- Email: [email protected]
Light microscopy of GTP-binding protein (Go) immunoreactivity within the retina of different vertebrates
Brain Research, 436 (1987] 384-389 Elsevier
384 BRE 22647
Light microscopy of GTP-binding protein (Go) immunoreactivity within tho retina of different vertebrates Toshio Terashima I , Toshiaki Katada 2, Eikichi Okada 3, Michio Ui 2'* and Yoshiro Inoue 1 tDepartment of Anatomy, Hokkaido UniversitySchool of Medicine, 2Department of Physiological Chemistry, Facultyof Pharmaceutical Sciences, Hokkaido University, Sapporo (Japan) and 3Departmentof Pathology, School of Medicine, Toyama Medicaland Pharmaceutical University, Toyama (Japan) (Accepted 8 September 1987)
Key words: Guanosine triphosphate (GTP)-binding protein; Islet-activating protein (pertussis toxin); Retina; Species difference; Immunohistochemistry
To examine species differences in the distribution pattern of guanosine triphosphate (GTP)-binding protein (Go) within the vertebrate retina, paraffin-embedded retinae from a number of vertebrate species, including the goldfish, frog, turtle, chicken, monkey, and human, were immunohistochemically stained with affinity-purified antibody against the a-subunit of Go. Go-immunoreactive products were found to be located in the neuropil, but not in the cell bodies of neurons, in the retina of all these species. However, some species differences were observed. In the frog, monkey and human, the inner plexiform layer (IPL) was homogeneously stained with this antibody, but in the goldfish, turtle and chicken, the IPL was heterogeneously stained. In the frog, chicken, turtle and human, the outer plexiform layer (OPL) was densely stained with this antibody, but in the goldfish and monkey, the OPL was rather faintly immunoreactive to the antibody. In the goldfish, monkey and human, the outer nuclear layer (ONL) was not immunoreactive to the G oantibody, whereas in the frog, turtle and chicken, the ONL was immunoreactive to it. The implications of these species differences in GOlocalization in the vertebrate retina are discussed.
Guaaosine triphosphate (GTP)-binding proteins (G proteins) have a heterotrimeric structure (aft),) and act to regulate the concentration of cyclic adenosine monophosphate (cAMP), or in some cases, regulate channels directly 6,25. Stimulation or inhibition of adenylate cyclase by hormones or neurotransmitters is mediated by stimulatory (Gs) or inhibitory (Gi) GTP-binding regulatory protein, respectively. Recently, a GTP-binding protein, which is sensitive to islet-activating protein (lAP; pertussis toxin) and differs from G i only in the a-subunit, was purified from brain tissue 1°,15A6,21. This G protein, termed G O ('0' = other), is prominent in the brain, and is related to calcium-dependent intracellular processes via the membrane signal-transducing pathway in the central nervous system 27. Previous immunohistochemical
analysis of the localization of G O within the rat retina 22 revealed that (1) dense Go-immunoreactive products are localized in the inner (IPL) and outer (OPL) plexiform layers, both of which are considered to represent synapse-rich zones within the retina, (2) rather weak Go-immu.noreactive products are localized in the neuropil of the ganglion cell layer (GCL) and inner nuclear layer (INL), and (3) no immunopositive products are seen in the nerve fiber layer (NFL), outer nuclear layer (ONL), and rods and cones (RC). However, it remains unclear whether or not the same pattern of Go immunoreactivities exists in the retina of different vertebrate species ranging from fish to primates. Although all vertebrate retinas are constructed to the same basic plan, some species differences in the structure of the retina
* Present address: Department of Physiological Chemistry, Faculty of Pharmaceutical Sciences, University of Tokyo, Tokyo, Japan. Correspondence: T. Terashima, Department of Anatomy, Hokkaido University School of Medicine, Sapporo 060, Japan. 0006-8993/87/$03.50 O 1987 Elsevier Science Publishers B.V. (Biomedical Division)
385 are present 26. For example, in fishes and reptiles, the bipolar, amacrine, and ganglion cells are not confined to their proper nuclear layer. The amphibian retina is characterized by its long and slender photoreceptor cell. The elaboration of retinal layering reaches its peak in birds, and their INL and IPL are more clearly differentiated into sublayers than in other vertebrates. The mammalian retina is fairly well generalized and presents no bizzare features. The aim of the present study was to elucidate the distribution pattern of Go within the retina of a broad range of vertebrate species at the light microscopic level. The animals examined were the goldfish (Carassius auratus), frog (Xenopus laevis), turtle (Clemmys japonica), chicken (GaUus gallus var. domesticus), monkey (Macacus fuscata) and human (Homo sa,~,,~)o Adult animals c" these various species were obtained from the following sources. Goldfish were purchased from a local pet shop. Frogs, turtles, and chickens were purchased from Hokkaido Laboratory Animals (Sapporo). The eyeball of a monkey was kindly given by Prof. J. Tanji. Human eyeballs were obtained during surgery from two patients (retinoblastoma, uvuitis). The reagents for the avidin-biotin complex (ABC) procedure, i.e. normal goat serum, biotinylated goat lgG, and avidin-biotin-peroxidase complex, were obtained from Vector (Burlingame, CA). The affinity-purified antibody against the a-subunit of Go (Gnu) employed in the presen't study was prepared, characterized by western immunobiot analysis of purified Gnu, and recognized only Goa, as described previously 11"19"23. Goldfish, frogs and turtles were killed by decapitation. Chickens, rats and a monkey were anesthetized with pentobarbitai sodium (30 mg/kg) or ether and perfused via the left ventricle with a solution of 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). Eyeballs of these experimental animals were removed from the orbit and placed in a solution of 4% p-,raformaldehyde. Following fixation for 2 h, the retinae were dissected and immersed in the same fixative for 12-24 h at 4 °C. The human eyeballs were surgically enucleated and immersed in 4% formaldehyde solution. The fixed tissues were dehydrated, embedded in paraffin, serially sectioned at a thickness of 5/~m on a sliding microtome, and mounted on
microscopic slides pretreated with albumin. The paraffin sections were deparaffinized rapidly and stained by a modification of the ABC method of Hsu et al. 8. The schedule for the ABC procedure employed in the present study was the same as that described previously23. Strong Go-immunoreactive products were localized in the IPL of all retinae examined. In the frog, monkey and human, the IPL was homogeneously stained with this antibody (Figs. 1B and 2A,B). However.:, in the goldfish, turtle and chicken, the IPL was heterogeneously staip.~.d with the antibody, and specific band patterns of strong G, immunoreactivity were observed against the rather weakly stained background. In the goldfish, two narrow bands densely stained with Go-antibody were localized in the middle third of the IPL (arrowheads a,b in Fig. 1A); in the turtle, 3 bands of strong G Oimmunoreactivity lay evenly spaced across the IPL (white arrows a-c in Fig. 1C); and in the chicken, 5 prominent bands were observed within the IPL, one at the uppermost, 3 at the n:iddle, and ene at the bottom of this layer (arrows a - e in Fig. 1D). In the frog, turtle, chicken and human, the OPL was densely immunoreactive to the antibody (Figs. 1B-D and 2B), whereas the OPL of the goldfish retina was faintly immunoreactive to it (Fig. 1A). In the monkey, very weak immunoreactive dots were sporadically ob,.,Aj. served in the OPL (Fig. "~ x Species diffeJrences in Go immunoreactivity were also present in 3 nuclear (cellular) layers of the retina, the GCL, ]NL, and ONL. Weak immunoreactivity was observed in the neuropil of the GCL of the frog retina (Fig. 1B), whereas dense, linearly arranged Go-imraunoreactive patches were seen in the neuropil of the GCL of the turtle retina (black arrows in Fig. 1C). "~'he GCL of other species examined showed no significant staining. In the frog and human, moderately immunoreactive products were distributed evenly in the neuropil of the INL (Figs. 1B and 2B), whereas in the turtle and chicken, they were distributed predominantly in the outer half of this layer (Fig. 1C, D). Ill the goldfish and monkey, very weak immunopositive products were sporadically observed in the neuropil of the INL (Figs. 1A and 2A). In the frog, turtle and chicken, the neuropil of the ONL was densely stained with this antibody (Fig. 1B-D), whereas no immunoreactive products were
386
iN
"
"
GC
RC RC
db
I
N
.
1
"
I
I
"
:
",.
.....
~,~,~
I
T
'
"
Z .¢
lOpm
),
.
"
.
•
J
D
Fig. 1. Photomicrographs showing G Oimmunoreactivity in non-mammalian retinae. A, goldfish; B, frog (Xenopus laevis): C, turtle; D, chicken. PE, pigmented epithelium; RC, rods and cones; ON(L), outer nuclear layer; OP(L), outer plexiform layer; IN(L), inner nuclear layer: IP(L), inner plexiform layer; GC(L), ganglion cell layer; NF(L), nerve fiber layer. The magnificatioli of all the photographs of Figs. 1 and 2 are the same to allow comparison of some species differences in the distribution pattern of G,, within the retina.
noted in this layer in the retinae of other species. The organization of the Go immunoreactivity within the retinae of these animals is illustrated schematically in Fig. 3. What do these species differences in distribution pattern of Go within the retina imply? Although this question cannot yet be answered adequa-
tely, because it remains unknown what kind of ~eceptor is regulated by the G O protein, a previour~ biochemical study I indicating that ~,-aminobutyric acid B (GABAB) receptors are coupled to lAP-sensitive GTP-binding proteins (G i or Go), prompts us to speculate that Go may be related to G A B A - m e d i a t e d
387
.r •
RC
ON
-
,,
'
RC
ON OP-.. '~,
= ,5
t
~
",.
.
.
~,j
~
A
"
B
lO~m
Fig. 2. Photomicrographs showing G Oimmunoreactivity in mammalian retinae. A, monkey; B, human. Abbreviations, see Fig. 1.
synaptic mechanisms. In fact, the band patterns of Go immunoreactivity within the IPL of the goldfish, turtle and chicken appear to correspond to those of GABA- or glutamic acid decarboxylase (GAD)-immunopositive terminals within this layer 2"3. Nevertheless, the distnbution patterns of Go in the IPL of the rat, monkey, and human differ from_ those of the G A B A - or GAD-immuno~-eactive terminals: the 5 well-defined laminae based on the G A B A - or G A D immunohistochemistry of these species 2"t4"17 were not recognized from the Go immunohistochemistry. The distribution pattern of G o cannot therefore be explained on the basis of the GABA-mediated receptoc only. It was recenti~ demonstrated that, in addition to G A B A , opioid, noradrenaline and dopamine inhibit the voltage-dependent calcium channels in neurons; and such inhibition is mimicked by intracellular application of guanine nucleotides and blocked by IAP 4"7. Furthermore, Go was also coupled to pu-
,E
I
1
RC
I
000o...oc =oo_or,,o o o_~o o ,OOooou tO O~O o OQooo
I
0 O C 0%°00 OoOo~o OO~-,°OO~OOo-OOooo~°a°~' O 0 I0oo oO
L
II
IPL
~) O ~
GCL
~
m
~
w=m~= i
OOC,~
I
%o000o oooo,~ ~%o oo
!11°/i
I
NFL Goldfish
Frog
Turtle Chicken Rat
Monkey Human
Fig. 3. Schematic illustration of the G o immunoreactivity of
vertebrate retinae, i'he G O immunoreactivity of the rat retina
reported prevmusl~ is also shown in this illustration. Two synapse-rich layers, the IPL and OPL, are densely stained with Go except for the OPL of the monkey retina. The Go-immunoreactive products in the 3 nuclear layers of the retina, GCL, INL, and ONL, appear to be localized on the cell membrane of the neuronal somata. Abbreviations, see Fig. 1. •
"~2
388 rified 13 or partially purified 5 muscarinic receptors in phospholipid vesicles. Conceivably, all of these neurotransmitters could functionally interact with the G Oprotein, rendering it difficult to correlate the distriL'~ution pattern of Go-immunoreactive products with any single neurotransmitter candidate. In summary, the species differences in Go immunoreactivity within the vertebrate retina may reflect overall differences in the localizations of various Go-mediated receptors within the vertebrate retina. Particular attention should be paid to the species differences observed in the ONL: in the frog, turtle and chicken, Go-immunoreactive products were seen in the neuropil of the ONL, whereas in the goldfish, rat, monkey and human, they were not. In the frog, turtle and chicken, the morphological structure of the ONL is unique: this layer in the retina of these species contains somata of photoreceptor cells and displaced bipolar ceils and Landolt's clubs 12. Bipolar cells whose somata are not in their usual location, the IN t.,, are termed 'displaced bipolar cells' and dendritic processes arising from the somata of bipolar cells an :t penetrating through the overall ONL to the external limiting membrane are termed 'Landolt's clubs'. The Go immunoreactivity of the ONL in these species could conceivably correspond to synaptic interplays between photoreceptors and Landolt's clubs of normally localized and displaced bipolar cells, since Go-immunoreactive products were commonly localized in the OPL which is the region of synaptic interplays between photoreceptor, bipolar, and horizontal cells. The present antibody against the a-subunit of Go
purified from the rat brain was thus found to crossreact broadly over different species. Such a broad immunohistochemical cross-reactivity is not surprising, since the antigen dcterrninant of Go appears to be common among different vertebrate species ~. G proteins are currently thought to be present in every cell type of higher organisms and to function as essential molecules in the transmembrane signalling system. It is natural to expect, therefore, that the amino acid sequence of a G protein would be highly conserved in evolution, giving rise to a broad immunohistochemical cross-reactivity of the Go antibody over different vertebrate species. This hypothesis is not without supporting evider, ce, since the amiao acid residues of rat Gs protein are known to have a 99% homology with those of bovine G~ protein (392 out of 394 identities plus only two conservative substitutions) 9'~8'2°. More interestingly, structurally and functionally homologous G proteins are present not only in vertebrates but also in invertebrates. For example, G proteins in sea urchin eggs serving as a specific substrate of lAP cross-react with the antibody raised against G i or G O purified from rat brain 19. In addition, an antiserum prepared against bovine transducin (Gt) , a member of the G family in the photoreceptor, cross-reacts with octopus photoreceptor G protein, indicating that octopus G protein may be akin to vertebrate transducin 24. These findings suggest that invertebrate G proteins may share common antigen determinants with vertebrate G proteins. The validity of this speculation is currently under investigation in our laboratories.
1 Asano, T., Ui, M. and Ogasawara, N., Prevention of the agonist binding to y-aminobutyric acid B receptors by guanine nucleotides and islet-activatingprotein, pertussis toxin, in bovine cerebral cortex, J. Biol. Chem., 260 (1985) 12653-12658. 2 Brandon, C., Retinal GABA neurons: localizationin vertebrate species using an antiserum to rabbit brain glutamate decarboxylase, Brain Research, 344 (1985) 286-295. 3 Brandon, C., Cholinergic neurons in the rabbit retina: immunocytochemical localization, and relationship to GABAergic and cholinesterase-containingneurons, Brain Research, 401 (1987)385-391. 4 Dunlap, K. and Fischbach, G.D., Neurotransmitters decrease the calcium conductance activated by depolarization of embryonic chick sensory neurons, J. Physiol. (Lond.), 317 (1981) 519-535. 5 Florio, V.A and Sternweis, 2.C., Reorganization of re-
solved muscarinic cholinergic receptors with purified GTPbinding proteins, J. Biol. Chem., 2!60(1985) 3477-3483. 6 Gilman, A.G., Receptor-regulated G proteins, Trends Neurosci., 9 (1986) 460-463. 7 Hescheler, J., Rosenthai, W., Trau~'wein, W. and Schuitz. G., The GTP-binding protein, G,,. regulates neuronal calcium channels, Nature (Lond.), 325 (1987) 445-447. 8 Hsu, S.-M., Raine, L. and Fanger, H.. Use of avidin-biotinperoxidase complex (ABC) in immunoperoxidase techniques, J. Histochem. Cytochem.. 29 (1981) 577-580. 9 ltoh, H., Kozasa, T., Nagata, S.. Nakamura, S., Katada, T., Ui, M., lwai, S., Ohtsuka, E., Kawasaki, H.. Suzuki, K. and Kaziro, Y., Molecular citifyingand sequence determination of cDNAs for a subumts of the guanine nucleotide-binding proteins. G,. G i and G,, from rat brain. Proc. Natl. Acad. Sci. U.S.A., 83 (1986)3766-3780. I0 Katada, T., Oinuma. M. and Ui. M., Two guanine nucleo-
389 tide-binding p~'oteins in rat brain serving as the specific substrate of islet-activating protein, pertussis toxin, J. BioL Chem., 261 (1986) 8182-8191. 11 Katada, T., Oinuma, M., Kusakabe, K. and Ui, M., A new GTP-binding protein in b~ain tissues serving as the specific substrate of islet-activating protein, pertussis toxin, FEBS Left., 213 (1987) 353-358. 12 Kouyama, N. and Ohtsuka, T., Quantitative morphologica| study of the outer nuclear layer in the turtle retina, Brain Research, 345 (1985) 200-203. 13 Kurose, H., Katada, T., Haga, T., Haga, K., Ichiyama, A. and Ui, M., Functional interaction of purified muscarinic receptors with purified inhibitory guanine nucleotide regulatory proteins reconstituted in phospholipid vesicles, J. Biol. Chem., 261 (1986) 6423-64.28. 14 Mariani, A.P. and Caserta, M.T., Electron microscopy of glutamate decarboxylase (GAD) immunoreactivity in the inner plexiform layer of the rhesus monkey retina, J. Neurocytol., 15 (1986) 645-655. 15 Milligan, G. and Klee, W., The inhibitory guanine nucleotide binding protein (Ni) purified from brains is a high affinity GTPase, I. Biol. Chem., 260 (1985) 2057-2063. 16 Neer, E.J., Lok, J.M. and Wolf, L.G., Purification and properties of the inhibitory guanine nucleotide regulatory unit of brain adenylate cyclase, J. Biol. Chert,., 259 (1984) 14222-14229. 17 Nishimura, Y., Schwartz, H.L. and Rakic, P., Localization of 7-aminobutyric acid and decarboxylase in rhesus monkey retina, Brain Research, 359 (1985) 351-355. 18 Nukada, T., Tanabe, T., Takahashi, H., Noda, M., Hirose, T., inayama, S. and Numa, S., Primary structure of the a subunit of bovine adenylate cyclase-stimulating G-protein deduced from the cDNA sequence, FEBS Lett., 195 (1986) 220-224. 19 Oinuma, M., Katada, T., Yokosawa, H. and Ui, M., Guanine nucleotide-binding protein in sea urchin eggs serving
as the specific substrate of islet-activating protein, pertussis toxin, FEBS Lett., 207 (1986) 28-34. 20 Robishaw, J.D., Russel, D.W., Harris, B.A., Smigel, M.D. and Gilman, A.G., Deduced primary structure of the a subunit of the GTP-binding stimulatory protein of adenylate cyclase, Proc. Natl. Acad. Sci. U.S.A., 83 (1986) 1251-1255. 21 Sternweis, P.S. and Robishaw, J.D., Isolation of two proteins with high affinity for guanine nucleotides from membranes of bovine brain, J. Biol. Chem., 259 (1984) 13806-13813. 22 Terashima, T., Katada, T., Oinuma, M., Inoue, Y. and Ui, M., Immunohistochemical localization of guanine nucleotide-binding protein in rat retina, Brain Research, 410 (1987) 97-100. 23 Terashima, T., Katada, T., Oinuma, M., Inoue, Y. and Ui, M., Endocrine cells in pancreatic islets of Langerhans are immunoreactive to antibody against guanine nucleotidebinding protein (Go) purified from rat brain, Brain Research, 417 (1987) 190-194. 24 Tsuda, M., Tsuda, T., Terayama, V., Fakada, Y., Akino, T., Yamanaka, G., Stryer, L., Katada, T., Ui, M. and Ebrey, T., Kinship of cephalopod photoreceptor G-protein with vertebrate transducin, FEBS Lr~,., !98 (1986) 5-10. 25 Ui, M., Islet-activating protein, pertussis toxin: a probe for functions of the inhibitory guanine nucleotide regulatory component of adenylate cyclase, Trends Pharmacol. Sci., 5 (1984) 277-279. 26 Walls, G.L., The Vertebrate Eye and Its AdaF',ive Radiation, Hafner, New York, 1967, 785 pp. 27 Worley, P.F., Baraban, J.M., Van Dop, C., Neer, E.J. and Snyder, S.H., G o, a guanine nucleotide-binding protein: immunohistochemical localization in rat brain resembles distribution of second messenger systems, Proc. Natl. Acad. Sci. U.S.A., 83 (1986) 4561-4565.
384 BRE 22647
Light microscopy of GTP-binding protein (Go) immunoreactivity within tho retina of different vertebrates Toshio Terashima I , Toshiaki Katada 2, Eikichi Okada 3, Michio Ui 2'* and Yoshiro Inoue 1 tDepartment of Anatomy, Hokkaido UniversitySchool of Medicine, 2Department of Physiological Chemistry, Facultyof Pharmaceutical Sciences, Hokkaido University, Sapporo (Japan) and 3Departmentof Pathology, School of Medicine, Toyama Medicaland Pharmaceutical University, Toyama (Japan) (Accepted 8 September 1987)
Key words: Guanosine triphosphate (GTP)-binding protein; Islet-activating protein (pertussis toxin); Retina; Species difference; Immunohistochemistry
To examine species differences in the distribution pattern of guanosine triphosphate (GTP)-binding protein (Go) within the vertebrate retina, paraffin-embedded retinae from a number of vertebrate species, including the goldfish, frog, turtle, chicken, monkey, and human, were immunohistochemically stained with affinity-purified antibody against the a-subunit of Go. Go-immunoreactive products were found to be located in the neuropil, but not in the cell bodies of neurons, in the retina of all these species. However, some species differences were observed. In the frog, monkey and human, the inner plexiform layer (IPL) was homogeneously stained with this antibody, but in the goldfish, turtle and chicken, the IPL was heterogeneously stained. In the frog, chicken, turtle and human, the outer plexiform layer (OPL) was densely stained with this antibody, but in the goldfish and monkey, the OPL was rather faintly immunoreactive to the antibody. In the goldfish, monkey and human, the outer nuclear layer (ONL) was not immunoreactive to the G oantibody, whereas in the frog, turtle and chicken, the ONL was immunoreactive to it. The implications of these species differences in GOlocalization in the vertebrate retina are discussed.
Guaaosine triphosphate (GTP)-binding proteins (G proteins) have a heterotrimeric structure (aft),) and act to regulate the concentration of cyclic adenosine monophosphate (cAMP), or in some cases, regulate channels directly 6,25. Stimulation or inhibition of adenylate cyclase by hormones or neurotransmitters is mediated by stimulatory (Gs) or inhibitory (Gi) GTP-binding regulatory protein, respectively. Recently, a GTP-binding protein, which is sensitive to islet-activating protein (lAP; pertussis toxin) and differs from G i only in the a-subunit, was purified from brain tissue 1°,15A6,21. This G protein, termed G O ('0' = other), is prominent in the brain, and is related to calcium-dependent intracellular processes via the membrane signal-transducing pathway in the central nervous system 27. Previous immunohistochemical
analysis of the localization of G O within the rat retina 22 revealed that (1) dense Go-immunoreactive products are localized in the inner (IPL) and outer (OPL) plexiform layers, both of which are considered to represent synapse-rich zones within the retina, (2) rather weak Go-immu.noreactive products are localized in the neuropil of the ganglion cell layer (GCL) and inner nuclear layer (INL), and (3) no immunopositive products are seen in the nerve fiber layer (NFL), outer nuclear layer (ONL), and rods and cones (RC). However, it remains unclear whether or not the same pattern of Go immunoreactivities exists in the retina of different vertebrate species ranging from fish to primates. Although all vertebrate retinas are constructed to the same basic plan, some species differences in the structure of the retina
* Present address: Department of Physiological Chemistry, Faculty of Pharmaceutical Sciences, University of Tokyo, Tokyo, Japan. Correspondence: T. Terashima, Department of Anatomy, Hokkaido University School of Medicine, Sapporo 060, Japan. 0006-8993/87/$03.50 O 1987 Elsevier Science Publishers B.V. (Biomedical Division)
385 are present 26. For example, in fishes and reptiles, the bipolar, amacrine, and ganglion cells are not confined to their proper nuclear layer. The amphibian retina is characterized by its long and slender photoreceptor cell. The elaboration of retinal layering reaches its peak in birds, and their INL and IPL are more clearly differentiated into sublayers than in other vertebrates. The mammalian retina is fairly well generalized and presents no bizzare features. The aim of the present study was to elucidate the distribution pattern of Go within the retina of a broad range of vertebrate species at the light microscopic level. The animals examined were the goldfish (Carassius auratus), frog (Xenopus laevis), turtle (Clemmys japonica), chicken (GaUus gallus var. domesticus), monkey (Macacus fuscata) and human (Homo sa,~,,~)o Adult animals c" these various species were obtained from the following sources. Goldfish were purchased from a local pet shop. Frogs, turtles, and chickens were purchased from Hokkaido Laboratory Animals (Sapporo). The eyeball of a monkey was kindly given by Prof. J. Tanji. Human eyeballs were obtained during surgery from two patients (retinoblastoma, uvuitis). The reagents for the avidin-biotin complex (ABC) procedure, i.e. normal goat serum, biotinylated goat lgG, and avidin-biotin-peroxidase complex, were obtained from Vector (Burlingame, CA). The affinity-purified antibody against the a-subunit of Go (Gnu) employed in the presen't study was prepared, characterized by western immunobiot analysis of purified Gnu, and recognized only Goa, as described previously 11"19"23. Goldfish, frogs and turtles were killed by decapitation. Chickens, rats and a monkey were anesthetized with pentobarbitai sodium (30 mg/kg) or ether and perfused via the left ventricle with a solution of 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). Eyeballs of these experimental animals were removed from the orbit and placed in a solution of 4% p-,raformaldehyde. Following fixation for 2 h, the retinae were dissected and immersed in the same fixative for 12-24 h at 4 °C. The human eyeballs were surgically enucleated and immersed in 4% formaldehyde solution. The fixed tissues were dehydrated, embedded in paraffin, serially sectioned at a thickness of 5/~m on a sliding microtome, and mounted on
microscopic slides pretreated with albumin. The paraffin sections were deparaffinized rapidly and stained by a modification of the ABC method of Hsu et al. 8. The schedule for the ABC procedure employed in the present study was the same as that described previously23. Strong Go-immunoreactive products were localized in the IPL of all retinae examined. In the frog, monkey and human, the IPL was homogeneously stained with this antibody (Figs. 1B and 2A,B). However.:, in the goldfish, turtle and chicken, the IPL was heterogeneously staip.~.d with the antibody, and specific band patterns of strong G, immunoreactivity were observed against the rather weakly stained background. In the goldfish, two narrow bands densely stained with Go-antibody were localized in the middle third of the IPL (arrowheads a,b in Fig. 1A); in the turtle, 3 bands of strong G Oimmunoreactivity lay evenly spaced across the IPL (white arrows a-c in Fig. 1C); and in the chicken, 5 prominent bands were observed within the IPL, one at the uppermost, 3 at the n:iddle, and ene at the bottom of this layer (arrows a - e in Fig. 1D). In the frog, turtle, chicken and human, the OPL was densely immunoreactive to the antibody (Figs. 1B-D and 2B), whereas the OPL of the goldfish retina was faintly immunoreactive to it (Fig. 1A). In the monkey, very weak immunoreactive dots were sporadically ob,.,Aj. served in the OPL (Fig. "~ x Species diffeJrences in Go immunoreactivity were also present in 3 nuclear (cellular) layers of the retina, the GCL, ]NL, and ONL. Weak immunoreactivity was observed in the neuropil of the GCL of the frog retina (Fig. 1B), whereas dense, linearly arranged Go-imraunoreactive patches were seen in the neuropil of the GCL of the turtle retina (black arrows in Fig. 1C). "~'he GCL of other species examined showed no significant staining. In the frog and human, moderately immunoreactive products were distributed evenly in the neuropil of the INL (Figs. 1B and 2B), whereas in the turtle and chicken, they were distributed predominantly in the outer half of this layer (Fig. 1C, D). Ill the goldfish and monkey, very weak immunopositive products were sporadically observed in the neuropil of the INL (Figs. 1A and 2A). In the frog, turtle and chicken, the neuropil of the ONL was densely stained with this antibody (Fig. 1B-D), whereas no immunoreactive products were
386
iN
"
"
GC
RC RC
db
I
N
.
1
"
I
I
"
:
",.
.....
~,~,~
I
T
'
"
Z .¢
lOpm
),
.
"
.
•
J
D
Fig. 1. Photomicrographs showing G Oimmunoreactivity in non-mammalian retinae. A, goldfish; B, frog (Xenopus laevis): C, turtle; D, chicken. PE, pigmented epithelium; RC, rods and cones; ON(L), outer nuclear layer; OP(L), outer plexiform layer; IN(L), inner nuclear layer: IP(L), inner plexiform layer; GC(L), ganglion cell layer; NF(L), nerve fiber layer. The magnificatioli of all the photographs of Figs. 1 and 2 are the same to allow comparison of some species differences in the distribution pattern of G,, within the retina.
noted in this layer in the retinae of other species. The organization of the Go immunoreactivity within the retinae of these animals is illustrated schematically in Fig. 3. What do these species differences in distribution pattern of Go within the retina imply? Although this question cannot yet be answered adequa-
tely, because it remains unknown what kind of ~eceptor is regulated by the G O protein, a previour~ biochemical study I indicating that ~,-aminobutyric acid B (GABAB) receptors are coupled to lAP-sensitive GTP-binding proteins (G i or Go), prompts us to speculate that Go may be related to G A B A - m e d i a t e d
387
.r •
RC
ON
-
,,
'
RC
ON OP-.. '~,
= ,5
t
~
",.
.
.
~,j
~
A
"
B
lO~m
Fig. 2. Photomicrographs showing G Oimmunoreactivity in mammalian retinae. A, monkey; B, human. Abbreviations, see Fig. 1.
synaptic mechanisms. In fact, the band patterns of Go immunoreactivity within the IPL of the goldfish, turtle and chicken appear to correspond to those of GABA- or glutamic acid decarboxylase (GAD)-immunopositive terminals within this layer 2"3. Nevertheless, the distnbution patterns of Go in the IPL of the rat, monkey, and human differ from_ those of the G A B A - or GAD-immuno~-eactive terminals: the 5 well-defined laminae based on the G A B A - or G A D immunohistochemistry of these species 2"t4"17 were not recognized from the Go immunohistochemistry. The distribution pattern of G o cannot therefore be explained on the basis of the GABA-mediated receptoc only. It was recenti~ demonstrated that, in addition to G A B A , opioid, noradrenaline and dopamine inhibit the voltage-dependent calcium channels in neurons; and such inhibition is mimicked by intracellular application of guanine nucleotides and blocked by IAP 4"7. Furthermore, Go was also coupled to pu-
,E
I
1
RC
I
000o...oc =oo_or,,o o o_~o o ,OOooou tO O~O o OQooo
I
0 O C 0%°00 OoOo~o OO~-,°OO~OOo-OOooo~°a°~' O 0 I0oo oO
L
II
IPL
~) O ~
GCL
~
m
~
w=m~= i
OOC,~
I
%o000o oooo,~ ~%o oo
!11°/i
I
NFL Goldfish
Frog
Turtle Chicken Rat
Monkey Human
Fig. 3. Schematic illustration of the G o immunoreactivity of
vertebrate retinae, i'he G O immunoreactivity of the rat retina
reported prevmusl~ is also shown in this illustration. Two synapse-rich layers, the IPL and OPL, are densely stained with Go except for the OPL of the monkey retina. The Go-immunoreactive products in the 3 nuclear layers of the retina, GCL, INL, and ONL, appear to be localized on the cell membrane of the neuronal somata. Abbreviations, see Fig. 1. •
"~2
388 rified 13 or partially purified 5 muscarinic receptors in phospholipid vesicles. Conceivably, all of these neurotransmitters could functionally interact with the G Oprotein, rendering it difficult to correlate the distriL'~ution pattern of Go-immunoreactive products with any single neurotransmitter candidate. In summary, the species differences in Go immunoreactivity within the vertebrate retina may reflect overall differences in the localizations of various Go-mediated receptors within the vertebrate retina. Particular attention should be paid to the species differences observed in the ONL: in the frog, turtle and chicken, Go-immunoreactive products were seen in the neuropil of the ONL, whereas in the goldfish, rat, monkey and human, they were not. In the frog, turtle and chicken, the morphological structure of the ONL is unique: this layer in the retina of these species contains somata of photoreceptor cells and displaced bipolar ceils and Landolt's clubs 12. Bipolar cells whose somata are not in their usual location, the IN t.,, are termed 'displaced bipolar cells' and dendritic processes arising from the somata of bipolar cells an :t penetrating through the overall ONL to the external limiting membrane are termed 'Landolt's clubs'. The Go immunoreactivity of the ONL in these species could conceivably correspond to synaptic interplays between photoreceptors and Landolt's clubs of normally localized and displaced bipolar cells, since Go-immunoreactive products were commonly localized in the OPL which is the region of synaptic interplays between photoreceptor, bipolar, and horizontal cells. The present antibody against the a-subunit of Go
purified from the rat brain was thus found to crossreact broadly over different species. Such a broad immunohistochemical cross-reactivity is not surprising, since the antigen dcterrninant of Go appears to be common among different vertebrate species ~. G proteins are currently thought to be present in every cell type of higher organisms and to function as essential molecules in the transmembrane signalling system. It is natural to expect, therefore, that the amino acid sequence of a G protein would be highly conserved in evolution, giving rise to a broad immunohistochemical cross-reactivity of the Go antibody over different vertebrate species. This hypothesis is not without supporting evider, ce, since the amiao acid residues of rat Gs protein are known to have a 99% homology with those of bovine G~ protein (392 out of 394 identities plus only two conservative substitutions) 9'~8'2°. More interestingly, structurally and functionally homologous G proteins are present not only in vertebrates but also in invertebrates. For example, G proteins in sea urchin eggs serving as a specific substrate of lAP cross-react with the antibody raised against G i or G O purified from rat brain 19. In addition, an antiserum prepared against bovine transducin (Gt) , a member of the G family in the photoreceptor, cross-reacts with octopus photoreceptor G protein, indicating that octopus G protein may be akin to vertebrate transducin 24. These findings suggest that invertebrate G proteins may share common antigen determinants with vertebrate G proteins. The validity of this speculation is currently under investigation in our laboratories.
1 Asano, T., Ui, M. and Ogasawara, N., Prevention of the agonist binding to y-aminobutyric acid B receptors by guanine nucleotides and islet-activatingprotein, pertussis toxin, in bovine cerebral cortex, J. Biol. Chem., 260 (1985) 12653-12658. 2 Brandon, C., Retinal GABA neurons: localizationin vertebrate species using an antiserum to rabbit brain glutamate decarboxylase, Brain Research, 344 (1985) 286-295. 3 Brandon, C., Cholinergic neurons in the rabbit retina: immunocytochemical localization, and relationship to GABAergic and cholinesterase-containingneurons, Brain Research, 401 (1987)385-391. 4 Dunlap, K. and Fischbach, G.D., Neurotransmitters decrease the calcium conductance activated by depolarization of embryonic chick sensory neurons, J. Physiol. (Lond.), 317 (1981) 519-535. 5 Florio, V.A and Sternweis, 2.C., Reorganization of re-
solved muscarinic cholinergic receptors with purified GTPbinding proteins, J. Biol. Chem., 2!60(1985) 3477-3483. 6 Gilman, A.G., Receptor-regulated G proteins, Trends Neurosci., 9 (1986) 460-463. 7 Hescheler, J., Rosenthai, W., Trau~'wein, W. and Schuitz. G., The GTP-binding protein, G,,. regulates neuronal calcium channels, Nature (Lond.), 325 (1987) 445-447. 8 Hsu, S.-M., Raine, L. and Fanger, H.. Use of avidin-biotinperoxidase complex (ABC) in immunoperoxidase techniques, J. Histochem. Cytochem.. 29 (1981) 577-580. 9 ltoh, H., Kozasa, T., Nagata, S.. Nakamura, S., Katada, T., Ui, M., lwai, S., Ohtsuka, E., Kawasaki, H.. Suzuki, K. and Kaziro, Y., Molecular citifyingand sequence determination of cDNAs for a subumts of the guanine nucleotide-binding proteins. G,. G i and G,, from rat brain. Proc. Natl. Acad. Sci. U.S.A., 83 (1986)3766-3780. I0 Katada, T., Oinuma. M. and Ui. M., Two guanine nucleo-
389 tide-binding p~'oteins in rat brain serving as the specific substrate of islet-activating protein, pertussis toxin, J. BioL Chem., 261 (1986) 8182-8191. 11 Katada, T., Oinuma, M., Kusakabe, K. and Ui, M., A new GTP-binding protein in b~ain tissues serving as the specific substrate of islet-activating protein, pertussis toxin, FEBS Left., 213 (1987) 353-358. 12 Kouyama, N. and Ohtsuka, T., Quantitative morphologica| study of the outer nuclear layer in the turtle retina, Brain Research, 345 (1985) 200-203. 13 Kurose, H., Katada, T., Haga, T., Haga, K., Ichiyama, A. and Ui, M., Functional interaction of purified muscarinic receptors with purified inhibitory guanine nucleotide regulatory proteins reconstituted in phospholipid vesicles, J. Biol. Chem., 261 (1986) 6423-64.28. 14 Mariani, A.P. and Caserta, M.T., Electron microscopy of glutamate decarboxylase (GAD) immunoreactivity in the inner plexiform layer of the rhesus monkey retina, J. Neurocytol., 15 (1986) 645-655. 15 Milligan, G. and Klee, W., The inhibitory guanine nucleotide binding protein (Ni) purified from brains is a high affinity GTPase, I. Biol. Chem., 260 (1985) 2057-2063. 16 Neer, E.J., Lok, J.M. and Wolf, L.G., Purification and properties of the inhibitory guanine nucleotide regulatory unit of brain adenylate cyclase, J. Biol. Chert,., 259 (1984) 14222-14229. 17 Nishimura, Y., Schwartz, H.L. and Rakic, P., Localization of 7-aminobutyric acid and decarboxylase in rhesus monkey retina, Brain Research, 359 (1985) 351-355. 18 Nukada, T., Tanabe, T., Takahashi, H., Noda, M., Hirose, T., inayama, S. and Numa, S., Primary structure of the a subunit of bovine adenylate cyclase-stimulating G-protein deduced from the cDNA sequence, FEBS Lett., 195 (1986) 220-224. 19 Oinuma, M., Katada, T., Yokosawa, H. and Ui, M., Guanine nucleotide-binding protein in sea urchin eggs serving
as the specific substrate of islet-activating protein, pertussis toxin, FEBS Lett., 207 (1986) 28-34. 20 Robishaw, J.D., Russel, D.W., Harris, B.A., Smigel, M.D. and Gilman, A.G., Deduced primary structure of the a subunit of the GTP-binding stimulatory protein of adenylate cyclase, Proc. Natl. Acad. Sci. U.S.A., 83 (1986) 1251-1255. 21 Sternweis, P.S. and Robishaw, J.D., Isolation of two proteins with high affinity for guanine nucleotides from membranes of bovine brain, J. Biol. Chem., 259 (1984) 13806-13813. 22 Terashima, T., Katada, T., Oinuma, M., Inoue, Y. and Ui, M., Immunohistochemical localization of guanine nucleotide-binding protein in rat retina, Brain Research, 410 (1987) 97-100. 23 Terashima, T., Katada, T., Oinuma, M., Inoue, Y. and Ui, M., Endocrine cells in pancreatic islets of Langerhans are immunoreactive to antibody against guanine nucleotidebinding protein (Go) purified from rat brain, Brain Research, 417 (1987) 190-194. 24 Tsuda, M., Tsuda, T., Terayama, V., Fakada, Y., Akino, T., Yamanaka, G., Stryer, L., Katada, T., Ui, M. and Ebrey, T., Kinship of cephalopod photoreceptor G-protein with vertebrate transducin, FEBS Lr~,., !98 (1986) 5-10. 25 Ui, M., Islet-activating protein, pertussis toxin: a probe for functions of the inhibitory guanine nucleotide regulatory component of adenylate cyclase, Trends Pharmacol. Sci., 5 (1984) 277-279. 26 Walls, G.L., The Vertebrate Eye and Its AdaF',ive Radiation, Hafner, New York, 1967, 785 pp. 27 Worley, P.F., Baraban, J.M., Van Dop, C., Neer, E.J. and Snyder, S.H., G o, a guanine nucleotide-binding protein: immunohistochemical localization in rat brain resembles distribution of second messenger systems, Proc. Natl. Acad. Sci. U.S.A., 83 (1986) 4561-4565.