Pharmacology, Biochemistry and Behavior 101 (2012) 602–608

Contents lists available at SciVerse ScienceDirect

Pharmacology, Biochemistry and Behavior

journal homepage: www.elsevier.com/locate/pharmbiochembeh

Gamma butyrolactone (GBL) and gamma valerolactone (GVL): Similarities and

differences in their effects on the acoustic startle reflex and the conditioned
enhancement of startle in the rat
Laureen J. Marinetti, Bonita J. Leavell, Calleen M. Jones, Bradford R. Hepler,
Daniel S. Isenschmid, Randall L. Commissaris ⁎
Department of Pharmaceutical Sciences, College of Pharmacy & Health Sciences, Wayne State University, Detroit, MI 48202, United States
Wayne County Medical Examiner's Office, 1300 East Warren, Detroit, MI 48207, United States
Montgomery County Coroner's Office and Miami Valley Regional Crime Laboratory, 361 West Third Street, Dayton, OH 45402, United States

a r t i c l e i n f o a b s t r a c t

Article history: Gamma butyrolactone (GBL) is metabolized to gamma hydroxybutyrate (GHB) in the body. GHB is a DEA
Received 29 June 2011 Schedule 1 compound; GBL is a DEA List 1 chemical. Gamma valerolactone (GVL) is the 4-methyl analog of
Received in revised form 24 January 2012 GBL; GVL is metabolized to 4-methyl-GHB; GVL is NOT metabolized to GBL or GHB. The effects of GBL
Accepted 28 January 2012
(18.75–150 mg/kg), GVL (200–1600 mg/kg) or vehicle on the acoustic startle reflex (ASR), and the
Available online 12 February 2012
classically-conditioned enhancement of startle, the Startle Anticipated Potentiation of Startle (SAPS) re-
Keywords:
sponse were studied in male rats. Both compounds produced a dose-dependent reduction of ASR, with GBL
Acoustic startle 5–7 times more potent than GVL. In contrast, GBL treatment significantly reduced SAPS at doses that exerted
Potentiated startle response only moderate effects on ASR, whereas GVL exerted little or no effect on the SAPS, except at doses that pro-
Gamma hydroxybutyrate duced pronounced reductions in Noise Alone ASR. In a second experiment, rats were tested for Noise Alone
4-Methyl-gamma hydroxybutyrate ASR behavior following treatment with a single mid-range dose of GBL (75 mg/kg), GVL (400 mg/kg) or ve-
Gamma valerolactone hicle; immediately following startle testing the animals were sacrificed and their brains and blood were col-
Gamma butyrolactone lected for determination of GHB, 4-methyl-GHB, GBL and GVL. GHB was found in measurable concentrations
Gamma-amino butyric acid (GABA)
in all of the blood specimens and 6 (of 8) of the brain specimens from the GBL-treated subjects. 4-Methyl-
GHB was found in measurable concentrations in all of the blood and brain specimens of the GVL-treated sub-
jects; the change in startle amplitude was inversely correlated to the brain concentrations of these com-
pounds. These findings confirm the differences in the metabolic fate of GBL and GVL as pro-drugs for the
formation of GHB and 4-methyl-GHB, respectively. Moreover, the dissimilarity in effect profile for GBL and
GVL on ASR versus SAPS behaviors suggests that different receptor(s) may be involved in mediating these be-
havioral effects.
© 2012 Published by Elsevier Inc.

1. Introduction
1.1. GBL and GHB effects and pharmacology
After gaining notoriety in some high-profile ‘date rape’ cases in the
1990s, the drug gamma hydroxybutyrate (GHB) was made a Schedule
1 drug in federal legislation in the form of the Hillory J. Farias and
Samantha Reid Date-Rape Drug Prohibition Act of 2000 to control
and penalize use and distribution of GHB, gamma butyrolactone
(GBL) and 1,4 butanediol (1,4 BD) (United States Drug Enforcement
Agency, 2000a). GBL is metabolized to GHB in the body following sys-
temic administration (Roth and Giarman, 1965, 1966); as a GHB pro-
⁎ Corresponding author at: 3126 Applebaum, Department of Pharmaceutical Sci-
ences, College of Pharmacy and Health Sciences, Wayne State University, Detroit, MI
48202, United States. Tel.: + 1 313 577 0813; fax: + 1 313 577 2033.
E-mail address: Commissaris@wayne.edu (R.L. Commissaris).
drug, GBL has therefore been identified by the DEA as List 1 chemical
(United States Drug Enforcement Agency, 2000b).
Studies on the receptor mechanism(s) for the effects of GHB have
focused primarily on two receptors: the GHB receptor and the GABAB
receptor. GHB binds to both receptors and they are functionally inde-
pendent (Kaupmann et al., 2003; Wu et al., 2004). GHB binds with
high affinity to the GHB receptor; this binding is displaced by the pu-
tative GHB receptor antagonist NCS-382. GHB also binds to the
GABAB receptor, although with somewhat lower affinity than at the
GHB receptor.
Many of the behavioral and physiological effects of administered
GHB appear to relate to actions at the low-affinity sites, i.e., GABAB re-
ceptors, rather than the higher affinity GHB receptor sites. GHB-
induced hypothermia in mice is reversed by treatment with the
GABAB antagonist SCH 5091, but not by the GHB antagonist NCS-382
(Carai et al., 2005). Similarly, GHB-induced hypothermia is prevented
(−/−)
in GABAB1 knockout mice (Queva et al., 2003). In Drosophila, GHB-

0091-3057/$ – see front matter © 2012 Published by Elsevier Inc.

doi:10.1016/j.pbb.2012.01.023
 L.J. Marinetti et al. / Pharmacology, Biochemistry and Behavior 101 (2012) 602–608
induced motor impairment in flying behavior is reduced by GABAB an-
tagonist treatment (CGP54626; the putative GHB receptor antagonist
NCS-382 was not evaluated) or by RNA interference (RNAi)-induced
knockdown of GABAB(1) receptors (Dimitrijevic et al., 2005). Finally,
in drug discrimination studies, the effects of GHB generalize to the
baclofen cue in both rats and pigeons (Carter et al., 2005).
There is, however, evidence that actions other than GABAB recep-
tor actions may be involved in some of the behavioral effects of GHB.
For example, although treatment with the GABAB antagonist
CGP35348 effectively, and in a dose-dependent manner, antagonized
the discriminative stimulus and rate-decreasing effects of baclofen
and the GABAB agonist SKF97541, this same CGP35348 treatment
was far less effective antagonist against the discriminative stimulus
and rate-decreasing effects of GHB (Carter et al., 2006).
1.2. GVL and 4-methyl-GHB effects and pharmacology
The compound gamma valerolactone (GVL) is the 4-methyl analog
of GBL that, until recently, was legally available via the internet as
‘Tranquili-G’. GVL has been touted on line as ‘an excellent Valium al-
ternative’ and also as a GHB alternative. GVL is converted to 4-
methyl-GHB in the body after administration and is not further me-
tabolized to GHB (Fishbein and Bessman, 1966). Fig. 1 depicts the
structures and metabolic relationships among the compounds GBL,
GHB, GVL and 4-methyl-GHB.
Compared to GHB, 4-methyl-GHB has been far less extensively
studied, but it, too binds to the GHB receptor with relatively high af-
finity; 4-methyl GHB also binds to GABAB receptors, but with a lower
affinity than GHB (Carter et al., 2005). Behaviorally, GHB and 4-
methyl GHB have both similarities and differences, depending upon
the behavioral measure being evaluated. Carter et al. (2005) have
reported that GHB and 4-methyl-GHB reduce locomotor activity, pro-
duce ataxia and catalepsy in a dose-dependent manner, but only GHB
produces a significant reduction in the righting reflex. Moreover, al-
though both GHB and 4-methyl-GHB treatment reduce overall oper-
ant response rates in the drug discrimination paradigm, 4-methyl-
GHB produces at best only weak generalization to the GHB ‘cue’ in
rats or pigeons trained to discriminate GHB (100 mg/kg in pigeons;
200 mg/kg in rats) from vehicle (Carter et al., 2005). These investiga-
tors also demonstrated that behaviorally active doses (i.e., doses that
reduce overall operant responding rates) of GHB, but not 4-methyl
GHB, generalized to the baclofen cue in rats trained to discriminate
3.2 mg/kg baclofen versus saline (Carter et al., 2004, 2005). Whether
these similarities and differences in GHB versus 4-methyl-GHB effects
relate to differences in GHB versus GABAB receptor binding has not
been determined.
Fig. 1. Chemical structures and metabolic relationships between gamma butyrolactone
(GBL), gamma hydroxybutyrate (GHB), gamma valerolactone (GVL) and 4 methyl
gamma hydroxy butyrate (4-methyl GHB). GBL is known to be metabolized to GHB
and GVL is known to be metabolized to 4-methyl GHB; there is no evidence for conver-
sion of GVL to GBL, nor from 4-methyl GHB to GHB.
603
1.3. The acoustic startle reflex and the potentiated startle response
The acoustic startle reflex in the rat has been used extensively as a
model for studying central nervous system (CNS) depressant com-
pounds (Helton et al., 1998; Davis, 1980; Pohorecky et al., 1976). In ad-
dition, the startle reflex can be augmented (‘potentiated’) by classical
conditioning. The potentiated startle response most commonly studied
is the ‘fear-potentiated startle’ response. In this classically-conditioned
fear behavioral procedure, the subject is ‘taught’ to anticipate an aver-
sive stimulus (electric shock) through multiple light + shock pairings
during training, and during testing, the magnitude of the startle re-
sponse is greater in the presence of the stimulus (light) that was previ-
ously paired with the delivery of the aversive stimulus. (Brown et al.,
1951; Davis, 1991; Davis et al., 1993; Helton et al., 1998).
A similar conditioned enhancement of startle can be observed
when the subject is trained to anticipate the startle noise burst itself
(Winston et al., 2001; McQueen et al., 2001). This startle anticipated
potentiated startle (SAPS) effect is based upon a report by Leaton
and Cranney (1990), who demonstrated that anticipation of the star-
tle stimulus increased the response to the noise burst. In the SAPS test-
ing procedure, rats are tested repeatedly, with each startle session
consisting of 10 Noise Alone trials and 10 Light + Noise trials
(McQueen et al., 2001; Winston et al., 2001). Initially there is no differ-
ence in startle amplitude for the two trial types. Within the course of
a single 20-trial test session, however, and continuing throughout
many weeks of repeated startle test sessions, startle amplitude on
the Light + Noise trials is consistently higher than on the Noise
Alone trials.
1.4. Purpose of the present studies
Both GHB-type and 4-methyl-GHB type compounds have been
found to reduce the Noise Alone acoustic startle response (Marinetti,
2003). The comparative effects of these compounds on the conditioned
enhancement of the startle response have not been determined. Thus,
the present studies evaluated the effects of the GHB pro-drug GBL and
4-methyl-GHB pro-drug GVL on ASR and also the SAPS procedure. In
addition, a second experiment was conducted to determine the meta-
bolic fate of GBL versus GVL as pro-drugs for GHB and 4-methyl-GHB,
respectively, and the possible relationship between their blood or
brain concentrations and the magnitude of the startle response.
2. Methods
2.1. Subjects
Male Sprague–Dawley rats (Charles River Laboratories, Wilmington,
MA) were used in all experiments. The rats weighed approximately 200
to 250 g upon arrival at the Department of Laboratory Animal Research
facilities in the Eugene Applebaum College of Pharmacy and Health Sci-
ences at Wayne State University. The rats were housed in groups of two
to four in a climate controlled room maintained on a 12 h light:12 h
dark cycle (lights on 0700 to 1900 h), with free access to food and
water, except during behavioral testing. All rats were allowed a one-
week acclimation time prior to any testing procedure. All procedures in-
volving experimental animals were reviewed and approved by the WSU
Animal Investigation Committee and followed all applicable NIH and
USDA guidelines.
2.2. Drugs
GVL was obtained from Fluka Chemical; GBL and GHB were
obtained from Sigma Chemical Company, St. Louis, MO. GVL was di-
luted to 10% by volume with de-ionized water for intraperitoneal
(IP) injections. GBL was diluted to 5% by volume with de-ionized
water for IP injections. Due to the chemical instability of GBL and
604 L.J. Marinetti et al. / Pharmacology, Biochemistry and Behavior 101 (2012) 602–608
the possible chemical instability of GVL, these aqueous solutions were
prepared fresh daily.
4-Methyl-GHB was synthesized for use in the analysis of 4-methyl-
GHB concentrations following GBL or GVL treatment. The synthesis of
4-methyl-GHB was undertaken utilizing the reaction scheme described
by Bourguignon et al. (1988). This reaction scheme involves the hydroly-
sis of gamma valerolactone with sodium hydroxide and ethanol, which
was then washed with 2-propanol to remove excess base and dried
with ethyl ether to yield the sodium salt of gamma methyl gamma
hydroxybutyric acid. The Na-methyl-GHB was then stored under dessi-
cation. This product was verified by nuclear magnetic resonance spec-
troscopy (NMR) and gas chromatography/mass spectrometry (GC/MS).
2.3. Apparatus—acoustic startle training and testing
The apparatus used for startle and potentiated startle testing and
training was the SR-LAB Startle Response System-R, purchased from
San Diego Instruments Inc. (San Diego, CA). The test chambers were
clear PlexiglasR cylinders, 15 cm long and 9 cm inside diameter, each
resting on a solid PlexiglasR base. An accelerometer was attached to
the base of the startle test cage cylinder and served as the transducer.
Cage movement resulted in a change in voltage output from the acceler-
ometer; the resultant voltage change was converted to a startle ampli-
tude value. The test cage and PlexiglasR support were enclosed in a
sound-attenuating cubicle. This cubicle was equipped with (1) a 15-
watt lamp which was used for potentiated startle training and testing
and (2) a speaker (Radio Shack Super TweeterR) for background noise
and for presentation of the acoustic startle stimuli. All parameters of
the acoustic startle test sessions (acoustic stimulus intensity, inter-
stimulus interval [ISI], presence or absence of the conditioned stimuli,
sampling interval, etc.) were controlled by a microprocessor using
‘PSR2’ software purchased from San Diego Instruments (SDI). The sensi-
tivity of each startle test chamber was standardized at the beginning of
each test day using the SR-LAB Standardization Unit purchased from SDI.
2.4. Behavior procedure—ASR and SAPS training and testing
The procedure for ASR and SAPS training and testing was a modi-
fication of that described previously (Winston et al., 2001; McQueen
et al., 2001). For each startle test session, subjects were placed indi-
vidually in the Plexiglas R test cylinder. All test sessions were preceded
by a 5-minute acclimation period (background noise only = 70 dB).
Each startle test session consisted of twenty (20) startle stimuli
(115 dB noise bursts; 40 ms in duration). Half of these startle stimuli
(Noise Alone) were presented in darkness and half (Light + Noise)
were presented during the final 40 ms of a 3540 ms presentation of
a 15-watt incandescent light. The order of presentation of the two
trial types was constant across all startle test sessions and was
based on a randomization scheme determined in earlier experiments
(Winston et al., 2001). The startle test chamber was dark during the
inter-trial interval (ITI). ITI values over the course of the 20 stimulus
presentations ranged from 26 to 34 s in 2-second increments. As
noted above, half of the trials in each session were of the ‘Light
+ Noise’ type and half of the trials were of the ‘Noise Alone’ type.
These daily startle sessions served two purposes. First, for the exper-
imental subjects, these test sessions established and maintained the
conditioned association between the light and the startle stimulus.
At the same time, these test sessions provided information regarding
the effects of prior conditioning (i.e., startle amplitude on the Noise
Alone versus Light + Noise trials) in the absence of extinction trials.
Thus, as with the Estes–Skinner fear-conditioned emotional response
(CER) operant conflict paradigm (Estes and Skinner, 1941), subjects
in these studies were tested repeatedly for baseline startle (Noise
Alone ASR) and for the effects of classical conditioning on the startle
response (SAPS), but they were never tested under conditions of ex-
tinction, i.e., presenting the light without the noise burst.
2.5. Drug treatments and ASR and SAPS startle behavior—procedure
All subjects were tested Noise Alone ASR and for the Startle-
Anticipated Potentiated Startle (SAPS) effect on 14 occasions over the
course of 7 weeks. These test sessions were approximately 15 min in
duration. Two test sessions were conducted each week, typically on
Monday and Thursday; these test sessions were separated by a period
of 2–4 days. For the first four sessions (SAPS acquisition and baseline
ASR determinations), no drug or vehicle treatments were administered.
For the last 10 sessions, i.e., over the course of the last 5 weeks of testing
(one treatment per week), the effects of acute challenges with GBL, GVL
or distilled water (DW) vehicle were determined. The effects of GBL
versus GVL were determined in a different group of subjects; GBL-
treated subjects received DW vehicle and a log-related range of GBL
doses (18.75, 37.5, 75, 150 mg/kg) over the course of 5 weeks; similarly,
GVL-treated subjects received DW vehicle and a log-related range of
GVL doses (200, 400, 800, 1600 mg/kg) over a similar time. All treat-
ments were administered intraperitoneally (IP), 10 min prior to the
start of the startle test session. Allowing for the 5-min pre-test accom-
modation period in the startle chamber, the effects of the drugs were
determined during the interval 15–25 min following IP injection. The
drug doses and the pretreatment interval were selected because they
produced mild to moderate sedation and ataxia at the indicated dose
ranges (Marinetti, 2003; Carter et al., 2005; Deichmann et al., 1945).
For each subject, the order of the five treatments received was random-
ized across the 5 weeks of SPS testing. In addition to the drug treatment
sessions, subjects were tested in one additional control (i.e., non-drug)
SAPS sessions each week.
2.6. GBL and GVL dispositional study
In a second experiment, groups of rats were pre-tested for their
Noise Alone ASR (average over 20 Noise Alone trials); they were
then treated with vehicle or a mid-range dose of GBL (75 mg/kg) or
GVL (400 mg/kg), and 10 min later Noise Alone ASR was re-assessed
for 20 trials. Immediately after testing, the subjects were sacrificed
by decapitation and trunk blood and brain tissue were collected on
ice for analysis of GHB, 4-methyl-GHB, GBL and GVL concentrations.
Both blood and brain samples were preserved with sodium fluoride,
2 to 3% by weight. Brain specimens were weighed and homogenized
with 3 mL of de-ionized water. After collection the blood and homog-
enized brain specimens were frozen until analysis.
2.7. GHB and 4-methyl-GHB quantitation
The method used for gas chromatography/mass spectrometry
(GC/MS) quantitation of GHB and 4-methyl-GHB in blood and brain
homogenates was a solid phase method adapted from Crifasi and
Telepchak (2000). 200 μL of rat blood or brain homogenate was
placed in small plastic test tubes containing 100 μL of deuterated
GHB (D6GHB) internal standard. 1.0 mL of acetone was added to
each tube to precipitate proteins and the tubes were vortexed for
15 s, then centrifuged. The acetone layer was removed and placed in
a clean glass test tube, then evaporated to dryness under nitrogen at
70 °C. The extract was reconstituted with 250 μL of 100 mM phos-
phate buffer, pH = 6.0, then vortexed for 15 s. The clean screen®
GHB solid phase extraction (SPE) column was conditioned as follows:
(1) pass 3 mL methanol through the column, then 3 mL de-ionized
water and finally 3 mL phosphate buffer from above. Low pressure
(b1 mm Hg) or no aspiration pressure was applied so as not to dry
the SPE columns. Each sample was applied to a conditioned SPE col-
umn; columns were not re-used. The solution was aspirated at
1 mm Hg or less. Clean glass test tubes were placed in the collection
rack under each SPE column. 1 mL methanol:ammonium hydroxide
(99:1) was added to each SPE column and the effluent was collected.
The contents of the collection tubes were transferred to individual
 L.J. Marinetti et al. / Pharmacology, Biochemistry and Behavior 101 (2012) 602–608
screw cap centrifuge tubes. The contents of the tubes were dried
under nitrogen at 70 °C. 100 μL BSTFA with 1% TMS was added to
the tube and it was placed in a heating block at 55 °C for 30 min.
Then 100 μL of chloroform was added to each tube and vortexed.
The tubes were then centrifuged for 10 min and the contents were
transferred from the centrifuge tubes to auto sampler vials and the
cap was crimped. 1.0 μL was injected onto the GC/MS and the follow-
ing ions were monitored: GHB-diTMS—233, 234 and 235, 4-Me-GHB-
diTMS—247, 248 and 231 and GHBd6-diTMS—239, 240 and 241.
The assay was performed in conjunction with GHB and 4-Me-GHB
calibrators (10, 50 and 100 mcg/mL), blanks (aqueous and biological)
and a 4-Me-GHB and GHB positive control. The limit of detection
(LOD) for this assay was 5 μg/mL and the limit of quantitation
(LOQ) was 5 μg/mL for 4-methyl-GHB and 10 μg/mL for GHB.
2.8. GBL and GVL quantitation
The method used to quantitate GBL and GVL was a one step liquid:
liquid extraction. 200 μL of the blood or brain homogenate was com-
bined with 4 mL of chloroform and a final concentration of 50 mcg/
mL internal standard (GBL for GVL quantitation and GVL for GBL
quantitation) in a 15 mL screw top test tube and placed on a rotator
for 10 min. The tubes were then centrifuged for 10 min and the
upper aqueous layer was aspirated to waste. The chloroform layer
was evaporated very carefully to approximately 1 mL. The remaining
chloroform layer was placed in a GC/MS auto-sampler vial and sealed
with a crimp cap. 1.0 μL was injected onto the GC/MS and the follow-
ing ions were monitored: GBL—86, 42 and 56 and GVL—85, 41 and 56.
The assay was performed in conjunction with the GBL and GVL cali-
brators (10, 50 and 100 mcg/mL), blanks (aqueous and biological)
and a GBL and GVL positive control in each matrix. The limit of detec-
tion (LOD) for this assay was 1.25 μg/mL and the limit of quantitation
(LOQ) was 1.25 μg/mL. The internal standard for this method was GBL
for the GVL analysis and GVL for the GBL analysis. This results in each
specimen being analyzed two times, once for GBL and again for GVL.
2.9. Dependent variables and statistical analyses
The primary dependent variable in this experiment was the max-
imum startle response and was measured as the maximum change
in velocity of cage movement as measured by the accelerometer
during the 100 ms interval beginning at the onset of the 40 ms
noise burst. Noise Alone ASR was defined as the average startle am-
plitude on the last nine Noise Alone trials for a given test session.
Similarly, startle amplitude across Light + Noise trials was defined
as the average startle amplitude across the last nine Light + Noise
trials for a given test session. Trial #1 (Light + Noise) and Trial #2
(Noise Alone) were excluded to minimize the effects of within-
session habituation responses in these measures. The magnitude of
the SAPS Startle effect was calculated for each subject and each ses-
sion and was defined as the difference between startle amplitude on
the Light + Noise trials minus startle amplitude on the Noise Alone
trials for that rat.
Statistically, the data from Experiment 1 were analyzed using a
3-way Factorial ANOVA, with Main Effects of Trial Type (two
levels, within subjects—Noise Alone versus Light + Noise), Drug
Dose, including vehicle (five levels, within subjects) and Drugs
(two levels, between subjects—GBL versus GVL). A similar analysis
that excluded vehicle-treatment data was also conducted (i.e.,
2 × 4 × 2 Factorial ANOVA); this second analysis was conducted be-
cause similarities in vehicle treatment outcomes might be expected
to ‘wash out’ potential differences in drug treatment responses. In
addition, the effects of treatment with individual doses of GBL or
GVL on the SAPS response were assessed using paired t-tests com-
paring Noise Alone Startle Amplitude to Light + Noise Startle Am-
plitude for each treatment.
605
In Experiment 2, the change in Noise Alone ASR (Pre-Treatment
minus Post-Treatment) following Vehicle, 75 mg/kg GBL or 400 mg/
kg GVL were compared using one-way ANOVA followed by post hoc
Student Newman Keuls (SNK) comparisons. In this Experiment, the
measured concentrations of GBL, GVL, GHB and 4-methyl GHB in
blood and brain were additional dependent variables. These drug
and metabolite concentrations in blood and brain tissue were com-
pared using ANOVA, followed by post hoc SNK comparisons. Finally,
correlations between the concentration of metabolite measured
(GHB following GBL treatment; 4-methyl-GHB following GVL treat-
ment) and the change in Noise Alone ASR behavior (post-drug
minus pre-drug) were evaluated using Pearson's correlation coeffi-
cient. In all statistical comparisons, p b 0.05 was used as the criterion
for statistical significance (Steele and Torrie, 1985).
3. Results
3.1. Non-drug noise alone ASR and SAPS behavior
There was no difference between startle amplitude on Noise Alone
versus Light + Noise trials on Test Day 1 (Noise Alone ASR: 292 ± 55
[Mean ± SEM]; Light + Noise ASR: 308 ± 66; t = 1.12, df = 14, n.s.).
Consistent with previous reports (Winston et al., 2001; McQueen et
al., 2001), over the course of 4 days of control (i.e.. non-drug) SAPS
startle test sessions, startle amplitude on Light + Noise trials grew to
be significantly greater than on Noise Alone trials. After these initial
sessions of SAPS testing, the magnitude of the SAPS response was rel-
atively stable for the remainder of the study (Noise Alone: 268 ± 45
[Mean ± SEM], Light + Noise: 339 ± 42; F[1,14] = 3.43, p b 0.05). The
magnitude of the SAPS enhancement (Light + Noise trials versus
Noise Alone trials) across non-drug sessions was approximately
25%. Overall, there was no difference in non-drug baselines for
Noise Alone ASR responses (F[1,14] = 1.61, n.s.) or for the magnitude
of the SAPS response (F[1,14] = 1.22, n.s.) in the two groups of rats
(GBL Study versus GVL Study).
3.2. GBL and GVL effects on noise alone ASR and SAPS behavior
Fig. 2 depicts the effects of GBL or GVL challenges on ASR and SAPS
behaviors. As can be seen, both drugs reduced Noise Alone ASR in a
dose-related manner. In addition, it can be seen that GBL treatment
reduced the SAPS response across a range of doses, whereas GVL
treatment failed to reduce the SAPS response, even at the doses that
dramatically reduced Noise Alone Startle. Statistically, 3-way Factori-
al ANOVA revealed a significant Main Effect for Trial Type (F[1,14] =
33.80, p b 0.05), as well as a significant Main Effect for Drug Dose
(F[4,56] = 17.03, p b 0.05); the Main Effect for Drug was not statisti-
cally significant (F[1,14] = 2.89, n.s.). The Drug × Dose interaction
(F[4,56] b 1.0), the Trial Type× Drug interaction (F[1,14] = 3.63)
and the Dose × Trial Type interaction (F[4,56] = 2.11) were not statisti-
cally significant. Finally, the 3-way interaction of Drug× Dose × Trial
Type (F[4,45] = 2.11, p b 0.05) was statistically significant. A second sta-
tistical analysis was conducted excluding the Vehicle treatment data,
since those data would tend to ‘wash out’ potential Drug treatment ef-
fect differences. This 3-way Factorial ANOVA again revealed significant
Main Effects for Trial Type (F[1,14] = 37.92, p b 005) and Drug Dose
(F[3,42] = 2.17, p0.05), with the Main Effect for Drug (F[1,14] =
3.34, n.s.) not being significant. The Dose ×Drug interaction (F[3,42] =
1.45) was not significant and the Dose × Trial Type (F[3,42] = 2.72,
p =0.056) was marginally significant. Perhaps most important, with
this analysis, the Trial Type × Drug interaction was statistically sig-
nificant (F[1,14] = 7.14, p b 0.05), meaning that GBL treatment
more dramatically affected the SAPS response (Noise Alone versus
Light + Noise) when compared to GVL treatment. Finally, the 3-
way interaction also was statistically significant (F[3,42] = 3.33,
p b 0.05). Paired t-tests for the SAPS response for the different
606 L.J. Marinetti et al. / Pharmacology, Biochemistry and Behavior 101 (2012) 602–608
Fig. 2. Effects of GBL or GVL treatment on Noise Alone ASR and the SAPS response. Plot-
ted are the Mean ± SEM values (N = 8) for startle amplitude on Noise Alone trials
(open bars) versus Light + Noise trials (cross-hatched bars). Factorial ANOVA revealed
significant Main Effects for Trial Types and for Drug Doses, plus a significant Drug (GBL
versus GVL) × Trial Type interaction; see text for details. #Noise Alone Startle Ampli-
tude is significantly different from Vehicle control for the indicated dose of GBL or
GVL, post hoc SNK comparison following Factorial ANOVA. *Significant SAPS effect for
the indicated treatment, i.e., Startle Amplitude on Light + Noise trials is significantly
different from Noise Alone trials for the indicated dose of GBL or GVL, paired t-test.
doses of the different drugs revealed a differential effect of GBL ver-
sus GVL on this measure, with a significant SAPS effect present for all
but the highest dose of GVL, whereas GBL treatment blocked the
SAPS effect for all except the lowest GBL dose.
Fig. 2 also depicts the effects of these GBL or GVL challenges on the
SAPS response (Light + Noise trials minus Noise Alone trials). GBL
treatment reduced the SAPS response at all doses at or above
37.5 mg/kg, whereas GVL treatment reduced the SAPS response only
at the highest dose tested (1600 mg/kg). At the 37.5 mg/kg dose,
GBL treatment selectively reduced the SAPS response, i.e., exhibited
no effect on Noise Alone ASR. In contrast, GVL treatment did not pro-
duce a selective reduction in the SAPS response at any dose. Indeed,
the SAPS response was present and robust following treatment with
400 mg/kg GVL and 800 mg/kg GVL, doses that reduced Noise Alone
startle by approximately 50% and 70%, respectively. Thus, the behav-
ioral effects of GBL and GVL on the Noise Alone ASR were qualitatively
similar, yet the effects of these two drugs were quite different for the
SAPS response.
Fig. 3. Effects of GBL and GVL treatment on Noise Alone acoustic startle. Plotted are the
Mean ± SEM values (N = 8) for the change in startle amplitude (from pretreatment
baselines) for rats treated with Vehicle (distilled water), 75 mg/kg GBL or 400 mg/kg
GVL. *Treatment-induced change in startle is significantly different from pretreatment
baseline, post hoc SNK test following Factorial ANOVA.
Table 1
Blood and brain concentrations following acute treatment with GBL or GVL.
Treatment Blood concentration (μg/mL)
GBLa GHBb GVLc 4-methyl GHBd
GBL (75 mg/kg) n.d 73.9 ± 11.7 n.d n.d
GVL (400 mg/kg) n.d n.d 1.6 ± 0.1e(n = 4) 822 ± 30
Brain concentration (μg/g)
GBL (75 mg/kg) n.d 8.8 ± 0.6e(n = 6) n.d n.d
GVL (400 mg/kg) n.d n.d 2.1 ± 0.3e(n = 5) 95 ± 6
Values represent mean ± SEM;
n.d—compound of interest not detected above LOD in any subject in that treatment
condition;
a
LOD for GBL assay = 1.25 μg/mL or μg/g
b
LOD for GHB assay 5.0 μg/mL or μg/g
c
LOD for GVL assay = 1.25 μg/mL or μg/g
d
LOD for 4-Me-GHB assay = 2.5 μg/mL or μg/g
e
GHB detected in 6 of 8 brain sample following GBL treatment; detected in 4 of
8 blood sample, and in 5 of 8 brain sample, following GVL treatment; all other values
represent Mean ± SEM derived from 8 subjects.
3.3. GBL and GVL startle behavior and disposition study
Fig. 3 and Table 1 depict the results of the GBL, GVL or vehicle
treatment on Noise Alone startle and metabolic disposition of GBL
and GVL. Pretreatment startle baselines did not differ across the
three treatment conditions (F[2, 21] = 1.63, n.s.). As seen in Fig. 3, ve-
hicle treatment did not affect startle amplitude relative to this pretreat-
ment baseline, but treatment with either 75 mg/kg GBL or 400 mg/kg
GVL treatment significantly reduced Noise Alone startle amplitude.
Statistically there was a significant Main Effect (F[2,21] = 7.32,
p b 0.05) for treatments (Veh/GBL/GVL) on the change in Noise Alone
ASR (pre-treatment minus post-treatment). Post hoc SNK tests
revealed that both 75 mg/kg GBL and 400 mg/kg GVL treatment
were significantly different from vehicle treatment. The difference
between GVL treatment and GBL treatment on the change in Noise
Alone ASR was not significant.
Table 1 illustrates that GBL-treated rats (all but two) had measur-
able concentrations of GHB in both blood and brain, but no measur-
able levels of 4-methyl GHB or GVL in either tissue. Conversely, all
GVL-treated rats had detectable concentrations of 4-methyl-GHB in
both blood and brain, but no measurable levels of GHB or GBL in ei-
ther tissue. The blood-to-brain ratio for GHB (GBL treatment) was
7.5 ± 3.7 (Mean ± SEM; n = 6), and the blood-to-brain ratio for 4-
methyl-GHB (GVL treatment) was 8.9 ± 1.5 (Mean ± SEM; n = 8).
Low levels of GVL were found in blood and brain of several animals
following GVL treatment, but there were no measurable GBL levels
following GBL treatment in any of the rats. The failure to detect GBL
in any of the GBL-treated rats, and the presence of GVL in the blood
(4 of 8 subjects) and brain (5 of 8 subjects) of GVL-treated rats likely
is due to the approximately five-times higher dose of GVL adminis-
tered compared to GBL.
There was a modest, but statistically significant, correlation be-
tween the GVL-induced reduction in startle amplitude and 4-
methyl-GHB brain concentrations (r = 0.64, p b 0.05) in GVL treated
subjects; the corresponding correlation between the GBL-induced re-
duction in startle amplitude and GHB brain concentrations in GBL-
treated subjects (r = 0.18, n.s.) was not statistically significant, per-
haps because of the reduced number of subjects (two GBL-treated
subjects did not have detectable GHB concentrations).
4. Discussion
GBL treatment resulted in a reduction in Noise Alone ASR startle
amplitude and also resulted in measurable levels of GHB in blood
and brain, but no formation of 4-methyl-GHB in either tissue. Simi-
larly, GVL treatment resulted in a reduction in Noise Alone ASR
 L.J. Marinetti et al. / Pharmacology, Biochemistry and Behavior 101 (2012) 602–608
startle amplitude and an increase in 4-methyl-GHB in both blood and
brain, but no formation of GHB in either tissue. These data confirm the
independence of the metabolic pathways for the compounds GBL and
GVL; moreover, the correlation between 4-methyl-GHB concentra-
tion and the GVL-induced change in startle amplitude further sup-
ports the 4-methyl-GHB pro-drug actions of GVL.
Both GBL and GVL treatment affected the startle response, but
there were some important differences between the effects of the
two drugs. GBL and GVL produced similar dose-dependent reductions
in Noise Alone startle; although GBL was approximately 5–7 times
more potent in this regard, both drugs exhibited full efficacy, as evi-
denced by the reduction of Noise Alone startle amplitude to near
zero levels following the highest doses examined. These doses were
associated with observed moderate levels of ataxia and incoordina-
tion; again GHB was approximately 5–10 times more potent than 4-
methyl-GHB in producing these effects. This finding is consistent
with the report by Carter et al. (2005) that GHB and 4-methyl-GHB
produced similar degrees of locomotor activity reduction, ataxia,
and also a similar degree of reduction of overall operant level pressing
in the drug discrimination procedure, at doses at or near those used in
the present study. This finding is also consistent with the well-known
CNS depressant effects of GBL and GHB and the anecdotal reports on
the effects of GVL in humans following the use of ‘Tranquili-G’ pur-
chased via the internet and the effects of GVL on animal behavior as
noted by Deichmann et al. (1945). The present findings of qualitative-
ly similar effects on Noise Alone startle with GBL being 5–7 times
more potent than GLV also are consistent with similar effects of
GHB and 4-methyl GHB to displace 3H NCS-382 binding at GHB recep-
tors (Carter et al., 2005).
With respect to the conditioned enhancement of startle (SAPS ef-
fect) however, GBL and GVL exerted qualitatively very different ef-
fects. GBL reduced the SAPS response across a wide range of doses,
beginning at a dose (37.5 mg/kg) that did not significantly reduce
Noise Alone startle amplitude. In contrast, GVL treatment failed to re-
duce the SAPS response even at several doses that significantly re-
duced Noise Alone startle; indeed it was only at the highest dose
examined (1600 mg/kg) that GVL also reduced the SAPS response.
Thus, the effect of GBL to reduce the SAPS conditioned increase in
startle at low to moderate doses was not shared by GVL.
This qualitative GBL versus GVL difference in the SAPS response is
consistent with a previous study by Carter et al. (2005) with respect
to GHB versus 4-methyl-GHB effects in drug discrimination. In both
rats (200 mg/kg GHB) and pigeons (100 mg/kg GHB) trained to dis-
criminate GHB from saline, challenges with 4-methyl-GHB significantly
reduced operant response rates, yet exhibited little if any generalization
to the GHB cue. Moreover, although GHB treatment would generalize
to a baclofen (3.2 mg/kg) cue in drug discrimination in rats, 4-
methyl-GHB treatment again failed to reliably generalize to the bac-
lofen cue. Together, these data are consistent with the hypothesis
that changes in the SAPS response produced by GBL, but not GVL in
the present study, may relate to GABAB receptor actions of GHB,
but not 4-methyl-GHB. Clearly, further studies are needed to test
this hypothesis.
5. Conclusions
The present studies demonstrate that GBL and its 4-methyl
substituted analog GVL have different metabolic fates in the rat,
with GBL serving as a pro-drug for GHB and GVL serving as a pro-
drug for 4-methyl-GHB. Moreover, although there are qualitative
similarities in the behavioral effects of GBL and GVL on the Noise
Alone Acoustic Startle response, there are notable differences be-
tween these compounds with respect to the SAPS response. The ex-
tent to which these similarities and differences may relate to
differential interactions at GHB and/or GABAB receptors, and possibly
other receptors, remains to be determined.
607
Acknowledgements
These studies were supported in part by GM08167 and AA12435
to RLC and by the Department of Pharmaceutical Sciences and the Ro-
land T. Lakey Research Fund, WSU College of Pharmacy & AHP. BJL
and CMJ were supported in part by GM08167 and also by the Depart-
ment of Pharmaceutical Sciences.
References
Bourguignon JJ, Schoenfelder A, Schmitt M, Wermuth CG, Hechler V, Charlier B, et al.
Analogues of gamma-hydroxybutyric acid. Synthesis and binding studies. J Med
Chem 1988;31(5):893–7.
Brown JS, Kalish HI, Farber IE. Conditioned fear as revealed by magnitude of startle re-
sponse to an auditory stimulus. J Exp Psychol 1951;41:317–28.
Carai MAM, Colombo G, Gessa GL. Resuscitative effect of gamma aminobutyric acid B
receptor antagonist on gamma-hydroxybutyric acid mortality in mice. Ann
Emerg Med 2005;45(6):614–9.
Carter LP, Chen W, Coop A, Koek W, France CP. Discriminative stimulus effects of GHB
and GABAB agonists are differentially attenuated by CGP35348. Eur J Pharmacol
2006;538:85–93.
Carter LP, Chen W, Wu H, Mehta AK, Hernancdez RJ, Ticku MK, et al. Comparison of the
behavioral effects of gammahydroxybutyric acid (GHB) and its 4-methyl-
substituted analog, gamma-hydroxyvaleric acid. Drug Alcohol Depend 2005;78:
91–9.
Carter LP, Wu H, Chen W, Cruz CM, Lamb RJ, Koek W, et al. Effects of gamma-
hydroxybutyrate (GHB) on schedule-controlled responding in rats: role of GHB
and GABAB receptors. J Pharmacol Exp Ther 2004;308:182–8.
Crifasi JA, Telepchak M. A solid phase method for gamma-hydroxybutyrate (GHB) in
blood, urine, vitreous or tissue without conversion to gamma-butyrolactone
(GBL) using United Chemical Technologies' ZSGHB020 or CSGHB203 solid phase
extraction columns — personal communication and method brochure; 2000.
Davis M. Neurochemical modulation of sensory-motor reactivity: acoustic and tactile
startle reflexes. Neurosci Biobehav Rev 1980;4(2):241–63.
Davis M. Animal models of anxiety based on classical conditioning: the conditioned
emotional response and the potentiated startle effect. In: File SE, editor. Psycho-
pharmacology of anxiolytics and antidepressants. New York: Pergamon Press;
1991. p. 187–212.
Davis M, Falls WA, Campeau S, Kim M. Fear-potentiated startle: a neural and pharma-
cological analysis. Behav Brain Res 1993;58:175–98.
Deichmann WB, Hirose R, Witherup S. Observations on the effects of gamma-
valerolactone upon experimental animals. J Ind Hyg Toxicol 1945;27(9):263–8.
Dimitrijevic N, Dzitoyeva S, Satta R, Imbesi M, Yildiz S, Manev H. Drosophila GABAB re-
ceptors are involved in behavioral effects of gamma-hydroxybutyric acid. Eur J
Pharmacol 2005;519:246–52.
Estes WK, Skinner BF. Some quantitative properties of anxiety. J Exp Psychol 1941;29:
390–400.
Fishbein WN, Bessman SP. Purification and properties of an enzyme in human blood
and rat liver microsomes catalyzing the formation and hydrolysis of gamma-
lactones. J Biol Chem 1966;241(21):4835–41.
Helton DR, Tizzano JP, Monn JA, Schoepp DD, Kallman MJ. Anxiolytic and side-
effect profile of LY354740: a potent, highly selective, orally active agonist for
group II metabotropic glutamate receptors. J Pharmacol Exp Ther 1998;284(2):
651–60.
Kaupmann K, Cryan JF, Wellendorph P, Mombereau C, Sansig G, Klebs K, et al. Specific
gamma-hydroxybutyrate binding sides but loss of pharmacological effects of
gamma-hydroxybutyrate in GABAB(1)-deficient mice. Eur J Neurosci 2003;18:
2722–30.
Leaton RN, Cranney J. Potentiation of the acoustic startle response by a conditioned
stimulus paired with acoustic startle stimulus in rats. J Exp Psychol Anim Behav
Process 1990;16(3):279–87.
Marinetti , L.J.: (2003). The pharmacology of gamma valerolactone (GVL) as compared
to gamma hydroxybutyrate (GHB), gamma butyrolactone (GBL), 1,4 butanediol
(1,4BD), ethanol (EtOH) and baclofen (BAC) in the rat. LJM doctoral dissertation.
ETD collection for Wayne State University.
McQueen DA, Overstreet DH, Ardayfio PA, Commissaris RL. Acoustic startle, condi-
tioned startle potentiation and the effects of 8-OH-DPAT and buspirone in rats se-
lectively bred for differences in 8-OH-DPAT-induced hypothermia. Behav
Pharmacol 2001;12:509–16.
Pohorecky LA, Cagan M, Brick J, Jaffe LS. The startle response in rats: effect of ethanol.
Pharmacol Biochem Behav 1976;4(3):311–6.
Queva C, Bremmer-Danielsen M, Edlund A, Ekstarand AJ, Elg S, Erickson S, et al. Effects
of GABA agonists on body temperature regulation in GABAB(1) (−/−)
mice. Br J Pharma-
col 2003;140:315–22.
Roth RH, Giarman NJ. Preliminary report on the metabolism of gamma-butyrolactone
and gamma-hydroxybutyric acid. Biochem Pharmacol 1965;14:177–8.
Roth RH, Giarman NJ. Gamma-butyrolactone and gamma-hydroxybutyric acid-I distri-
bution and metabolism. Biochem Pharmacol 1966;15:1333–48.
Steele RGD, Torrie JH. Principles and procedures of statistics. Inc, New York: McGraw-
Hill Book Co.; 1985.
United States Drug Enforcement Agency. Controlled Substances Act (21 U.S.C.
802(34)): US Department of Justice. Final rule. Fed Regist 2000a;65:21645.
608 L.J. Marinetti et al. / Pharmacology, Biochemistry and Behavior 101 (2012) 602–608

United States Drug Enforcement Agency. Placement of gamma-butyrolactone in List I
of the Controlled Substances Act (21 U.S.C. 802(34)): United States Department
of Justice. Final rule. Fed Regist 2000b;65:21645.
Winston CR, Leavell BJ, Ardayfio PA, Beard C, Commissaris RL. A nonextinction proce-
dure for long-term studies of classically conditioned enhancement of acoustic star-
tle in the rat. Physiol Behav 2001;73:9-17.
Wu Y, Ali S, Ahmadian G, Liu CC, Want YT, Gibson KM, et al. Gamma-hydroxybutyric
acid (GHB) and gamma-aminobutyric acidB receptor (GABABR) binding sites are
distinctive from one another: molecular evidence. Neuropharmacology 2004;47:
1146–56.