1
Du a ion selec i e neu ons in he in e io colliculus o he a :
opog aphic dis ibu ion and ela ion o du a ion sensi i i y o o he
esponse p ope ies.
D. Pé ez-González1,2, M. S. Malmie ca1, J. M. Moo e1, O. He nández1, and E. Co ey1,2
(1) Audi o y Neu ophysiology Uni . Labo a o y o he Neu obiology o Hea ing. The Ins i u e
o Neu oscience o 'Cas illa y León' (INCyL) and Facul y o Medicine. Uni e si y o Salamanca
Campus 'Miguel de Unamuno'
37007 Salamanca - Spain
(2) Depa men o Psychology
Box 351525
Uni e si y o Washing on
Sea le, WA 98195 USA
Keywo ds: Du a ion uning; sound du a ion; audi o y midb ain; il e ing p ope ies;
Co esponding au ho : Ellen Co ey; email: [email p o ec ed]
2
ABSTRACT
Many animals use du a ion o help hem iden i y he sou ce and meaning o a sound. Du a ion
sensi i e neu ons ha e been ound in he audi o y midb ain o mammals and amphibians, whe e
hei selec i i y appea s o co espond o he leng hs o species-speci ic ocaliza ions. In his
s udy, single neu ons in he a in e io colliculus (IC) we e es ed o sensi i i y o sound
du a ion. Abou hal (54%) o he uni s sampled showed some o m o du a ion selec i i y. The
majo i y o hese (76%) we e long pass neu ons ha esponded o sounds exceeding some
du a ion h eshold ( ange: 5 o 60 ms). Band pass neu ons, which only esponded o a es ic ed
ange o du a ions, made up 13% o du a ion sensi i e neu ons (bes du a ions: 15 o 120 ms).
O he uni s displayed sho pass (2%), o mixed (9%) esponse pa e ns. The majo i y o du a ion
sensi i e neu ons we e localized ou side he cen al nucleus o he IC, especially in he do sal
co ex, whe e o e hal o he neu ons sampled had long pass selec i i y o du a ion. Band pass
du a ion uned neu ons we e only ound ou side he cen al nucleus. Cha ac e is ics o du a ion
sensi i e neu ons in he a suppo he idea ha his il e ing a ises h ough an in e ac ion o
exci a o y and inhibi o y inpu s ha con e ge in he IC. Band pass neu ons ypically esponded
a sound o se , sugges ing ha hei uning is c ea ed h ough he same mechanisms ha ha e
been desc ibed in echoloca ing ba s. The inding ha he i s spike la encies o all long pass
neu ons we e longe han he sho es du a ion o which hey esponded suppo s he idea ha
hey ecei e ansien inhibi ion p io o, o simul aneously wi h a sus ained exci a o y inpu . The
anges o selec i i y in a IC neu ons a e wi hin he ange o du a ions o a ocaliza ions.
These da a sugges ha a popula ion o neu ons in he a IC ha e e ol ed o ansmi
3
in o ma ion abou beha io ally ele an sound du a ions using mechanisms ha a e common o
all mammals, wi h an emphasis on long pass uning cha ac e is ics.
INTRODUCTION
Neu ons ha a e selec i e o he du a ion o a sound ha e been desc ibed in he cen al
audi o y sys em o se e al e eb a es including ogs (Po e 1965; Na ins and Cap anica 1980),
ba s (Pinhei o e al. 1991; Casseday e al. 1994, 2000; Eh lich e al. 1997; Fuzesse y 1994; Mo a
and Kössl 2004), ca s (He e al. 1997), chinchillas (Chen 1998) and mice (B and e al. 2000).
Du a ion selec i e neu ons ha e only been ound a o abo e he le el o he in e io colliculus
(IC) o i s homolog in any o he species s udied, so his o m o neu al il e ing appea s o be an
eme gen p ope y ha esul s om ci cui y ope a ing wi hin he midb ain. The IC is inne a ed
by mul iple lowe b ains em s uc u es wi h exci a o y o inhibi o y inpu s ha a y in la ency,
empo al pa e n, and sensi i i y o mul iple sound pa ame e s including equency and ampli ude
(e.g., Casseday e al. 2002). P e ious s udies ha e sugges ed ha du a ion selec i i y is gene a ed
h ough he in e ac ion o such inpu s, in pa icula , he combina ion o con e gen exci a o y
and inhibi o y inpu s wi h speci ic empo al ela ionships (Eh lich e al. 1997; Casseday e al.
1994, 2002; B and e al. 2000; Fau e e al. 2003).
In ce ain species wi h specialized audi o y beha io , he anges o sound du a ions o
which neu ons a e uned co espond closely o he ange o du a ions ound in beha io ally
impo an sounds. This ma ching o neu al du a ion sensi i i y o he du a ion o ocaliza ions
has been desc ibed in ogs, whe e neu ons’ du a ion uning ma ches he du a ions o
communica ion calls (Na ins and Cap anica 1980) and in ba s, whe e neu ons’ du a ion uning
ma ches he du a ions o echoloca ion signals (Pinhei o e al. 1991; Eh lich e al. 1997; Fau e e
4
al. 2003). Mo e ecen ly, du a ion sensi i e neu ons ha e been ound in animals ha lack
ob ious audi o y specializa ions. Al hough he ad an ages o du a ion sensi i i y in mo e
gene alized species ha e no been ho oughly in es iga ed, i po en ially se es o iden i y
speci ic sounds. Fo example, he IC o mice con ains a high p opo ion o long pass neu ons
selec i e o sounds longe han se e al ens o ms. The ew obse ed band pass neu ons
ypically ha e bes du a ions on he o de o ens o ms (B and e al. 2000). Bo h o hese
il e ing p ope ies a e cong uen wi h he du a ion o adul and pup ul asound ocaliza ions
(Liu e al. 2003).
Many calls in a s ha e highly s e eo yped empo al pa e ns (Kal wasse 1990) in which
call du a ion could con ey impo an in o ma ion such as indi idual iden i y, social s a us, o
con ex ual si ua ions. Fo example, a ul asonic ocaliza ions include long (>300 ms) calls ha
a e emi ed du ing a nega i e a ec i e s a e when animals a e an icipa ing punishmen o
engaging in a oidance beha io s and sho (<300 ms) calls ha a e p oduced du ing a posi i e
a ec i e s a e when hey a e an icipa ing a ewa d o displaying app oach beha io s (Knu son e
al. 2002). Isola ion calls o a pups ange om abou 80 o 140 ms, dec easing in du a ion as a
unc ion o he pup’s age. They also di e in leng h be ween indi iduals and show g ea e
simila i y wi hin han ac oss li e s (B udzynski e al. 1999). Thus, du a ion could be used by
ecei e s o ecognize kin and o sp ing (B udzynski e al. 1993). Ra s ha e a ich epe oi e o
sounds ha a e p oduced du ing explo a ion, agonis ic encoun e s, sexual beha io , and o he
social in e ac ions o which du a ion could con ey meaning o conspeci ics. To da e, he e ha e
been no s udies o de e mine whe he he a cen al audi o y sys em con ains du a ion sensi i e
neu ons and, i so, how hese ela e o du a ions o ocal communica ion signals. The aim o his
s udy was o de e mine whe he he e a e du a ion selec i e neu ons in he midb ain o he a ,
5
one o he mos commonly used labo a o y animals and, i so, how he ange o du a ion
selec i i y co esponds o he du a ion o hei known epe oi e o ocaliza ions. P elimina y
epo s ha e been p esen ed in abs ac o m elsewhe e (Pé ez-González e al. 2004a,b).
METHODS
All expe imen al p ocedu es we e app o ed by he Uni e si y o Salamanca Animal Ca e and
Use Commi ee, and con o med o he guidelines o he U.S. Na ional Ins i u es o Heal h. Six y-
one adul male a s (Ra us no egicus; 47 Long E ans pigmen ed, 14 Wis a albino s ain),
weighing be ween 166 and 365 g ams, we e used in his s udy. Anes hesia was induced wi h an
in ape i oneal injec ion o u e hane (1.5 g/kg) and main ained wi h supplemen a y doses (0.5
g/kg) as needed. U e hane is a s anda d anes he ic used in he a , and was chosen because i is
known o a ec inhibi o y p ocesses less han ba bi u a e anes he ics (e.g., Ha a and Ha is
2002). A acheo omy was pe o med in o de o assu e adequa e en ila ion, and a opine sul a e
was adminis e ed subcu aneously (0.05 mg/kg) o educe b onchial sec e ions. Body empe a u e
was main ained a 38 ± 1 ºC wi h a he mos a ically con olled elec ic blanke (Malmie ca e al.
2003, He nández e al. 2005). The a ’s head was immobilized by placing i in a s e eo axic
ame in which he ea ba s had been eplaced by hollow specula ha accommoda ed a sound
deli e y sys em (Rees e al. 1997, He nández e al. 2005). A c anio omy was pe o med o
expose he issue o e he eco ding si e, usually o e he igh IC, and he du a was e lec ed o
allow en y o he elec ode. The su ace o he b ain was i iga ed egula ly wi h saline o
p e en desicca ion.
Acous ic s imula ion and elec ophysiological eco ding was pe o med inside a sound
a enua ed boo h. A ungs en elec ode (Me ill and Ainswo h 1972) was placed o e he

6
exposed co ex and mo ed along he do so en al axis using a piezoelec ic mic od i e (Bu leigh
6000, EXFO Bu leigh, Quebec, Canada) ha was ad anced by emo e con ol om ou side he
boo h. Once he IC was eached, single neu ons we e isola ed, using whi e noise and pu e ones
as sea ch s imuli, and sound-e oked ac ion po en ials we e eco ded ex acellula ly.
S imuli we e syn hesized by a TDT Sys em II wo ks a ion (Tucke -Da is Technologies,
Alachua, FL, USA) using cus om so wa e, and deli e ed h ough a closed ield deli e y sys em
(Rees 1990) ia wo elec os a ic speake s (TDT EC1) con olled by an elec os a ic speake
d i e (TDT ED1). The ou pu o he sys em a each ea was calib a ed in si u using a condense
mic ophone (B üel and Kjæ 4134, Næ um, Denma k) and a DI-2200 spec um analyse
(Diagnos ic Ins umen s L d., Li ings on, Sco land, UK). The maximum ou pu o he sys em
was la om 0.3-5 kHz (≈100 ± 7 dB SPL), om 5-40 kHz (90 ± 5 dB SPL) wi h a no ch a
abou 22-23 kHz wi h a slope o app oxima ely 15 dB/oc a e. Second and hi d ha monic
componen s in he signal we e 45 dB o mo e below he le el o he undamen al a he highes
ou pu le el. Because o he nonlinea i ies o he loudspeake ou pu , all alues a e exp essed as
dB SPL. The highes equency p oduced by ou sys em was limi ed o 40 kHz. Ac ion po en ials
we e ampli ied (x10000) wi h a Bioamp ampli ie (TDT) and il e ed (0.5 – 3 kHz, TDT DB4)
be o e being p ocessed in a spike disc imina o (TDT SD1). The spike imes we e hen s o ed on
a compu e .
Once a single uni was isola ed, i s cha ac e is ic equency (CF) and h eshold we e
de e mined by an au oma ed p ocedu e ha consis ed o he andomized p esen a ion o pu e
ones ac oss a ma ix o equency and in ensi y alues ha ex ended beyond he isually
es ima ed esponse a ea. The s anda d du a ion o hese s imuli was 75 ms. Fo du a ion sensi i e
neu ons, i was se a he neu on’s es ima ed bes du a ion. Fo he expe imen s in which du a ion
7
was a ied, s imuli we e pu e ones a he neu on’s CF, ypically 10 o 20 dB abo e h eshold,
and a ise- all ime o 1 o 2 ms. S imuli we e p esen ed o he ea con ala e al o he IC om
which eco dings we e ob ained. S imulus du a ion was a ied andomly among 10 o 20
s epwise alues s a ing a 2-5 ms and ex ending up o 100-200 ms. To a oid con ounds
associa ed wi h spec al a i ac s, we did no es du a ions below 2 ms. Each du a ion was
p esen ed 10 imes. The da a ob ained we e analyzed and plo ed using comme cial so wa e
(Mic oso Excel, Sigmaplo , and SPSS).
To de e mine whe he a neu on was sensi i e o sound du a ion, we plo ed esponse
p obabili y as a unc ion o du a ion. Response p obabili y (also e e ed o as spike p obabili y)
was de ined as he numbe o p esen a ions (o ials) o a gi en s imulus o which he e was a
leas one s imulus-e oked ac ion po en ial, di ided by he o al numbe o p esen a ions o ha
s imulus. Fo neu ons wi h spon aneous ac i i y, esponse p obabili y was calcula ed a e
sub ac ing he spon aneous a e. This measu e was used ins ead o spike coun s in o de o
dis inguish long pass du a ion sensi i i y om simple sys ema ic inc eases in spike coun wi h
inc eased du a ion in uni s wi h sus ained discha ge pa e ns. A neu on was conside ed o be
du a ion sensi i e i , a any poin , he esponse p obabili y unc ion d opped below 50% o he
maximum. Fo long pass o sho pass neu ons he cu o du a ion was de ined as he poin a
which he p obabili y unc ion d opped below 50% o he maximum; i his poin was be ween
wo es ed du a ions, i was calcula ed by linea in e pola ion. Fo band pass neu ons, which had
cu o s a bo h long and sho du a ions, he bes du a ion was de ined as ha a which spike
coun was maximal. Because we we e unable o es du a ions sho e han 2 ms, i is possible
ha hose neu ons ha we classi ied as sho pass we e ac ually band pass, o ha hey migh
ha e had band pass cha ac e is ics a some sound le els (F emouw e al. 2005).
8
Selec ed eco ding si es and o he landma ks we e ma ked wi h elec oly ic lesions.
These ma ks we e used o econs uc elec ode acks and o con i m he loca ions o he
eco ded uni s. Fo his ological examina ion o he b ain, he animal was adminis e ed a le hal
dose o sodium pen oba bi al (Nembu al; 60 mg/kg in saline, in ape i oneal), and pe used
ansca dially wi h Ringe ’s solu ion ollowed by ixa i e (1% pa a o maldehyde and 1%
glu a aldehyde in 0.1 M phospha e bu e , pH: 7.4). The b ain was imme sed in a 30% suc ose
solu ion o 2-3 days be o e being cu in o 40 µm hick ans e se sec ions using a eezing
mic o ome. Sec ions we e s ained wi h c esyl iole . Neu ons we e assigned o he IC
subdi isions acco ding o he pa cella ion scheme o Malmie ca e al. (1993).
RESULTS
We eco ded esponses o 160 neu ons o sounds o di e en du a ions. O hese, 5
neu ons showed some le el o habi ua ion o epea ed s imuli o esponded weakly o all s imuli
(Co ey e al. 2003; Malmie ca e al. 2004) and we e consequen ly excluded om all analyses. O
he emaining 155 neu ons, 84 (54.2%) showed some o m o du a ion sensi i i y. By a he
mos common o m o du a ion sensi i i y was long pass il e ing. Long pass neu ons we e
de ined as hose wi h a 50% cu o on he sho du a ion side, bu a highe p obabili y o i ing a
all longe du a ions es ed. Neu ons wi h long pass sensi i i y comp ised 41.3% o all IC neu ons
om which we eco ded and 76.2% o du a ion sensi i e neu ons. Long pass neu ons’ cu o
du a ions anged om 5 ms o 60 ms. An example o a long pass neu on is shown in Figu e 1A
and B. This neu on had no spon aneous ac i i y. I was comple ely un esponsi e o du a ions less
han abou 35 ms. F om his poin , he esponse p obabili y g adually inc eased un il eaching he
50% cu o a abou 60 ms and a pla eau a abou 80 ms. This neu on’s esponse ook he o m o
a spa se sus ained discha ge. Figu e 1B shows ha he esponse p obabili y and sound du a ion
9
we e posi i ely co ela ed un il he neu on eached i s maximum p obabili y o i ing. Fi s spike
la ency was ex emely a iable ac oss ials, bu was always mo e han 35 ms. Figu e 1C shows
he esponse om ano he neu on wi h a sus ained discha ge pa e n, bu his neu on was no
du a ion sensi i e (all pass). Al hough he numbe o spikes pe ial was posi i ely co ela ed
wi h sound du a ion, as would be he case i eco ding om an audi o y ne e ibe , he e was no
du a ion es ed a which he neu on ailed o espond eliably. The spike p obabili y unc ion o
his neu on (Fig. 1D) was high (>70%) and essen ially la ac oss all o he du a ions es ed.
Figu e 2A and B shows ano he example o a long pass du a ion sensi i e neu on, in his
case one wi h a obus ansien esponse o no mo e han 2 spikes pe ial. This neu on was
comple ely un esponsi e o 5 ms s imuli, bu esponded eliably o du a ions g ea e han o
equal o abou 15 ms. Once he du a ion exceeded he cu o , he numbe o spikes pe s imulus
emained cons an and esponse p obabili y emained la wi h u he inc eases in du a ion.
Response la ency was consis en ly abou 15 ms, abou he same as he cu o du a ion. Fo
compa ison, Figu e 2C and D show he esponse o an onse esponde ha was no du a ion
sensi i e (all pass). This neu on esponded wi h one spike on e e y ial, e en a he sho es
du a ion es ed. The spike p obabili y was 1.0 o all du a ions.
Because he e appea ed o be a co ela ion be ween he cu o du a ion and esponse
la ency o long pass neu ons, we compa ed he i s spike la encies o long pass neu ons o hose
o non-du a ion sensi i e neu ons a 20 dB abo e h eshold (Fig. 3). The a e age la ency o long
pass neu ons wi h sus ained discha ge pa e ns (30.5 ms, SD 15.1) was signi ican ly longe han
ha o he o he classes o neu ons ( wo- ailed - es , P < 0.001). Non-du a ion sensi i e neu ons
wi h sus ained esponses had a mean la ency o jus 14.0 ms (SD 5.0). Long pass neu ons wi h
ansien esponses had a mean la ency o 16.8 ms (SD 3.9) compa ed wi h non-du a ion
16
echoloca ion calls sugges s ha he la ge numbe o band pass neu ons in he ba ep esen s a
specializa ion o echoloca ion, whe eas long pass neu ons may be used o p ocessing
communica ion sounds, and he e o e be he mo e p imi i e cha ac e is ic.
In a s, he cu o du a ions o sho pass neu ons and he bes du a ions o band pass
neu ons a e consis en wi h he longe du a ions o a s’ ocaliza ions, and all mainly in he
du a ion ange o audible calls emi ed du ing explo a ion, agonis ic beha io , and sexual
beha io . Some a e sensi i e o du a ions co esponding o he ange o pup calls (B udzynski e
al. 1999; Kal wasse 1990). Long pass neu ons’ cu o du a ions we e dis ibu ed o e a ai ly
wide ange, om abou 5-60 ms, indica ing ha a subse o hese neu ons would be un esponsi e
o sho du a ion calls such as he “b oadband” signals emi ed du ing explo a ion o he “sho ”
and “wa e-like modula ed” calls emi ed du ing sexual beha io . These cells would p esumably
be esponsi e o in an calls as well as he “sc eams” and “ equency s ep” calls emi ed du ing
agonis ic beha io (Kal wasse 1990).
Ou inding ha mos band pass and long pass neu ons we e loca ed ou side he cen al
nucleus o he IC also seems di e en om he si ua ion in he ba , whe e mos du a ion sensi i e
neu ons, a leas band pass ones, a e ound in he caudal hal o he cen al nucleus (Eh lich e al.
1997). The inding ha he a has a high p opo ion o long pass neu ons in he do sal pa o he
IC is consis en wi h he idea ha neu ons in his egion ha e longe in eg a ion imes due o he
in luence o cascaded in insic p ojec ions wi hin he IC i sel (Mille e al. 2005) and/o
descending inpu om he co ex (Caicedo and He be 1993; Saldaña e al. 1996; e iewed in
Malmie ca 2003).
P e ious s udies in which inhibi ion was blocked (Casseday e al. 1994, 2000; Fuzesse y
and Hall 1999) p o ide con incing e idence ha du a ion sensi i i y o all ypes can be c ea ed

17
h ough he in e ac ion o exci a ion and inhibi ion wi hin he IC. Howe e , he e is also e idence
ha inhibi ion may no be di ec ly esponsible o he du a ion uning o all IC neu ons
(Fuzesse y and Hall 1999). The ac ha la encies o long pass neu ons we e always longe han
hei cu o du a ion, and he a e age la ency o long pass neu ons was longe han he a e age
la ency o hose ha we e no sensi i e o du a ion, is consis en wi h he idea ha he du a ion o
a single onse -e oked inhibi o y inpu de e mines bo h he cu o du a ion and esponse la ency
o long pass neu ons (Fig. 14 A,B) (B and e al. 2000; Fau e e al. 2003). Fo neu ons wi h
sus ained esponses and pa adoxical la ency shi (Sulli an 1982), whose cu o du a ions shi ed
o longe alues wi h inc easing sound le el, his clea ly seems o be he case. Fo hese neu ons,
he ela i e s eng h and du a ion o a ansien , onse -e oked inhibi o y inpu appa en ly g ew a
a as e a e han ha o a sus ained, onse -e oked exci a o y inpu (Fig. 14C). P e ious s udies
ha e shown ha in many IC neu ons, inhibi ion p ecedes exci a ion (e.g., Co ey e al. 1996;
Kuwada e al. 1997), and ha such inhibi ion can cause pa adoxical la ency shi (Co ey e al.
1996).
Fo hose neu ons whose cu o s shi ed o sho e du a ions wi h inc eases in ampli ude,
his shi could be explained i he si ua ion we e e e sed so ha he s eng h and du a ion o
inhibi ion g ew a a slowe a e han ha o exci a ion (Fig. 14D). Consis en wi h his idea is he
inding ha inhibi ion can ha e a lowe h eshold han exci a ion (Co ey e al 1996; Kuwada e
al. 1997) and ha blocking inhibi ion causes a lowe ing o h eshold in some IC neu ons (Va e
e al. 1992, LeBeau e al. 2001). An equally plausible al e na i e explana ion would be ha hese
neu ons we e in eg a ing sub h eshold exci a o y inpu s o e a pe iod as long as 90 ms, possibly
h ough p ojec ions om a cascaded sys em o delay lines wi hin he IC (Mille e al. 2005) o
con e gence o ascending inpu s wi h descending exci a o y p ojec ions om ou side he IC. In
18
his case, inc eases in sound ampli ude would esul in la ge , longe las ing EPSPs, allowing
empo al summa ion o occu . The inding ha blocking inhibi ion in some IC neu ons in he
pallid ba did no abolish hei du a ion uning is consis en wi h his hypo hesis, and sugges s
ha du a ion sensi i i y may a ise h ough mul iple mechanisms. Mo eo e , he ela i e
p e alence o each mechanism may a y ac oss species. Fo long pass neu ons wi h ansien
esponses, a ansien onse -e oked inhibi ion could no easily c ea e a long pass esponse,
because i would p esumably ha e he same e ec ega dless o s imulus du a ion. Howe e , i
ansien inhibi ion we e e oked by he o se o he s imulus, i could elimina e esponses o
sounds wi h du a ions sho e han he la ency o he exci a o y inpu , while pe mi ing esponses
o sounds a all longe du a ions. The la ency o an o se esponse ( ela i e o he onse o he
s imulus) is equal o he du a ion o he s imulus plus he la ency o he inhibi ion ela i e o
sound o se . Based on he da a in Fig. 6, he a e age alue o he heo e ical 50% cu o
du a ion a which he la ency o he o se inhibi ion becomes long enough o pe mi a esponse
o onse -e oked ansien exci a ion is 7.3 ms (SD 3.1), a pe ec ly plausible alue o la ency in
he a IC.
The obse a ion ha band pass neu ons we e ypically o se esponde s is consis en
wi h p e ious s udies showing ha ansien onse -e oked exci a ion and sus ained inhibi ion
ollowed by an exci a o y ebound a sound o se is esponsible o his esponse pa e n
(Casseday e al. 1994, 2000; Fau e e al. 2003). To u he suppo his iew, all o he o se
esponde s in his s udy showed some kind o du a ion sensi i i y. Al hough he numbe o o se
neu ons in ou sample is ela i ely small (7), he same inding has been published in o he
species, including he mouse (B and e al. 2000) and chinchilla (Chen 1998). In he ba , up o
80% o du a ion sensi i e neu ons a e o se esponde s (F emouw e al. 2005), a inding ha is
19
consis en wi h he obse a ion ha he majo i y o du a ion sensi i e neu ons in he ba ha e
band pass uning (Eh lich e al. 1994). I seems his ela ionship also holds in egions ou side he
IC. Fo example, in he do sal zone o he audi o y co ex o he ca , 24/28 o se neu ons we e
desc ibed as du a ion sensi i e (He e al. 1997). The e o e, i seems o be a highly consis en
inding, bo h ac oss species and ac oss sound le els, ha a majo i y o o se esponde s a e
sensi i e o s imulus du a ion. The conclusion o be d awn om conside ing all o hese da a
oge he is ha du a ion sensi i i y in he IC akes many o ms, and may a ise h ough di e se
mechanisms, bu ha hese mechanisms a e consis en ac oss species.
Al hough i is emp ing o hink o du a ion sensi i i y as an independen il e ing
mechanism ha c ea es “ ea u e de ec o s” selec i e o biologically ele an sound du a ions,
we mus also conside he possibili y ha i is bu one componen o a la ge scale mechanism
unde lying he il e ing and analysis o complex empo al sequences o na u al sounds. The ac
ha du a ion uned neu ons in he ba expe ience supp ession analogous o o wa d and
backwa d masking when p esen ed wi h sounds o di e en du a ions (Fau e e al. 2003)
sugges s ha hei esponses a e also de e mined by he empo al con ex in which a gi en sound
occu s.
In he a , we ound ha mos du a ion sensi i e neu ons we e loca ed in he co ical a eas
o he IC. This inding was su p ising because in he ba hey ha e been epo ed only in he
CNIC, mainly in he caudal po ion (Eh lich e al. 1997). In he o he s udies, ei he he
eco dings we e pe o med only in he CNIC (Eh lich e al. 1997; Fau e e al. 2003; F emouw e
al. 2005) o he e was no p ecise indica ion o he loca ion (Pinhei o e al. 1991; Chen 1998;
Fuzesse y and Hall, 1999; B and e al. 2000; Mo a and Kössl 2004), so he oppo uni ies o
compa ison a e limi ed. The e a e some ana omical di e ences ha may explain he di e en
20
dis ibu ion o he du a ion sensi i e neu ons in he a and ba . In Ep esicus uscus, he ba in
which du a ion uned neu ons we e localized, he “co ical” a eas su ounding he CNIC a e no
nea ly as p ominen as hey a e in a s. In Ep esicus he “pe icen al a ea” is qui e small (less
han 200 µm hickness) and includes egions ha would p esumably co espond o bo h he
ECIC and DCIC in a s (Casseday and Co ey 1992; Co ey and Ca 2005). E en hough we
canno ule ou a sampling bias due o he ela i ely la ge size o he co ical a eas in he a , i is
s ill signi ican ha only 20% o he CNIC uni s in ou sample had any kind o du a ion
sensi i i y while he pe cen age was much highe in he co ical a eas. This inding sugges s ha
du a ion sensi i e neu ons in he a may be impo an o o ien ing o sounds o speci ic
du a ions.
The inding ha all o he di e en ypes o du a ion sensi i e neu ons ha ha e been
desc ibed in o he species a e p esen in he a IC ein o ces he idea ha du a ion uning is an
eme gen p ope y in he midb ain o all e eb a e species, and ha i a ises h ough mul iple
mechanisms and akes mul iple o ms. The selec i i y o a IC neu ons o ela i ely long
du a ions ha co espond o hose wi hin he a ’s ocaliza ion epe oi e sugges s ha he e a e
species-speci ic adjus men s in la encies, discha ge pa e ns, and synap ic s eng hs ha op imize
he ange o sound du a ion sensi i i y in each species o p oduce il e ing p ope ies ha a e
impo an o beha io s ha depend on analyzing he empo al pa e n o sound.
Acknowledgmen s: Resea ch suppo ed by he Spanish JCyL-UE (SA0400/04, MSM),
DGES (BFI-2003-09147-02-01, MSM) and NIH (DC00607, EC).
P esen add ess o J.M. Moo e: Depa men o Neu obiology & Beha io , Co nell
Uni e si y, Mudd Hall, Towe Rd., I haca, NY 14853 USA.
21
REFERENCES
B and A, U ban A, and G o he B. Du a ion uning in he mouse audi o y midb ain. J
Neu ophysiol 84: 1790-1799, 2000.
B udzynski SM, Biha i F, Ociepa D and Fu XW. Analysis o 22 kHz ul asonic ocaliza ion
in labo a o y a s: long and sho calls. Physiol Beha 54: 215-221, 1993.
B udzynski SM, Kehoe P and Callahan M. Sonog aphic s uc u e o isola ion-induced
ul asonic calls o a pups. De Psychobiol 34: 195-204, 1999.
Caicedo A and He be H. Topog aphy o descending p ojec ions om he in e io colliculus o
audi o y b ains em nuclei in he a . J Comp Neu ol 328: 377-392, 1993.
Casseday JH and Co ey E. F equency uning p ope ies o neu ons in he in e io colliculus o
an FM ba . J Comp Neu ol 319: 34-50, 1992.
Casseday JH, Eh lich D and Co ey E. Neu al uning o sound du a ion: Role o inhibi o y
mechanisms in he in e io colliculus. Science 264: 847-850, 1994.
Casseday JH, Eh lich D and Co ey E. Neu al measu emen o sound du a ion: Con ol by
exci a o y-inhibi o y in e ac ions in he in e io colliculus. J Neu ophysiol 84:
1475-1487, 2000.
Casseday JH, F emouw T and Co ey E. The in e io colliculus: Hub o he audi o y sys em.
In: In eg a i e Func ions in he Mammalian Audi o y Pa hway, edi ed by D. Oe el, A.N.
Poppe and R.R. Fay. New Yo k: Sp inge -Ve lag, 2002, pp 238-318.
Chen GD. E ec s o s imulus du a ion on esponses o neu ons in he chinchilla in e io
colliculus. Hea Res 122: 142-150, 1998.

22
Co ey E and Ca CE. The in e io colliculus in ba s and bi ds. In: The In e io Colliculus,
edi ed by J.A. Wine and C.E. Sch eine . New Yo k: Sp inge -Ve lag, 2005, pp 493-536.
Co ey E, Kaue JA and Casseday JH. Whole-cell pa ch-clamp eco ding e eals sub h eshold
sound-e oked pos synap ic cu en s in he in e io colliculus o awake ba s. J Neu osci
16(9): 3009-18, 1996.
Co ey E, Pé ez-González D, He nández O and Malmie ca MS. No el y uni s in he in e io
colliculus o he a . Neu osci Abs 29, 2003.
Eh lich D, Casseday JH and Co ey E. Neu al uning o sound du a ion in he in e io
colliculus o he big b own ba , Ep esicus uscus. J Neu ophysiol 77: 2360-2372, 1997.
Fau e PA, F emouw T, Casseday JH and Co ey E. Tempo al masking e eals p ope ies o
sound-e oked inhibi ion in du a ion- uned neu ons o he in e io colliculus. J Neu osci
23: 3052-3065, 2003.
F emouw T, Fau e P, Casseday JH and Co ey E. Du a ion selec i i y o neu ons in he
in e io colliculus o he big b own ba : ole ance o changes in sound le el. J
Neu ophysiol 94: 1869-1878, 2005.
Fuzesse y ZM. Response selec i i y o mul iple dimensions o equency sweeps in he pallid
ba in e io colliculus. J Neu ophysiol 72: 1061-1079, 1994.
Fuzesse y ZM and Hall JC. Sound du a ion selec i i y in he pallid ba in e io colliculus. Hea
Res 137: 137-154, 1999.
Galazyuk AV and Feng AS. Encoding o sound du a ion by neu ons in he audi o y co ex o
he li le b own ba , Myo is luci ugus. J Comp Physiol [A] 180: 301-311, 1997.
Gould E. S udies o ma e nal-in an communica ion and de elopmen o ocaliza ions in he
ba s Myo is and Ep esicus. Comm Beha Biol 5: 263-313, 1971.
23
Ha a K, Ha is RA. The anes he ic mechanism o u e hane: he e ec s on neu o ansmi e -
ga ed ion channels. Anes h Analg 94: 313-318, 2002.
He J, Hashikawa T, Ojima H and Kinouchi Y. Tempo al in eg a ion and du a ion uning in he
do sal zone o ca audi o y co ex. J Neu osci 17: 2615-2625, 1997.
He nández O, Espinosa N, Pé ez-González D and Malmie ca MS. The in e io colliculus o
he a : a quan i a i e analysis o monau al equency esponse a eas. Neu oscience 132:
203-217, 2005.
Kal wasse MT. Acous ic signaling in he black a (Ra us a us). J Comp Psychol 104: 227-
232, 1990.
Knu son B, Bu gdo J and Panksepp J. Ul asonic ocaliza ions as indices o a ec i e s a es
in a s. Psychol Bull 128: 961-977, 2002.
Kuwada S, Ba a R, Yin TC, Oli e DL, Habe ly LB and S an o d TR. In acellula
eco dings in esponse o monau al and binau al s imula ion o neu ons in he in e io
colliculus o he ca . J Neu osci 17: 7565-7581, 1997.
LeBeau FE, Malmie ca MS and Rees A. Ion opho esis in i o demons a es a key ole o
GABA(A) and glycine gic inhibi ion in shaping equency esponse a eas in he in e io
colliculus o guinea pig. J Neu osci 21: 7303-7312, 2001.
Liu RC, Mille KD, Me zenich MM and Sch eine CE. Acous ic a iabili y and
dis inguishabili y among mouse ul asound ocaliza ions. J Acous Soc Am 114: 3412-
3422, 2003.
Malmie ca MS. The s uc u e and physiology o he a audi o y sys em: an o e iew.
In Re Neu obiol 56: 147-211, 2003.
24
Malmie ca MS, Blacks ad TW, Osen KK, Ka agulle T and Molowny RL. The cen al
nucleus o he in e io colliculus in a : a Golgi and compu e econs uc ion s udy o
neu onal and lamina s uc u e. J Comp Neu ol 333: 1-27, 1993.
Malmie ca MS, He nández O, Falconi A, Lopez-Po eda EA, Me chán M and Rees A. The
commissu e o he in e io colliculus shapes equency esponse a eas in a : an in i o
s udy using e e sible blockade wi h mic oinjec ion o kynu enic acid. Exp B ain Res
153: 522-529, 2003.
Malmie ca MS, Pé ez-González D, He nández O and Co ey E. No el y esponses in single
neu ons o he audi o y midb ain a e gene a ed by apid s imulus-speci ic habi ua ion.
FENS Abs 2A15721, 2004.
Me ill EG and Ainswo h A. Glass-coa ed pla inum-pla ed ungs en mic oelec odes. Med
Biol Eng 10: 662-672, 1972.
Mille KE, Casseday JH and Co ey E. Rela ion be ween in insic connec ions and
iso equency con ou s in he in e io colliculus o he big b own ba , Ep esicus uscus.
Neu oscience, In p ess, 2005.
Mo a EC and Kössl M. Ambigui ies in sound-du a ion selec i i y by neu ons in he in e io
colliculus o he ba Molossus molossus om Cuba. J Neu ophysiol 91: 2115-2126, 2004.
Na ins PM and Cap anica RR. Neu al adap a ions o p ocessing he wo-no e call o he
Pue o Rican ee og, Eleu he odac ylus coqui. B ain Beha E ol 17: 48-66, 1980.
Pé ez-González D, Malmie ca MS, Moo e J and Co ey E. Du a ion sensi i e neu ons in he
in e io colliculus o he a . Assoc Res O ola yngol Abs: 232, 2004a.
Pé ez-González D, Malmie ca MS, Moo e J and Co ey E. Sound du a ion selec i i y in he
a audi o y midb ain. FENS Abs ol 2, A226.6, 2004b.
25
Pinhei o AD, Wu M and Jen PHS. Encoding epe i ion a e and du a ion in he in e io
colliculus o he big b own ba , Ep esicus uscus. J Comp Physiol [A] 169: 69-85, 1991.
Po e HD. Pa e ns o acous ically e oked discha ges in he mesencephalon o he bull og. J
Neu ophysiol 28: 1155-1184, 1965.
Rees A. A closed ield sound sys em o audi o y neu ophysiology. J Physiol Lond 430: 6, 1990.
Rees A, Sa baz A, Malmie ca MS and LeBeau FEN. Regula i y o i ing o neu ons in he
in e io colliculus. J Neu ophysiol 77: 2945-2965, 1997.
Saldaña E, Feliciano M and Mugnaini E. Dis ibu ion o descending p ojec ions om p ima y
audi o y neoco ex o in e io colliculus mimics he opog aphy o in acollicula
p ojec ions. J Comp Neu ol 371: 15-40, 1996.
Simmons JA. A iew o he wo ld h ough he ba 's ea : he o ma ion o acous ic images in
echoloca ion. Cogni ion 33: 155-99, 1989.
Sulli an WE 3 d. Possible neu al mechanisms o a ge dis ance coding in audi o y sys em o
he echoloca ing ba Myo is luci ugus. J Neu ophysiol 48: 1033-1047, 1982.
Va e M, Habbich H, Kossl M and G o he B. The unc ional ole o GABA and glycine in
monau al and binau al p ocessing in he in e io colliculus o ho seshoe ba s. J Comp
Physiol [A] 171: 541-553, 1992.
FIGURE CAPTIONS
Figu e 1. Long pass du a ion sensi i e neu on compa ed wi h a sus ained esponde ha
was no sensi i e o du a ion (all pass). A: Do as e showing he esponse o he long pass
neu on o di e en du a ions. Sounds we e pu e ones p esen ed a 10 dB abo e h eshold. In his
32

33
34
35
36
37

38
39
40
41