hr of initial exposure to Cd. There were also dose-related volume loss from the fetal vasculature (greater than 6 ml/hr) and ultrastructural changes (subsyncytiotrophoblastic vesiculations, stromal edema, vacuoles in Hofbauer cells), with necrosis at 100 nmol Cd/ml occurring between 5 and 8 hr. Cadmium (10 nmol/ml) reduced the placental transfer of zinc into the fetal circuit. Thus, the human placenta is a site for toxic action of cadmium and is at least as sensitive as the rodent placenta to the actions of cadmium. In addition, these human studies demonstrated a selectivity in the toxic effects with a maintenance of carbohydrate metabolism and amino acid uptake even after 12 hr of exposure with placental Cd burdens of 151 +/- 37 nmol/g, but with the earliest (within 4 hr) dose-related functional alterations occurring in protein hormone production and zinc transfer followed by later changes in morphology with a tissue Cd burden of 46.5 +/- 4.0 nmol/g.

Evaluation of the teratogenic potential of octyl acetate in rats.

Octyl acetate (CAS RN 108419-32-5) was administered via oral gavage to pregnant Sprague-Dawley rats on Gestation Days 6 through 15 at dose levels of 0, 0.1, 0.5, and 1.0 g/kg. The dams were weighed and observed for clinical signs of toxicity during pregnancy, and food consumption was measured. On Gestation Day 20 the dams were sacrificed and the fetuses were examined for external, visceral, and skeletal malformations and variations. The mid- and high-dose levels resulted in maternal toxicity as evidenced by reductions in body weight gain and food consumption. There were no statistically significant effects on embryo-fetal lethality or fetal growth for any treatment group. The number of litters with at least one malformed fetus and the mean percentage of the litter malformed were significantly (p less than 0.05) elevated in the high-dose group only. The results of the present study demonstrate that octyl acetate produced some evidence of developmental toxicity at a dose (1.0 g/kg) that was maternally toxic. Developmental toxicity was not observed at the maternally toxic 0.5 g/kg dose level or the maternally nontoxic dose level (0.1 g/kg). Therefore, these data indicate that octyl acetate is not a selective developmental toxicant in the rat.

Implementation of a primary screen for developmental neurotoxicity.

A battery of tests to evaluate physical growth/development and neurobehavioral function was conducted with 78 litters of control Sprague-Dawley rats given purified water by intubation. The objectives of this study were to optimize test methods and to document the range and variability of the experimental endpoints. Data are presented for maternal evaluations (body weight gain, food consumption), gestation length, litter size, and postnatal survival. Pup body weight was used to assess postnatal somatic growth rate from birth to 85 days of age, while whole and regional brain weight measurements at 7, 28, and 85 days provided a more specific measure of physical growth relevant to a neurobehavioral study. Physical landmarks of development evaluated were pinna unfolding, incisor eruption, and eye opening while reflex landmarks of development evaluated were the negative geotaxis and pupillary reflexes. The mean percentage of litters acquiring a physical trait or reflex increased sigmoidally with age, and the data suggest that the potential to detect developmental delays would be optimal when ca. 90% of control litters reach the test landmark. Functional evaluations were arranged according to four testing subsets so that each litter was evaluated in each test (1 pup/sex/litter), but repeated testing on pups was minimized. Auditory and tactile startle reflexes, as well as prepulse inhibition, were measured at ages 22 days and 60-64 days and found to increase with age. A passive avoidance paradigm (age 40-43 days) was used to assess exploratory behavior (approach) and memory (avoidance). Swimming performance in a water maze was used to evaluate learning. In this test, escape times and error rates improved to their highest level by five or six trials and showed acceptable degrees of variability. Spontaneous motor activity was monitored for 23 hr at age 54-61 days to evaluate exploratory activity, photoperiod entrainment, and catecholamine-induced locomotion (amphetamine challenge). Finally, landmarks of sexual maturation (balanopreputial separation evident at 45 days of age, vaginal perforation evident at 33 days) and estrous cyclicity (4.8 cycles per 21 days) were evaluated as measures of reproductive neuroendocrine function. In sum, the test battery provided an efficient yet comprehensive screen for evaluating effects on physical growth/development and neurobehavioral function which meets practical criteria for preclinical testing of pharmaceutical agents.

Design of a primary screen for developmental neurotoxicity.

A battery of tests was devised for routine use as a primary screen for developmental neurotoxicity. The battery was divided into preweaning and postweaning tests using rats as subjects. Since the rat CNS is structurally incomplete at birth, preweaning tests were predominantly physical, using specific landmarks of somatic maturation and regional brain growth as indices of normal development. Pupillary responses to light and positional responses to gravity (negative geotaxis) were also included in the preweaning battery to monitor reflex behavior, a relatively simple CNS function. The postweaning battery predominantly contained functional tests to evaluate higher order behaviors that develop after completion of neurogenesis. The postweaning tests were divided into 4 subsets designed to evaluate (I) neuromuscular function, (II) memory, (III) problem solving, and (IV) neuroendocrine function, respectively. Curiosity, rhythmicity, patency of monoamine neurons, and physical measures of brain growth were included within the subsets so as to evaluate a spectrum of CNS functions. Preliminary findings suggest that tests and instrumentation selected for the proposed battery provide an informative, objective, comprehensive and cost-efficient means to screen for developmental neurotoxicity.

A comparison of developmental toxicity evident at term to postnatal growth and survival using ethylene glycol monoethyl ether, ethylene glycol monobutyl ether and ethanol.

This study was designed to compare an abbreviated evaluation of uterine contents at term (teratology probe) with a modified Chernoff-Kavlock assay (postnatal study), [Chernoff N, Kavlock RJ: J Toxicol Environ Health 10:541-550, 1982]. Mice were intubated during gestation and were evaluated for signs of toxicity. In the teratology probe, uterine contents were examined at term. In the postnatal study, offspring were examined and weighed through day 22 postpartum. Ethylene glycol monoethyl ether (EGEE) produced embryo lethality and malformations, and decreased fetal weight at a dose level which was not maternally toxic in the teratology probe. In the postnatal study, EGEE decreased litter size and neonatal body weight; while litter size continued to decrease beyond the neonatal period, body weights of surviving pups were not significantly different from control. Pups exposed prenatally to EGEE developed kinked tail which was not apparent in fetuses or neonates. Maternally toxic dose levels of ethylene glycol monobutyl ether and ethanol were associated with increased embryo lethality in teratology probe studies. In postnatal studies, there were no significant effects on pup growth or survival at maternally toxic dose levels. Preliminary conclusions regarding maternal and developmental toxicity were comparable based on the teratology probe or postnatal study. Both assays measure litter size and offspring weight, but the teratology probe measures resorption incidence which may be a more sensitive index of prenatal death than number of live born. Neither fetal weight nor neonatal weight reliably predict permanent alteration of growth. A postnatal study permits detection of internal malformations or functional defects which reduce postnatal survival and gross abnormalities which appear postnatally.

Actions of the anticodon arm in translation on the phenotypes of RNA mutants.

In previous publications, we have shown that it is practical to study the translational activity of tRNAs by replacement and alteration of the anticodon arm sequence of the genus on a plasmid clone. Experiments in which the anticodon arm sequence is transplanted between tRNA genes suggest that the translational activity is determined by these sequences. We have therefore made every variant of the anticodon loop and the three base-pairs of the stem proximal to the loop, in order to resolve the relation between the structure of Su7Am tRNATrp, and its function. All derivatives conserved the normal secondary structure of the molecule, which was known to be essential for translational activity. The probability of translation of the amber codon by these suppressors is measured in this work. This translational activity in vivo is rationalized in terms of data on the copy numbers of the plasmid clones, the nucleotide modifications of the tRNAs, the steady-state level of the mature tRNA, and the aminoacylation of these molecules. Nucleotide modification levels vary among these tRNAs, giving information about the specificities of modification systems that make O-methylribose, pseudouridine, and modified A in the anticodon arm. However, for this series of tRNAs, none of these modifications has a strong effect on translational efficiency of the tRNAs. A few of the substitutions reduce aminoacylation of the tRNAs with glutamine, as determined by comparison of suppression in normal strains and related strains, which have 25-fold elevated levels of the glutaminyl-tRNA synthetase (GlnRS). The substitutions that have the largest effect on GlnRS action are, unexpectedly, purines for conserved pyrimidines on the 5' side of the anticodon loop. Data on the concentrations of tRNA in vivo suggest that the anticodon loop and helix contribute similarly to the determination of the steady-state level of the tRNAs. This level varies sevenfold, though all tRNAs are processed from a homologous precursor made from the same transcription unit. Effects on levels appear to be mediated by changes in anticodon arm structure. A robust equation that relates aminoacyl-tRNA levels to suppressor efficiency is developed in order to resolve effects on tRNA levels and on ribosomal steps: E = A/(K + A), where E is efficiency, A is aminoacyl-tRNA concentration, and K is the effective concentration, or cellular tRNA content required for an individual tRNA to have an efficiency of 0.50. The tRNAs vary in their intrinsic ability to function on the ribosome (represented by K), after other influences have been normalized.(ABSTRACT TRUNCATED AT 400 WORDS)

The translational efficiency of tRNA is a property of the anticodon arm.

We have reciprocally transplanted the anticodon arm sequences of a set of amber suppressor tRNA genes, using recombinant DNA techniques. By this means, a very efficient suppressor may be converted to a poor one, and the poorest tRNA to the efficiency of the best one. In tRNA molecules of normal 2 degrees and 3 degrees structure, the suppressor efficiencies of different composite tRNAs having the same anticodon arm sequence are approximately the same. Large numbers of simultaneous changes throughout the rest of the molecule do not affect the efficiency. Selective nucleotide modification as a result of varied anticodon arm sequences cannot explain these efficiencies. Efficiencies are also unlikely to differ because of selective aminoacylation. Measurement of in vivo tRNA shows, however, that tRNA levels do vary if the anticodon arm sequence is changed. If tRNA levels are normalized, the anticodon arm effect on the translational efficiency remains. Therefore, different anticodon arms, all of normal secondary structure, are not equivalent in translation. The most efficient sequences in this series resemble those found in natural tRNAs associated with similar anticodons, as is proposed in the extended anticodon theory (Yarus, M. (1982) Science 218, 646-652). These molecules also provide some information on the specificity of nucleotide modification enzymes and on determinants of the steady-state tRNA level.

Construction of a systematic set of tRNA mutants by ligation of synthetic oligonucleotides into defined single-stranded gaps.

A series of mutant tRNA genes has been constructed by site-directed mutagenesis in pOP203, a colE1 derivative carrying a transcription unit under control of the lacUV5 promoter. These mutant genes include all possible amber suppressing variants of tRNATrp with single nucleotide substitutions at anticodon loop positions 32, 37, and 38 (numbered from the 5' end), and all possible paired base substitutions in the three base pairs nearest the anticodon loop. G at position 38 was not recovered as a single mutation, but rather in conjunction with an undirected mutation to T at position 32. The singly mutated G38 tRNA may not be active, though all the other tRNA derivatives are functional in the translation of amber codons. To construct the mutants, we ligated a synthetic deoxyoligonucleotide into a precisely formed single-stranded gap covering the anticodon arm region DNA, in an otherwise double-stranded fragment containing the tRNATrp gene. The resulting heteroduplex was then ligated into the plasmid and introduced into Escherichia coli. This method of mutagenesis is simple, reproducible, and highly tolerant of varying degrees of heteroduplex in the gap, variations in temperature of ligation, and changes in the oligonucleotide concentration. Mutagenesis does not require a 5'-phosphorylated oligonucleotide. These qualities suit the gap method for intensive study of a region by site-directed mutagenesis.