Salivary prolactin as a marker for central serotonin turnover.

Central nervous system (CNS) serotonin deficits have been linked to many pathological behaviors in both human and nonhuman primates. The plasma prolactin response to fenfluramine has been widely used to assess CNS serotonin functioning in humans. Prolactin is also found as an integrated measure in saliva. We hypothesized that salivary prolactin concentrations would correlate positively with cerebrospinal fluid (CSF) 5-hydroxyindoleacetic acid (5-HIAA) in rhesus monkeys. Twenty-seven adult male and female rhesus macaques (Macaca mulatta) were sampled for concurrent saliva, blood, and CSF. Saliva and blood serum were assayed for prolactin concentrations, and CSF was assayed for 5-HIAA, homovanillic acid (HVA), and 3-methoxy-4-hydroxyphenylglycol (MHPG). Salivary prolactin concentrations were positively correlated with CSF 5-HIAA concentrations. No other relationships between any of the measures, including that between salivary prolactin and serum prolactin, were found to be statistically significant. These findings suggest the possibility of using salivary prolactin concentrations as an index of CNS serotonin turnover in humans.

A replication study of violent and nonviolent subjects: cerebrospinal fluid metabolites of serotonin and dopamine are predicted by plasma essential fatty acids.

BACKGROUND: Among an independent group of subjects selected for their history of violent, impulsive behaviors and nonviolent control subjects, we attempted to replicate the finding that plasma docosahexaenoic acid concentrations were negatively correlated with cerebrospinal fluid 5-hydroxyindoleacetic acid (CSF 5-HIAA) concentrations. METHODS: CSF 5-HIAA and homovanillic acid (HVA), fasting total cholesterol, and plasma fatty acid concentrations were examined in violent and nonviolent subjects matched for their severity of alcohol dependence. RESULTS: Violent subjects had significantly higher lifetime violence and hostility ratings and lower concentrations of CSF 5-HIAA than nonviolent subjects. Plasma docosahexaenoic acid was negatively correlated with CSF 5-HIAA only among violent subjects. CONCLUSIONS: This observational study suggests that dietary essential fatty acids may change neurotransmitter concentrations. Prospective dietary intervention trials will be required to determine if increasing dietary intake of docosahexaenoic acid will increase or decrease either CSF 5-HIAA concentrations or impulsive and violent behaviors.

The suitability of [11C]-alpha-methyl-L-tryptophan as a tracer for serotonin synthesis: studies with dual administration of [11C] and [14C] labeled tracer.

The tracer [11C]-alpha-methyl-L-tryptophan (alphaMTP) has been used to measure brain serotonin synthesis rates with positron emission tomography (PET). To address questions about the accuracy of the kinetic model, [14C]alphaMTP was used to directly measure conversion to [14C]-alpha-methyl-serotonin (alphaM5HT) in monkeys that had been previously studied with PET and [11C]alphaMTP. Four male, fasted, isoflurane-anesthetized rhesus monkeys were studied with [11C]alphaMTP and PET. Immediately after the initial 3-hour scan, a second dose of [11C]alphaMTP was coinjected with 1 mCi of [14C]alphaMTP, and additional PET data were collected. Approximately 90 minutes after the second alphaMTP administration, the animals were killed with an overdose of phenobarbital, and brain samples from 21 regions were taken and analyzed by HPLC. Minimal conversion of alphaMTP to alphaM5HT occurred; HPLC analysis of 14C radioactivity showed that greater than 96% of the total counts were in fractions corresponding to the alphaMTP peak. Brain concentrations of serotonin, tryptophan, 5-hydroxyindole-3-acetic acid, and alphaMTP also were determined fluorometrically using external quantification. Patlak plots generated from PET images acquired over 3 hours showed no time period of linear increase, and final slopes were not significantly different from zero, consistent with the finding of minimal conversion to [14C]alphaM5HT. These data indicate that in the 3-hour period after injection, [11C]alphaMTP is acting predominantly as a tracer of tryptophan uptake, not serotonin synthesis.

Plasma total cholesterol concentrations do not predict cerebrospinal fluid neurotransmitter metabolites: implications for the biophysical role of highly unsaturated fatty acids.

Low concentrations of a metabolite of serotonin found in cerebrospinal fluid (CSF), 5-hydroxyindolacetic acid (5-HIAA), are strongly associated with suicidal and violent behaviors. Although lowering of plasma total cholesterol has been suggested to increase mortality from suicide and violence by decreasing concentrations of CSF 5-HIAA via changes in membrane biophysical properties, highly unsaturated fatty acids may play a more important role. Violent and nonviolent comparison groups, early- and late-onset alcoholics, and healthy comparison subjects were studied to control for alcohol use and predisposition to violence. Fasting plasma total cholesterol and CSF were assayed under stringently controlled conditions. When all groups were combined (n = 234), plasma cholesterol concentrations had a weak positive correlation with CSF 5-HIAA (r = 0.18, P < 0.01). However, age correlated with both plasma total cholesterol and CSF 5-HIAA concentrations. When age was included in multiple regression models, the correlation between cholesterol and CSF 5-HIAA concentrations was not significant. Cholesterol correlated weakly with CSF 5-HIAA concentrations only in late-onset alcoholics after age was controlled for, but the relation was not significant after correction for multiple testing. CSF homovanillic acid did not correlate with plasma total cholesterol in any group. Plasma total cholesterol had no apparent relation to CSF neurotransmitter metabolites in any group of subjects. Highly unsaturated essential fatty acids, which are also critical determinants of membrane biophysical properties and may be linked to brain serotonin concentrations, should also be considered in studies examining the effect of lowering fat intake on the incidence of suicide and violence.

Brain serotonin synthesis rates in rhesus monkeys determined by [11C]alpha-methyl-L-tryptophan and positron emission tomography compared to CSF 5-hydroxyindole-3-acetic acid concentrations.

Twelve male, fasted, anesthetized rhesus monkeys were studied with positron emission tomography (PET) and [11C]alpha-methyl-L-tryptophan (alpha MTP) to determine serotonin synthesis rates as described by Diksic et al. (1991). It was expected that the serotonin synthesis rates determined for the whole brain would be correlated with CSF 5-hydroxyindole-3-acetic acid concentrations, a measure of central serotonin turnover, because both measures were obtained at steady state. However, no significant correlation was found. During data analysis, it was noticed that the calculated serotonin synthesis rates were significantly correlated to free plasma tryptophan (TP) concentrations (r = 0.88, p < .001). From repeat scans conducted in six monkeys, it was determined that day-to-day variability in free plasma TP and the percentage of protein binding (average percent difference was 48 and 37%, respectively) produced most of the variability in the calculated serotonin synthesis rates (50%); repeat K images, obtained from the PET data alone, differed by only 11%. Calculated serotonin synthesis rates reported for [11C]alpha MTP PET studies of humans (Nishizawa et al. 1997) and dogs (Diksic et al. 1991) were also highly correlated to reported differences in plasma free TP concentrations. It seems that the [11C]alpha MTP model for the computation of serotonin synthesis rates is very dependent on plasma free TP concentration and that it may not accurately determine actual serotonin synthesis rates.

Seasonal variation in CSF 5-HIAA concentrations in male rhesus macaques.

Seasonal changes in cerebrospinal fluid (CSF) 5-hydroxyindoleacetic acid (5-HIAA) concentrations were assessed on multiple occasions in 103 free-ranging male rhesus macaques (Macaca mulatta). At the time of sampling subjects ranged between the ages of 2 and 6 years. CSF samples were collected between the hours of 0900 and 1600 throughout the Fall, Winter, and Spring from 1990 through 1994. Data were analyzed in a general linear mixed model with random intercepts. Results indicated that CSF 5-HIAA concentrations decreased with age. CSF 5-HIAA concentrations were significantly increased in the Fall (October and November), which is the height of the breeding season, with no evidence of differences between Winter and Spring. There was also some evidence that the seasonal variation in CSF 5-HIAA concentrations was blunted in younger, more immature subjects.

Rapid changes in cytochrome P4502E1 (CYP2E1) activity and other P450 isozymes following ethanol withdrawal in rats.

This study describes the effects of chronic ethanol (ETOH) treatment and withdrawal on the rat hepatic mixed-function mono-oxygenase system. Male Sprague-Dawley rats (150-200 g, 10 per group) were administered ETOH as part of the Lieber-deCarli liquid diet for 3 weeks. Ethanol was removed, and the animals were euthanized at 0, 24, 48, 72 and 168 hr post-withdrawal. Microsomes were prepared, and ethanol-inducible cytochrome P4502E1 (CYP2E1) activity was measured using the enzyme markers N-nitrosodimethylamine demethylase (NDMAd), p-nitrophenol hydroxylase (PNPH) and aniline hydroxylase (AH). Activities were found to be induced significantly after chronic ETOH feeding using all three assays (NDMAd, 5-fold; PNPH, 3.5-fold; AH, 9-fold). Upon ETOH withdrawal, all three activities dropped markedly, with NDMAd and PNPH at control values at 24 hr and all subsequent time points. AH activity remained 3-fold higher than controls at 24, 48 and 72 hr. Western blot analyses showed that immunoreactive CYP2E1 returned to control at 24 hr, consonant with NDMAd and PNPH activities. The prolonged induction of AH activity following ETOH withdrawal indicates that it is not a specific marker of CYP2E1-catalyzed reactions. Collectively, these data are suggestive of a rapid mechanism of CYP2E1 degradation in the rat liver. Of the other parameters investigated in this study, total cytochrome P450 content was increased 2.5-fold after ETOH feeding, with levels dropping markedly 24 hr post-withdrawal. NADPH-dependent cytochrome c reductase activity was unchanged throughout the course of the study. CYP1A1, CYP2B1 and CYP3A activities were assessed by the substrate probes ethoxyresorufin O-dealkylase (EROD), pentoxyresorufin O-dealkylase (PROD) and erythromycin N-demethylase (ERNd). EROD and PROD were induced significantly by ETOH administration (2-fold) at 0 hr, with EROD remaining elevated over controls 24 hr post-withdrawal. Quantitative western blot analysis of CYP1A1 and CYP2B1 revealed a pattern of immunostaining generally consistent with but less variable than levels predicted by the respective substrate markers. Both proteins were induced significantly by chronic ethanol administration (CYP1A1, 1.9-fold; CYP2B1, 4-fold). Induction of these P450 isoforms persisted for several days following withdrawal. In contrast, immunoreactive CYP1A2 was found to decrease significantly (by 30-40%) during ethanol withdrawal (24, 48, 72, 168 hr). ERNd activity was induced significantly by chronic ETOH feeding (2.5-fold) and remained so for 24 hr into the withdrawal period (2-fold). Immunoreactive CYP3A1 was also induced significantly following ETOH administration (0 hr) and 24 hr following withdrawal.(ABSTRACT TRUNCATED AT 400 WORDS)

Ondansetron modulates pharmacodynamic effects of ketamine on electrocardiographic signals in rhesus monkeys.

Electrocardiographic signal dynamics were examined in rhesus monkeys (Macaca mulatta) before and after treatment with ketamine and/or ondansetron. Ketamine exerts differential pharmacodynamic effects on behavior in animals stratified according to a measure of central serotonergic turnover. We hypothesized that measures of serotonergic turnover might explain some of the variance in the electrocardiographic (ECG) response to ketamine. Electrocardiographic recordings of animals were obtained at baseline, after administration of either saline or ondansetron (0.125 mg/kg), and after administration of ketamine (15 mg/kg). Electrocardiographic signal dynamics were measured using an algorithm that extracts the Hurst parameter (H) of the interbeat interval (IBI) time-series. H decreased after ketamine administration, (mean+/-S.E.M.), 0.33+/-0.04 vs. 0.12+/-0.02, P

Measurement of lingual and palatine somatosensory evoked potentials.

OBJECTIVE: A technique is presented for generating and recording lingual and palatine nerve somatosensory evoked potentials (SEPs). METHODS: Pairs of thin, stainless steel disk electrodes were mounted onto mandibular or maxillary acrylic splints, similar to orthodontic retainers. Mandibular splint electrodes were oriented to contact the under surface of the tongue along the course of the right and left lingual nerves and maxillary splint electrodes were oriented to contact the hard palate bilaterally along the course of the palatine nerves. SEP recording electrodes were placed on the scalp 1 cm posterior to C5 and C6 (C5' and C6', respectively) using the combinatorial nomenclature of the International 10-20 system. Two reference electrode locations, Fz and C5' or C6', over the cortical hemisphere opposite that of the recording electrode, were used. RESULTS: Right and left lingual and palatine nerve SEPs were recorded from five normal adults. SEP latencies were similar to the N13 and P18 cortical peak latencies recorded in previous studies of trigeminal nerve branches to the lips regardless of reference electrode position. CONCLUSIONS: A more precise method of stimulating the intraoral lingual and palatine nerves was accomplished using dental splints. SEPs were easier to obtain using a contralateral cortex reference electrode location.

Effects of a single oral dose of desipramine on postoperative morphine analgesia.

Drugs that block norepinephrine reuptake offer promise as opioid potentiators, because norepinephrine mediates opioid analgesia but not side effects such as sedation or nausea. In a two-by-two factorial design, we randomized 62 inpatients with pain following major surgery to receive either desipramine, 50 mg by mouth, or placebo at 6 a.m. on the first day after surgery. At their first request of pain medication after 8 a.m., they were given intravenous morphine, either 0.033 mg/kg or 0.10 mg/kg. Pain relief and side effects were assessed for 4 hr; peak relief on the visual analog scale (VAS) was the primary outcome variable. Pain relief, side effect scores, and time to remedication were significantly greater with the higher dose than with the lower dose of morphine, verifying assay sensitivity, but desipramine pretreatment did not significantly enhance morphine analgesia. The mean increase in peak VAS relief score after desipramine pretreatment, relative to placebo, was 6%; the 95% confidence interval for this estimate ranged from a 21% reduction to a 34% increase in pain relief. These results differ from a previous report that 1 week of pretreatment with desipramine, 75 mg per day, potentiated postoperative morphine analgesia. We conclude that if desipramine potentiation of opioid analgesia occurs in humans, its demonstration may require higher doses or chronic treatment.

Distribution of free and conjugated dopamine in rats and mice.

Free and conjugated dopamine occurs in many tissues and fluids of mammals. We have used reversed phase liquid chromatography with electrochemical detection in sequence with an alumina adsorption procedure to separate and then detect amounts of free and conjugated dopamine in the major visceral organs and brain of the rat and mouse. We report those results here. Values for free and conjugated dopamine are given for all tissues studied. Conjugated DA was found in the rat kidney, liver and adrenal gland and in the kidney of the mouse. Those tissues which have a significant amount of conjugated dopamine will be studied further for the presence of dopamine sulfate.

Vascular reactivity in alcoholic rat aortas: in vitro interactions between catecholamines and alcohol.

To understand the nature of vascular problems of hypertensive and alcoholic subjects, an in vitro interaction between catecholamine and alcohol was investigated in male Sprague-Dawley rats. The descending aortas were isolated from either the control or alcohol treated rats. The aortic strips were cooled rapidly in ice-chilled Krebs-Henseleit (K-H) solution and the aorta was cut into a series of rings of approximately 1 mm width. The rings were fixed under 1 g of resting tension between a force-displacement transducer and anchoring electrodes at 37 degrees C. The rings were equilibrated for an hour with frequent changes of the K-H solution before testing. There was norepinephrine (NE) dose-dependent contraction of the rings, which showed the maximum tension with 1 microM NE. Acetylcholine or carbachol produced slow relaxation. Pretreatment of the rings with 10 microM prazocin prevent the contraction induced by 1 microM NE. Even in the presence of prazocin, 60 mM KCI was able to generate the maximum tension. NE-induced contraction was analyzed in the control K-H solution or in the presence of 1.5, 3.0 and 5.0% of ethanol. Ethanol (1.5%) slowed the rate of rise of the aortic tension without a remarkable compromise of the maximum tension. But, there was a significant reduction in contraction with 3% ethanol. A complete inhibition of contraction was noted with 5% ethanol. However, the effect of ethanol was fully reversible upon washings the aortic rings with K-H solution. Aortic rings prepared from the rats that were fed with alcohol for 30 days were not able to generate the maximum tension with 1 microM NE.(ABSTRACT TRUNCATED AT 250 WORDS)

Inverse relationship between plasma epinephrine and testosterone levels during acute glucoprivation in healthy men.

In healthy men, a decrease in plasma testosterone levels was observed in the context of metabolic stress. While physiological mechanisms underlying this response are unclear, there are several lines of evidence suggesting circulating epinephrine's influence on plasma testosterone levels. The purpose of this study was to directly relate stress-induced changes in plasma testosterone and epinephrine. The stressor used was blockade of glucose metabolism with pharmacological doses (40 mg/kg) of 2 deoxyglucose (2DG). Arterial plasma samples from 10 healthy males were assayed at 20 minutes intervals for 60 minutes for the concentrations of testosterone, epinephrine and related biochemicals. Bolus administration of 2DG resulted in progressive decline in testosterone and increases in epinephrine and norepinephrine plasma levels (mean change from baseline: 29, 2530 and 186%, respectively). Inverse correlation was detected between both absolute (r(s)=-0.72; df=8; p=0.017) and baseline-corrected testosterone concentrations at the 60 minute time point and epinephrine area under the curve values. Our results suggest that adrenomedullary activation may be involved in stress-induced testosterone effects. The implications of these data for the understanding of the role of catecholamines in glucoprivic stress response are discussed.

Serotonin, testosterone and alcohol in the etiology of domestic violence.

In a previous study we administered the panicogenic agent sodium lactate to a select group of perpetrators of domestic violence and comparison groups. Results of that study showed that perpetrators exhibited exaggerated lactate-induced fear, panic and rage. In this current study, we compared the cerebral spinal fluid (CSF) concentrations of 5-hydroxyindoleacetic acid (5-HIAA) and testosterone obtained from perpetrators of domestic violence and a group of healthy comparison subjects. All subjects were assessed for DSM-III-R diagnoses. Perpetrators with alcohol dependence (DV-ALC) (n=13), perpetrators without alcohol dependence (DV-NALC) (n=10) and healthy comparison subjects (HCS) (n=20) were clinically assessed using the Spielberger Trait Anxiety, Brown-Goodwin Aggression Scale, Buss Durkee Hostility Inventory and Straus Conflict Tactics. Following an overnight fast and bed rest, subjects received a lumbar puncture to obtain CSF concentrations of 5-HIAA and testosterone. Perpetrators scored significantly higher on measures of aggression than HCS. DV-NALC had significantly lower concentrations of CSF 5-HIAA and higher Straus Conflict Tactics (CT) physical violence scores than DV-ALC and HCS. DV-ALC had significantly higher concentrations of CSF testosterone than DV-NALC. DV-ALC also had significantly higher Straus CT physical violence scores than HCS. DV-NALC and DV-ALC differed on 5-HIAA concentrations, testosterone concentrations, Straus CT physical violence scores and alcohol dependence. These results suggest that DV-NALC and DV-ALC groups could have different biological mechanisms mediating domestic violence.

Pharmacokinetics of sulfadiazine/trimethoprim in neonatal male calves: effect of age and penetration into cerebrospinal fluid.

Sulfadiazine (SDZ)/trimethoprim (TMP; 30 mg of SDZ/TMP/kg of body weight) was given IV to the same 6 male calves at 1, 7, and 42 days of age and to 2 additional calves at 7 days of age. Serum concentrations of SDZ and TMP were best represented by a 2-compartment open model, but in 42-day-old calves, CSF concentrations of both drugs were best represented by a 1-compartment open model with first-order input. Between 1 and 42 days of age, the elimination half-life (t1/2(beta)) of SDZ decreased from 5.7 to 3.6 hours, and total body clearance (CLtot) increased from 1.43 to 1.88 ml/min/kg; the area under the curve (AUC0----infinity) decreased from 291.5 to 225.4 mg/L.h. The distribution coefficient (Vd(area)/kg of body weight) decreased with age, changing from 0.72 to 0.59 L/kg, between 1 and 42 days of age. Therapeutic concentrations of SDZ in serum (greater than 2 micrograms/ml) were maintained for 24 hours in 1-day-old calves and for about 15 hours in 7- and 42-day-old calves. The elimination rate of TMP increased about 9-fold; t1/2(beta) was 8.4, 2.1, and 0.9 hours, respectively, at 1, 7, and 42 days of age. Other values also reflected an increase in TMP elimination rate with age: CLtot increased from 2.8 to 12 to 28.9 ml/min/kg, k13 increased from 0.336 to 0.654 to 1.664/h and AUC0----infinity decreased from 32.8 to 7.9 to 3.1 mg/L.h, respectively. Therapeutic concentrations (greater than 0.1 microgram/ml) were maintained for 15 hours, 8 hours, and about 6 hours in 1-, 7-, and 42-day-old calves, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)

Interaction of ethanol and smoking on the pharmacokinetics and pharmacodynamics of psychotropic medications.

Ethanol and smoking are known to affect the pharmacokinetics and pharmacodynamics of many medications. This review focuses on the interactions between ethanol and smoking and psychotropic medications, in particular, antidepressants, anxiolytics, and neuroleptics. A discussion of the various mechanisms by which ethanol and smoking may exert their effects precedes a summary of specific interactions.

Pharmacokinetics of alpha-methyl-L-tryptophan in rhesus monkeys and calculation of the lumped constant for estimating the rate of serotonin synthesis.

We have determined several kinetic and pharmacokinetic parameters of L-tryptophan (Trp) and alpha-methyl-L-tryptophan (alphaMTrp) in the rhesus monkey from which the lumped constant for the alphaMTrp method of estimating serotonin synthesis rates is calculated. AlphaMTrp was isolated from DL-alphaMTrp using a chiral separation column with high performance liquid chromatography. AlphaMTrp (50 microgram/kg) was administered i.v. to four adult male rhesus monkeys and arterial blood samples were collected for a 4-hr period. Plasma concentrations, as determined by high performance liquid chromatography with electrochemical detection, were best fitted by a tri-exponential equation. Plasma protein binding of Trp and alphaMTrp was determined by measuring concentrations in ultrafiltrates obtained at 30 degrees C. After a 2-hr adjusted rate infusion of alphaMTrp designed to establish steady-state plasma concentrations, three adult male rhesus monkeys were killed by exsanguination with perfused ice-cold saline. Brain/arterial plasma concentration ratios of Trp and alphaMTrp and the Michaelis-Menten parameters for tryptophan hydroxylase, EC 1.14.16.4, with Trp and alphaMTrp as initiating substrates, were determined for seven brain regions. The lumped constants determined for the different brain regions were not significantly different from each other and indicate, that for modeling purposes, the brain may be treated as a homogeneous area and the lumped constant given a single value, 0.18 +/- 0.05.

The effect of age and diet on sulfadiazine/trimethoprim disposition following oral and subcutaneous administration to calves.

Thirty milligrams per kilogram of sulfadiazine/trimethoprim (SDZ/TMP, Tribrissen) was given orally and subcutaneously (s.c.) to two groups of male, Holstein calves. One group was fed milk-replacer throughout the 13-week period of the study while the second group was weaned onto a chopped grain-fiber mixture when 5 weeks old. Serum and urine were assayed for concentrations of unchanged drug. Trimethoprim bioavailability, following oral administration at 1, 6 and 12 weeks of age, is higher in milk-fed calves (non-ruminants) than in grain-fiber-fed calves (ruminants); bioavailability decreases with increasing age in both groups of calves. Serum concentrations above 0.1 micrograms/ml (the level of sensitivity of the assay) could not be obtained in ruminating calves. The rate of SDZ absorption following oral administration, as determined by the Wagner-Nelson method, was very slow in all the calves in this study with average half-life values ranging from 8.2-12.67 h; absorption was slightly faster in ruminating calves. Absorption of SDZ is rate-limiting and determines the biological half-life of the drug; SDZ serum concentrations above 2 micrograms/ml were maintained in all calves for at least 24 h. Following s.c. administration of Tribrissen to 7-and 13-week-old calves, urinary excretion patterns indicated that TMP was slowly released from the injection site; serum concentrations were below 0.1 micrograms/ml. In contrast, absorption of SDZ was very rapid; values for tmax were 1.5-1.8 h. The pharmacokinetic parameters for SDZ were calculated according to a one-compartment open model; neither diet nor age had a significant effect on SDZ disposition following s.c. injection. Subcutaneous administration of 30 mg/kg Tribrissen, b.i.d., may be the best therapeutic regimen; even though measureable concentrations of TMP cannot be achieved in the serum following a single s.c. dose, TMP concentrations should accumulate and, because of its sustained release, provide almost continual potentiation of SDZ.

Application of reaction-diffusion models to cell patterning in Xenopus retina. Initiation of patterns and their biological stability.

We have examined the behavior of two reaction-diffusion models, originally proposed by Gierer & Meinhardt (1972) and by Kauffman, Shymko & Trabert (1978), for biological pattern formation. Calculations are presented for pattern formation on a disc (approximating the geometry of a number of embryonic anlagen including the frog eye rudiment), emphasizing the sensitivity of patterns to changes in initial conditions and to perturbations in the geometry of the morphogen-producing space. Analysis of the linearized equations from the models enabled us to select appropriate parameters and disc size for pattern growth. A computer-implemented finite element method was used to solve the non-linear model equations reiteratively. For the Gierer-Meinhardt model, initial activation (varying in size over two orders of magnitude) of one point on the disc's edge was sufficient to generate the primary gradient. Various parts of the disc were removed (remaining only as diffusible space) from the morphogen-producing cycle to investigate the effects of cells dropping out of the cycle due to cell death or malfunction (single point removed) or differentiation (center removed), as occur in the Xenopus eye rudiment. The resulting patterns had the same general shape and amplitude as normal gradients. Nor did a two-fold increase in disc size affect the pattern-generating ability of the model. Disc fragments bearing their primary gradient patterns were fused (with gradients in opposite directions, but each parallel to the fusion line). The resulting patterns generated by the model showed many similarities to results of "compound eye" experiments in Xenopus. Similar patterns were obtained with the model of Kauffman's group (1978), but we found less stability of the pattern subject to simulations of central differentiation. However, removal of a single point from the morphogen cycle (cell death) did not result in any change. The sensitivity of the Kauffman et al. model to shape perturbations is not surprising since the model was originally designed to use shape and increasing size during growth to generate a sequence of transient patterns. However, the Gierer-Meinhardt model is remarkably stable even when subjected to a wide range of perturbations in the diffusible space, thus allowing it to cope with normal biological variability, and offering an exciting range of possibilities for reaction-diffusion models as mechanisms underlying the spatial patterns of tissue structures.