Auditory-responsive units in the midbrain vocal nuclei in the ring dove (Streptopelia risoria)

0361-9230/93 $6.00 + .OO Copyright 0 1993 Pergamon Press Ltd.

Brain Research Bulletin, Vol. 30, pp. I I 1-7 15,1993 Printed in the USA. All rights reserved.

BRIEF COMMUNICATION

Auditory-Responsive Units in the Midbrain Vocal Nuclei in the Ring Dove (Streptopeliarisoria) MEI-FANG

CHENG*’

AND

MICHAEL

H. HAVENS?

*Institute of Animal Behavior, Rutgers-The State University, 101 Warren Street, Newark, NJ 07102 fDepartment of Psychology, University of Wyoming, Laramie, Wyoming 82071 Received

19 September

199 1; Accepted

22 July 1992

CHENG, M.-F. AND M. H. HAVENS. Auditory-responsive units in fhe midbrain vocal nucleus of fhe ring dove (Streptopelia risoria). BRAIN RES BULL 30(5/6) 71 I-715, 1993.-The avian midbrain vocal control nucleus, n. intercollicularis (ICo),

receives inputs from midbrain auditory nucleus and from a subdivision of auditory thalamus, suggesting a possibility of auditory response units in the ICo. Using single-unit recordings, we explored auditory response units throughout the dorsomedial midbrain of female ring doves under deep general anesthesia (acute preparation). We found exclusively in the ICo, units that responded preferentially to taped courtship coos of conspecitics (male or female coos) and units that responded to specific frequencies present in coos. Units in the midbrain auditory nucleus also responded to these auditory stimulation in a tonotopic fashion, and were responsive to tone burst as well. The results, along with data from other experiments, suggest that species-specific sound responsive units within the ICo may mediate acoustically facilitated female coos and endocrine responses of the ring dove. Coo-responsive units

Midbrain

Ring dove

VOCALIZATION, from simple calls to complex song, is a conspicuous feature of courtship in many species of birds. Development of normal vocalization in many songbirds depends on acoustic information (16). Song-specific cells have been identified in the forebrain song control nucleus, the caudal n. of the hyperstriatum ventral (HVc) of the white crowned sparrow (14). It has been suggested that the close proximity of song-control units and auditory units may have evolved in association with song learning. If this were the case, one would not expect to find species-specific vocalization responsive units in the vocal control pathway of a bird such as the ring dove that emits simple call and does not learn its vocalization. Nottebohm and Nottebohm ( 17) found that deafened ring dove squabs developed vocalizations that are normal in structure and temporal delivery. It was concluded then that acoustic stimulation plays no role in the ring dove’s coo production. It also follows that auditory responsive units are presumably absent in the vocal control regions of the ring dove. However, our working on the breeding behavior of ring doves has led us to suspect the existence of auditory responsive units in the midbrain vocal control nucleus, the ICo. In the ring dove, female nest-cooing is initiated by male coos, and her own coos stimulate more cooing which, in turn, trigger

endocrine changes reflected in follicular development. Devocalization by lesioning the intercollicular nuclei (7), by cutting the hypoglossal nerves (6), or by puncturing the air space around the syrinx (3) resulted in a deficit in the female’s cooing and blockage of follicular growth in females paired with courting males. Playback tapes of the female’s own coo or the coos of other females were sufficient to stimulate endocrine responses in the devocalized females. Male coos, on the other hand, were effective only in promoting female cooing (3). These observations led us to reason that there may be auditory responsive units in the coo control nuclei to regulate coo production and endocrine responses. In the present study, female doves were deeply anesthetized, so that the birds were absolutely still in order to record units responsive to specific acoustic stimuli. We recorded and compared responses of single units in the auditory nucleus and vocal control nucleus in the dorsal medial region of the midbrain: the nucleus mesencephalicus lateralis dorsalis (MM) and the ICo. The ICo in the ring dove, as in the canary (18) sends a projection to the hypoglossal motor neurons (nXIIts) (11). The medial portion of ICo (mICo), which has a concentration of estrogen binding sites ( 15) plays an obligatory role in mediating the production

’ To whom requests for reprints should be addressed.

711

CHENG

712

MALE

NEST COO

FEMALE NEST COO

AND

HAVENS

of the estrogen-dependent nest-coo in female ring doves, as demonstrated by ablation and replacement studies (7). In contrast, the MLd, which receives auditory inputs from lower-order auditory nuclei (lo), does not concentrate estrogen, and its ablation does not impair coo production (unpublished observation). Data from the MLd, thus, enable us to make an interesting and meaningful comparison with the mICo. METHOD

FIG. I. Oscillographs of the male and female nest-coos. The male coo is characterized by three segments (arrows show beginning of each segment), whereas the female coo contains only two. The male coo averages I .4 s in duration compared with 1. I s for the female. On average. the additional segment of the male accounts for the longer duration. Time calibration: 0.5 s.

All of the ring doves used in single-unit experiments were adult females who were bred at the Institute of Animal Behavior and had been through at least one breeding cycle during which they successfully courted, incubated, and raised squabs. The animals were allowed free access to food, water, and grit at all times and were housed in rooms on a 14L: 10D lighting schedule.

FIG. 2. Locations of the recorded midbrain cells. The diagrams represent coronal sections of the ring dove at the plane of anterior ICo and the bottom at the plane of medial ICo: A = 3.50 and 2.75, respectively, in Karten and Hodo’s (12) atlas of the pigeon. Cells were recorded on both sides of the brain but are located on left side of the diagram for simplicity. Closed circles denote cells that did not respond to auditory stimulation; open circles show the locations of responsive cells. Abbrev: Cb = cerebellum: FRL = formatio reticularis lateralis mesencephah; ICo = nucleus intercollicularis; IP = nucleus interpeducularis; Ipc = nucleus isthmi, pars parvocellularis; LoC = locus ceruleus; MM = nucleus mesencephalicus lateralis, pars dorsalis; MLv = nucleus mesencephalicus, pars ventrallis; OM = tractus occipitomesencephalicus: SGF = stratum grisseum et fibrosum superticiale.

ACOUSTIC

UNITS IN MIDBRAIN

713

FIG. 3. Representative samples of cells in the midbrain of female doves that responded to dove nest-coo calls. In all of the pictures the top trace reflects the firing rate of the neuron, and the bottom trace the osciliographs of the male or the female coo. In this ~pres~ntatio~ the line deviates downward in inverse proportion to the interspike interval. The scale at the IeR shows spikes per second. (A) The resmnse of an MLd cetl to female nest-coo. Notice that the ceil responded during both com~n~nts of the coo. This cell responded during many sounds, including an ascending series of tones. (B) The response of an 1Co cell to two male nest-coos. Notice that the last note ofthe coo (a note absent in femaie coos) elicited the strongest response from the ~4. (C,t>t examples of a neuron in the mKo that responded to male and female nest-coo. The male nest-coo (C) elicited a small but reliable response from this neuron, although the female nest-coo {D) caused a larger response. Time caiibratio~ for aif pictures = 0.5 s.

SurgerJi and recording. Surgical preparation for recording and the recording procedures were performed under general anesthesia (chioropent, 0.2 ml/IO0 g, IM). The bird was placed in a stereotaxic device with a pair of hollow ear bars, the skin incised, and the skull removed above the midbrain. Mineral oil was placed on the surface of the brain to prevent desiccation. Epoxyiite-coated tungsten electrodes with impedances of 1~~~2.5 Qohm were lowered into the midbrain, and extracellular potentials of neurons were recorded throughout the dorsomedial midbrain on both sides of the brain in each animal. Conventional ~j~trophysio~ogica~ equipment was used to amplify, display, count, and store the spikes from single neurons. Standard criteria were used to determine whether a potential from a given recording site was being generated by a single neurona ceil body ( t 0). Each isolated cell was recorded for at least several minutes to establish its spontaneous firing rate and then auditory stimulation was delivered. The spontaneous activity and response to auditory stimulation of the neurons were stored on magnetic tape for later analysis, ~~~i~orys~i~~iati~~. For each unit, auditory stim~t~on was delivered using a Nagra IV-S tape recorder, and co~sjsted of male and female nest-coos and an ascending series of tones from

1~0-1~~ Hz, arranged in 100 Hz increments, This range of frequencies was used to cover the narrow frequencies (450-650 Hz} that make up the coo patterns (9). The coos were taped from breeding birds in the dove colony. The stimuli consisted of approximately 30 s of male and female nest coos and of the same coos played backwards. The dove vocalizations and tones were delivered at 70 dB through a speaker 1 m from the &r&s head. Data analysis. A t~me/am~litude window dj~~minator and a digital counter were used to compute average spike discharge frequencies, and an inte~pjk~ interval timer was used to create frequency histograms, A neural response was defined as a 30% increase or decrease in firing rate as compared with that cell’s spontaneous firing rate recorded for 60 s. This arbitrary criterion has proven useful in other midbrain recording experiments ( 19). Only responses that occurred inde~~dently of EEG ~uctuations were included in the data analysis. This step was taken to ensure that recorded unit responses were specific to auditory ~m~ation. EEG recording was done by implanting stainless steel screws in the skull over the anterior striatum. FoIfowin~ each recording ex~~ment the birds were given an overdose ofanesthetic and transcardially perfused with forma&n. The brains were postfixed overnight in a sucro~/fo~~i~

CHENG

AND

HAVENS

and second segment of the nest-coo is 0.3 s, and the average duration of the second segment is 0.5 s. The male nest-coo contains, in addition, a third phrase that is absent in the female nest-coo. Because of this third note, the duration of the male coo is longer than the female coo (1.4 s for male, I. I s for females).

Neural Responses to Dove Vocalizations

A

A

A

200

300

A

A

400

500

In total, there were 108 cells recorded that met the criteria of neural response. Only 16 of these cells responded to auditory stimulation (see Fig. 2: left hand side). It should be noted that four of the responsive cells were located in the MM and that four out of five cells recorded in this area responded to auditory stimulation. The MLd cells responded to many ofthe test stimuli. Figure 3A shows one of these cells: this cell responded to female nest-coos as well as clicks and tones. Notice that the firing pattern of the cell generally followed the pattern of amplitude modulation in the coo, accelerated during all parts of the coo, and was nearly silent between the first and second phrases of the coo. In contrast, neurons in the dorsomedial areas inclusive of the ICo region responded maximally only to certain segments of vocalizations and to certain frequencies of tones. Figure 3B shows an mICo cell that responded maximally to the third phrase of male nest-coo. This cell did not respond to female nest-coos or tones. The male nest-coo played backwards elicited the same maximal response to the third note (not shown in the figure). Therefore, this cell was responsive to the frequency of sound contained in the third phrase of the coo, and the order of the phrases did not appear to affect its response, as is the case with some HVc neurons in the white-crowned sparrow (14). The responses of one neuron to male and female nest-coos are shown in Figs. 3C and D. This cell, located in the medial portion of ICo, responded very weakly to male nest-coos (3C) but fired very strongly to female nest-coos (3D). This cell did not respond during tonal stimulation. Figure 4 shows an example of cells responsive to both coos and tones. The range of frequencies to which this cell responded are, interestingly, the frequencies contained in dove vocalization.

A

600

TOOA

HZ FIG. 4. Example of neuron that responds selectively to frequencies contained in dove vocalizations. This neuron responds to dove vocalizations as shown in the top figure (details are as in Fig. 3) and responds to 400, 500, and slightly to 600 Hz tones, as shown in the bottom picture (arrows show the beginning of each 10-s tone). The middle trace shows superimposed waveforms of the spike. Time calibration = 0.5 s, top; 0.2 s. middle; 5.0 s, bottom.

solution, after which they were sectioned on a freezing microtome. Sections were mounted on slides and stained for cell bodies with the cresyl violet method. The locations of recorded cells were plotted with reference to microlesions placed during recording. RESULTS

Dove Vocalization Characteristics Female and male nest-coos consist of segments or phrases. The first segment in each coo is a short duration, single note followed by a second longer phrase (see Fig. I). For both sexes, based on six birds each, the average duration between the first

DISCUSSION

The vast majority of cells we recorded in the dorsomedial midbrain (inclusive of mICo) are not auditory. Of 16 responsive cells, none responded to tone bursts but showed preferential responsiveness in one of the following categories: a) responsive to female nest-coo only, b) responsive to both male and female nest-coos but preferentially responsive to female nest-coos, and c) responsive to both male and female nest-coos but with different firing patterns for each. Recent studies have shown that small numbers of cells, such as we report here, may play a significant role in the life of animals. In an elegant study by Margoliash (14). 93 1 units were isolated in one of the telencephalic nuclei, the HVc (hyperstriatum ventrale, pars caudale) of white crown sparrow. Of these units, 27 were song-specific units and only 22 of them were clearly responsive to the birds’ own song. By studying responses of these units in wild-caught and laboratory-raised birds with tutored song, it was strongly suggested that the development of song involves modification of responses of these song-specific units in the HVc. Margoliash proposed three criteria as the basis for deciding that units’ responses were song specific: a) the units do not respond to single tone bursts, b) they only respond to a subset within a song or vocal repertoire, and c) they respond to artificial stimuli that share acoustic parameters with natural song. We be-

ACOUSTIC

UNITS

IN MIDBRAIN

715

lieve most of our responsive units satisfy the first two criteria: the units responded only to certain segments of sounds and did not respond to single tone bursts. Although most of the auditory responsive units recorded in the ICo did not respond to the tonal frequencies, the possibility exists that perhaps these cells are responsive to tones of frequencies not used in our experiment. If such cells exist, we have a complicated situation in which a cell responds specifically to coos within their frequency range but behaves like an ordinary auditory responsive unit outside of this range. Based on available data we feel the most parsimonious explanation would be that these cells are selectively responsive to coos. Of the I6 neurons that exhibited clear auditory responses to taped coos of conspecifics, four units were also responsive to tonal stimulation at specific frequencies contained in the dove vocalization. These units, therefore, are selectively responsive to specific frequencies contained in coos. Because one of these neurons responded to segments of a coo played either forward or backwards, one might argue that these cells were not coo specific. Playing the coo backwards still leaves spectral information that is more coo specific than pure tones of specific frequencies. In any event, the fact that cells responsive to specific frequencies contained in a coo does not necessarily rule out their biological significance. That all the preferentially coo-responsive units are located exclusively in the estrogen-sensitive dorsomedial region of the

midbrain is intriguing in light of a recent PHAL anterograde study in the mICo. Neurons in the mICo send axonal projections to the estrogen-sensitive end zone of shell region surrounding the avian thalamic auditory relay, the Ov (n. ovoidalis) (4). The coo-responsive shell region, in turn, sends massive projections to the posterior medial hypothalamus (PMH) (8) that regulates pituitary function (20). We would like to suggest, therefore, that coo-responsive units in the midbrain vocal control nucleus are involved in distinguishing male from female coos; the male coo stimulates the female to perform coo via a descending pathway, although the female coo promotes LHRH output via one of the two ascending pathways: one terminating in the anterior hypothalamic area (1,4), the other terminating in the PMH by way of Ov shell (8). In conclusion, our results demonstrate for the first time that there are auditory response units in the vocal control nucleus in the midbrain. Moreover, some of these units show a primitive form of selectivity for conspecific sounds. Our study further implies that the close proximity of vocal units and auditory units is not a unique property of vocal learner. ACKNOWLEDGEMENTS 1 would

like to thank MarthaLeah Chaiken for editorial comments on this paper. This work was supported by grant number MH 47010 from NIMH. This is contribution number 541 from the Institute of Animal Behavior.

REFERENCES 1. Akesson. T. R.; de Lanerolle, N.; Cheng, M.-F. Efferent projections of the nucleus intercollicularis in the ring dove. Exp. Neurol. 95: 34-43; 1987. 2. Cheng, M.-F. Female cooing promotes ovarian development in ring doves. Physiol. Behav. 37:37 l-374; 1986. 3. Cheng, M.-F.; Desiderio, C.; Havens, M.; Johnson, A. Behavioral stimulation of ovarian growth. Horm. Behav. 22:388-402; 1988. 4. Cheng, M.-F.; Zuo. M. Connectivities of midbrain vocal control nucleus and hypothalamus: Contribution of enkephalinergic pathways. Sot. Neurosci. Abstr. 18: 1992. 5. Cohen, J. Hormones and brain mechanisms of vocal behaviour in non-vocal learning birds. In: Balthazart, J.; Prove, E.; Gilles, R., eds. Hormones and behaviour in higher vertebrates. Berlin: SpringerVerlag; 1983. 6. Cohen, J.; Cheng, M.-F. Role of vocalization in the reproductive cycle of ring doves (Strepfopelia risoria): Effect of hypoglossal nerve section on the reproductive behavior and physiology of the female. Horm. Behav. 13:113-127. 7. Cohen, J.; Cheng, M.-F. The role of the midbrain in courtship behavior of the female ring dove (Streptopelia risoria): Evidence from radiofrequency lesion and hormone implant studies. Brain Res. 207: 279-301; 1981. 8. Durand, S. E.; Tepper, J. M.; Cheng, M.-F. Novel efferent pathways of the avian auditory thalamus: The nucleus ovoidalis and shell region. J. Comp. Neurol. (in press). 9. Gaunt, A. S.; Gaunt, S. L.; Casey, R. M. Syringeal mechanics reassessed: Evidence from Streptopelia. Auk 99:474-494; 1982. 10. Humphrey, D. R. Extracellular, single-unit recording methods. In: Electrophysiological techniques, a short course. Sot. Neurosci. Abstr. 5: 1979.

1 I. Karten, H. J. The organization of the ascending auditory pathway in the pigeon (Columba livia). I. Diencephalic projections of the inferior colliculus (nucleus mesencephalic lateralis, pars dorsalis). Brain Res. 6409-427; 1967. 12. Karten, H. J.; Hodos, W. A stereotaxic atlas of the brain of the pigeon. Baltimore, MD: John Hopkins Press; 1966: 193. 13. Kroodsma, D. E. Reproductive development in a female songbird: Differential stimulation by quality of male song. Science 192:574575; 1976. 14. Margoliash, D. Acoustic parameters underlying the responses of songspecific neurons in the white-crowned sparrow. J. Neurosci. 3: 10391057; 1983. 15. Martinez-Vargas, M. C.; Stumpf, W. E.; Sar, M. Anatomical distribution of estrogen target cells in the avian CNS: A comparison with mammalian CNS. J. Comp. Neurol. 167:83-104; 1976. 16. Nottebohm, F. The origins of vocal learning. Am. Nat. 106: 116140; 1972. 17. Nottebohm, F.; Nottebohm, M. E. Vocalizations and breeding behaviour of surgically deafened ring doves (Sfreptopelia risoria). Anim. Behav. 19:3 13-327; 197 I. 18. Nottebohm, F.; Stokes, T. M.; Leonard, C. M. Central control of song in the canary, Serinus canarius. J. Comp. Neurol. 165:457486; 1976. 19. Rose, J. D.; Havens, M. H. Lordosis-disrupting tectal lesions alter midbrain unit somatosensory responsiveness in hamsters. Brain Res. Bull. 16:39-45; 1986. 20. Sharp, P. J. Hypothlamic control of gonadotrophin secretion in birds: In: Nistico, Cl; Bolis, L., eds. Progress in non-mammalian brain research. vol. 3. Boca Raton, FL: CRC Press; 1983:124176.