Intraocular Injections of Nipecotic Acid Produce a Preferential Block of Neuronal 3H-GABA Accumulation in Adult Rabbit Retina

Home , Nipecotic acid

Paul Madres, Jr. and Dianna A. Redburn

A procedure by which the activity of the retinal GABA uptake system can be manipulated in vivo has been developed. Intraocular injections of nipecotic acid, a proported GABA uptake blocker, were administered to adult rabbits every 48 hours for a two-week period. No behavioral or systemic changes were observed. Injections were well-tolerated with less than 10% loss of the tissue caused by physical damage or injection. Biochemical analyses demonstrated a dose-dependent inhibition of 14C-GABA uptake into retinal tissue. No effect on uptake was observed for saline-treated tissue. Autoradiographic analyses showed that in vivo treatment with nipecotic acid preferentially blocked accumulation of 3H- GABA into the amacrine cell bodies and processes in the inner plexiform layer. This treatment may be especially useful in assessing the functional significance of GABA transport in vivo. Invest Ophthal- mol Vis Sci 24:886-892, 1983

A specific, high affinity uptake system for 7-ami- rons, we wished to develop a procedure by which the nobutyric acid (GABA) has been well-characterized activity of the uptake system could be manipulated in a variety of neuronal tissues and is believed to serve in vivo. We chose to study a chronic treatment of several important functions. A portion of the GABA nipecotic acid, a proported GABA uptake blocker. transport sites is located on presynaptic terminals Our results demonstrated that multiple intraocular and serve to replenish releaseable pools of the neu- injections of nipecotic acid cause a preferential in- rotransmitter. In many instances, uptake also serves hibition of neuronal accumulation of 3H-GABA with to terminate the action of GABA after there has been no observable systemic side effects. This preparation a termination of release, thus allowing the system to may be potentially useful for assessing the functional return a resting state.1"3 significance of GABA transport systems in vivo. In the rabbit retina, GABA has been identified as a major inhibitory neurotransmitter45 and is released Materials and Methods from a population of amacrine cells to inhibit do- Adult (2-kg) New Zealand white rabbits were ob- paminergic6 and cholinergic transmission.7 Some of tained from a local breeder and housed singly for two the GABA appears to be released tonically in rabbit weeks in a light-controlled (12-hour light/dark cycle) retina,8 and thus the role of the GABA uptake system room. Each animal was maintained on commercial may be especially important. The concentration of rabbit chow. Fresh water was given daily. For ap- GABA in the synaptic cleft reflects the balance be- proximately 30 minutes each day, animals were han- tween release and uptake, and thus the activity of dled gently to increase environmental interaction. the uptake system can directly influence the duration and size of the postsynaptic response. Injection Because the uptake system appears to be a critical component of normal function in GABA-ergic neu- Each animal was anesthetized every 48 hours with an acepromazine-ketamine-rompun cocktail (0.7, 21.4, and 4.3 mg/kg, respectively) via an intramus- From the Department of Neurobiology and Anatomy, The Uni- versity of Texas Medical School, Houston, Texas. cular injection in the hind limb. Thirty minutes post- Supported by the National Institutes of Health grants 5 F32 EY injection, the orbit of the eye to receive nipecotic acid 05475-02 (P.M.), EY 0-1655 (DAR), and RCDA 1K04 EY 00088 treatment was bathed with 1% lidocaine hydrochlo- (DAR). ride. The eyelids were restrained by a lid speculum. Submitted for publication July 23, 1982. Using a 0.12 Castra Viejo forceps, the superior rectus Reprint requests: Paul Madtes, Jr., PhD, Department of Neu- robiology and Anatomy, The University of Texas Medical School, muscle was grasped gently to stabilize the globe. A Houston, TX 77025. 27 gauge X 0.5 in gauge needle was inserted caudal

0146-0404/83/0700/886/$ 1.15 © Association for Research in Vision and Ophthalmology

886

Downloaded from iovs.arvojournals.org on 09/25/2021 No. 7 NIPECOTIC ACID CLOCK OF NEURONAL 3H-GADA UPTAKE IN VIVO / Modres and Redburn 887

to the muscle insertion, thus assuring that the lens buffer, 221.3 mM choline Cl was added in place of was not pierced during injection. One hundred mi- NaCl). AOAA was included in order to inhibit trans- croliters of one of the following were injected: saline, amination of GABA. 3H-dopamine (final concentra- 1 raM, 10 mM, or 100 raM nipecotic acid (Sigma) tion = 1 /iM; 20.1 Ci/mmole, Amersham) and 14C- dissolved in saline. The non-injected eye served as GABA (final concentration = 10 nM; 267 Ci/mmole, the control for each experimental animal. The orbit Amersham) were added and the samples were incu- was rinsed with gentamicin sulfate to reduce infection bated for 10 min. The samples then were poured onto after injection. The final concentration in the vitreal glass fiber niters (Whatman GF/A) and rinsed three chamber was calculated to be approximately a 10- times with the appropriate buffer (sodium-free or so- fold dilution of the injection solution. dium-containing). The filters containing the samples were removed, soaked in 1% SDS and 20 mM EDTA Observation solution overnight, and counted in a toluene-based solution with Triton X-100, by double label liquid A daily record of water and food intake, morning scintillation counting. activity level, feces appearance, response to sound Since specific uptake of GABA has been shown to and touch, and any irregular behavior was main- be sodium- and temperature-dependent, results were tained for each animal. Three hours after anesthesia, calculated by correcting for background and protein each animal was tested for pupillary light reflex and content. Specific uptake of 3H-GABA in retinal ho- contralateral closure. Each animal also was tested for mogenates was determined by subtracting nonspecific visual reflexes on days during which they did not re- uptake (the amount of uptake at 0 C in sodium-free ceive an injection. The pupillary light reflex was buffer) from total uptake (the amount of uptake at checked by placing the animal in total darkness and, 37 C in sodium-containing buffer) and dividing by after allowing acclamation, flashing a penlight source the amount of protein in each sample. The data are in one eye. Contralateral closure was followed under expressed as percentage ratios of experimental to con- identical conditions, shining a light in one eye and trol values. observing the change in pupil diameter in the contralateral eye. Autoradiography Retinal Isolation The second retinal piece was bisected and both In order to minimize the influence of acute effects halves were incubated at 37 C for 15 min in an ox- of nipecotic acid on this chronic study, the animals ygenated (95% O2/5% CO2) Ringer's buffer (118.3 were killed by decapitation under normal laboratory mM NaCl, 25.2 mM NaHCO3, 4.7 mM KC1, 2.5 mM CaCl2, 1.2 mM MgSO4, 1.2 mM KH2PO4, and illumination 24 hours after the final injection, at 3 which time most of the 3H-nipecotic acid is cleared 11.1 mM glucose), pH 7.4 containing 10 nM. H- from the retina (Madtes, unpublished observations). GABA (57 Ci/mmole, Amersham). One-half then The eyes were enucleated, hemisected, and the vit- was fixedimmediatel y in 2.5% glutaraldehyde in 0.05 reous discarded. Each eyecup then was divided into M sodium cocadylate buffer, pH 7.2. The remaining two pieces with a no. 21 scalpel blade for analysis. half was incubated an additional 30 min in fresh buffer containing no GABA, fixed for 30 min at room Each injected globe was inspected for signs of phys- temperature, and stored overnight at 15 C. Each sam- ical damage such as scleral scars and infection. Sim- ple then was post-fixed in 1% osmium tetroxide for ilarly, retinas were scrutinized for scarring, detach- 1 1/2 hr at 4 C. After osmication, the samples were ment, or hemorrhage. Any defective sample was dis- dehydrated in a graded alcohol series followed by carded. propylene oxide and embedded in Epon epoxy resin for sectioning. Thick sections (1 n) were cut with glass Biochemical Analysis knives and placed on glass slides cleaned with nitric The first retinal piece was homogenized in 10-fold acid. Representative sections were stained with Ri- 0.32 M sucrose with a Teflon® pestle in a glass ho- chardson's stain and assessed for morphologic integ- mogenizer tube (Thomas). Homogenate samples (ap- rity. The remaining slides were coated with Kodak® proximately 0.2 mg per tube) were incubated for 6 NTB-2 emulsion, allowed to dry, and placed in light- min in either a sodium-free (0 C) or sodium-con- tight boxes that contained dessicant for 4 weeks. The taining (37 C) modified HEPES buffer (20 mM sections then were developed in 1:1 Dektol for 2 min, HEPES, 150 mM NaCl, 6.2 mM KC1, 1.2 mM rinsed, fixed in Kodak® Fixer for 5 min, washed, and dried at 60 C. Some sections were stained with to- MgSO4, 3 mM CaCl2, 10 mM glucose, and 1 mM amino-oxyacetic acid (AOAA), pH 7.4; for Na-free luidine blue for comparison with unstained sections.

Downloaded from iovs.arvojournals.org on 09/25/2021 888 INVESTIGATIVE OPHTHALMOLOGY 6 VJSUAL SCIENCE / July 1983 Vol. 24

Fig. I. Lack of effect of chronic nipecotic acid treatment on the morphology of adult rabbit retina. Adult rabbits were given intraocular injections of nipecotic acid (final concentrations = 10 mM) every 48 hr for a 2-week period. Animals were decapitated and the eyes were enucleated, hemisected, and the vitreous discarded. Retinas were fixed, sectioned, and stained with Richardson's stain. A(left). Control retina (untreated) B(right). Treated retina. No difference in the morphologic appearance was observed (Magnification X400).

Protein Determination sembly (0.2 /*m; Gelman) and into a sterilized vial Protein was determined according to the method fitted with a rubber septum. The solutions were stored of Lowry et al.9 Bovine serum albumin was used as at 15 C until used. Saline solutions for injections also were sterilized and stored as above. the reference standard. Results Injection Solution Preoparation Tolerance of Injection Procedures Nipecotic acid (SIGMA) was dissolved in saline Adult rabbits received intraocular injections every and passed through, a disposable Acrodisc® filter as- 48 hours during a two-week period. There was little

Downloaded from iovs.arvojournals.org on 09/25/2021 No. 7 NIPECOTIC ACID BLOCK OF NEURONAL 3H-GABA UPTAKE IN VIVO /Modres ond Redburn 889

evidence for any significant adverse effect caused by 100 the anesthesia or the injection. No changes in water r = 0.830 intake, food intake, urine output, feces output, re- 90 . sponse to sound or touch, or general behavior were KEY observed. Similarly, no change in visual reflexes were A 14C-GABA

noted, except for the animals receiving 10 mM ni- 3 pecotic acid, in which a dimunition of pupillary light a H-DOPAMINE reflex was seen for the injected eye and almost total CO A SALINE INJECTED loss of contralateral closure for both eyes, these observations being found only on the day of injection. Inspection of enucleated eyes revealed little physical damage to the globe, the only observable damage being scarring of the sclera around the injection site. This resulted from repeated intrusion into the tissue by the needle. Ocassionally, a hemorrhage that resulted from puncture of a blood vessel during injection was found. Small vitreal scars were found along the tract of injection. The vast majoriy of retinal samples showed no damage resulting from the injection. No retinal detachment was observed. Less than 10% of the samples were eliminated because of physical damage or infection. Microscopic analyses of control LOG (NIPECOTIC ACID) (mM) and injected retinas did not reveal any significant Fig. 2. Uptake of 14C-GABA and 3H-dopamine by adult rabbit morphologic damage caused by the injections retina. Adult rabbits were given intraocular injections of nipecotic (Fig. 1). acid (final concentration = 10 mM) every 48 hr for a two-week period. Retinal homogenates were incubated for 10 min in 10 /xM Biochemical Analysis I4C-GABA and 1 pM 3H-dopamine and uptake was determined by liquid scintillation counting. Data are expressed as specific up- Rabbits received intraocular injections of 0, 1, 10, take (uptake at 37 C in the presence of Na+ minus uptake at 0 C, or 100 ^niol of nipecotic acid in 100 p\ saline every Na+-free), divided by protein content. Points given are percentage 48 hr. Assuming uniform dispersion of the injected ratios of experimental-to-control values. A dose-dependent de- 14 fluid, the final concentrations of the compound were crease in C-GABA uptake was found. No change in the uptake of 3H-dopamine was found. Similarly, no effect on uptake was approximately 0.1, 1, and 10 mM. Retinas were iso- found for saline-injected animals, the uptake for I4C-GABA being lated, homogenized in isotonic sucrose, and then in- 100% of control values. The correlation coefficient (r) was obtained l4 cubated in the presence of 10 nM C-GABA. After by linear regression analysis. incubation, the samples were washed and counted in order to determine uptake activity. The amount of uptake present at 0 C in Na-free medium was taken mine was not changed significantly after nipecotic as a measure of nonspecific binding, and was sub- acid treatment. tracted from the uptake observed at 37 C in Na-containing medium in order to obtain a measure of spe- Autoradiographic Analysis cific uptake activity. The results are plotted in Fig. We examined the effect of chronic in vivo treat- 2. As seen in the graph, nipecotic acid caused a dose- ment with 10 mM nipecotic acid on the accumulation 14 dependent inhibition of C-GABA uptake. Uptake patterns of 3H-GABA. These results are shown in Fig. activity was 70% of control values at 0.1 mM nipe- 3. Control tissue that received no injections showed cotic acid, 40% of control values at 1 mM nipecotic a predominantly glial accumulation after a 15-min acid, and 24% of control values at 10 mM nipecotic in vitro incubation in 10 pM 3H-GABA. Intense ac- 14 acid. No effect on C-GABA uptake was found for cumulation into Mtiller cell bodies and footpads was saline-injected tissue, the uptake being 100% of con- evident. However, after a 30-min post-incubation in trols. buffer, the neuronal labelling pattern was found. The In one series of experiments, 3H-dopamine was in- characteristic amacrine cell body accumulation and cluded in the incubation medium along with 14C- trilaminar distribution of axon terminal labelling can GABA, in order to determine if nipecotic acid af- be seen. Some displaced amacrine cells also were la- fected the transport of dopamine. Pargyline (10 /iM) beled. Identical labelling patterns were observed in and ascorbate (1 nM) also were added to the incu- saline-injected samples. bation medium in order to prevent metabolism and In vivo treatment with nipecotic acid had pro- oxidation of the 3H-dopamine. Uptake of 3H-dopa- nounced effects on neuronal accumulation of 3H-

Downloaded from iovs.arvojournals.org on 09/25/2021 890 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / July 1983 Vol. 24

A2 Fig. 3. Preferential Loss of Neuronal 3H-GABA Labeling after Chronic Nipecotic Acid Treatment. Adult rabbits were given intraocular injections of nipecotic acid (final concentration = 10 mM) every 48 hr for a 2-week period. Animals were decapitated and the eyes were enucleated, hemisected, and the vitreous discarded. Retinas were processed for 3H-GABA autoradiography. For each experimental group, autoradiographic results are shown (1, top panels) at X400 magnification, stained with toluidine blue in order to localize the labeling, and (2, bottom panels) at X250 magnification, unstained to emphasize the labeling pattern. A, control retina incubated 15 min in 3H-GABA; B, control retina incubated 15 min in 3H-GABA and post-incubated 30 min in buffer; C, nipecotic acid-treated retinal incubated 15 min in 3H-GABA; D, nipecotic acid-treated retina incubated 15 min in 3H-GABA and post-incubated 30 min in buffer. During a 15-min in vitro incubation, 3H-GABA is accumulated preferentially by glial cells in both control and treated tissues (A, C). After a 30-min post- incubation in buffer, labeling of neurons (amacrine cells) is seen only in control retina (B). There is a preferential loss of neuronal labeling in the nipecotic acid-treated retina (D). (Unstained; original mag. magnification = X25O).

GABA. A typically glial labelling pattern was seen a method for manipulation of that system to analyze after a 15-min in vitro incubation of the tissue in 10 the factors that regulate it in the rabbit retina. The fiM 3H-GABA that is similar to the control pattern. current investigation was limited to the analysis of However, the 30-min post-incubation in buffer re- chronic treatment with nipecotic acid over a 2-week vealed a pattern that was markedly distinguishable period in order to identify the long-term effects of from control. While some residual labelling in Muller altering GABA uptake. Subsequent studies will be cells is present, neuronal accumulation was virtually required to examine acute effects. The use of multiple absent in the treated tissue. intraocular injections of nipecotic acid was found to be highly successful. Injections were well-tolerated. Discussion Similarly, no significant damage to the retinas was seen either macroscopically or microscopically. The The importance of a high affinity uptake system appearance of vitreal scars along the line of injection for GABA has been well-established. We report here was evident. These scars resulted from the damage

Downloaded from iovs.arvojournals.org on 09/25/2021 No. 7 NIPECOTIC ACID CLOCK OF NEURONAL °H-GADA UPTAKE IN VIVO / Madres ond Kedburn 891

>•*•".

C1 D1

C2 D2

Fig. 3. (Continued)

that occurred in the vitreous from intrusion by the imals receiving the highest concentration. These ob- needle since it was evident in all animals receiving servations show that the effects of nipecotic acid are injections, both saline and nipecotic acid. The oc- direct effects on the retina itself and are not caused currence of tissue loss caused by infection was low- by an indirect behavioral or systemic effect. ered to approximately 10% of the eyes injected by The effects of in vivo treatment of nipecotic acid prophylactic treatment with gentamicin sulfate. on the GABA uptake system were assessed biochem- Careful observation was made for behavioral or ically and autoradiographically. Retinal homogenates systemic effects of nipecotic acid. No changes were were analyzed by liquid scintillation counting to found during the experimental period in water intake, quantify the nipecotic acid inhibition of GABA up- food intake, urine output, feces output, response to take. A dose-dependent decrease in 14C-GABA up- touch or sound, general activity level, or general be- take was found with maximal block being at the high- havior, except for the animals that contracted Pas- est concentration of nipecotic acid injected. An ap- teurella multocida. These animals were treated daily proximate ICso of 0.68 mM was calculated from these with penicillin until recovery, or destroyed and not data. This inhibition appears to be specific as no dose- utilized in the experiments. dependent effect was observed for 3H-dopamine up- Similarly, no changes were found in visual reflexes, take. Similarly, no effect on uptake was found for except, on the day of injection only, a diminution of saline-injected retinas. Preliminary results suggest pupillary light reflex in the nipecotic acid-injected eye that the inhibition of GABA uptake is reversible in and almost total loss of contralateral closure for an- that 5-10 days after termination of treatment, the

Downloaded from iovs.arvojournals.org on 09/25/2021 892 INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / July 1983 Vol. 24

effect appears to decrease. We conclude that the aration in which neuronal GABA transport is blocked GABA uptake system in the adult retina is sensitive preferentially. This preparation demonstrates that to in vivo nipecotic acid treatment and that this treat- nipecotic acid treatment is a useful tool in directly ment is specific to the GABA system. affecting the uptake system for GABA. This treat- Intact retinas were analyzed using 3H-GABA au- ment may be especially useful in assessing the func- toradiography to localize the inhibitory effect of in tional significance of GABA transport in vivo. vivo nipecotic acid treatment. The highest concentration was chosen to maximize the visibility of the 3 3 Key words: amacrine cells, GABA uptake, H-GABA au- effect. The normal labelling pattern for H-GABA in toradiography, intraocular injections, in vivo, Miiller cells, mammalian retina demonstrates that both Miiller nipecotic acid cells and amacrine cells accumulate GABA. Under different incubation conditions, either Miiller or Acknowledgments amacrine cell accumulation will predominate. A The authors wish to thank Y. Blocker for her excellent short incubation time in radiolabeled GABA results technical assistance with the autoradiography and photog- in a predominantly glial pattern, whereas a post-in- raphy and A. Patterson for her excellent technical assistance cubation period in label-free medium will lead to a in the treatment and handling of the animals. predominantly amacrine cell labelling pattern.410 These differences are thought to reflect the relatively References high metabolic rate for GABA in Miiller cells as com- pared with neurons. Our studies here support those 1. Hertz L: Functional interactions between neurons and astroy- cytes. I. Turnover and metabolism of putative amino acid findings. In addition, we note that chronic in vivo transmitters. Prog Neurobiol 13:277, 1979. treatment with nipecotic acid preferentially blocked 2. Schousboe A: Significance of the glial transport system for the accumulation into the amacrine cell bodies and pro- inactivation of GABA and effect of structural analysis on the cesses in the inner plexiform layer. Careful survey of uptake. In: GABA-Neurotransmitters. Pharmacochemical, the morphologic features of the treated tissue reveals Biochemical and Pharmacological Aspects, Proceedings of the Alfred Benzon Symposium Krogsgaard-Larsen P, Scheel-Kruger no significant necrosis which suggests that GABA J, and Kofod H, editors. Copenhagen: Munksgaard; New accumulating cells are not destroyed by the treat- York, Academic Press, 1979, p. 263. ment. These studies clearly demonstrate that nipe- 3. Schousboe A: Transport and metabolism of glutamate and cotic acid treatment preferentially inhibited neuronal GABA in neurons and glial cells. Int Rev Neurobiol 22:1, uptake of 3H-GABA. 1981. 11 3 4. Neal MJ: Amino acid transmitter substances in the vertebrate In a recent report it was demonstrated that H- retina. Gen Pharmacol 7:321, 1976. nipecotic acid labels amacrine cell bodies and pro- 5. Redburn DA: GABA and glutamate as retinal neurotrans- cesses as well as glial cells. The most strongly labeled mitters in rabbit retina. Adv Biochem Psychopharmacol 27:79, cells appear to be neurons. These observations of 3H- 1981. 6. Morgan WW and Kamp CW: GABAergic influence on the nipecotic acid binding, along with our findings of light-induced increase in dopamine turnover in the dark- nipecotic acid inhibition of the neuronal GABA up- adapted rat retina in vivo. J Neurochem 34:1082, 1980. take site, suggest a preferential interaction of nipe- 7. Massey SC and Neal MJ: The light evoked release of acetyl- cotic acid with amacrine cells. Using autoradiogra- choline from the rabbit retina in vivo and its inhibition by phy, we are unable to detect a noticeable inhibition gamma-aminobutyric acid. J Neurochem 32:1327, 1979. of glial uptake of 3H-GABA after chronic treatment 8. Massey SC and Redburn DA: A tonic gamma-aminobutyric acid-mediated inhibition of cholinergic amacrine cells in rabbit with nipecotin acid, although Agardh and Ehinger 3 retina. J Neurosci 2:1633, 1982. reported Miiller cell labeling with H-nipecotic acid.'' 9. Lowry OH, Rosebrough NJ, Farr AL, and Randall RJ: Protein Since autoradiographic findings are not readily ame- measurement with the Folin phenol reagent. J Biol Chem nable to quantitative interpretations, a partial, nipe- 193:265, 1951. cotic acid-induced inhibition of glial uptake of 3H- 10. Ehinger B: Glial and neuronal uptake of GABA, glutamic acid, glutamine and glutathione in the rabbit retina. Exp Eye Res GABA could be present, but not detectable under our 25:221, 1977. conditions. It also is possible that, in contrast to neu- 11. Agardh E and Ehinger B: (3H)-muscimol, (3H)-nipecotic acid, ronal sites, the binding of nipecotic acid to glial sites and (3H)-isoguvacine as autoradiographic markers for GABA may not necessarily result in an inhibition of GABA tieurotransmission. J Neurol Transm 54:1, 1982. 12. Jafle EH and Cuello AC: Neuronal and glial release of uptake. This interpretation would suggest a funda- 3 mental difference between the two GABA systems, [ H]GABA from the rat olfactory bulb. J Neurochem 37:1457, 31213 1981. as has been proposed previously. 13. Krogsgaard-Larsen P: Inhibitors of the GABA uptake systems. In summary, we have established an in vivo prep- Mol Cell Biochem 31:105, 1980.

Downloaded from iovs.arvojournals.org on 09/25/2021