Department of Biological Sciences, Wellesley College;
Megan Libbey
Department of Biological Sciences, Wellesley College
Jill Arters
Department of Biological Sciences, Wellesley College
Mehnaz Junagadhwalla
Department of Biological Sciences, Wellesley College
Christine F. Hohmann
Department of Biology, Morgan State University
Acknowledgement: This research was supported by National Science Foundation Grant IBN94585101, by the Whitehall Foundation, and through Morgan State University. We thank U. Berger and K. Frick for critically reviewing the manuscript, D. Smith for help in designing the odor discrimination protocol, E. Caliguri for aid in neurochemical analyses, and L. Baldwin for statistical advice. Portions of the research reported in this article have appeared previously in abstract form (
The organization of the mature cerebral cortex is influenced by several subcortical afferent systems. In rodents, noradrenergic, serotonergic, dopaminergic, and cholinergic fibers innervate the cortex just prior to birth, the most dynamic period of cortical differentiation and synapse formation. A wealth of studies suggest that these afferents regulate cortical maturation and establish neural circuitry that is important for cognition (
Norepinephrine (NE) fibers from the locus coeruleus and serotonin (5HT) fibers from the raphe nuclei arrive at the cortical anlage around Embryonic Day 17 and penetrate all cortical layers within the 1st week of life (
Indirect evidence suggests that alterations in 5HT around the time of birth may have different effects on the two sexes: (a) Maternal stress that alters cortical 5HT receptors in the offspring reportedly feminizes male rats' cerebral cortex such that a characteristic female laterality pattern is displayed (
The present study examined the developmental role of monoaminergic projections on cognitive development and neurochemistry. The BALB/cByJ mice used in our previous studies are slow to learn spatial navigation tasks because of their limited visual capabilities (
Groups of female and male mice received bilateral 5,7 dihydroxytryptamine (DHT) lesions to the nBM/MFB region on PND 1 to deplete hippocampally and cortically projecting NE and 5HT fibers. Mice grew to adulthood and were tested on a battery of tasks: SOD and DNMS, passive avoidance (PA), and locomotor activity. Other mice were killed for neurochemical analyses.
BALB/cByJ mice (Mus musculus) were bred at the Wellesley College colony and placed randomly into the following groups: unoperated controls, sham-operated mice, and bilateral DHT-lesioned mice. All mice were maintained on a 12-hr light–dark cycle (lights on at 7 a.m.), and behavioral testing was conducted during the light cycle. Food and water were available ad libitum during activity and PA testing. During SOD and DNMS, mice were deprived of food 4–5 hr before testing.
Fifty-one mice were used for behavioral testing: 21 unoperated controls (10 females and 11 males), 16 ascorbic acid sham-lesioned mice (10 females and 6 males), and 14 bilateral DHT-lesioned mice (9 females and 5 males). These mice underwent SOD, DNMS, PA, and general motor activity measurements. Thirty mice were used for neurochemical analyses: 10 unoperated controls, 10 bilateral ascorbic acid sham-lesioned mice, and 10 bilateral DHT-lesioned mice. Equal numbers of females and males were used. The ascorbic acid sham-operated mice were not significantly different from unoperated controls on any of the behavioral tasks or in neurochemical measures. The data from these two groups, therefore, were combined, and they are referred to as controls.
Pups were removed from their mothers 12 to 24 hr after birth. Mice were anesthetized by hypothermia (placed on ice for 5 min until movement ceased) and were injected in the nBM/MFB (ventromedial globus pallidus) region with the monoaminergic toxin 5,7-DHT (Sigma Chemical, St Louis, MO; 0.5 μl of a 5 μg/μl solution of 5,7-DHT containing 0.1 mg/ml ascorbic acid) or ascorbic acid (sham-operated controls). The lesion coordinates, as previously reported, were 1 mm anterior to bregma, 1.5 mm lateral to bregma, and lowered 3.5 mm, 4.0 mm, and 4.5 mm ventral from the dura (
Odor discrimination
Shaping and testing for SOD and DNMS tasks were conducted in standard (30 × 19 cm) clear mouse cages (Allentown Caging, Allentown, PA) with bedding. Either one or two bait cups were located at one end of the rectangular cage about 3 cm apart (see
Passive avoidance
The test enclosure (52.5 × 52.5 × 30 cm) had an automatic gate that separated a light (white walls) and a dark (black walls) chamber. The grid floor of the black chamber was electrified to give a 0.2-mA footshock for 3 s. The data were collected with GEMINI software (
Motor activity
The Cage Rack activity system (San Diego Instruments, San Diego, CA) consisted of a clear acrylic cage (43 × 22 cm) surrounded by a bracket containing three photobeam projectors. The beams were oriented to transect the short axis of the cage and were located 2.5 cm from the cage floor. The photobeam breaks were recorded with PAS software (
Testing began when the mice reached adulthood, 8–12 weeks after the surgery. Before testing began, all mice were handled by the experimenter for 1 week.
Odor discrimination
Mice were deprived of food 4–5 hr prior to odor discrimination shaping and testing procedures. Baited cups contained 1–3 small pieces of chocolate (≈ 15 mg); unbaited cups did not contain chocolate. Mice were placed in a holding cage for 30 s between trials.
Shaping
On Day 1, the mouse was presented with a baited cup in which one piece of chocolate was visible, one was semivisible, and another was buried. The mouse was allowed a maximum of 1 hr to retrieve all three of the chocolate pieces. If the mouse did not begin to dig within 15 min, another piece of chocolate was placed on top of the sand, and the mouse was given an additional 5 min. On Day 2, the mouse was presented with two unscented cups, one baited and one unbaited. In the baited cup, one piece of chocolate was visible to the mouse, and one was buried. Each mouse underwent four trials with a maximum trial length of 15 min. If the mouse did not begin digging for the buried piece of chocolate within 10 min, then an additional piece of chocolate was placed on top. The position of the baited cup (left side vs. right side) was randomized. On Day 3, the mouse was introduced to two scents, cinnamon and curry. Each mouse was assigned a baited scent of either cinnamon or curry, and this association was maintained throughout the SOD task. Assignments were counterbalanced across groups.
Simple odor discrimination
The location of the baited cup was varied so that it was never in the same location for more than two trials. A dig was defined by displacement of sand in the cup by the paws or snout; sniffing or standing on the cup was not counted as a dig. For each trial, a mouse was presented with two cups (one cinnamon, one curry) and given a maximum of 5 min to retrieve the buried chocolate; if the mouse had not begun digging within 5 min, a piece of chocolate was dropped on the top of the correct cup, and the mouse was given an additional 2 min to find the chocolate. Choice accuracy (percentage correct on the first choice), number of errors (total number of digs in the unbaited cup), and latency to retrieve the chocolate were recorded. Each mouse was given four trials per day for 5 consecutive days. On the day after SOD testing, two probe trials were conducted to ensure that the mouse was using odor cues to guide the choice. For the probe trials, each mouse was presented with two scented cups, both unbaited. The first cup choice was recorded. Once the correct cup was chosen, a piece of chocolate was placed on the sand, and the mouse was allowed to retrieve it. For each test, two experimenters performed the task. After initial training, observations between the two experimenters were identical.
Delayed nonmatch-to-sample
This task began 1 day after completion of the SOD. Each trial began with the presentation of one baited cup (sample presentation) filled with sand mixed with a scent. The mouse was allowed to retrieve the chocolate from this cup and was then removed from the test cage for a 30-s delay. At the end of the delay, the mouse was placed in the choice cage and presented with two cups (choice presentation). One cup (match) was identical to the sample but was unbaited. The other cup (nonmatch) contained a novel scent and was baited. The mouse was allowed to dig until the buried chocolate was found in the nonmatch cup. For each choice presentation, the cup in which the mouse dug first (choice accuracy), the latency to retrieve the chocolate piece, and the errors (digs in the unrewarded match cup) were recorded. Scents were reused across the trials but were never paired with the same scent as before. Mice were given three trials per day for 4 consecutive days.
Passive avoidance
This protocol has been previously reported (
Motor activity
Animals were placed in individual activity cages at 4:30 p.m., and activity, as assessed by photobeam breaks, was recorded and expressed as beam break totals per hour.
At 12–17 weeks of age, groups of lesioned and control mice were killed by cervical dislocation after carbon dioxide anesthesia (
Behavioral and neurochemical data were analyzed by a two-factor, repeated measures analysis of variance (ANOVA). Lesion status and sex were examined as the main effects, and the tasks (SOD and DNMS), acquisition trials (PA), or hours (activity) were examined as the repeated measures. PA retention trials, neurochemical, and probe data were analyzed in two-factor ANOVAs with lesion status and sex as main effects.T tests were used to compare pairwise means between groups. Significance level was set atp < .05 for all analyses.
The measure of errors was not recorded during the SOD of the first group of mice tested. Therefore, there were 6 mice (3 female control, 3 male control) that were excluded from the SOD analysis. Errors were recorded during the DNMS portion of testing for those 6 mice and are included in the DNMS analysis. In addition, 1 mouse did not cross over to the black chamber during PA and was therefore dropped from the analysis.
Simple odor discrimination
Choice accuracy increased significantly across the 5 days of testing for all groups,F(4, 164) = 6.7,p
= .001 (see
Errors for all groups decreased significantly across the 5 days of testing,F(4, 164) = 8.7,p
= .001 (see
Latency to retrieve the chocolate also decreased significantly across the 5 days of testing for all groups,F(4, 164) = 11.9,p
= .0001 (see
On the probe trials, all groups performed with 82–100% accuracy, which suggests that all mice were using odor cues to guide SOD performance. There were no significant differences in probe accuracy between the lesioned groups or the sexes.
Delayed nonmatch-to-sample
Overall, the lesioned mice performed the DNMS with higher choice accuracy than did control mice,F(1, 47) = 5.6, p = .02 (see
The lesioned mice also made significantly fewer errors than did control mice,F(1, 47) = 5.8,p
= .02 (see
During the choice presentation, control mice had significantly longer latencies to find the piece of chocolate than did DHT-lesioned mice,F(1, 47) = 7.8,p
= .007, on the DNMS task (see
Passive avoidance
Control and lesioned mice of both sexes acquired the PA task similarly; there were no significant differences in acquisition among the groups. The latency to enter the dark chamber was significantly higher on the second acquisition trial for all groups,F(1, 46) = 7.3,p
= .009. At the 24-hr retention trial, however, a significant main effect of lesion was apparent. The control mice had significantly longer latencies to enter the dark compartment (better performance) than did the DHT-lesioned mice,F(1, 46) = 26.6,p
= .01 (see
Motor activity
Locomotor activity in control and lesioned mice was similar; there were no significant differences between the groups. In all groups, activity decreased significantly over the 12 hr of testing,F(11, 517) = 50.6,p
= .0001 (see
Bilateral neonatal DHT lesions resulted in significant decreases in NE and 5HT levels in both the frontoparietal cortex and hippocampus in lesioned mice relative to controls,F(1, 16) ranged from 22.1 to 74.5, allp
s < .01 (see
The results of this study show that DHT lesions on PND 1 interrupt monoaminergic projections to the cortex and hippocampus and have long-lasting effects on behavior and neurochemistry. Neonatal DHT-lesioned mice performed SOD with similar accuracy, perseverance, and speed as compared with control mice. In contrast, the lesioned mice performed DNMS faster, more accurately, and with less perseverance than controls. This lesion-induced improvement in performance is not attributable to increased locomotor activity or different accuracy on probe (unrewarded) trials. The improved performance in the lesioned mice relative to controls could result from improved mnemonic or attentional abilities or differences in digging strategies. In addition to persistent behavioral alterations, the neonatal DHT lesions lead to significant depletions in cortical and hippocampal NE and 5HT in the adult mice. This study provides one of the first demonstrations of the long-lasting effects of such lesions and provides insights into the developmental role of monoaminergic input to the cortex and hippocampus.
Choice accuracy and errors on SOD were similar for the control and lesioned mice. All mice performed SOD well (more than 70% accuracy) on the 1st day and improved to more than 90% accuracy by the 5th day of the task. Moreover, errors (total number of incorrect digs) decreased over the 5 days of SOD, which suggests that the mice did not continue to persevere in the incorrect cup as testing continued. There were no overall differences between control and lesioned mice on latencies on SOD trials; however, the rate of latency changes over the 5 days did differ between the groups. Control and lesioned mice also had a similar sensitivity to odor cues (J. Berger-Sweeney, unpublished, observations, & May, 1998), performed similarly during unrewarded probe trials, and appeared to have a similar motivation to associate an odor cue with a reward, as shaping was similar in the two groups. The control and lesioned mice appeared to be able to learn a “win–stay” strategy with similar speed and accuracy. In contrast, control and lesioned mice differed significantly in accuracy and perseverance in the performance on the DNMS task; lesioned mice of both sexes significantly outperformed (better choice accuracy and fewer errors) the controls. In addition, the DHT-lesioned mice performed DNMS faster than their control littermates of both sexes. General locomotor activity, however, did not differ between control and lesioned groups, which makes this an unlikely explanation for their performance differences on DNMS. As such, the lesioned mice appeared to adopt the “win–shift” strategy more readily than controls.
The DHT-lesioned mice were significantly impaired in 24-hr retention of PA, although they acquired the PA task similarly to controls and responded to shock stimuli similarly to controls. What could account for the improved performance of the lesioned mice on DNMS and their impaired performance on PA retention trials? One possibility is a response inhibition deficit in the DHT-lesioned mice. Response inhibition refers to interrelated processes that permit a delay in the decision to respond, allow for distraction by competing events, and inhibit a response that is immediately reinforced (
Consistent with the results of the current study, performance improvements have been reported after serotonergic lesions in adulthood. In one study, rats (Rattus norvegicus) were required to switch from pressing one lever to pressing another lever at a specific time for reinforcement (
The noradrenergic and serotonergic systems have also been associated with contextual learning and attention. PCA-treated adult rats, for example, respond differently to contextual cues than control rats (
Another possible explanation for the improved performance of the lesioned mice on DNMS is an indirect cholinergic-induced improvement in working memory. Serotonergic lesions can reportedly increase acetylcholine turnover in the cortex (
In addition to altering behavior, the 5,7-DHT lesions affected cortical and hippocampal neurochemistry, which suggests that MFB fibers projecting to these structures were affected. Our previous results after 5,7-DHT lesions (which used the same lesion protocol) showed that 4 days postlesion, NE and 5HT levels in the cortex were 53% and 75%, respectively, relative to controls (
Sex differences in performance have been reported on a wide variety of behavioral tasks (
Researchers have shown previously that electrolytic lesions to the nBM on PND1 cause significant deficits in spatial learning and PA retention and alter cortical morphology in adult mice (
The present study provides convincing evidence that serotonergic and noradrenergic fibers play an important role in the normal development of cognitive behaviors. Further studies will be necessary to determine whether 5HT, NE, or both are responsible for the observed performance alterations. Developmental monoaminergic lesions likely lead to alterations in the cortex and hippocampus that result in deficits in response inhibition and alterations in integration of contextual cues. Interestingly, poor behavioral inhibition is considered to be a central deficiency in ADHD (
Alonso, J., Castellano, M. A., & Rodriguez, M. (1991). Behavioral lateralization in rats: Prenatal stress effects on sex differences. Brain Research, 539, 45–50.
Altman, H. J., Ogren, S. O., Berman, R. F., & Normile, H. J. (1989). The effects of ρ-Chloroamphetamine, a depletor of brain serotonin, on the performance of rats in two types of positively reinforced complex spatial discrimination tasks. Behavioral and Neural Biology, 52, 131–144.
Arnsten, A. F. T., Steere, J. C., & Hunt, R. D. (1996). The contribution of α2-noradrenergic mechanisms to prefrontal cortical cognitive function. Archives of General Psychiatry, 53, 448–455.
Arters, J., Hohmann, C. F., Mills, J., Olaghere, O., & Berger-Sweeney, J. (in press). Sexually dimorphic responses to neonatal basal forebrain lesions in mice: I. Behavior and neurochemistry. Journal of Neurobiology.
Bachman, E. S., Berger-Sweeney, J. E., Coyle, J. T., & Hohmann, C. F. (1994). Developmental regulation of adult cortical morphology and behavior: An animal model for mental retardation. International Journal of Developmental Neuroscience, 12, 239–253.
Barkley, R. A. (1997). Behavioral inhibition, sustained attention, and executive functions: Constructing a unifying theory of ADHD. Psychological Bulletin, 121, 65–94.
Beatty, W. W. (1979). Gonadal hormones and sex differences in nonreproductive behaviors in rodents: Organizational and activational influences. Hormones and Behavior, 12, 112–163.
Berger-Sweeney, J., Arnold, A., Gabeau, D., & Mills, J. (1995). Sex differences in learning and memory in mice: Effects of sequence of testing and cholinergic blockade. Behavioral Neuroscience, 109, 859–873.
Berger-Sweeney, J., Berger, U. V., Sharma, M., & Paul, C. A. (1994). Effects of carbon dioxide-induced anesthesia on cholinergic parameters in rat brain. Laboratory Animal Science, 44, 369–371.
Berger-Sweeney, J., & Hohmann, C. F. (1997). Behavioral consequences of abnormal cortical development: Insights into developmental disabilities. Behavioural Brain Research, 86, 121–142.
Blue, M. E., Erzurumlu, R. S., & Jhaveri, S. (1991). A comparison of pattern formation by thalamocortical and serotonergic afferents in the rat barrel field cortex. Cerebral Cortex, 1, 380–389.
Breese, G. R., Vogel, R. A., & Mueller, R. A. (1978). Biochemical and behavioral alterations in developing rats treated with 5,7-dihydroxytryptamine. Journal of Pharmacology and Experimental Therapeutics, 205, 587–595.
Bunsey, M., & Eichenbaum, H. (1995). Selective damage to the hippocampal region blocks long-term retention of a natural and nonspatial stimulus–stimulus association. Hippocampus, 5, 546–556.
Decker, M. W., & McGaugh, J. L. (1989). Effects of concurrent manipulations of cholinergic and noradrenergic function on learning and retention in mice. Brain Research, 477, 29–37.
Fleming, D. E., Anderson, R. H., Rhees, R. W., Kinghorn, E., & Bakaitis, J. (1986). Effects of prenatal stress on sexually dimorphic asymmetries in the cerebral cortex of the male rat. Brain Research Bulletin, 16, 395–398.
Fuster, J. M. (1995). Memory in the cerebral cortex. Cambridge, MA: MIT Press.
Gaffan, D. (1974). Recognition impaired and association intact in the memory of monkeys after transections of the fornix. Journal of Comparative and Physiological Psychology, 86, 1100–1109.
GEMINI Active Avoidance System [Computer software]. (1993). San Diego, CA: San Diego Instruments.
Hanna, G. L., Ornitz, E. M., & Hariharan, M. (1996). Urinary catecholamine excretion and behavioral differences in ADHD and normal boys. Journal of Child and Adolescent Psychopharmacology, 6, 63–73.
Ho, M., al-Zahrani, S., Velazquez Martinez, D., Lopez Cabrera, M., Bradshaw, C., & Szabadi, E. (1995). The role of the ascending 5-hydroxytryptaminergic pathways in timing behavior: Further observations with the interval bisection task. Psychopharmacology, 120, 213–219.
Hohmann, C. F., Brooks, A. R., & Coyle, J. T. (1988). Neonatal lesions of the basal forebrain cholinergic neurons result in abnormal cortical development. Developmental Brain Research, 43, 253–264.
Hohmann, C. F., & Ebner, F. F. (1985). Development of cholinergic markers in mouse forebrain: I. Choline acetyltransferase enzyme activity and acetylcholinesterase histochemistry. Developmental Brain Research, 23, 225–241.
Hohmann, C. F., Richardson, C. M., Redding, C., Kaufman, W. E., Sanwal, I. B., Arters, J., & Berger-Sweeney, J. (1997). Neonatal lesions of cholinergic and monoaminergic afferents to neocortex produce different alterations in cortical morphology. Society for Neuroscience Abstracts, 23, 78.
Kalsbeek, A., Buijs, R. M., Hofman, M. A., Matthijssen, M. A., Pool, C. W., & Uylings, H. B. (1987). Effects of neonatal thermal lesioning of the mesocortical dopaminergic projection on the development of the rat prefrontal cortex. Brain Research, 429, 123–132.
Kalsbeek, A., DeBruin, J. P. C., Feenstra, M. G. P., Matthijssen, M. A. H., & Uylings, H. B. M. (1988). Neonatal thermal lesions of the mesolimbocortical dopaminergic projection decrease food-hoarding behavior. Brain Research, 475, 80–90.
Kalsbeek, A., DeBruin, J. P. C., Matthijssen, M. A., & Uylings, H. B. (1989). Ontogeny of open field activity in rats after neonatal lesioning of the mesocortical dopaminergic projection. Brain Research, 32, 115–127.
Kalsbeek, A., Matthijssen, M. A., & Uylings, H. B. (1989). Morphometric analysis of prefontal cortical development following neonatal lesioning of the dopaminergic mesocortical projection. Experimental Brain Research, 78, 279–289.
Kolb, B., & Sutherland, R. J. (1992). Noradrenaline depletion blocks behavioral sparing and alters cortical morphogenesis after neonatal frontal cortex damage in rats. Journal of Neuroscience, 12, 2321–2330.
Lauder, J. M. (1990). Ontogeny of the serotonergic system in the rat: Serotonin as a developmental signal. Annals of the New York Academy of Sciences, 600, 297–314.
Libbey, M., Berger-Sweeney, J., & Hohmann, C. F. (1996). Neonatal 5,7 DHT lesions alter performance of passive avoidance and a novel odor discrimination task in adult mice. Society for Neuroscience Abstracts, 26, 270.2.
Lidov, H. G. W., & Molliver, M. E. (1982). The structure of cerebral cortex in the rat following prenatal administration of 6-hydroxydopamine. Brain Research, 255, 81–108.
Markowska, A. L., Price, D., & Koliatsos, V. E. (1996). Selective effects of nerve growth factor on spatial recent memory as assessed by a delayed nonmatching-to-position task in the water maze. Journal of Neuroscience, 16, 3541–3548.
Mefford, I. N., & Potter, W. Z. (1989). A neuroanatomical and biochemical basis for attention deficit disorder with hyperactivity in children: A defect in tonic adrenaline mediated inhibition of locus coeruleus stimulation. Medical Hypotheses, 29, 33–42.
Olton, D. S., & Werz, M. A. (1978). Hippocampal function and behavior: Spatial discrimination and response inhibition. Physiology and Behavior, 20, 597–605.
Parent, A., Descarries, L., & Beaudet, A. (1981). Organization of ascending serotonin systems in the adult rat brain: A radioautographic study after intraventricular administration of tritiated 5-hydroxytryptamine. Neuroscience, 6, 115–138.
Photobeam Activity System [Computer software]. (1992). San Diego, CA: San Diego Instruments.
Pliszka, S. R., McCracken, J. T., & Maas, J. W. (1996). Catecholamines in attention-deficit hyperactivity disorder: Current perspectives. Journal of the American Academy of Child and Adolescent Psychiatry, 35, 264–272.
Posner, M., & Petersen, S. E. (1990). The attention system of the brain. Annual Review of Neuroscience, 13, 25–42.
Saari, M., & Pappas, B. A. (1978). Behavioural effects of neonatal systemic 6-hydroxydopamine. Neuropharmacology, 17, 863–871.
Sara, S. J. (1985). Noradrenergic modulation of selective attention: Its role in memory retrieval. In D.Olton, E.Gamzu, & S.Corkin (Eds.), Memory dysfunctions: An integration of animal and human research from clinical and preclinical perspectives (Vol. 444, pp. 178–193). New York: New York Academy of Sciences.
712 HPLC System Controller 1.0 [Computer software]. (1992). Middleton, WI: Gilson Medical Electronics.
Shekim, W. O., Dekirmenjian, H., Chapel, J. L., Javaid, J., & Davis, J. M. (1979). Norepinephrine metabolism and clinical response to dextroamphetamine in hyperactive boys. Journal of Pediatrics, 95, 389–394.
Steckler, T., & Sahgal, A. (1995). The role of serotonergic–cholinergic interactions in the mediation of cognitive behaviour. Behavioural Brain Research, 67, 165–199.
Stewart, J., & Kolb, B. (1988). The effects of neonatal gonadectomy and prenatal stress on cortical thickness and asymmetry in rats. Behavioral and Neural Biology, 49, 344–360.
Sullivan, R. M., Wilson, D. A., & Leon, M. (1989). Norepinephrine and learning-induced plasticity in infant rat olfactory system. Journal of Neuroscience, 9, 3998–4006.
Wallace, J. A., & Lauder, J. M. (1983). Development of the serotonergic system in the rat embryo: An immunocytochemical study. Brain Research Bulletin, 10, 459–479.
Zola-Morgan, S., & Squire, L. (1986). Memory impairment in monkeys following lesions limited to the hippocampus. Behavioral Neuroscience, 100, 155–160.
Submitted: December 12, 1997 Revised: May 6, 1998 Accepted: June 16, 1998