Three Subtypes of Ar-Bungarotoxin-Sensitive Nicotinic Acetylcholine Receptors Are Expressed in Chick Retina

The Journal of Neuroscience, February 1993, 13(2): 442-454

Three Subtypes of ar-Bungarotoxin-Sensitive Nicotinic Acetylcholine Receptors Are Expressed in Chick Retina

Kent T. Keyser,’ Luiz R. G. Britto,3 Ralf Schoepfer,4 Paul Whiting,5 John Cooper,6 William Conroy, Agnieszka Brozozowska-Prechtl,’ Harvey J. Karten,’ and Jon Lindstrom6 Departments of ‘Neurosciences and 2Biology, University of California, San Diego, California 92093-0608, 3Department of Physiology and Biophysics, Institute of Biomedical Sciences, Sao Paulo State University, 05508 Sao Paulo, Brazil, 4ZMBH, 6900 Heidelberg, INF 282 Germany, 5Merck, Sharp and Dohme Laboratories, Harlow, CM20 2QR, England, and 6The Institute of Neurological Sciences, University of Pennsylvania, Philadelphia, Pennsylvania 19104-6142

A recent report described the isolation of cDNA clones en- units are differentially expressed by different populations of coding a7 and a8 subunits of a-bungarotoxin-sensitive nic- retinal neurons. In addition, the distribution of aBgtAChR otinic ACh receptors (aBgtAChRs) from chick brain and dem- subunit immunoreactivity was found to differ from that of onstrated that they were related to, but distinct from, the a a-Bgt-insensitive nAChR subunits. subunits of nicotinic ACh receptors (nAChRs) from muscles [Key words: ACh, bungarotoxin, nicotinic ACh receptors, and neurons. Monoclonal antibodies against the’ two receptors, neurotransmitters] aBgtAChR subunits were used to demonstrate that at least two subtypes are present in embryonic day 18 chicken brain. The predominant brain subtype contains a7 subunits, while a minor subtype contains both a7 and a8 subunits. Both ACh is known to act as a neurotransmitter in the vertebrate subtypes may also contain other subunits. Here we report retina. Physiological studies have shown that nicotine, ACh, the results of immune precipitation studies and immunohis- and AChE inhibitors can have profound affects on the response tochemical studies of aBgtAChRs in the chick retina. In ad- properties of many ganglion cells, including those that display dition to the two subtypes found in brain, a new aBgtAChR directional selectivity (Noel1 and Lasansky, 1959; Straschill, subtype that contains a8 subunits, but not a7 subunits, was 1968; Ames and Pollen, 1969; Straschill and Perwein, 1973; identified and was found to be the major subtype in chick Masland and Ames, 1976; Ariel and Daw, 1982a,b; McRey- retina. This subtype has a lower affinity for a-bungarotoxin nolds and Miyachi, 1986). The availability of antiseradirected (aBgt) than does the subtype containing only a7 subunits. against ChAT, the enzyme responsiblefor ACh synthesis, has allowed the precise identification and characterization of cho- Small amounts of this a8 subtype were also detected in brain by labeling with higher concentrations of Y-aBgt than had linergic cells in the inner nuclear layer (INL) and ganglion cell been used previously. The subtype containing only a7 sub- layer (GCL) of the retina (for a review, see Vaney, 1990). In units comprised 14% of the aBgtAChRs in hatchling chick contrast, the identification and study of the cholinoceptiveneu- retina. The subtype containing a8 subunits (but no a7 sub- rons in the retina has been more problematic. units) accounted for 69%, and the a7a8 subtype accounted Initial attempts to identify the cholinoceptive cellsrelied upon for 17%. Amacrine, bipolar, and ganglion cells displayed a8 the localization of radioactively labeled or peroxidase-conju- subunit immunoreactivity, and a complex pattern of labeling gated oc-bungarotoxin(“Bgt). otBgt is a protein from snake ven- was evident in both the inner and outer plexiform layers. In om that binds with high affinity to the ACh binding site of the contrast, only amacrine and ganglion cells exhibited a7 sub- nicotinic ACh receptors (nAChRs) from muscle. Thus, when unit immunoreactivity, and the pattern of a7 subunit labeling high-affinity cvBgtbinding was found in the retina, this binding in the inner plexiform layer differed from that of a8 subunit was assumedto represent functional nAChRs. Various inves- labeling. These disparities suggest that the aBgtAChR sub- tigators reported that aBgt bound specifically and with high affinity to sitesin both the inner plexiform layer (IPL) and outer plexiform layer (OPL) of goldfish, chick, pigeon, turtle, and rabbit retina (Vogel and Nirenberg, 1976; Yazulla and Schmidt, Received May 1, 1992; revised July 15, 1992; accepted July 21, 1992. 1976,1977; Vogel et al., 1977; Schwartz and Bok, 1979; Zucker This work was supported by EY07845 (K.T.K.) and EY06890 (H.J.K.). J.L. and Yazulla, 1982). However, the resolution of these methods was supported by NS I 1323 and grants from the Muscular Dystrophy Association, was too low to permit the identification of all of the cell types the Council for Tobacco Research, and the Council for Smokeless Tobacco Re- search. L.R.G.B. was supported by FAPESP, CNPq, and Contract USP/BID (Bra- that contained the otBgt binding sites. Subsequently, other in- zil), and by the Fogarty International Center. R.S. was supported by a fellowship vestigators showed that (1) there were neuronal nAChRs that from Boehringer Ingelheim Fonds. We thank Kevin Cox for reviewing the manu- did not bind cuBgt,and (2) cells that had both nAChRs and otBgt script and assistance with illustrations and Thomas Hughes for his comments on the manuscript. We also thank Dania Hammassaki-Britto for assistance with the binding sitesexhibited ACh-gated cation channelsthat were not figures. blocked by otBgt (Patrick and Stallcup, 1977; Carbonetto et al., Correspondence should be addressed to Kent T. Keyser, Department of Neu- 1978; Ravdin and Berg, 1979; Chiappinelli, 1985; Whiting and rosciences, 0608, University of California at San Diego, San Diego, CA 92093- 0608. Lindstrom, 1986a,b, 1987; Boulter et al., 1987; Lipton et al., Copyright 0 1993 Society for Neuroscience 0270-6474/93/130442-13$05.00/O 1987; Ballivet et al., 1988). Furthermore, cYBgtbinding siteson The Journal of Neuroscience, February 1993, 73(2) 443

Table 1. Specificities of mAbs to aBgtAChRs

Specificity0 Immunoglobulin mAb Subunit Synthetic peptide binding Immunogenb type 306 a7 a7(380-400) Purified chicken brain Mouse IgG 1 olBgtAChR 318 a7 017(380-400) Bacterially expressed Rat IgG chicken ~y7 321-412 308 018 cr8(323-342) Bacterially expressed Rat IgG2B chicken 018293-435 305 ~y8 None (absolutely Purified chicken brain Rat IgG2C conformation aBgtAChR dependent)

” McLane et al. (I 992). h Schoepfer et al. (1990). chick ciliary ganglion neurons were found only on pseudoden- We describe below the results of immune precipitation and drites rather than at synaptic siteswhere nAChRs were localized immunohistochemical studies of orBgtAChRs in the retinas of (Jacob and Berg, 1983; Jacob et al., 1984). These findings sug- E 18 and newly hatched chicks. We report the identification of gested that olBgt binding sites did not represent functional a new cvBgtAChR subtype containing ot8, but not ot7, subunits nAChRs. that is present in small quantities in chick brain but that con- The availability of antibodies against, and cDNA probesfor, stitutes the major subtype in chick retina. In addition, our im- structural and ligand binding subunits of nAChRs has greatly munohistochemical data demonstrate that (1) the distribution enhanced our ability to localize these proteins in the retina. of ot7 subunit immunoreactivity differs from that of ot8 subunit Studies using monoclonal antibodies (mAbs) againstthe struc- immunoreactivity, and (2) the distribution of both olBgtAChR tural and ACh binding subunits have shown that populations subunits in the retina differs from the distribution of of amacrine, ganglion, and displacedganglion cells in the chick oc-Bgt-insensitivenAChRs. retina contain nAChRs and that the dendrites of thesecells are in registerwith the dendrites of cholinergic amacrine cells (Key- Materials and Methods ser et al., 1988; Whiting et al., 199 1). White leghorn chicken eggs were obtained from MacIntyre Poultry Recently, chicken brain cDNA clones that encodetwo 01sub- (Lakeside, CA) and were housed in a humidified incubator at 37°C. units of the otBgt-sensitivenAChRs (aBgtAChRs) were isolated Following hatching, chicks were housed in a brooder and provided with (Schoepferet al., 1990). Thesewere originally designatedcul and food and water ad libitum. Twenty-five chickswere usedfor immu- or2but are now referred to asot7 and ot8 (seeBritto et al., 1992). nohistochemical analyses. Tissue preparation. One or two days after hatching, the chicks were On the basisof the deduced amino acid sequences,Schoepfer given an overdose of ketamine and xylazine and decapitated and the et al. (1990) reported that otBgtAChRs are membersof the li- head was placed in a shallow dish of ice-cold 2% paraformaldehyde in gand-gatedion channelfamily that includesnAChRs from mus- 0.1 M phosphate buffer (pH 7.4). The eyes were immediately removed cle and neuronsand the GABA, and glycine receptors. The two from the head, the anterior pole of each eye was cut away, and the (Y subunits were shown to be ligand binding subunits by the vitreous was removed as completely as possible. The eyes were im- mersedin ice-coldfixative for 4-6 hr andthen placed in 0.1 M phosphate ability of correspondingsynthetic peptidesto bind otBgt(M&me buffer containing 30% sucrose for at least 14 hr to minimize damage et al., 1991). mAbs were raised againstnative oiBgtAChRs,and resulting from freezing. two bacterially expressedsubunit fragments and the epitopes Some of the eyecups were immersed in embedding medium, frozen, recognized by the mAbs were mapped with synthetic peptides and sectioned (10 pm) on a cryostat. Sections were collected on gelatin- coated glass slides. Alternatively, the retina was removed from the eye- (McLane et al., 1992). Studies using thesemAbs demonstrated cup, radial incisions were made, and the retina was flattened. The retina that in embryonic day 18 (E18) chick brain more than 75% of was then frozen and 25-40 pm sections were cut parallel to the vitreal all otBgtAChRs have ot7 subunits, but no ot8 subunits, while surface on a sliding microtome. Sections cut in this plane were processed roughly 15% of all cYBgtAChRshave both (~7 and ot8 subunits “free floating” in small vials. (Schoepfer et al., 1990). Immunohistochemistry. Four monoclonal antibodies (mAbs) were used in this studv: mAb 305. mAb 306. mAb 308. and mAb 3 18. A comnlete Several lines of evidence suggestthat someof the otBgtAChRs description-of the immunogen used to prohuce each antibody is-pre- found in the nervous system may representa new classof ACh- sented in Schoepfer et al. (1990), and synthetic peptide mapping of their gated ion channels. For example, aBgtAChRs affinity purified epitopes is reported in McLane et al. (1992). Table 1 summarizes the from chicken brains and reconstituted into planar lipid bilayers properties of these mAbs. form carbamylcholine-gated cation channels that are blocked Tissue sections were washed in three changes of phosphate buffer and then incubated with mAb diluted in a 0.1 M sodium phosphate buffer by d-tubocurarine (Gotti et al., 1991). Furthermore, Couturier solution containing 0.3% Triton X- 100 for 12-24 hr. The primary mAbs et al. (1990) demonstrated that the ot7 subunit, when injected were used at concentrations ranging from 10 to 20 nM. into Xenopus oocytes, assemblesinto a homo-oligomeric chan- Avidin-biotin labeling. Following incubation in mAb, the sections nel that respondsto ACh and nicotine. Chick ciliary ganglion were washed in three changes of phosphate buffer and incubated in a solution of biotinylated goat anti-rat or mouse IgG (Jackson Labs) di- neuronsexpress large amounts of ot7 (Vernallis et al., 1991) and luted 1:200 in buffer containing 0.3% Triton X- 100 for 1 hr. The sections exhibit ACh-induced CaZ+influx that can be blocked by otBgt were then washed and incubated in a mixture of biotin and avidin- (Vijayaraghavan et al., 1992). bound horseradish peroxidase diluted 1:lOO in 0.3% Triton X-100 in 444 Keyser et al. . a-Bungarotoxin-Sensitive Acetylcholine Receptors in Chick Retina phosphate buffer for 1 hr. After washing, the tissue was incubated in of beads to extract of 1:7 were incubated overnight at 4°C with gentle 0.05% diaminobenzidine in buffer for 15 min. Hydrogen peroxide was agitation, and then the beads were removed by brief centrifugation. then added to make a final concentration of 0.0 l%, the dish was gently Assays of a7-containing olBgtAChRs used duplicate wells coated with shaken, and after 15 min the tissue was washed in several changes of mAb 3 18 and blank wells from which the mAb was omitted. In each buffer and coverslips were applied. well, 50 ~1 of extract depleted of a8 was mixed with 50 ~1 of 0.1 M NaCl, Controls. The specificity of staining was verified by omission of the 10 mM NaN,, and 0.5% Triton X-100 containing sufficient 1251-olBgtto mAb or secondary antibody. In addition, in some instances the mAb yield final concentrations of 0. l-l 6 nM. After overnight incubation at was replaced by normal serum from the same species. In both cases 4°C the wells were washed three times with 200 ~1 of 10 rnr+r Na phos- specific staining was abolished. phate (pH 7.5), 0.05% Tween 20 and then bound 12sIwas determined Radioimunoassays of olBgtAChR subtypes. Detergent extracts of by gamma counting. More than 97% of a7-containing olBgtAChR was olBgtAChRs from brains and retinas of embryonic day 18 and newly depleted from El8 retina extract by using mAb 318 coupled at 3 mg/ hatched chicks were prepared in the following manner. Brains and ret- ml of ael to Affiael Hz (Bio-Rad) according to the manufacturer’s di- inas were dissected from the animals and then frozen. The frozen tissues rections. The mAb-coated beads ‘were used-for absorption in the same were thawed in 10 vol of 50 mM Na phosphate buffer (pH 7.5) 1 M way as were the mAb 305-costed beads. Assays of a8-containing NaCl. 5 mM EDTA. 5 mM EGTA. 2 mM nhenvlmethvlsulfonvl fluoride. aBgtAChRs used duplicate wells coated with mAb 305. In each well, 5 m& benzamidinet and 5 mM iddoacetamide, and then homogenized 25 ~1 ofextract depleted ofa was mixed with 12+~Bgt in 100 ~1 reaction on ice for 2 min at 15,000 rpm with a Brinkman Polytron. After dilution mixes similar to those for a7 assays but containing l-75 nM Y-aBgt. with another 10 vol of buffer, the homogenate was centrifuged at 50,000 The relative abundance of each olBgtAChR subtype was determined rpm in a Beckman 50.2 ti rotor for 30 min at 4°C. The pellet was by immune precipitation. Extracts were diluted to 0.5 nM in total homogenized again in an equal volume of buffer, this time containing cvBgtAChR using 0.5% Triton X-100, 100 mM NaCl, 10 mM NaN,, and 50 mM NaCl, and centrifuged to extract soluble components further. 10 mM Na phosphate buffer (pH 7.5) and labeled with 35 or 50 nM The pellet was then extracted with 4 vol of 2% Triton X-100, 50 mM IzSI-olBgt. Triplicate 200 ~1 aliquots containing the indicated amounts NaCl, 50 mM Na phosphate buffer (pH 7.5), and the same mix of of mAb 308, mAb 318, or both, and 4 ~1 of normal rat serum were chelating agents and protease inhibitors used in the homogenization incubated overnight at 4°C. Then, 100 ~1 of an appropriate dilution of buffer. After initiallv disruntina for 15 set at 10.000 ram with a Brink- goat anti-rat IgG was added for 4 hr. The samples were diluted with 1 man Polytron, the resuspended pellet was shaken for-2 hr at 4°C. The ml of 0.5% Triton X- 100 buffer and centrifuged in a Beckman Microfuge mix was centrifuged at 50,000 rpm in a Beckman 50.2 ti rotor at 4°C 12 for 8 min. The pellets were washed twice using 1 ml of buffer and for 30 min. The concentration of aBgtAChR in the resulting supima- recentrifugation before gamma counting. Blank values determined using tants was typically 3-5 nM. only normal serum and no mAb were subtracted from each assay value. The total amount of olBgtAChR in extracts was determined by an assay that depends upon the negative charge of aBgtAChRs to bind to DEAE while positively charged Y-aBgt does not. lZ51-olBgt was iodi- Results nated to specific activities of > 6 x 101’ cpm/mol as previously described Immunohistochemical localization of cu7and ot8 subunits (Lindstrom et al., 198 1). Schmidt and Raftery (1973) initially used DEAE to measure AChRs, and the technique used here is a modification revealed more ot8 immunoreactivity than expected of that previously described by Whiting and Lindstrom (1986b). DE52 mAbs 306 and 318, which recognize two different epitopes on (Whatman) equilibrated with- 10 mMNa phosphate buffer (pH 7.5) 017subunits of otBgtAChRs, yielded identical patterns of labeling 0.5% Triton X-l 00 (150 ul nacked volume) was loaded in USB Comoact in hatchling chick retina (Fig. lA,C). Similarly, mAbs 308 and Reaction Columns in triplicate. Extracts were labeled overnight at-CC with 50 nM lZ51-aBgt with or without 10 mM nicotine as a negative 305, which recognize two different epitopes on a8 subunits, control. Then, 10 ~1 aliquots were diluted to 100 ~1 in the 10 mM buffer yielded identical patterns of labeling (Fig. l&D). Examination and applied to the DE52. After 2 mitt, the columns were washed four of sectionsof hatchling retina stainedwith thesemAbs revealed times with 500 ~1 of buffer. The whole column was then counted and that more 018immunoreactivity waspresent than expectedbased specific binding was determined by subtracting the negative control value (typically about 20-30% of total binding). upon previous immune precipitation studies of E 18 brain. In The total amount of olBgtAChRs and the affinity of the ot7 and a8 these studies, only 15% of aBgtAChRs were found to contain subtypes for Y-olBgt were determined by a solid-phase radioimmu- c~8subunits (Schoepfer et al., 1990). Indeed, more a8 immu- noassay using mAb-coated microwells. Protein G affinity column-pu- noreactivity was evident than (117immunoreactivity. In view of rified mAb 305, mAb 318, or a mixture of the two, was coupled to theseobservations, immune precipitation assaysto quantify the Immulon 4 Removawells (Dynatech) by incubating 100 pi/well of 40 mg/ml mAb in 10 mM Na bicarbonate buffer (pH 8.8) overnight at room fraction of olBgtAChRs incorporating a7, (~8, or both subunits temperature in a closed, humidified box. After three washes (200 ~1 were conducted with hatchling and embryonic retinasand brains. each), excess protein binding capacity of the wells was quenched with 200 ~1 of 3% bovine serum albumin in 100 mM NaCl, 10 mM Na or7-containingaBgtAChRs have a higher ajinity for olBgt than phosphate buffer (pH 7.5) and 0.05% Tween 20 for 2 hr. The wells do those that contain a8 subunits were then washed three times with 200 ~1 of the same buffer without albumin. Binding of 1251-aBgtto aBgtAChRs in detergent extracts from The total amount of crBgtAChR containing either a7 or a8 subunits E 18 retina was studied usinga solid-phaseradioimmunoassay. was measured using a solid-phase assay in which the microwells were The ot7 subtype and the newly discovereda8 subtype were stud- coated with a mixture of both mAb 305 and mAb 3 18. Trinlicate ali- ied separately. cvBgtAChR subtypescontaining only (r7 subunits quots of 0.1-10 ~1 of extract in a total volume of 50 ml of i0 mM Na phosphate (pH 7.5), 100 mM NaCl, 10 mM NaN,, and 0.5% Triton or only ot8 subunits (in conjunction with some unknown com- X- 100 were mixed in the wells with 50 ~1 of buffer containing 100 nM plement of structural subunits), but not the subtype containing lZSI-otBgtto give a final concentration of 50 nM Y-(uBgt. After overnight both 017and a8 subunits, were prepared by absorbing detergent incubation at 4°C in a humidified box, the wells were washed three times extracts of El 8 retinas with either mAb 305 coupled to agarose with 200 ~1 of 10 mM Na phosphate (pH 7.5) 0.05% Tween 20 and then bound rZ51was determined by gamma counting. In the solid-phase to remove otBgtAChRs containing a8 subunits or mAb 318 assays, binding of cuBgtAChRs to the wells was linear to >2.5 x lo-” coupled to agaroseto remove olBgtAChRs containing (~7 sub- mol/well and background values were < 1% of the maximum labeling. units. Figure 2 showsthat the or7 subtype had relatively high Affinity for cvBgtwas measured separately for the or7 and (~8 subtypes affinity for lZ51-olBgt(K, = 2.1 x 1O-9 M) whereasthe lZ51-orBgt by solid-phase assays using extracts depleted of either all LvBgtAChRs affinity of the cu8subtype was substantially lower (& = 17.2 x containing cu8 or all of those containing 017. Extract from El8 retinas was depleted of > 99% of ol8-containing aBgtAChRs using purified mAb 1Om9 M). Thus, the amount of o18could be easily underestimated 305 coupled at 2 mg/ml of gel to Actigel ALD (Sterogene) according to at the concentrations of lZ51-otBgttypically usedin such assays. the manufacturer’s instructions. The mAb 305-costed beads at a ratio In subsequentexperiments lZ51-cuBgtwas usedat concentrations The Journal of Neuroscience, February 1993, 73(2) 445

Figure I. Photographs of transverse sections of chick retina illustrating the distribution of cllBgtAChR immunoreactivity. The sections shown in A and C were processed using mAb 306 and mAb 3 18, respectively. These antibodies recognize two different epitopes on (r7 subunits. The sections shown in B and D were processed using mAb 305 and mAb 308, respectively. These antibodies recognize different epitopes on a8 subunits. More (~8 immunoreactivity was present than expected. Scalebar, 50 mm. sufficiently high to ensure complete labeling of the cu8 subtype. cifically measures the immunologically identified subtypes and The binding of 12*I-aBgt to the a7or8 subtype was not specifically has a very high signal:noise ratio. Table 2 shows that even when studied. 1251-olBgtwas used at concentrations as high as 50 nM, essentially all of the ‘25I-arBgt binding in extracts can be accounted for by All specific binding of IzsI-cyBgt in brain and retina extracts is olBgtAChRs containing a7 subunits, (~8 subunits, or a combi- associated with a7 or (~8 subunits nation of the two. The total number of binding sites for 1251-(uBgtin detergent extracts of both retina and brain were determined using two Determination of the fraction of cYBgtAChRs in brain and assays. One assay used small DEAE columns to bind labeled retina that contain a7, c& or both subunits aBgtAChRs. This assay was independent of the immunological The relative abundance of olBgtAChR subtypes in both E 18 and properties of the protein and depended only upon the rather hatchling chick brains and retinas was determined by immune low isoelectric point (i.e., rather negative charge) of all known precipitation assays. The maximum fraction of the 12sI- AChRs as compared to the highly positive charge of cyBgt. This cwBgt-labeled aBgtAChR that could be immunoprecipitated with is a very general assay that is not characterized by a high signal: mAbs specific for or7 or CY~subunits was determined as shown noise ratio. The other assay was a solid-phase radioimmuno- in Figure 3. The subunit composition calculated from such ex- assay in which individual mAbs or a mixture of mAbs to 017or periments is shown in Table 3. Because ofthe high concentration (~8 was used to bind cuBgtAChRs to microwells. This assay spe- of ‘251-cyBgt used, we were now able to detect low levels of a 446 Keyser et al. * ol-Bungarotoxin-Sensitive Acetylcholine Receptors in Chick Retina

0 5 10 15

Figure 2. The al subtype of aBgt- AChR has higher affinity for cuBgtthan does the 018subtype. As described in the Materials and Methods, detergent extracts of El 8 retina containing the 017 subtype were prepared by absorbing with mAb to c&, and extracts containing the or8 subtype were prepared by absorbing with mAb to a7. The binding of lz51-aBgtto cvBgtAChRs from these extracts was then measured using with mAbs specific for a7 or a8 subunits to bind the aBgtAChRs to microwells. The top panel illustrates binding data for the al subtype, and the bottom panel shows binding data for the 018subtype. Both panels show the binding curve and an inset with a Scatchard plot of the binding curve used to determine the KD for 0 10 20 30 40 50 60 70 80 binding. l%xBgt ( x lO+‘M) subtype containing only cu8 subunits in brain as well as some- Specifically, there was a marked decrease in the amount of the what larger amounts of the subtype containing a7 and a8 sub- 017subtype and a substantial increase in the amount of the c~8 units. However, the preponderant subtype in brain contained subtype in retina from E 18 to hatching (Table 3). These changes only ot7 subunits, as previously reported (Schoepfer et al., 1990). were not studied further at this time. In contrast, the predominant subtype in retina contained only cu8 subunits, while the other two subtypes were present in substantially lower amounts. Thus, the immune precipitation ex- Immunohistochemical localization of aBgtAChRs in the INL periments were consistent with the immunohistochemical evi- c~7 subunit immunoreactivity was observed in neurons in the dence that aBgtAChRs containing a8 subunits predominate in inner portion of the INL (Fig. lA,C). Labeling was detected in hatchling chick retina and demonstrated the existence of a pre- neurons that ranged in size from 4.5 to 10 pm in diameter (Fig. viously undetected subtype that includes ot8 subunits not in 4A). In contrast, a8 subunit immunoreactivity was observed in association with ot7 subunits. An unexpected finding was that neurons within two distinct regions of the INL (Fig. l&D). In while the relative amounts ofthe different subtypes were similar the inner third of the INL, a8 staining was confined to cells that in E 18 and hatchling in brain, this was not the case in retina. ranged from 4.5 to 10 pm in diameter (Fig. 4B). An analysis of The Journal of Neuroscience, February 1993, 13(2) 447

Table 2. Measurement of the concentration of aBgtAChR using two different assays

Concentration of aBgtAChR in extracts Immunoassayusing both DEAE Assay mAb 305 and mAb 3 18 Specific Specific lZSI-aBgtbound binding 1251-aBgtbound binding Sample (cpm ? SE) (nM f SE) (cpm & SE) (nM + SE) E 18 brain extract 15,369 k 292 2.93 + 0.02 17,863 k 626 3.05 f 0.11 Blank 2903 k 90 115 * 15 E 18 brian extract 22,078 f 176 3.40 k 0.06 17,597 k 677 3.05 + 0.12 Blank 7132 & 257 148 f 10

the size and relative frequency of the a8-immunoreactive neu- found within three or four tiers of somata in the outer half of rons revealed that cells in the 6-7 pm and 7.5-8.5 pm size ranges the INL but was absent from the cells at the outer margin of were more numerous than cells in other size classes.However, the INL (Figs. l&D; 6). The labeled cells were generally small, there were relatively fewer ot7-immunoreactive neurons in the ranging from 4.5 to 6.5 wrn in diameter, and were present in 6-7 pm and 7.5-8.5 pm diameter size ranges(Fig. 4A) than were large numbers (Fig. 5B). These cells were identifiable as bipolar detected in sectionsexposed to a8 subunit-specific mAbs. cells since many of them gave rise to processesthat projected c~8subunit immunoreactivity was also evident in neurons outward to the OPL and inward to the IPL. ~y8immunoreac-

Anti-a7 . c I m I I 40-1 I Anti-a7 rm n - n I 20-' - E-18 RETINA H-l RETINA 100 I Anti-a7 8 I, . -r * m - @ I Anti-a7 80 m

60t Anti-d g 40 6-~---1---,--~---“------

20 H-l BRAIN

0 2 4 6 8 IO 0 2 4 6 8 IO fmoles of mAb Figure 3. Determination of the fraction of cYBgtAChRs that contain al, 018, or both subunits. The maximum fraction of lZ51-otBgt that could be immunoprecipitated from Triton X- 100 extracts of chick retinas and brains with mAb 308 (specific for olBgtAChR ot8 subunits), mAb 3 18 (specific for a7 subunits), or a combination of the two mAbs was determined as described in Materials and Methods. Total olBgtAChR was defined as the maximum that could be precipitated by a combination of both mAbs. The percentage containing only LYIsubunits was calculated as the total minus the percentage precipitated by mAb to (~8 subunits. The percentage containing only a8 subunits was calculated as the total minus the percentage precipitated by mAb to a7 subunits. The percentage containing both a7 and 018 subunits was calculated as the total minus 018 plus a7 subunits. The results of three to four repeats of each panel in this figure were averaged in Table 3. In this experiment and one other, 50 nM IY-(rBgt was used, while in two other instances 35 nM 1251-aBgt was used. All of the experiments yielded similar results. This contrasts with the 4 nM lz51-otBgt used in previous experiments by Schoepfer et al. (1990) with El 8 brain. cvBgtAChRs containing only a8 subunits were not detected in these earlier experiments. 448 Keyser et al. l a-Bungarotoxin-Sensitive Acetylcholine Receptors in Chick Retina

Table 3. Subtypes of aBgt binding proteins

Percentage compositioi+ Subunit Possible subunit Brain Retina Subtype composition arrangement0 El8 Hl El8 Hl

I a, + ? 68 74 36 14

III lY* + ? 9 9 41 69

H 1, first day of hatching. a This assumes that olBgt binding proteins are composed of five subunits, as are AChRs of muscles and nerves that are homologous in sequence, and that there are two a-nonadjacent a subunits as in muscle AChRs. * This is calculated from data in Figure 3 as described in the caption.

Figure 5. Photographs of chick retina that had been sectioned parallel to the vitreal surface at the level of the outer INL. The section in A was processed using mAb 306, while the section in B was processed using Figure 4. Photographs of chick retina that had been sectioned parallel mAb 305. Nomarski optics allow visualization of both labeled and to the vitreal surface at the level of the inner INL. The section in A was unlabeled cells. Few, if any, cells in this portion of the retina exhibited processed using mAb 306, while the section in B was processed using (~7 immunoreactivity (A), while cells that displayed (~8 immunoreactiv- mAb 305. Scale bar, 25 mm. ity were present in high density (B). Scale bar, 25 mm. The Journal of Neuroscience, February 1993, 13(2) 449

Figure 6. Photograph of a transverse section of chick retina illustrating the distribution of a8 olBgtAChR immunoreactivity in the outer retina. The sectionwas processed using mAb 305. The labeled processes that project outward from the OPL are known as Lan- dolt clubs (arrows).These features were not labeled in tissue processed with mAbs against cu7subunits (not shown). Scale bar, 20 mm. tivity within the OPL proper appeared to be restricted to round pattern. There were differencesbetween the sizesof someof the or oval structures that were distributed in two or three strata cr7-and cY8-immunoreactiveneurons in the GCL. In the central (Fig. 6). Many of the bipolar cell processes that projected to the retina, cells that exhibited ot8 subunit immunoreactivity ranged OPL gave rise to small extensions known as Landolt clubs that from 5 to 16 pm in diameter while those that displayed ot7 continued outward to the outer limiting membrane (Fig. 6, ar- subunit labeling ranged from 5 to 21 pm in diameter. In the rows). The heavy labeling throughout the IPL obscured the ar- peripheral retina, the or7-immunoreactive cells ranged from 5 borizations of individual immunoreactive bipolar cells. ot7 im- to 27 pm in diameter while the a8-immunoreactive neurons munoreactivity was not detectable in neurons in the outer half were 7-30 wrn in diameter. of the INL (Fig. 4A) when the mAbs directed against the ar7 In some instances,processes of the labeled neuronscould be subunit were used at the same concentrations as those used for traced to the inner margin of the IPL. However, becauseof the the ~y8subunit antibodies. However, if the concentration of mAb intenselabeling throughout this layer, we were unable to discern directed against 017subunit was increased, weak staining of some the arborization patterns of individual cells in the GCL. In neurons in the outer half of the INL cells was detectable. (~7 horizontal sectionsthat included the GCL and optic fiber layer, immunoreactivity in the OPL in these instances was always less many labeled cells could be seento give rise to labeled axons intense and more diffuse than the c~8 immunoreactivity. (Fig. 7B, arrows) that entered the optic fiber layer and projected to the optic nerve head. Immunohistochemical localization of cuBgtAChRs in the GCL ot7 and a8 subunit immunoreactivity was detected within cells Immunohistochemical localization of otBgtAChRs in the IPL of the GCL, although the intensity of staining was highly vari- otBgtAChR ~y8immunoreactivity wasdistributed throughout the able (Figs. 1, 7A,B). Because of this variability, we adopted a IPL (Fig. 1B) but was particularly intense in a band coincident conservative approach when counting labeled neurons. There with the innermost portion of lamina 1 and all of lamina 2. were at least 3500-4000 ot8-immunoreactive cells/mm2 in the Moderately heavy labeling wasalso observed in laminae 3 and central retina, while in the peripheral retina there were about 4. In contrast, while a7 subunit immunoreactivity wasalso dis- 1500-2000 labeled cells/mm2. The neurons that exhibited or7 tributed throughout the IPL (Fig. lA), labelingwas more intense immunoreactivity were present in lower density than those that in lamina 3, in the innermost portion of lamina 4, and in outer displayed ot8 immunoreactivity, but were distributed in a similar portions of lamina 5. 450 Keyser et al. l a-Bungarotoxin-Sensitive Acetylcholine Receptors in Chick Retina

Figure 7. Photographs of chick retina that had been sectioned parallel to the vitreal surface at the level of the GCL. The section in A was processed using mAb 306, while the section in B was processed using mAb 305. The cells that displayed (~7 immunoreactivity were present in lower density (A) than those that exhibited than a8 subunit immunoreactivity (B). Many of the labeled cells gave rise to labeled axons that entered the optic fiber layer (arrows).Scale bar, 25 mm.

Discussion prisesroughly 86% of all arBgtAChRsfound in hatchling chick cYBgtAChRsubtypes in the retina retina. Thus, the retina and brain differ not only in the expres- All of the crBgtAChRs detected in detergent extracts of chick sion of the cu8subtype, but alsoin the relative abundanceof the brains and retinas were shown to contain (~7subunits, (~8 sub- (r7 and a7a8 subtypes.This finding is reminiscent of previous units, or both. Schoepferet al. (1990) reported that more than reports that the major nAChR subtype in chick brain has the 75% of all aBgtAChRs in El8 chick brain have (~7subunits but subunit composition of (ar4)&32),(Anand et al., 1991) and a no (~8subunits while roughly 15%have both a7 and 018subunits. minor brain nAChR subtype includes(~3 subunits (Wada et al., Thus, two subtypes were defined: the “a7” subtype and the 1989; Morris et al., 1990;Whiting et al., 1991). In contrast, the “a7c~8” subtype. A third subtype that includes(~8, but no cu7, nAChR subtype that incorporates (~3subunits is the predomi- subunits was not detected in chick brain. By using a higher nant subtype found in chick retina and ciliary ganglion (Hal- concentration of 1251-aBgt,we have now been able to detect a vorsen and Berg, 1990; Whiting et al., 1991). The a8 subtype subtype,termed “LYE,”which contains(~8 subunits, either alone is characterized by a much lower affinity for aBgt than the (~7 or in conjunction with one or more other, as yet unidentified subtype. This differencein pharmacologicalcharacteristics (and subunits. This subtype constitutes 9% of the aBgtAChRs in others) is also exhibited by aBgtAChRs expressedfrom (~7and hatchlingchick brain and 69% of aBgtAChRs in hatchling chick a8 cDNAs (R. Anand and J. Lindstrom, unpublished obser- retina. The (~8subtype, together with the a7a8 subtype, com- vations). The Journal of Neuroscience, February 1993, 73(2) 451

Labeled neurons in the INL and GCL populations of large neurons in the central retina may express The somata and dendrites of selected populations of bipolar, 017,but not 018,subunits. amacrine, and ganglion cells displayed otBgtAChR immuno- The observation that bipolar, amacrine, and ganglion cells in reactivity. The observation that the receptor immunoreactivity the chick retina exhibit aBgtAChR immunoreactivity is con- was distributed throughout the neurons is consistent with pre- sistent with the results of previous studies of aBgt binding in vious findings in retina and in other portions of the nervous the retinas of goldfish, turtle, chick, and mouse. For example, system. For example, nAChR and otBgtAChR subunit immu- James and Klein (1985) reported that in turtle retina only bipolar noreactivity has been described in cell bodies, as well as den- cells bound aBgt. In contrast, Vogel et al. (1977) observed that drites, of chick brain and retinal neurons (Keyser et al., 1988; bipolar and amacrine synapses in the chick retina were labeled Whiting et al., 1991; Britto et al., 1992). Similarly, Jacob et al. with peroxidase-labeled clBgt. Zucker and Yazulla (1982) found (1986) and Stollberg and Berg (1987) reported that many mAb that peroxidase-conjugated aBgt binding in the IPL of goldfish 35 binding sites (mAb 35 is an antibody that recognizes the retina was always associated with synapses in which an amacrine main immunogenic region of muscle nAChRs) in chick ciliary cell process was the presynaptic element while the postsynaptic ganglion were intracellular. These and other examples suggest element was either an amacrine, ganglion, or bipolar cell. How- that the distribution of receptors throughout nonsynaptic por- ever, peroxidase-conjugated olBgt binding in the IPL of mouse tions of neurons is widespread. Such observations may reflect retina was confined to amacrine-bipolar cell contacts and ama- the detection of AChRs during synthesis, transport, or destruc- crine-amacrine cell contacts (Pourcho, 1979). Labeling was not tion or, alternatively, extrasynaptic AChRs whose physiological detected at synapses in which a bipolar cell was the presynaptic roles are, at present, unknown. element. In none of these studies was the identity of the pre- The neurons in the outer half of the INL that displayed ot8 synaptic amacrine cells known. immunoreactivity were identifiable as bipolar cells because of their position and the appearance of the inwardly directed axons Labeling in the OPL and the outwardly directed dendrites. The staining throughout There are reportedly six types of photoreceptors in the avian the IPL obscured the axons and their terminations, thus pre- retina, and their terminals are distributed in three layers in the venting a determination of which of the eight bipolar cell sub- OPL (Mariani, 1987). The round or oval structures that were types described in the avian retina (Mariani, 1987) display visible in the OPL in transverse sections of retina following cvBgtAChR immunoreactivity. In many instances, the labeled exposure to antibodies against ot8 subunits appeared to be dis- dendrites gave rise to Landolt clubs that extended from the OPL tributed in two or three layers, suggesting that these features to the outer limiting membrane. These processes, whose func- may represent clusters of labeled bipolar cell dendrites. Previous tion is unknown, are found primarily in the retinas of nonmam- studies by Yazulla (1979) who found that otBgt binding sites in malian vertebrates, although there is one report of such pro- the OPL were associated with a plexus of small processes bound- cesses in the retina of a prototherian mammal (Young and Vaney, ed distally by photoreceptor terminals and proximally by a large 1990). The fact that cells in this portion of the retina exhibited dendrite, provide support for this suggestion. However, the otBgt weak ot7 subunit immunoreactivity only when the antibodies binding sites in the OPL were not consistently associated with were used at much higher concentrations is consistent with the identifiable synaptic structures (Yazulla, 1979) rendering the possibility that small amounts of this subunit may be present interpretation of these findings difficult at best. in some of these cells. Other findings reported by Yazulla (1979) demonstrate that The position of the ot7- and ol8-immunoreactive somata in cvBgt binding in the OPL of pigeons differs in some respects from the inner INL and the absence of outwardly directed processes that observed in other species. The binding of aBgt in the IPL indicate that the majority of these neurons were probably ama- and OPL of goldfish and turtle retina was eliminated by treat- crine cells. The observation that cells in particular size classes ment with 1O-5 M nicotine. However, while otBgt binding in the exhibited cu8 immunoreactivity than cu7 immunoreactivity sug- IPL of avian retina was blocked by similar concentrations of gests that partially overlapping, but not identical, populations nicotine, the binding in the OPL was unaffected by nicotine at of cells express the c~7 and (~8 subunits. concentrations as high as 1O-3 M (Yazulla, 1979). Drug com- Many of the cells in the GCL that exhibited 017 or cu8 im- petition assays indicated that this resistance to nicotine block munoreactivity were identifiable as ganglion cells since they was not a result of nicotine uptake or enzymatic degradation gave rise to labeled axons. However, we cannot exclude the but rather reflected a property of the binding site itself (Yazulla, possibility that some of the labeled cells in this layer were ama- 1979). This implied that there are at least two pharmacologically crine cells, nor do we know precisely what proportion of the distinct forms of the binding site in pigeon retina. Our obser- labeled cells were ganglion cells. The cuBgtAChR-positive cells vation of intense ot8 subunit immunoreactivity in the OPL sug- were present in highest density in the central retina near the tip gests that one of the known otBgtAChR subtypes that incorporate of the pecten but decreased in density near the periphery. The a8 subunits, or an as yet unidentified ot8 subtype, may display neurons that exhibited 017 immunoreactivity were present in this resistance to nicotine block. lower density than those that displayed a8 immunoreactivity, but were distributed in a similar pattern. Such a change in cell Labeling in the IPL density with increasing eccentricity is characteristic of ganglion The observation that cuBgtAChR immunoreactivity was dis- cells as a whole in the chick retina (Ehrlich, 198 1). tributed throughout the IPL demonstrates that many different The ot7-immunoreactive neurons in the GCL in the central cell types with diverse patterns of arborization express these retina ranged from 5 to 21 km in diameter, while the ot8-im- proteins. The differences between the patterns of ot7 and ot8 munoreactive cells ranged from 5 to 16 km in diameter. In the immunoreactivity within the IPL reflect the differences in ar- peripheral retina, the range of sizes of the ot7- and cu8-immu- borization patterns of the subpopulations of neurons that ex- noreactive cells was essentially the same. This suggests that some press each subunit. Intense cu8 labeling was coincident with the 452 Keyser et al. * a-Bungarotoxin-Sensitive Acetylcholine Receptors in Chick Retina

sitesin the IPL of goldfish retina were associatedwith synapses while the remaining 84% were not associated with synaptic structures. Similarly, the a7 and ot8 immunoreactivity distributed throughout much of the IPL of the chick retina may represent aBgtAChRs not associated with synapses. The category of “nonsynaptic” otBgtAChRs could include proteins in transit from cell bodies to their dendritic terminals. aBgtAChRs and nAChRs in the retina The distribution of nAChR ACh binding subunit immunoreactivity in the chick retina is different, in some respects,than the distribution of aBgtAChR subunit immunoreactivity, and this is summarized diagrammatically in Figure 8. The shadedcells representcholinergic amacrine cells, and examplesof the possible contacts between them and the receptor-containing cells are illustrated. Whereas amacrine, ganglion, and bipolar cells exhibit olBgtAChR immunoreactivity, amacrine, ganglion, and displaced ganglion cells (Fig. 8, large cell labeled ~y3) were labeled following exposure to antibodies againstnAChR subunits(Key- ser et al., 1988; Whiting et al., 1991). In contrast to the wide- spreaddistribution of cYBgtAChR labeling in the IPL, nAChR immunoreactivity was restricted to the two bandswithin which the cholinergic amacrine cellsarborize (seeFig. 1). Furthermore, nAChR immunoreactivity was absent from the OPL. Differ- encesbetween the distribution of orBgtbinding sitesand nAChRs in both avian and rodent brains have been reported as well GCL (Marks and Collins, 1982; Clarke et al., 1985; Swansonet al., 1987; Watson et al., 1988; Britto et al., 1992).

Conclusion Previous studies have demonstrated that otBgtAChRs in the Figure 8. Summary diagram illustrating the distribution of nAChR retina are pharmacologically heterogeneous.For example, only ACh binding subunit and LvBgtAChR subunit immunoreactivity in the 10% of the olBgtbinding in the chick retina is competed for by chick retina. Amacrine, ganglion, and bipolar cells exhibit aBgtAChR neuronal bungarotoxin (Loring et al., 1989). Moreover, as dis- immunoreactivity. In contrast, amacrine, ganglion, and displaced gan- cussedpreviously, nicotine blocks aBgt binding in the IPL but glion cells (large cell labeled 013) are labeled following exposure to antibodies against nAChR subunits. The cholinergic amacrine cells are in- not in the OPL of avian retina (Yazulla, 1979). The correlation dicated by shading, and possible contacts between them and the receptor- between this pharmacological heterogeneity and the three im- containing cells are illustrated. munologically defined cYBgtAChR subtypes identified in chick retina and brain is not clear. While the cu7aBgtAChR subunit inner portion of lamina 1 and all of lamina 2, while a second, has been shown to form a channel that responds to ACh in less intensely labeled band included most of lamina 4. In con- Xenopus oocytes (Couturier et al., 1990), similar studieswith trast, (~7 immunoreactivity was intense in lamina 4, moderate the a7a8 subtype and the cu8subtype have, to date, not detected in laminae 3 and 5, and weaker throughout the remainder of functional channels (R. Anand and J. Lindstrom, unpublished the IPL. The intense a8 immunoreactivity in laminae 1, 2, and observations). Data from such studies and subsequentphar- 4 and a7 immunoreactivity in lamina 4 were in register with macologicalcharacterization will be necessaryfor us to address the arborizations of cells that exhibit nAChR immunoreactivity the question of whether specific aBgtAChRs subtypes subserve (Keyser et al., 1988) and those of ChAT-containing cells (Millar unique functional roles in the nervous system. et al., 1987; Spira et al., 1987). This finding is consistent with the possibility that otBgtAChRsin someretinal neuronsfunction References as nAChRs and that the aBgtAChR-containing cells that ar- Ames A, Pollen DA (1969) Neurotransmission in central nervous borize in theselaminae receive cholinergicinput. However, there tissue: a study of isolated rabbit retina. J Neurophysiol 32:424442. is evidence that otBgtAChRs may bind other ligands in some Anand X, Conroy W, Schoepfer R, Lindstrom J (1991) Neuronal portions of the nervous system. For example, Quik et al. (199 1) nicotinic acetylcholine receptors expressed in Xenopus oocytes have demonstratedthat the peptide thymopoietin competesfor cvBgt a pentameric quaternary structure. J Biochem 266: 11192-l 1198. Ariel M, Daw NW (1982a) Effects of cholinergic drugs on receptive binding sitesin rat brain. In addition, otBgt has been shown to field properties of rabbit retinal ganglion cells. J Physiol (Lond) 324: compete for substanceP binding sitesin rat brain (Utkin et al., 13.5-160. 1989). However, the degreeto which these findings are physi- Ariel M, Daw NW (1982b) Pharmacological analysis of directionally ologically significant is uncertain. sensitive rabbit retinal ganglion cells. J Physiol (Lond) 324:16 1-185. Ballivet M, Nef P, Couturier S, Rungger D, Bader CR, Bertrand D, The significanceof the diffuse labeling observed throughout Cooper E (1988) Electrophysioiogy of a chick neuronal nicotinic the remainder of the IPL is unclear. Zucker and Yazulla (1982) acetylcholine receptor expressed in Xenopus oocytes after cDNA in- reported that 16% of the peroxidase-conjugatedaBgt binding jection. Neuron 1:847-852. The Journal of Neuroscience, February 1993, 13(2) 453

Boulter J, Evans K, Goldman D, Martin G, Treco D, Heinemann S, Millar TJ, Ishimoto I, Chubb IW, Epstein ML, Johnson CD, Morgan Patrick J (1987) Isolation of a cDNA clone coding for a possible IG (1987) Cholinergic amacrine cells of the chicken retina: a light neural nicotinic acetylcholine receptor a-subunit. Nature 3 19:368-374. and electron microscope immunocytochemical study. Neuroscience Britto LRG, Keyser KT, Lindstrom JM, Karten HJ (1992) Immu- 21~725-743. nohistochemical localization of nicotinic acetylcholine receptor sub- Morris B, Hicks A, Wisden W, Darlison M, Hunt S, Barnard E (1990) units in the mesencephalon and diencephalon of the chick (Gal/us Distinct regional expression of nicotinic acetylcholine receptor genes gullus). J Comp Nemo1 3 17:325-340. in chick brain. Mol Brain Res 7:305-3 15.

Carbonetto S. FambrouahI D. Muller K (1978). Nonequivalence of Noel1 WK, Lasansky A (1959) Effects of electrophoretically applied a-bungarotoxin receptors and acetylcholine receptors in chick sym- drugs and electric currents on ganglion cells of the retina. Fed Proc pathetic neurons. Proc Nat1 Acad Sci USA 75: 10 16-l 020. 18:115. Chiappinelli VA (1985) Actions of snake venom toxins on neuronal Patrick J, Stallcup WB (1977) Immunological distinction between nicotinic acetylcholine receptors and other neuronal receptors. Phar- acetylcholine receptor and the ol-bungarotoxin binding component on macol Ther 3 1: l-32. svmnathetic neurons. Proc Nat1 Acad Sci USA 74:4689-4692. Clarke PBS, Schwartz SM, Paul CD, Pert CD, Pert A (1985) Nicotinic Pour&o RG (1979) Localization of cholinergic synapses in mam- binding in rat brain: autoradiographic comparison of acetylcholine, malian retina with peroxidase-conjugated cr-bungarotoxin. Vision Res [3H]nicotine, [1251]-a-bungarotoxin~ J Neurosci 5: 1307-13 15. 19:287-292. Couturier S. Bertrand D. Matter J-M. Hemandez M-C, Bertrand S, Quik M, Babu U, Audhya T, Goldstein G (1991) Evidence for thy- Millar N, Valera S, Barkas T, Ballivet’M (1990) A neurbnal nicotinic mopoietin and thymopoietin/oc-bungarotoxin/nicotinic receptors acetylcholine receptor subunit (alpha 7) is developmentally regulated within the brain. Proc Nat1 Acad Sci USA 88:2603-2607. and forms a homo-oligomeric channel blocked by alpha bungarotox- Ravdin PM, Berg DK (1979) Inhibition of neuronal acetylcholine in. Neuron 5:847-856. sensitivity by a-toxins from Bungurus multicinctus venom. Proc Nat1 Ehrlich D (198 1) Regional specialization of the chick retina as revealed Acad Sci USA 7612072-2076. by the size and density of neurons in the ganglion cell layer. J Comp Schmidt J, Raftery M (1973) A simple assay for the study of solubilize Neurol 195~643-657. acetylcholine receptors. Anal Biochem 52:349-354. Gotti C, Esparis Ogando A, Hanke W, Schlue R, Moretti M, Clementi Schoepfer R, Conroy WC, Whiting P, Gore M, Lindstrom J (1990) F (199 1) Purification and characterization of an a-bungarotoxin Brain (Ybungarotoxin binding protein cDNAs and mAbs reveal sub- receptor that forms a functional nicotinic channel. Proc Nat1 Acad types of this branch of the ligand-gated ion channel gene superfamily. Sci USA 88:3258-3262. Neuron 5:35-48. Halvorsen S, Berg D (1990) Subunit composition of nicotinic acetyl- Schwartz IR, Bok D (1979) Electron microscopic localization of [125I] choline receptors from chick ciliaty ganglia: correlation with known a-bungarotoxin binding sites in the outer plexiform layer of the gold- gene products. J Neurosci 10: 17 1 l-l 7 18. fish retina. J Neurocytol 8:53-66. Jacob MH, Berg DK (1983) The ultrastructural localization of alpha- Snira AW. Millar TJ. Ishimoto I. Enstein ML. Johnson CD. Dahl JL. bungarotoxin binding sites in relation to synapses on chick ciliary I Morgan’IG (1987) Localization of choline acetyltransferase-like im: ganglion neurons. J Neurosci 3:260-27 1. munoreactivity in the embryonic chick retina. J Comp Neurol 260: Jacob MH, Berg DK, Lindstrom JM (1984) A shared antigenic de- 526-538. terminant between the Electrophorus acetylcholine receptor and a Stollberg J, Berg DK (1987) Neuronal acetylcholine receptors: fate of synaptic component on chick ciliary ganglion neurons. Proc Nat1 Acad surface and internal pools in cell culture. J Neurosci 7: 1809-l 8 15. Sci USA 81:3223-3227. Straschill M (1968) Action of drugs on single neurons in the cat’s Jacob MH, Lindstrom JM, Berg DK (1986) Surface and intracellular retina. Vision Res 8:3547. distribution of a putative neuronal nicotinic acetylcholine receptor. Straschill M, Perwein J (1973) The effect of iontophoretically applied J Cell Biol 103:205-214. acetylcholine upon the cat’s retinal ganglion cells. Pfluegers Arch 339: James WM, Klein WL (1985) a-Bungarotoxin receptors on neurons 289-298. isolated from turtle retina: molecular heterogeneity of bipolar cells. Swanson LW, Simmons DM, Whiting PJ, Lindstrom JM (1987) Im- J Neurosci 5:352-361. munohistochemical localization of neuronal nicotinic acetylcholine Keyser KT, Hughes TE, Whiting PJ, Lindstrom JM, Karten HJ (1988) receptors in the rodent central nervous system. J Neurosci 7:3334-3342. Cholinoceptive neurons in the retina of the chick: an immunohisto- Utkin YN, Lazakovich EM, Kasheverov IE, Tsetlin VI (1989) a-Bun- chemical study of the nicotinic acetylcholine receptors. Vis Neurosci garotoxin interacts with the rat brain tachykinin receptors. FEBS Lett 1:349-366. 255:111-115. Lindstrom JM, Einarson B, Tzartos S (198 1) Production and assay of Vaney DI (1990) The mosaic of amacrine cells in the mammalian antibodies to acetylcholine receptors. Methods Enzymol74:432460. retina. Prog Retinal Res 9:49-100. Lipton SA, Frosch MP, Phillips MD, Tauck DL, Aizenman E (1987) Vemallis AB, Conroy WG, Corriveau RA, Halvorsen SW, Berg DK Nicotinic antagonists enhance process outgrowth by rat retinal gan- (199 1) AChR gene products in chick ciliary ganglia: transcripts, sub- elion cells in culture. Science 239: 1293-l 295. units, and receptor subtypes. Sot Neurosci Abstr 17:23. Loring RH, Aizenman E, Lipton SA, Zigmond RE (1989) Character- Vijayaraghavan S, Pugh P, Zhang Z, Rathonz M, Berg D (1992) Nic- ization of nicotinic receptors in chick retina using a snake venom otinic receptors that bind (Ybungarotoxin on neurons raise intracel- neurotoxin that blocks neuronal nicotinic acetylcholine receptor func- lular free Ca++. Neuron 8:353-362. tion. J Neurosci 9:2423-243 1. Vogel Z, Nirenberg M (1976) Localization of acetylcholine receptors Mariani AP (1987) Neuronal and synaptic organization of the outer durina svnautoaenesis in retina. Proc Nat1 Acad Sci USA 73: plexiform layer of the pigeon retina. Am J Anat 179:25-39. 1806:18io. - - Marks MJ, Collins AC (1982) Characterization of nicotine binding in Vogel Z, Maloney GJ, Ling A, Daniels MP (1977) Identification of mouse brain and comparison with the binding of a-bungarotoxin and acetylcholine receptor sites in retina with peroxidase-labeled 01bun- quinuclidinyl benzilate. Mol Pharmacol 22:554-564. garotoxin. Proc Nat1 Acad Sci USA 74:3268-3272. Masland RH, Ames A (1976) Responses to acetylcholine of ganglion Wada E, Wada K, Boulter J, Deneris E, Heinemann S, Patrick J, Swan- cells in an isolated mammalian retina. J Neurophysiol 32:424-442. son L (1989) The distribution ofor2,013,014 and p2 neuronal nicotinic McLane KE, Wu X, Schoepfer R, Lindstrom JM, Conti-Tronconi BM receptor subunit mRNAs in the central nervous system: a hybridiza- (199 1) Identification of sequence segments forming the a-bungaro- tion histochemical study in the rat. J Comp Neurol 284:314-335. toxin binding sites on two nicotinic acetylcholine receptor 01subunits Watson JT, Adkins-Regan E, Whiting P, Lindstrom JM, Podleski TR from the a&n brain. J Biol Chem 266111192-l 1198. (1988) Autoradioaranhic localization of nicotinic acetvlcholine re- M&me KE. Wu X. Lindstrom JM. Conti-Tronconi BM (1992) Ep- ceptors in the brain of the zebra finch (Poephilu gut&a) J Comp itope mapping of polyclonal and monoclonal antibodies against two Neurol 274~255-264. LYbungarotoxin binding (Ysubunits from neuronal nicotinic receptors. Whiting P, Lindstrom J (1986a) Pharmacological properties of im- J Neuroimmunol 38: 115-128. munoisolated neuronal nicotinic receptor. J Neurosci 6:3061-3069. McReynolds JS, Miyachi E (1986) The effect of cholinergic agonists Whiting PJ, Lindstrom JM (1986b) Purification and characterization and antagonists on ganglion cells in the mudpuppy retina. Neurosci of a nicotinic acetylcholine receptor from chick brain. Biochemistry Res [Suppl] 4:S153-S161. 25:2082-2093. 454 Keyser et al. * u-Bungarotoxin-Sensitive Acetylcholine Receptors in Chick Retina

Whiting PJ, Lindstrom JM (1987) Purification and characterization Yazulla S, Schmidt J (1976) Radioautographic localization of 1*51-cu- of a nicotinic acetylcholine receptor from rat brain. Proc Nat1 Acad bungarotoxin binding sites in the retina of goldfish and turtle. Vision Sci USA 84:595-599. Res-16:878-880. - Whiting PJ, Schoepfer R, Conroy WG, Gore MJ, Keyser KT, Shimasaki, Yazulla S, Schmidt J (1977) Two types of receptors for or-bungarotoxin S, Esch F, Lindstrom J (1991) Differential expression of nicotinic in the synaptic layers of the niaeon retina. Brain Res 138:45-47. acetylcholine receptor subtypes in the brain and retina. Mol Brain Young HM, Haney DI (1990)- The retinae of prototherian mammals Res 10:6 l-70. possess neuronal types that are characteristic of non-mammalian ret- Yazulla S (1979) Synaptic layers of the retina: a comparative analysis inae. Vis Neurosci 5:61-66. with [‘Z51]-ol-bungarotoxin. In: Neural mechanisms of behavior in the Zucker C, Yazulla S (1982) Localization of synaptic and nonsynaptic pigeon (Granda AM, Maxwell JH, eds), pp 353-369. New York: nicotinic acetylcholine receptors in the goldfish retina. J Comp Neurol Plenum. 204:188-195.