Retinal Input Induces Three Firing Patterns in Neurons of the

RAPID COMMUNICATION

Retinal Input Induces Three Firing Patterns in Neurons of the Superficial Superior Colliculus of Neonatal Rats FU-SUN LO1 AND R. RANNEY MIZE1,2 1 Department of Cell Biology and Anatomy and 2Department of Ophthalmology and Neuroscience Center of Excellence, Louisiana State University, Medical Center, New Orleans, Louisiana 70112 Lo, Fu-Sun and Ranney R. Mize. Retinal input induces three firing patterns in neurons of the superficial superior colliculus of neonatal rats. J. Neurophysiol. 81: 954 –958, 1999. By using an in vitro isolated brain stem preparation, we recorded extracellular responses to electrical stimulation of the optic tract (OT) from 71 neurons in the superficial superior colliculus (SC) of neonatal rats (P1–13). At postnatal day 1 (P1), all tested neurons (n 5 10) already received excitatory input from the retina. Sixty-nine (97%) superficial SC neurons of neonatal rats showed three response patterns to OT stimulation, which depended on stimulus intensity. A weak stimulus evoked only one spike that was caused by activation of non–N-methyl-D-aspartate (NMDA) glutamate receptors. A moderate stimulus elicited a short train (,250 ms) of spikes, which was induced by activation of both NMDA and non-NMDA receptors. A strong stimulus gave rise to a long train (.300 ms) of spikes, which was associated with additional activation of L-type high-threshold calcium channels. The long train firing pattern could also be induced either by temporal summation of retinal inputs or by blocking g-aminobutyric acid-A receptors. Because retinal ganglion cells show synchronous bursting activity before eye opening at P14, the retinotectal inputs appear to be sufficient to activate L-type calcium channels in the absence of pattern vision. Therefore activation of L-type calcium channels is likely to be an important source for calcium influx into SC neurons in neonatal rats.

INTRODUCTION

During development, the retinotectal pathway of mammals undergoes a synaptic refinement to form the adult pattern of precise retinotopic connections in the superficial layers of the superior colliculus (SC). In the rat, axons of retinal ganglion cells enter the contralateral SC at embryonic day 16 (E16) and enter the ipsilateral SC at E17. The retinotectal synapses are first formed at E17 (Bunt et al. 1983; Lund and Bunt 1976). The crossed and uncrossed retinotectal projections initially distribute across the whole area of the SC. Subsequently most of the ipsilateral projections retract caused by binocular competition (Serfaty and Linden 1994) during the first 10 postnatal days (P10) (Land and Lund 1979). At prenatal ages, axons of retinal ganglion cells typically form side branches and often arborize in topographically inappropriate regions. During the refinement process, side branches and arbors from topographically inappropriate positions retract, and those at topographically correct positions increase. This remodeling of axon arbors leads to an adult-like topographic ordering of axonal The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 954

arborization by P11–12 (O’Leary 1986; Simon and O’Leary 1992a,b; Simon et al. 1994), although even in the adult there are still some aberrant ganglion cells that project to topographically inappropriate regions of the SC (Rahman and Yamadori 1994). The mechanisms underlying the synaptic refinement in another subcortical visual pathway—that from the retina to the lateral geniculate nucleus—were studied intensively (Shatz 1994, 1996). Before eye opening, retinal ganglion cells have spontaneous synchronous bursting activity (Galli and Maffei 1988; Maffei and Galli-Resta 1990; Meister et al. 1991; Wong et al. 1993). This presynaptic activity can be transmitted faithfully to the postsynaptic lateral geniculate neurons (Mooney et al. 1996). The remodeling of presynaptic projections depends on this retinal activity because blocking presynaptic activity and synaptic transmission by tetrodotoxin (TTX), a Na1 channel blocker, prevents the formation of eye-specific layers and ON/OFF sublamination in ferret lateral geniculate nucleus (LGN) (Cramer and Sur 1996; Shatz and Stryker 1988). In addition, blocking N-methyl-D-aspartic acid (NMDA) receptors or nitric oxide synthase (NOS) also disrupts the formation of ON/OFF sublaminae (Cramer and Sur 1994; Hahm et al. 1991). Similarly, in the retinotectal pathway, the synaptic refinement appears to be activity dependent. Blocking NMDA receptor or NOS activity in the postsynaptic neuron also disrupts synaptic refinement in the SC (Mize et al. 1998; Simon et al. 1992; Wu et al. 1994, 1996). Therefore the refinement requires an interaction between presynaptic and postsynaptic activity. However, in vivo studies (Itaya et al. 1995; Molotchnikoff and Itaya 1993) reported that retinotectal synaptic transmission in neonatal rats is not established until P10 –12, when the refinement is almost complete. This finding contradicts the notion that postsynaptic activity plays an important role in the refinement of the retinotectal connection. We used a novel in vitro isolated brain stem preparation (Xia and Lo 1996) to study retinotectal transmission in neonatal rats (P1–13). Both extracellular and whole cell recordings (Mize et al. 1998) demonstrate that all tested superficial SC neurons respond to electrical stimulation of the optic tract (OT) as early as P1. Throughout the period before eye opening (P1–13), almost all neurons show three different response patterns that are associated with activation of non-NMDA receptors, NMDA receptors, and L-type calcium channels, respectively. Activation of L-type calcium channels by retinal inputs is induced by either spatial summation or temporal summation of retinal input and also by blocking g-aminobutyric acid-A

0022-3077/99 $5.00 Copyright © 1999 The American Physiological Society

RESPONSE TO RETINAL INPUT IN NEONATAL RAT SC

(GABAA) receptors. We present our extracellular recording results. METHODS

The method for producing the in vitro isolated brain stem preparation was previously described (Xia and Lo 1996). In brief, SpragueDawley rat pups ranging in age from P1 to P13 were anesthetized with Fluothane (halothane) and killed by decapitation. The brain was removed and immersed in an artificial cerebrospinal fluid (ACSF) bubbled with 95% O2-5% CO2 at room temperature. The brain was cut along the midline and glued onto a silver plate. By using a dissection microscope, the surface of the thalamus and midbrain was exposed by gently removing the forebrain. This ‘‘isolated brain stem’’ was then placed into a submerged-type recording chamber perfused with ACSF of 33°C at a rate of 4 –5 ml/min. Extracellular recordings were made with patch electrodes filled with 1 M NaCl (2– 4 MV). We collected action potentials from cell bodies, all of which were located at a depth of ,140 mm as determined by measuring the depth of the electrode below the pial surface. Action potentials recorded from axons, identified by their characteristic monophasic response and slow decay, were discarded (Ahlsen et al. 1982). Because the SC is not laminated until P30 (Warton and Jones 1985), we named all recorded neurons as superficial SC neurons because of the uncertainty of the division into superficial gray and optic layers. A pair of stimulating electrodes was put on the OT fibers at the level of the thalamus. Single electrical pulses (0.1– 0.3 ms, 20 –1,000 mA) were delivered through stimulating electrodes to evoke postsynaptic responses in the superficial SC. Electrical signals were amplified with an Axoclamp 2B amplifier and collected by an Instrutek VR10B interface unit and stored on a Macintosh Power PC (9500/132) with the Pulses (HEKA) software program. The ACSF contained (in mM) 124 NaCl, 2.5 KCl, 1.25 NaH2PO4, 0.1 MgSO4, 26 NaHCO3, 10 glucose, and 2 CaCl2 (pH 7.4 after saturation by gases). We applied different antagonists such as D-2amino-5-phosphonopentanoic acid (D-APV, 100 mM), 6,7-dinitroquinoxaline-2,3-dione (DNQX, 25 mM), and bicuculline (10 mM) to block NMDA, nonNMDA, and GABA receptors. We also applied nitrendipine (10 –20 mM) to block L-type high-threshold calcium channels. RESULTS

All recorded superficial SC neurons (n 5 71 including 10 at P1) from neonatal rats at P1–13 responded to electrical stimulation of the OT. Among them, 69 (97%) neurons showed 3 different firing patterns in response to stimulation at different intensities. Single spike pattern As shown in the left column of Fig. 1, weak stimulation (1 pulse, 0.1– 0.3 ms, ,100 mA) only induced one spike from superficial SC neurons at P1–13 (Fig. 1, A–E). These results also show a tendency for shortened response latency during development. We defined the minimal intensity that constantly induced one spike as threshold (1T). Short train firing pattern When the stimulus intensity increased to two to five times that of threshold, superficial SC neurons gave a short train of spikes (Fig. 1, A–E, middle traces). The train duration was always ,250 ms. The amplitude of spikes either remained constant or decreased gradually. There was also a tendency of

955

decreasing train duration during development, most likely caused by the increase of GABA-mediated inhibition during development. The responses shown in Fig. 1 are typical of those of the remaining population from which we recorded. Long train firing pattern Increasing stimulus intensity to 5–10 times that of threshold resulted in a long train of spikes in superficial SC neurons. The train duration of spikes ranged from 300 to 1,000 ms. In the same neuron, the long train duration increased with increasing stimulus intensity. This firing pattern was characterized by a specific change in spike amplitude (Fig. 1, A–E, right traces). The first spike was always at the maximal amplitude. Later spikes decreased gradually in amplitude and even vanished. After that, even later spikes increased gradually. The concave envelope of spikes produced by this event implied a large and long-lasting depolarization of the membrane, the likely cause being Na1 channels that were partially or completely inactivated. The long spike train could be induced not only by increasing stimulus intensity but also by other manipulations. As shown in Fig. 1F, a single electrical pulse at 1T intensity elicited just one spike (left trace). Three pulses of the same intensity at 20 –50 Hz induced a long train of spikes (middle trace). This result indicates that the long train firing pattern can be produced either by spatial summation of retinal inputs from increasing stimulus intensity or by temporal summation of retinal inputs from repetitive stimuli. In addition, application of bicuculline, a GABAA receptor antagonist, usually transformed a single spike or short train firing pattern into a long train firing pattern (right trace). This effect was confirmed for seven tested neurons at P1–13. This indicates that GABAA receptor-mediated inhibition prevents the generation of the long train firing at low stimulus intensity. Pharmacological analyses of the three firing patterns The long train firing pattern transformed into a short train pattern at the same stimulus intensity after blocking L-type high-threshold calcium channels by application of nitrendipine, an L-type calcium blocker (Fig. 2, A vs. B and E vs. F). Because nitrendipine did not affect single spike or short train firing patterns induced by weak or moderate stimulation, the blockade of the long train pattern was unlikely to be caused by a decrease of presynaptic neurotransmitter release. Thus the long train firing pattern must be associated with activation of postsynaptic L-type calcium channels. Additional application of D-APV, an NMDA receptor antagonist, changed the short train pattern into a single spike pattern in response to strong stimulation (Fig. 2, B vs. C and F vs. G). The single spike was abolished by additional application of DNQX, a non-NMDA antagonist (Fig. 2, C vs. D and G vs. H). The results shown in Fig. 2 are virtually identical to those nitrendipine, APV, and DNQX results in eight other neurons from which we recorded. In summary, the first spike appears to be mediated by nonNMDA receptors, late spikes in the short train by NMDA receptors, and late spikes in the long train by L-type calcium channels. DISCUSSION

In adult rats, stimulation of the OT induces a single spike in most SC neurons (Sefton 1969). However, in neonatal rats

956

F.-S. LO AND R. R. MIZE

FIG. 1. Three firing patterns of neurons in superficial superior colliculus (SC) of neonatal rats in response to optic tract (OT) stimulation. A–E: extracellular recordings from superficial SC neurons at ages from postnatal day 1 (P1) to postnatal day 13 (P13). Stimuli at threshold (1T) intensity induce single spikes at P1–13 (A–E, left traces). Stimulation artifacts are indicated by arrowheads. Moderate stimuli (2–5 times of the threshold intensity, 2–5T) elicit short trains (,250 ms) of spikes (A–E, middle traces). Strong stimuli (5–10 times of the threshold intensity, 5–10T) evoke long trains (.300 ms) of spikes with specific changes in spike amplitude (A–E, right traces). F: single electrical pulse at threshold intensity induces 1 spike (left trace), whereas 3 pulses at 50 Hz induce a long train of spikes in the same neuron (middle trace). After blocking g-aminobutyric acid-A receptors by bicuculline, a single stimulus at threshold intensity evokes a long train of spikes (right trace).

RESPONSE TO RETINAL INPUT IN NEONATAL RAT SC

957

FIG. 2. Pharmacological analyses of 3 response patterns. Extracellular recordings from superficial SC neurons at ages P7 and P10. A and E: strong stimuli induce long trains of spikes. B and F: blocking L-type calcium channels by application of nitrendipine transfers the long train firing pattern into a short train pattern at the same intensity. C and G: additional blockade of N-methylD-aspartate (NMDA) receptors with D-2-amino-5-phosphonopentanoic acid changes the short train pattern into a single spike pattern. D and H: additional blockade of non-NMDA receptors by 6,7-dinitroquinoxaline-2,3-dione abolishes the single spikes. Note that stimulation intensity is kept constant for each neuron.

(P1–13), stimulation of the OT may elicit three firing patterns, depending on stimulus intensity. Among them, the L-type calcium channel-mediated long train firing pattern is the most striking event. It was not observed in adults. The following factors may account for the difference between neonatal and adult rats. 1) In neonatal rats, the aberrant retinotectal projections probably provide more retinal inputs to single SC neurons than those in adult rats. Namely, a SC neuron may receive inputs from topographically appropriate and inappropriate regions of both retinae. Strong stimulation of the OT recruits more OT fibers, which converge onto the single SC neuron. Synchronous retinal inputs produce spatial summation of excitatory postsynaptic potentials (EPSPs) and depolarize the neuron, which then reaches the threshold of L-type calcium channels (Mize et al. 1998). 2) In neonatal rats, GABA receptor-mediated inhibition is weaker than that in adults. In the rat SC, GABAergic neurons and GABAA receptors are present before birth (Warton et al. 1990), which is consistent with our results. That application of bicuculline transferred a single spike or short train firing pattern into a long train firing pattern even at P1 confirms that GABAA receptor-mediated inhibition already exists at P1. However, several lines of evidence indicate there is a postnatal increase of GABA receptor-mediated inhibition (Schliebs and Rothe 1988; Schliebs et al. 1986; Shi et al.1997; Warton et al. 1990). The weaker inhibition in neonatal rats makes temporal summation of retinal inputs possible. Thus repetitive stimuli at 20 –50 Hz may induce membrane depolarization large enough to activate L-type calcium channels. 3) In neonatal rats, the contribution of NMDA receptors in retinotectal transmission is larger than that in adults (Hestrin 1992; Shi et al. 1997). Our whole cell recordings show that strong stimulation of the OT induces a depolarizing plateau potential mediated by L-type calcium channels. This plateau potential is triggered by an NMDA receptor-mediated EPSP (Mize et al. 1998). Therefore a prominent NMDA-

mediated EPSP is a prerequisite of activation of L-type calcium channels. Although L-type calcium channels are activated by electrical stimulation of the OT, activation of L-type channels is most likely an endogenous event in superficial SC neurons of neonatal rats. Before eye opening, retinal ganglion cells have spontaneous synchronous bursting activity (Galli and Maffei 1988; Maffei and Galli-Resta 1990; Meister et al. 1991; Wong et al. 1993). The synchronization of retinal inputs leads to spatial summation, and high-frequency bursting causes temporal summation of retinal inputs on single SC neurons. Therefore spontaneous retinal activity before eye opening appears likely to be sufficient to activate L-type calcium channels of superficial SC neurons. Synaptic plasticity requires an increase of intracellular Ca21 concentration. Previously, studies emphasized that NMDA receptors are the main source of calcium influx into postsynaptic neurons during synaptic plasticity (Bear 1996; Cramer and Sur 1996; Simon et al. 1992). Our results suggest that activation of NMDA receptors not only causes calcium influx through receptors per se but also secondarily activates L-type high-threshold calcium channels because the long train of spikes is also blocked by APV. This dual activation of NMDA receptors and L-type calcium channels is likely to induce a much longer and larger calcium influx into superficial SC neurons in neonatal rats. It will be important in future studies to determine directly the extent of the contributions of these two channels to calcium influx in both developing and adult rats. It will also be useful to resolve the discrepancy with previous studies that failed to activate tectal neurons via retinal stimulation before P10. (Itaya et al. 1995; Molotchnikoff and Itaya 1993). The discrepancy is likely due to the in vitro versus in vivo preparations used. In in vivo experiments, it is more difficult to maintain synaptic transmission due to anesthesia, hypothermia, anoxia, and other unfavorable conditions. A real advantage of our isolated brain

958

F.-S. LO AND R. R. MIZE

stem preparation is the ability to maintain an intact, functional retinotectal circuitry at very early postnatal ages. We thank Drs. W. Guido for use of his electrophysiological facility and J. R. Cork for valuable help in this study. This work was supported by National Institutes of Health Grants EY-02973 and NS-36000. Address for reprint requests: F.-S. Lo, Dept. of Cell Biology and Anatomy, LSU Medical Center, 1901 Perdido St., New Orleans, LA 70112. Received 10 August 1998; accepted in final form 8 October 1998. REFERENCES AHLSEN, G., LINDSTROM, S., AND LO, F.-S. Functional distinction of perigeniculate and thalamic reticular neurons in the cat. Exp. Brain Res. 46: 118 –126, 1982. BEAR, M. F. NMDA-receptor-dependent synaptic plasticity in the visual cortex. Prog. Brain Res. 108: 205–218, 1996. BUNT, S. M., LUND, R. D., AND LAND, P. W. Prenatal development of the optic projection in albino and hooded rats. Brain Res. 282: 149 –168,1983. CRAMER, K. S. AND SUR, M. Inhibition of nitric oxide synthase disrupts on/off sublamination in the ferret lateral geniculate nucleus. Soc. Neurosci. Abstr. 20:1470, 1994. CRAMER, K. S. AND SUR, M. The role of NMDA receptors and nitric oxide in retinogeniculate development. Prog. Brain Res. 108: 235–244, 1996. GALLI, L. AND MAFFEI, L. Spontaneous impulse activity of rat retinal ganglion cells in prenatal life. Science 242: 90 –91, 1988. HAHM, J.-O., LANGDON, R. B., AND SUR, M. Disruption of retinogeniculate afferent segregation by antagonists to NMDA receptors. Nature 351: 568 – 570, 1991. HESTRIN, S. Developmental regulation of NMDA receptor-mediated synaptic currents at a central synapse. Nature 357: 686 – 689, 1992. ITAYA, S. K., FORTIN, S., AND MOLOTCHNIKOFF, S. Evolution of spontaneous activity in the developing rat superior colliculus. Can. J. Physiol. Pharmacol. 73: 1372–1377, 1995. LAND, P. W. AND LUND, R. D. Development of the ratı´s uncrossed retinotectal pathway and its relation to plasticity studies. Science 205: 698 –700, 1979. LUND, R. D. AND BUNT, A. H. Prenatal development of central optic pathways in albino rats. J. Comp. Neurol. 165: 247–264, 1976. MAFFEI, L. AND GALLI-RESTA, L. Correlation in the discharges of neighboring rat retinal ganglion cells during prenatal life. Proc. Natl. Acad. Sci. USA 87: 2861–2864, 1990. MEISTER, M., WONG, R.O.L., BAYLOR, D. A., AND SHATZ, C. J. Synchronous busts of action potentials in ganglion cells of the developing mammalian retina. Science 252: 939 –943, 1991. MIZE, R. R., CORK, R. J., AND LO, F.-S. Retinal input induces a calciummediated plateau potential in developing SC cells of the rat Abstr. Soc. Neurosci. 24: 811, 1998. MIZE, R. R., WU, H. H., CORK, R. J., AND SCHEINER, C. A. The role of nitric oxide in development of the patch-cluster system and retinocollicular pathways in the rodent superior colliculus. Prog. Brain Res. 118: 133–152, 1998. MOLOTCHNIKOFF, S. AND ITAYA, S. K. Functional development of the neonatal rat retinotectal pathway. Brain Res. Dev. Brain Res. 72: 300 –304, 1993. MOONEY, R., PENN, A. A., GALLEGO, R., AND SHATZ, C. J. Thalamic relay of spontaneous retinal activity prior to vision. Neuron 17: 863– 874, 1996.

O’LEARY, D. D., FAWCETT, J. W., AND COWAN, W. M. Topographic targeting errors in the retinocollicular projection and their elimination by selective ganglion cell death. J. Neurosci. 6: 3692–3705, 1986. RAHMAN, H. A. AND YAMADORI, T. A study on the projection patterns of aberrant retinal ganglion cells in the albino rat. Okajimas Folia Anat. Jpn. 71: 249 –257, 1994. SCHLIEBS, R. AND ROTHE, T. Development of GABAA receptors in the central visual structures of rat brain. Effect of visual pattern deprivation. Gen. Physiol. Biophys. 7: 281–291, 1988. SCHLIEBS, R., ROTHE, T., AND BIGL, V. Dark-rearing affects the development of benzodiazepine receptors in the central visual structures of rat brain. Brain Res. 389: 179 –185, 1986. SEFTON, A. J. The electrical activity of the anterior colliculus in the rat. Vision Res. 9: 207–222,1969. SERFATY, C. A. AND LINDEN, R. Development of abnormal lamination and binocular segregation in the retinotectal pathways of the rat. Brain Res. Dev. Brain Res. 82: 35– 44, 1994. SHATZ, C. J. Role for spontaneous neural activity in the patterning of connections between retina and LGN during visual system development. Int. J. Dev. Neurosci. 12: 531–546, 1994. SHATZ, C. J. Emergence of order in visual system development. Proc. Natl. Acad. Sci. USA 93: 602– 608, 1996. SHATZ, C. J. AND STRYKER, M. P. Prenatal tetrodotoxin infusion blocks segregation of retinogeniculate afferents. Science 242: 87– 89, 1988. SHI, J., AAMODT, S. M., AND CONSTANTINE-PATON, M. Temporal correlations between functional and molecular changes in NMDA receptors and GABA neurotransmission in the superior colliculus. J. Neurosci. 17: 6264 – 6276, 1997. SIMON, D. K. AND O’LEARY, D. D. Development of topographic order in the mammalian retinocollicular projection. J. Neurosci. 12: 1212–1232, 1992a. SIMON, D. K. AND O’LEARY, D. D. Influence of position along the mediallateral axis of the superior colliculus on the topographic targeting and survival of retinal axons. Brain Res. Dev. Brain Res. 69: 167–172, 1992b. SIMON, D. K., PRUSKY, G. T., O’LEARY, D. D., AND CONSTANTINE-PATON, M. N-Methyl-D-aspartate receptor antagonists disrupt the formation of a mammalian neural map. Proc. Natl. Acad. Sci. USA, 89:10593–10597, 1992. SIMON, D. K., ROSKIES, A. L., AND O’LEARY, D. D. Plasticity in the development of topographic order in the mammalian retinocollicular projection. Dev. Biol. 162: 384 –393, 1994. WARTON, S. S. AND JONES, D. G. Postnatal development of the superficial layers in the rat superior colliculus: a study with Golgi-Cox and KluverBarrera techniques. Exp. Brain Res. 58: 490 –502, 1985. WARTON, S. S., PEROUANSKY, M., AND GRANTYN, R. Development of GABAergic synaptic connections in vivo and in cultures from the rat superior colliculus. Brain Res. Dev. Brain Res. 52: 95–111, 1990. WONG, R.O.L., MEISTER, M., AND SHATZ, C. J. Transient period of correlated bursting activity during development of the mammalian retina. Neuron 11: 923–938, 1993. WU, H. H., WAID, D. K., AND MCLOON, S. C. Nitric oxide and the developmental remodeling of retinal connections in the brain. Prog. Brain Res. 108: 273–286, 1996. WU, H. H., WILLIAMS, C. V., AND MCLOON, S. C. Involvement of nitric oxide in the elimination of a transient retinotectal projection in development. Science 265: 1593–1596, 1994. XIA, J. AND LO, F.-S. IPSP-EPSP interaction on neurons in the lateral geniculate neurons of the rat. Chin. Sci. Bull. 41: 1116 –1123, 1996.