Neurotoxicology and Teratology, Vol. 11, pp. 395---403. © Pergamon Press plc, 1989. Printed in the U.S.A. 0892-0362/89 $3.00 + .00

Behavioral Effects of Prenatal Exposure to Lidocaine in the Rat:

Effects of Dosage and of Gestational Age at Administration

R O B E R T F. S M I T H , M A U R A F. K U R K J I A N , 1 KATHLEEN M. M A T I ' R A N 2 A N D S T E V E N L. K U R T Z

Department o f Psychology, George Mason University, Fairfax, VA 22030

Rece ived 11 N o v e m b e r 1988

SMITH, R. F., M. F. KURKIIAN, K. M. MATTRAN AND S. L. KURTZ. Behavioral effects of prenatal exposure to lidocaine in the rat: Effects of dosage and of gestational age at administration. NEUROTOXICOL TERATOL 11(4) 395--403, 1989.--Pregnant Long-Evans hooded rats were dosed via injections into the gum with 3, 6, or 9 mg/kg lidocaine, or vehicle, or were uninjected, on gestational day 4 (GD4), GD11, or GD18. Offspring (8-11 litters/group) were tested on a variety of tests of behavioral development and adult behavior. No effects of any dose at any time of administration were found upon maternal weight gain in gestation, litter size, or initial birth weight or weight gain of the pups. Administration at GD4 produced few effects; only footshock sensitivity showed a significant effect of dosing, although there were trends toward dosing effects on spontaneous alternation. For administration on GD 11, lidocaine was associated with slight but significant alterations in sex ratios, and a trend toward drug effects on development of spontaneous alternation. Vehicle administration at this age reduced barbiturate sleep time in offspring and slightly altered footshock sensitivity. Lidocaine dosing on GD18 was associated with a number of significant alterations of behavior, including visual discrimination, shuttlebox avoidance, tail flick, and water maze errors; there were also both vehicle and lidocaine effects on water maze latencies. These data reinforce our previous report that lidocaine may be a behavioral teratogen, and suggest that administration in later gestation in the rat may alter a broader range of behaviors than earlier in gestation.

Prenatal lidocaine Pregnancy Behavioral development Spatial learning Shock avoidance learning Schedule controlled behavior Prenatal epinephrine

WE previously reported that prenatal administration of a single acute dose of 6 mg/kg lidocaine hydrochloride on gestational day 11 (GD11) in the rat is followed by several behavioral changes in the offspring (16). Changes included delays in development of negative geotaxis and righting reflex, enhanced footshock sensi- tivity, increased errors in a water maze, increased suppression in a conditioned suppression task, and changes in response patterns in an operant visual discrimination task.

Teiling et al. (18) attempted a partial replication of our study, "using injections of 6 mg/kg lidocaine on GD10 and G D l l . Although procedural, statistical, dosing, and strain (Sprague- Dawley vs. our Long-Evans hooded rats) differences limit com- parisons with our study, they failed to confirm most of our findings. They did, however, find that male lidocaine-dosed offspring received more reinforcements than controls in a DRL-20 task and a visual discrimination task, Holson et al. (7) have also attempted to assess the effects of prenatal lidocaine on offspring behavior. Their work used doses similar to ours (although dosing was conducted on a daily basis throughout first, second or third

trimester), although their testing procedures varied considerably from ours. They reported that the number of significant effects found did not exceed the number expected on the basis of type I error alone, and therefore concluded that they could find no effect of lidocaine on behavioral development,

Although no other studies have examined possible neurobehav- ioral effects of a single dose of a local anesthetic during gestation in rats, other work had indicated that local anesthetics, including lidocaine, are capable of altering neural development in the chick (1,8). Correlational work in humans has found that exposure to a group of (primarily local) anesthetics in gestation is associated with strikingly altered visual attentional behavior in neonates (2), and with marginally reduced IQ test scores at four years of age (6). Lidocaine is widely used in pregnant women, particularly in dental settings (2,5). In addition, correlational work in humans with the related local anesthetic agent cocaine has indicated that maternal cocaine consumption during pregnancy may be associated with adverse outcomes for the offspring (although direct comparisons of chronic cocaine consumption with single acute doses of

IPresent address: Department of Psychology, University of Maryland, College Park, MD 20742. 2Present address: Hazelton Laboratories America, 1330-B Piccard Drive, Rockville, MD 20850.

395

396 SMITH, KURKJIAN, MATTRAN AND KURTZ

lidocaine may be inappropriate). Reported effects in humans include altered pregnancy outcome (spontaneous abortion, abrup- tio placentae, premature parturition), increased incidence of uro- genital defects, increased incidence of Sudden Infant Death Syndrome, and altered neurobehavioral function in the neonate [primarily increased irritability and poorer organization of arousal states--(3, 4, 11, 19)]. Finally, we have reported (15) that 10 mg/kg/day cocaine given to pregnant rats alters a number of behaviors in the offspring,

Our initial report on prenatal lidocaine failed to address some important issues regarding its effects. We used a single dose of lidocaine (6 mg/kg, within permissible limits for human dosing) injected into the massetter muscle of the jaw. We also injected at only a single time point (GD11). Finally, our previous report, as an initial report, utilized an injected control (saline), but no uninjected controls. The present study was undertaken to deter- mine a dose-effect relationship and the effects of gestational age at administration on neurobehavioral consequences in the offspring. Sperm-positive dams were dosed at GD4, G D l l , or GD18. Dosing groups included three doses of lidocaine (3, 6, and 9 mg/kg), an uninjected control, and a vehicle-injected control (containing 1:100,000 epinephrine). Since we had received infor- mal criticism for our massetter muscle injection technique utilized in the previous study (local tissue damage might interfere with chewing, and hence nutrition), we developed a technique for injection into the gum of the rat which we utilized in the present study. We also included assays of lidocaine in serum at various times after dosing pregnant animals. Tests were selected to include a range of measures of behavioral development, activity, learning, reflexive and operant sensory measures. In most cases, specific test selection was based on significant findings in our previous work with prenatal lidocaine, or upon inferences from that work (e.g., seizure susceptibility was assessed due to observations of seizures in lidocaine-dosed offspring in previous work).

METHOD

Subjects

Sperm-positive female Long-Evans derived rats were pur- chased from Blue Spruce Farms, Altamont, NY. Dams arrived in our laboratory on GD2, (GD1 =day found sperm positive), and were individually housed in hanging wire cages until GD 19, when they were transferred to breeding boxes with Beta Chip bedding. Litters were maintained in breeding boxes until weaning and transferred to hanging wire cages on PND21 (postnatal day 21; since litters were born on either GD22 or GD23, GD22 was designated as PND1). Dams were eartagged upon arrival in the laboratory; offspring were eartagged at weaning on PND21. Ear tags were coded so that all testers were unaware of the dosing history of all animals. Food and water were available ad lib, except when restriction was required for a particular behavioral test. Animals were maintained on a 12:12 light-dark cycle (0730- 1930 ON); testing occurred between 1000 and 1600 hours.

Dosing

Lidocaine (w/l:100,000 epinephrine) was purchased from TechAmerica, Elwood, KS. Vehicle was generously donated by TechAmerica and assayed for comparability of all components, including epinephrine. Administration was by injection into the gum of the lower jaw under manual restraint, with 1/2 of the dose injected into each side of the jaw. Pilot work with injections of food coloring via this route determined that no appreciable leakage occurred; in rare instances where an animal was able to dislodge the needle before the injection was finished, a second penetration

at a more posterior site was made to avoid leakage through the original puncture. Use of a very experienced animal handler insured that less than 30 sec of restraint was necessary to accomplish both injections, thus minimizing stress to the animal.

Assay Procedures

To determine levels of serum lidocaine at representative intervals postdosing, orbital sinus blood samples were taken from unanesthetized pregnant females (not the dams whose offspring were tested) under manual restraint. Samples were taken at 30 min, 2 hr, and 4 hr postdosing from animals dosed with 3, 6, or 9 mg/kg lidocaine on either GD4, GD11, or GD18. Samples were refrigerated in heparinized tubes until assay for lidocaine by American Medical Laboratories, Fairfax, VA, using the Abbott TDx immunoassay system.

Design

Sperm-positive females were randomly assigned to dosing groups. The design was a 3 (dosed on GD4, GD11, or GD18) × 5 [dosed with 3, 6, or 9 mg/kg lidocaine, vehicle (volume equal to that of the high dose group), or uninjected] design, Because of the large number of dams and litters in this design, dosing was staggered, with one dam dosed in each condition every two weeks. This procedure insured that circannual rhythms or other extraneous influences could not differentially affect any treatment group. For tests of postweaning behavior, only one male and one female per litter were tested, and the individual animal data were used as the unit of measurement for data analysis.

Birth and Growth

Within 24 hours of birth, all pups were weighed, sexed, and examined for facial or limb deformities. Litters were culled to four males and four females. Litters with fewer than four of either sex were culled to a total of eight. Litters with fewer than eight pups were retained for behavioral testing. Offspring were weighed weekly until weaning, and semiweekly thereafter until the begin- ning of adult testing,

Negative Geotaxis and Righting Reflex Development

Righting reflex was tested daily in all pups on PND2 through PND8. Subjects were placed on a flat surface covered with DACB cage board, in a supine position. Time to right to all four paws was measured with a stop watch to a maximum of 120 sec. Negative geotaxis was assessed daily from PND4 through PND10 in all offspring. Subjects were placed on a 25 ° inclined plane (covered with DACB) facing directly downhill. Time to complete a 180 ° turn was measured with a stop watch, to a maximum of 120 sec. On days when both tests were conducted, pups were returned to home cages for a minimum of one hour between tests.

Assignment to Test Conditions

At weaning, four sex pairs from each litter were randomly assigned to subsequent tests, with testing for each sex pair as indicated below. Unless otherwise indicated, sequential tests on a given sex pair were separated by an interval of five days.

Sex Pair 1

Open field. Open field activity was assessed on PND 25, 30, 35, and 40, and again on PND90. For each session, a 76 × 76 cm box was used as the open field, BRS/LVE infrared photosensors

PRENATAL LIDOCAINE AND OFFSPRING BEHAVIOR 397

trisected the open field in each horizontal dimension. The OPN computer control software (17) running on Apple II + computers was used to segregate beam breaks into ten consecutive three min periods, as well as for DRL, visual discrimination, and shuttlebox testing described below.

DRL-20 sec acquisition. This procedure, and all tests described below, commenced on PND90, except that a second test for a given subject occurred no sooner than five days following com- pletion of the first adult test, unless otherwise noted. Subjects were food deprived to 80% of ad lib body weight, shaped to bar press, and given four days of 100 reinforcements per day on a continuous reinforcement schedule, followed by 30 min per day on a DRL-20 sec schedule for 30 days. Apparatus consisted of Gerbrands G7321 Skinner boxes housed in BRS/LVE sound attenuation chambers, controlled by OPN on Apple computers.

Water maze. The water maze was constructed in house of Plexiglas. All alleys were 20 cm wide and 20 cm deep, with a 10 cm water depth. A 25 cm start box led to the first choice point. The incorrect (straight ahead) choice led to a blind T (25 cm leg, 25 cm arms). The correct (L) choice led to the second choice point after a 45 cm swim. The incorrect (straight) choice was a 25 cm blind alley, and the correct (R) choice was a 75 cm swim to the exit ramp. Water temperature was maintained at 27°C. Subjects were tested for ten trials per day for three days. On each trial the subjects were placed in the start compartment, and the numbers and directions of errors during the trial and the latency to reach the exit ramp (max = 300 sec) were recorded.

Audiogenic seizure susceptibility. Subjects were placed in a sound attenuation chamber equipped with an API model 108 alarm bell (114 dB SPL). The alarm was manually triggered and subjects observed through an observation port. Subjects were scored for presence or absence (and type/severity) of seizure during 60 sec of alarm activation. Seizures were scored on a scale similar to that of Schreiber and Schlesinger (12); levels were 1) wild running, 2) clonic seizure, 3) tonic seizure, and 4) lethal seizure. If a seizure occurred, the alarm was terminated and the type of seizure scored. This procedure has never produced lethality in our lab.

Sex Pair 2

Spontaneous alternation. On PND 21, 26, 31, 36, 41, and 46, spontaneous alternation was assessed. The apparatus consisted of a T-maze made of wood, painted flat grey, and having a start box 18 cm long and two goal arms 61 cm long. The subject was placed in the start box and doors to the start box and goal arms were opened simultaneously. Subjects were scored as choosing a side when the entire body (to the base of the tail) entered that arm. Subjects were given two trials on each test day, and were scored as either alternating (choosing each side once) or perseverating (choosing the same side on both trials).

Operant visual discrimination. Following food deprivation to 80% of ad lib body weight, animals were shaped to bar press and given four days of 100 reinforcements per day on a continuous reinforcement schedule. They then received 10 days of 30 min per day training on the visual discrimination task, using the Skinner boxes and control system identified above for DRL. On this schedule, responses are reinforced when the cue light is lit, while responses in the absence of the cue light are never reinforced. Light-on and light-off periods alternate on a variable time schedule (mean = 30 sec, S.D. = 10).

Conditioned suppression. A Lafayette Instruments shuttle box was used for testing. On test day 1, the subject was placed on one side of the box, and the house light flashed at a 2/sec rate. The latency to cross was recorded. On the following day, the subject was confined to the side with the 2/sec flashing light, and after 10

sec was administered a 3 sec 5 mA footshock. On the final day, latency to cross was again measured in the presence of the flashing cue light. Latency ceilings for days 1 and 3 were 600 sec.

Barbiturate sleep time. Subjects were given an intraperitoneal injection of 30 mg/kg sodium pentobarbital. Time from loss of the righting reflex to regain of the righting reflex was measured with a stop watch.

Sex Pair 3

Shuttle box avoidance. BRS/LVE shuttle cages mounted in sound attenuating chambers were used. For each trial, a 10 sec CS (consisting of Sonalert plus shift of the cue light) preceded shock. If the subject did not cross within 10 sec, a 0.6 mA shock (BRS/LVE SGS-003) was delivered for a maximum of 40 sec. Each trial terminated when the animal crossed the barrier. The program detected and compensated for interlrial crossings, and ignored false crossings (floor movement due to the animal jumping up and down without crossing).

Tail flick. Subjects were placed innto Wahmann LC-153/B restrainers. Approximately 2.5 cm of the tail was immersed into 55°C water at a 30 ° angle. Latency to raise the tail from the water (max = 10 sec to avoid injury) was recorded for the single trial.

Footshock sensitivity. Animals were placed in a Skinner box with a shock grid floor. A modified ceiling height of 8.5 cm prevented rearing, and the box was mounted on a Lafayette 86010 activity monitor, with output to a strip chart recording. Electro- mechanical equipment controlled shock timing and intensity through a BRS/LVE SG-903 shocker modified for external control. The subject received a series of 10 ascending shocks of 0.2 to 2.0 mA intensity, in 0.2 mA increments, followed by descending shocks of the same intensity. The intershock interval was 15 sec. After a two min period, the procedure was repeated, for a total of 40 shocks (four at each of ten intensities). The primary data were the mean peak heights to each intensity of shock. The design for this measure thus included three within subject factors: shock inten- sity, an ascending/descending order of intensity factor, and first/ second series.

Five days following footshock sensitivity testing, these sub- jects were assessed for audiogenic seizure susceptibility as de- scribed above.

Sex Pair 4

These sex pairs received testing identical to that of Sex Pair 3, except that no shuttle box avoidance training occurred. Sex Pairs 3 and 4 were compared on footshock sensitivity, tail flick, and seizure susceptibility to determine whether shock experience affected those behaviors and/or interacted with prenatal dosing condition.

Statistical analysis. For pregnancy and litter measures, a 3 (age at dosing) x 5 (dose) analysis of variance was used for all analyses. For other measures on the offspring, either a litter mean per sex, or actual data from only one male and one female per litter, was the unit of analysis. Data for each dosing age was analysed separately. For each measure a 5 x 2 analysis of variance was used, with dose and sex as the between-subject variables, respectively, and one or more nested levels of repeated measures where appropriate. When justified by the initial ANOVA, addi- tional significance tests were performed as indicated in the results section.

RESULTS

Assays

Serum lidocaine concentrations produced by our doses and

398 SMITH, K U R K J I A N , M A T T R A N A N D K U R T Z

T A B L E 1

SERUM LIDOCAINE VALUES AS A FUNCTION OF DOSAGE, STAGE OF GESTATION, AND TIME AFTER DOSING

Time After Dosing

30 Min 2 Hr 4 Hr

GD4 dosing 3 mg/kg 1.145 - 0.091 0.048 +- 0.009 0.090 _+ 0.015 6 mg/kg 1.165 +_ 0.080 0.180 +- 0.032 0.178 --- 0.023 9 mg/kg 2.171 _+ 0.421 0.275 + 0.040 0.251 +- 0.050

GDI 1 dosing 3 mg/kg 0.470 +_ 0.170 0.740 +- 0.326 0.120 _+ 0.031 6 mg/kg 2.410 +- 1.167 0.148 +- 0.130 0.135 _+ 0.024 9 mg/kg 2.905 _+ 0.067 0.275 ± 0.061 0.177 +- 0.050

GD18 dosing 3 mg/kg 0.752 +_ 0.075 0.528 +- 0.353 0.092 +_ 0.017 6 mg/kg 1.258 - 0.319 0.298 + 0.063 0.307 --- 0.385 9 mg/kg 2.090 -+ 0.611 0.632 +- 0.154 0.367 _+ 0.050

All values are Ixg/ml serum, mean _+ SEM.

route of adminis t ra t ion are summar i zed in Table 1, An analys is o f var iance indicated that, in addit ion to expected effects of dosage and t ime s ince injection, there was a s ignif icant effect o f the dose × t ime × t r imester factor, F(8,103) = 2 .09 , p < 0 . 0 5 . As Table I indicates, absorpt ion appeared s lower later in gestat ion, partic- ularly with the low dose o f l idocaine. Thir ty m in concent ra t ions are lower in G D l l and GD18 dosed an imals , while two hour concentra t ions are higher .

Pregnancy, Birth, and Growth

Measures o f weight gain dur ing gestat ion failed to d is t inguish be tween any of the t rea tment groups , as did birth weights , or any measu re s o f body weight gain. No facial or l imb deformit ies were noted. For offspr ing o f d a m s dosed in second t r imester , there was a s ignif icant dose effect on sex ratios o f the litters obtained, F ( 4 , 3 8 ) = 2 . 8 6 , p < 0 . 0 5 . Six mg /kg dos ing at G D l l was associ- ated with a h igh M:F ratio, while there was a low M:F ratio in the 9 m g / k g condit ion. These measu re s , as well as the reproduct ive success and cell sizes in each o f the dos ing condi t ions , are summar i zed in Table 2.

Effects of dosing at GD4

The only effect s ignif icant ly associa ted with l idocaine dos ing at

T A B L E 2

LITTER MEASURES

Dose (mg/kg) Uninjected Vehicle 3 6 9

Viable litters/no, dosed First trimester 9/12 8/11 11/12 9/10 l 1/11 Second trimester 8/12 10/11 8/11 9/12 10/11 Third trimester 8/11 11/15 8/11 9/10 8/11

Maternal weight gain (g) First trimester 118.67 114.33 106.60 117.20 117.70

+4.63 ___4.63 -+9.09 +5.93 _+5.19 Second trimester 109.89 116.70 99.98 115.11 109.50

___4.02 -+6.65 ___8.91 +6.16 _.5.72 Third trimester 115.90 107.89 106.93 119.56 116.88

+6.40 _+5.56 _+9.30 _+5.19 _+9.02

Number male pups First trimester 6.25 6.30 6.00 5.73 4.40

_+0.86 _+0.80 +0.75 +0.79 -+0.62 Second trimester 5.86 7.10 6.00 7.88 4.60

_+0.67 +- 1.12 +_0.60 -+0.99 -+0.37 Third trimester 6.00 5.88 5.88 5.70 5.86

_+0.80 +-0.84 +0~55 +_0.88 -+0.59

Number female pups First trimester 5.85 6.40 6.67 6.18 7.10

_+ 1.03 +_0.41 -+0.93 ___0.75 +_0.50 Second trimester 6.43 5.90 5.30 4.88 7.10

-+0.75 ---0.67 +-0.73 +0.41 +_0.62 Third trimester 6.78 6.25 6.25 5.30 5.57

+_0.49 +-0.65 _+0.56 +_0.86 +_0.72

Total litter weight (g) First trimester 76.67 86.16 75.86 75.66 76.87

-+7.15 +_ 3.77 -+6.23 -+5.16 _+3.68 Second trimester 77.57 84.79 73.81 84.01 76.82

_+6.07 +_5.69 _+6.49 +-5.45 +_3.98 Third trimester 8 I. 54 79.18 77.63 73.67 76.87

_+4.55 --_3.58 +_3.90 _+3.96 +_5.63

PRENATAL LIDOCAINE AND OFFSPRING BEHAVIOR 399

TABLE 3

BARBITURATE SLEEP TIME IN OFFSPRING OF DAMS DOSED AT GD11

Dosing Condition

Con Veh 3 mg/kg 6 mg/kg 9 mg~g

Males 70.80 + 17.29 32.33 --+ 6.12 36.86 ___ 7.10 40.85 __+ 5.00 38.00 _ 3.28 Females 147.29 ± 5.17 116.00 _ 7.84 133.33 ± 13.41 131.00 __- 6.19 129.67 _ 4.02

GD4 was on footshock sensitivity. The analysis of variance indicated significant interactions of dosing condition, sex, inten- sity of shock, and series of shock (first vs. second), F(36,1287) = 1.45, p<0.05, and of dosing condition, presence or absence of prior shuttle box experience, shock intensity, and ascending vs. descending series, F(36,1287), p<0.001, Separate analyses of variance for males and females revealed that only males were significantly affected by dosing at this age. For the males, there was a significant interaction of first vs. second series, shock intensity, and dosing condition, F(36,612)= 2.27, p<0.001. Al- though the magnitude of effect was slight, this effect was primarily due to 6 mg/kg males reacting more strongly to low shock intensities than did controls on the first and last of the four shock series. No effects were seen in females.

In addition to effects of lidocaine on footshock sensitivity, there was a trend toward a day x sex × dose interaction for latency in the spontaneous alternation task, F(20,360)= 1.58, p<0.10.

Effects of Dosing at GD11 In addition to the altered sex ratio associated with GD11 dosing

(noted above), dosing at this age was associated with alterations in footshock sensitivity. Significant interactions included: 1) the dose × ascending/descending x shuttle box experience interaction, F(4,140) = 2.57, p<0.05, 2) the first vs. second series × ascend- ing vs. descending × dose interaction, F(4,140) = 2.48, p<0.05, and 3) the shock intensity × dose x shuttle box history interaction, F(36,1260)= 1.63, p<0.05. To determine the contribution of experience with footshock in the shuttle box, we then analyzed the data of subjects with shuttle box experience separately from those of subjects with no such experience. For rats with experience, there was a significant interaction of shock intensity and dosing condition, F(36,279) = 1.47, p<0.05, while there were no signif- icant interactions involving dose for the animals with no prior experience. The magnitude of effect in animals with prior shuttle box experience was slight; it appears from inspection of the data that the major contributor to the significant effect was an enhanced reactivity of vehicle control animals at higher shock amplitudes.

Barbiturate sleep time was also altered in offspring of dams dosed at GD11, F(4,59) = 5.42, p<0.01. This effect, however, appeared to be due to the vehicle, as the vehicle group, as well as

15o • Un

145140 [ V e h ~ _ _ R - -~ 3 m g / k g / / ~ ~..__.~_

e 1 3 5 i 8 g/kg j

n 1 2 5

f lzo 0 1 1 5

r 1 1 0

C 1 0 5

e 1 0 0

IT1 95

e 90 n 85 t 8o S 75

7o I I I I I I I I I I I 0 1 2 3 4 5 6 7 8 9 1 0

Days FIG. 1. Visual discrimination, third trimester dosed offspring. Depicted are responses when the cue is lit, for males and females averaged together.

400 SMITH, K U R K J I A N , M A T T R A N A N D K U R T Z

M e a

n

T 0

t a

l

E F

r

o

F

S

26

2 4

2 2

20

18

16

14

12

10

8

6

4

E

0

0 1

Un -+ - Veh

. ~ - -~ 3 m g / k g 1:3 6 m g / k g

2

Days 3

FIG. 2. Errors per day in the water maze for third trimester dosed female animals.

M e

a n

A V

0

i d a

n

c

e

25 24 23 22 21 20 19 18 17 16 15 14 13 12 II I0 9

6 5 4

---- Un -+ - Veh

-~ - - ~ 3 m g / k g - - ~ 6 m g / k g

F

3 - 2 - 1 t t t t t

0 I 2 3 4 5

D a y s

FIG. 3. Shuttle box avoidances per day, third trimester dosed offspring, males and females averaged together.

PRENATAL LIDOCAINE AND OFFSPRING BEHAVIOR 401

TABLE 4

TAIL FLICK LATENCIES

Sex

Dose Male Female

Uninjected 3.18 ± 0.30 2.31 --- 0.25 Vehicle 2.77 ± 0.27 2.34 ± 0.19 3 mg/kg 2.42 ± 0.15 3.18 ± 0.40 6 mg/kg 2.72 ± 0.24 2.35 ± 0.24 9 mg/kg 2.77 ± 0.41 1.90 ___ 0.23

all dosed groups (except low dose females), had significantly shorter sleep times than undosed controls (see Table 3).

Although not statistically significant, GDI1 administration produced a trend toward alterations in the development of spon- taneous alternation. For latency data on spontaneous alternation, both the trial × dose interaction, F(4,78) = 2.09, p<0.10, and the trial × dose × day interaction, F(20,390)= 1.46, p<0.10, approached significance.

Finally, the analysis of variance for tail flick data on the subjects dosed at GD11 revealed a trend toward an effect of dose, F(4,160) = 2.10, p<0.10, with the differences between control and high dose subjects approaching significance (mean of 2.24 sec for controls vs. 3.13 for high dose subjects).

Effects of Dosing at GD18

Dosing at this age was associated with a number of significant alterations in offspring behavior. The analysis of variance for performance on the visual discrimination task indicated a signifi- cant interaction of dose × day × stimulus (light vs. dark), F(36,486) = 1.90, p<0.005. Analyses of variance for individual days indicated significant differences on days 5, 6, 7, 8, and 9; on each of those days there were significant cue × dose interactions (ps<0.05). As Fig. 1 indicates, 3 mg/kg offspring emitted fewer responses in the cue lit condition (responses reinforced) than did controls. There were no significant group differences in the cue dark (nonrein- forced) condition.

Water maze acquisition was significantly altered by GD18 dosing. An analysis of variance of the total errors per day on each of the three days of testing revealed a significant day × dose interaction, F(8,86)=2.20, p<0.05. As Fig. 2 indicates, this interaction was primarily due to the 6 mg/kg dose group having a different pattern of numbers of errors, beginning with more errors than controls on day 1, and finishing with fewer errors on day 3.

There were also significant dose-related differences in latencies in the water maze. There were substantial dose-related effects in the analysis comparing vehicle-injected offspring with uninjected offspring; these included the day×dose interaction, F(8,86)= 2.34, p<0.05, the trial×dose interaction, F(9,387)=2.25, p<0.05, and the trial × dose × sex interaction, F(8,86) = 2.33, p<0.05. Since the level of significance was lower in comparing dosed and uninjected controls than between dosed and vehicle controls, the former is reported here. Separate analyses were conducted on male and female animals. For males, there were no significant differences related to dosing. For females, there were significant dose, F(3,19) =4.07, p<0.05, dose × day, F(6,38) = 6.24, p<0.0001, trial× dose, F(27,171)=2.06, p<0.0001, and dose×day×trial , F(54,342)=2.24, p<0.0001, effects. Essen- tially, this reflected dose-related increases in latency on early trials, particularly on the first day of testing. For example, mean

latencies for Trial 1 of Day 1 were 53.48-+16.15 sec for uninjected controls, 87.06-+ 34.19 for 3 mg/kg offspring, 169.22 -4- 16.35 for 6 mg/kg, and 101.88-+18.13 for 9 mg/kgr Long latencies in the 6 mg/kg group were on trials associated with significantly elevated numbers of errors, as noted above.

For shuttle box avoidance learning, analysis of variance re- vealed a significant main effect of dosing, F(4,72) = 3.08, p<0.05. Figure 3 depicts those data. On Days 2-5 of testing, 6 mg/kg subjects had fewer successful avoidances than controls, while 9 mg/kg subjects had more avoidances than controls on Days 4 and 5.

The analysis of variance for tail flick data revealed a significant dose × sex interaction, F(4,144)=2.54, p<0.05. As indicated in Table 4, the principal effect was in the low dose group, with the effects directionally different in males and females.

DISCUSSION

The present data confirm that prenatal lidocalne administration may have neurobehavioral consequences for the offspring, al- though details of the types of effects found in the present study differ from those reported previously (16). We believe that our most substantial findings are 1) the significant effects of lidocaine were clustered in offspring of dams dosed on GD18, 2) many effects of GD11 dosing previously reported were not replicated in the present study, and 3) vehicle controls differed significantly from uninjected controls on three different measures.

Lack of substantial effect of GD4 dosing is perhaps not surprising, since dosing occurred prior to major development of the CNS. We were surprised, however, to find that GD11 dosing produced no statistically significant alteration in behavior of the offspring. In the rat, substantial proliferation of subcortical neural systems is occurring around GD11; in addition, our previous study (16) and other studies [e.g., (14) for halothane effects in rat] have reported effects of dosing with various substances near midgesta- tion. Although failure to replicate does not necessarily imply that a finding is spurious, it should prompt additional study and/or examination of procedural differences. Two potentially important procedural differences were implemented in the present study, as compared to the previous report. First, injection site was changed from massetter muscle to gum, which might result in changes in absorption kinetics (we collected no pharmacodynamic data in the original study). Secondly, we initiated testing (hence, daily handling) at a younger age in the current study, since maximal effects on behavioral development in the previous study occurred on the first day of testing (PND4). Earlier handling in the present study may have obscured effects seen previously. Meaney et al. (9) have recently reported that limited daily handling (15 rain removal from dams) in the same strain used in the present study essentially permanently reduced glucocorticoid responses to stress, and consequently retarded hippocampal cell loss and spatial memory deficits as the animals aged. Although one cannot infer from that report that differences between our present results and those of our previous study are in fact due to handling differences, clearly, further work is necessary to investigate that possibility.

GD11 dosing produced two significant behavioral changes, but those were associated with vehicle administration. Footshock sensitivity was slightly greater in vehicle controls than in unin- jected controls, and barbiturate sleep time was significantly shorter in all dosed groups (including vehicle), compared to uninjected controls. The footshock sensitivity effect was found only when subjects had prior shuttlebox avoidance experience, confn'ming our previous hypothesis (16) that dosing may interact with shock history to affect sensitivity.

In this study, then, significant GDl l dosing effects were

402 SMITH, KURKJIAN, M A T r R A N AND KURTZ

limited to vehicle effects, although we found trends toward drug effects on spontaneous alternation and tail flick data. Telling et al. (18) have reported effects on visual discrimination and DRL acquisition of 6 mg/kg lidocaine on GD10 and GD11. Holson et al. (7) concluded that their few significant effects of lidocaine were attributable to type I statistical error. Our earlier conclusion that G D l l dosing with lidocaine containing epinephrine alters offspring behavior has received a limited amount of support from the present paper and Teiling's work, although questions remain concerning the robustness of particular behavioral effects reported in that study, the relative contribution of lidocaine vs. vehicle to those effects, and definition of other variables (strain of subject, early handling, etc.) which may interact with lidocaine dosing to affect behavior.

Dosing at GD18 was associated with a number of alterations in offspring behavior, including visual discrimination, shuttle box avoidance, water maze and tail flick. Significant alterations of offspring behavior were more numerous after GD18 dosing than after dosing at GD4 or GD11, although testing was carried out simultaneously on all groups, using identical (and blind) testing conditions and procedures. These data suggest to us that neural structures which develop late in gestation in the rat may be more susceptible to lidocalne dosing than earlier developing structures. We are currently exploring the possibility of quantitative morpho- logical changes in late developing structures such as cerebellum and hippocampus [which is substantially affected by lidocaine administered to adult rats--(10,13)] following GD18 dosing. Others (1,8) have previously indicated that lidocaine alters neural development in chick embryos, and our behavioral data suggest the possibility of lidocaine-induced alterations in neural develop- ment in mammals.

It is interesting to note that, where we have found significant lidocaine effects upon offspring behavior, we have not found linear dose-effect relationships. In several cases, significant be- havioral effects apparent at 3 or 6 mg/kg were not evident at a higher dose. Although this is not uncommon in behavioral teratology, it underscores the notion that one cannot simply increase the dose of a compound in order to increase the magnitude of effect, as Teiling et al. (18) and Holson et al. (7) effectively

attempted to do in their attempts at replication of our first lidocaine paper.

Significant differences in footshock sensitivity, barbiturate sleep time and water maze latencies which were found between uninjected controls and vehicle-injected controls were not antici- pated. Although the vehicle contains methylparaben and sodium metabisulfite, we consider it most plausible that either the epi- nephrine in the vehicle or the stress of the injection procedure produced these effects (although the duration of the restraint and apparent stress to the animal were no greater than for subcutaneous injections).

Since work with prenatal lidocaine is in its infancy, further work is clearly necessary. Although we have found that lidocaine produces some neurobehavioral changes in the offspring of dosed rats, we have yet to determine whether this is due to a direct effect upon neural development, or whether behavioral changes are secondary to maternal changes (although gross maternal toxicity was not evident). Such a determination would require a cross- fostering study. We have not yet determined whether the behav- ioral effects of G D I I dosing are influenced by early postnatal handling, an issue which directly affects the robustness of our reported changes. Finally, we are beginning to examine possible effects on neural development at the light microscopic level, to determine whether our reported behavioral changes are correlated with changes in the central nervous system.

In summary, we have confirmed that prenatal lidocaine admin- istration in the rat does indeed alter some behaviors in the offspring, and that effects appear more numerous at GD 18 than at GD4 or GD11. We have also found some behavioral sequelae of the epinephrine-containing vehicle itself. These findings suggest that further work is required to characterize the parameters and mechanisms of lidocaine's effects on neurobehavioral devel- opment.

ACKNOWLEDGEMENTS

This work was supported by NICHHD grant No. RO1 HD20030-01A1 to Robert F. Smith. We thank Michael Lathan, Regina James and Chris Konczal for technical assistance, and TechAmerica for donation of vehicle for vehicle control injections.

REFERENCES

1. Anderson, P. L.; Bamburg, J. R. Effects of local anesthetics on nerve growth in culture. Rev. Neurosci. 4:273-290; 1981.

2. Blair, V. W.; Hollenbeck, A. R.; Smith, R. F.; Scanlon, J. W. Neonatal visual pattern preferences: modification by prenatal anes- thetic exposure? Dev. Med. Child Neurol. 2:476-483; 1984.

3. Chasnoff, I. J.; Bums, K. A.; Bums, W. J. Cocaine use in pregnancy: Perinatal morbidity and mortality. Neurotoxicol. Teratol. 9:291-293; 1987.

4. Chasnoff, I. J.; Bums, W. J.; Schnoll, S. H.; Bums, K. A. Cocaine use in pregnancy. N. Engl. J. Med. 313:666-669; 1985,

5. Hollenbeck, A. R.; Smith, R. F.; Edens, E. S.; Scanlon, J. W. Early trimester anesthetic exposure: incidence rates in an urban hospital population. Child Psychiatry Hum. Dev. 16:126-134; 1985.

6. Hollenbeck, A. R.; Grout, L. A.; Smith, R. F.; Scanlon, J. W. Neonates prenatally exposed to anesthetics: four year followup study. Child Psychiatry Hum. Dev. 17:66-70; 1986.

7. Holson, R. R.; Hansen, D. K.; LaBorde, J. B.; Bates, H. K. Prenatal lidocaine (L) exposure: effect upon selected behavioral measures in rat offspring. Teratology 37:521; 1988.

8. Lee, H. ; Nagele, R. G. Neural tube deficits caused by local anesthet- ics in chick embyros. Teratology 31:119-127; 1985.

9. Meaney, M. J.; Aitken, D. H.; van Berkel, C.; Bhatnagar, S.; Sopolsky, R. M. Effect of neonatal handling on age-related impair- ments associated with the hippocampus. Science 239:766-768; 1988.

10. Post, R. M.; Kennedy, C.; Shinohara, M.; Squillace, K.; Miyaoka, M.; Suda, S.; lngvar, D. H.; Sokoloff, L. Metabolic and behavioral

consequences of lidocaine-kindled seizures. Brain Res. 324:295-303; 1984.

11. Ryan, L.; Ehrlich, S.; Finnegan, L. Cocaine abuse in pregnancy: Effects on the fetus and newborn. Neurotoxicol. Teratol. 11:295- 299; 1987.

12. Schreiber, R. A.; Schlesinger, K. Circadian rhythms and seizure susceptibility: Effects of manipulations of light cycles on susceptibil- ity to audiogenic seizures and on levels on 5-hydroxytryptamine and norepinephrine in brain. Physiol. Behav. 8:699-703; 1972.

13. Schurr, A.; Spears, B.; Reid, K. H.; West, C. A.; Edmonds, H. L.; Rigor, B. M. Lidocaine depresses synaptic activity in the rat hippo- campal slice. Anesthesiology 64:501-503; 1986.

14. Smith, R. F.; Bowman, R. E.; Katz, J. Behavioral effects of exposure to halothane during early development in the rat: sensitive period during pregnancy. Anesthesiology 49:319-323; 1978.

15. Smith, R. F.; Mattran, K. M.; Kurkjian, M, F.; Kurtz, S. L. Alterations in offspring behavior induced by chronic prenatal cocaine dosing. Neurotoxicol. Teratol. 11:35-38; 1989.

16. Smith, R. F.; Wharton, G. G.; Kurtz, S. L.; Mattran, K. M.; Hollenbeck, A. R. Behavioral effects of mid-pregnancy administra- tion of lidocaine and mepivacaine in the rat. Neurobehav. Toxicol. Teratol. 8:61-68; 1986.

17. Spencer, D. G., Jr.; Emmett-Oglesby, M. W. Parallel processing strategies in the application of computers to the behavioral laboratory. Behav. Res. Methods Instrument. Comp. 17:294-300; 1985.

18. Teiling, A. K. Y.; Mohammed, A. K.; Minor, B. G.; Jarbe, T. U. C.;

PRENATAL LIDOCAINE AND OFFSPRING BEHAVIOR 403

Hiltunen, A. J.; Archer, T. Lack of effects of prenatal exposure to lidocaine on development of behavior in rats. Anesth. Analg. 66: 533-541; 1987.

19. Ward, S. L. D.; Schuetz, S.; Krishna, V.; Bean, X.; Wingert, W.;

Wachsman, L.; Keens, T. G. Abnormal sleeping ventilatory patterns in infants of substance-abusing mothers. Am. J. Dis. Child 140: 1015-1020; 1986.