Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 3

0006-8993/$ - see frohttp://dx.doi.org/10

Abbreviations: AD

torus semicircularis

magnocellularis; CR

MGB, medial genic

density; OS, oliva

PV, parvalbumin; RnCorresponding a

d’Anatomie CompaE-mail addresse

(H. Tostivint), roge

Research Report

Differences in parvalbumin and calbindin chemospecificityin the centers of the turtle ascending auditory pathwayrevealed by double immunofluorescence labeling

Tatiana V. Chudinovaa, Margarita G. Belekhovaa, Herve Tostivintb, Roger Wardb,c,Jean-Paul Riod, Natalia B. Kenigfesta,b,n

aLaboratory of Evolution of Neuronal Interactions, Sechenov Institute of Evolutionary Physiology and Biochemistry,

Russian Academy of Sciences; 44, Thorez Avenue, 194223 Saint-Petersburg, RussiabCNRS UMR 7221, MNHN USM 0501, Departement Regulations, Developpement et Diversite Moleculaire du Museum National d’Histoire

Naturelle; 7, rue Cuvier, 75005 Paris, FrancecLaboratoire de Neuropsychologie, Universite du Quebec, Trois-Rivi�eres, CanadadInstitut du Fer �a Moulin, INSERM UMR-S 839; 17, rue du Fer �a Moulin, 75005 Paris, France

a r t i c l e i n f o

Article history:

Accepted 12 July 2012

Using double immunofluorescence labeling, quantitative ratio between parvalbumin- and

calbindin-containing neurons, neurons that co-localize both peptides, as well as the

Available online 20 July 2012

Keywords:

Auditory system

Calcium-binding proteins

Reptiles

Double immunofluorescence

labeling

nt matter & 2012 Elsevie.1016/j.brainres.2012.07.0

VRvm, ventromedial a

; CoA, nucleus cochlea

, calretinin; GS, gangl

ulate body; MLd, nucle

superior; OSd, nucleus

e, nucleus reuniens; Tuthor at: CNRS UMR 722ree, 55, rue Buffon, 75005s: tvchudinova@gmail.co

r.ward@uqtr.ca (R. Ward)

a b s t r a c t

intensity of their immunoreactivities were studied in the brainstem, midbrain and

forebrain auditory centers of two chelonian species, Testudo horsfieldi and Emys orbicularis.

In the spiral ganglion and first-order cochlear nuclei, highly immunoreactive parvalbumin-

containing neurons predominated, and almost all neurons in these nuclei also exhibited

weak immunoreactivity to calbindin. The number of strongly calbindin-immunoreactive (-

ir) cells increased in the second-order brainstem auditory centers (the laminar cochlear

nucleus, superior olivary complex, lateral lemniscal nucleus), and co-localization with

parvalbumin in some of them was observed. In the midbrain, a complementary distribu-

tion of parvalbumin and calbindin immunoreactivity was found: the central (core) region

of the torus semicircularis showed strong parvalbumin immunoreactivity, while the

laminar (belt) nucleus was strongly calbindin-ir. In the thalamic nucleus reuniens, almost

complete topographic overlapping of the parvalbumin-ir and calbindin-ir neurons was

shown in its dorsomedial region (core), with the intensity of immunoreactivity to calbindin

being much higher than that to parvalbumin. The predominance of calbindin immunor-

eactivity in neurons of the dorsomedial region of the nucleus reuniens is correlated with

r B.V. All rights reserved.22

rea of the anterior dorsal ventricular ridge; CB, calbindin; Ce, nucleus centralis of the

ris angularis; CoL, nucleus cochlearis laminaris; CoM, nucleus cochlearis

ion spiralis; ir, immunoreactive; L, nucleus laminaris of the torus semicircularis;

us mesencephalicus lateralis, pars dorsalis; nLl, nucleus lemnisci lateralis; OD, optical

dorsalis of the oliva superior; OSv, nucleus ventralis of the oliva superior;

S, torus semicircularis1, MNHN USM 0501, Departement RDDM, Museum National d’Histoire Naturelle, Batiment

Paris, France. Fax: þ33 1 40 79 36 18.m (T.V. Chudinova), belekhova@yahoo.com (M.G. Belekhova), htostivi@mnhn.fr, riojp@yahoo.com (J.-P. Rio), kenig51@yahoo.com (N.B. Kenigfest).

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 388

the existence of the dense calbindin-ir terminal field in its projection area in the

telencephalon. We conclude that the turtle auditory pathway is chemically heterogeneous

with respect to calcium-binding proteins, the predominance of parvalbumin in the

brainstem and midbrain centers giving way to that of calbindin in the forebrain centers;

the portion of neurons co-localizing both peptides nonlinearly decreases from lower to

higher order centers.

& 2012 Elsevier B.V. All rights reserved.

1. Introduction

A variety of electrophysiological and neurochemical studies

(DiFiglia et al., 1989; Celio, 1990; Heizmann and Braun, 1990;

Andressen et al., 1993; Chard et al., 1993; Schwaller et al., 2002;

Camp and Wijesinghe, 2009; Schwaller, 2009) has shown that

several neuronal populations, each with a particular, distinct,

pattern of activity, may also be characterized by the different

calcium-binding proteins that they incorporate. Differences in

the biophysical properties of these proteins, and hence in their

functional roles, may thus render them selective markers of

functionally distinct neuronal populations.

Calcium-binding proteins (parvalbumin, PV; calbindin, CB)

have been widely used in investigations of sensory systems,

particularly for defining different neuronal subpopulations

(Jones and Hendry, 1989; Wong-Riley, 1989; Celio, 1990; Wild

et al., 1993; Puelles et al., 1994; Hevner et al., 1995; de Venecia

et al., 1995, 1998; Partata et al., 1999; Jones, 2003; Ashwell and

Paxinos, 2005; Anderson et al., 2007). Selective PV and CB

labeling of different morphofunctional types of neurons in

mammalian thalamic nuclei led Jones (1998, 2003) to postu-

late a core-matrix principle of the organization of mamma-

lian thalamus. According to the hypothesis, projection

neurons of central (core) divisions of sensory lemniscal

thalamic nuclei, being highly active, contain PV. At the same

time, metabolically less active neurons of their peripheral

non-lemniscal (belt or shell) divisions of relay sensory nuclei

which belong to the diffuse matrix system are CB-ir. However,

alternative distribution of PV and CB in the core and belt

subdivisions correspondingly appears to be more evident in

primates than in other mammalian species. Non-primate

mammals markedly differ in a strong diversity in the dis-

tribution of these proteins in the projection neurons of the

lemniscal, core, subdivisions of sensory thalamic nuclei, thus

containing either PV or CB or both proteins (Celio, 1990; Zettel

et al., 1991; Braun and Piepenstock, 1993; Vater and Braun,

1994; Ashwell and Paxinos, 2005). It has therefore been

suggested that, in the relay sensory nuclei, PV regulation of

neuronal calcium balance is a more recent feature, in con-

trast to the phylogenetically ancient CB regulation (Jones,

1998, 2007; Parvizi and Damasio, 2003). The Jones’ core-matrix

model of the thalamic organization is, however, much less

applicable to thalamic organization in non-mammalian ver-

tebrates including reptiles (Belekhova et al., 2003, 2010).

More specifically, calcium-binding proteins serve as effec-

tive markers of different neuronal populations in the verte-

brate auditory system. While data on the distribution of

calcium-binding proteins in the auditory centers of mammals

(Celio, 1990; Zettel et al., 1991; Braun and Piepenstock, 1993;

Vater and Braun, 1994) and birds (Braun et al., 1985;

Takahashi et al., 1987; Braun, 1990; Rogers et al., 1990;

Braun et al., 1991; Kubke et al., 1999) are considerable,

comparable information in reptilian species is limited

(lizards: Davila et al., 2000; Yan et al., 2010; turtles:

Belekhova et al., 2004, 2008, 2010). Even so, the results from

different species of reptiles are not completely congruent. At

the same time, our knowledge of morphofunctional and

neurochemical properties of the organization of the central

auditory system of reptiles, and particularly of turtles, pro-

vides a great opportunity to clarify both the basic mechan-

isms of the transmission of auditory information and the

phylogenesis of this system in amniotes.

In our recent investigations, we showed that all centers of the

turtle auditory system, including the sensory (spiral) ganglion,

contained both PV-ir and CB-ir neurons (Belekhova et al., 2004,

2008, 2010) and that high metabolic activity is a typical feature of

the lemniscal pathway. We also came to the conclusion that in

turtles, the distinction between the core and the belt of the

various auditory centers progressively diminishes as one

ascends the neuraxis. Though we have found both proteins in

projection neurons of turtles’ auditory centers, the data obtained

were not sufficient to assume which type of calcium-binding

proteins predominates in each center of the turtle auditory

pathway. As in many other studies, we also noted that the

intensity of immunolabeling either for PV or CB in neurons of

these centers strongly varied (Belekhova et al., 2004, 2008, 2010).

Usually the difference in the intensity of labeling is explained by

different concentration of corresponding protein in neurons,

and therefore, the predominance of any protein in each

auditory center may be estimated not only by the number of

immunoreactive neurons but also by the intensity of their

immunoreactivity.

To complement our previous findings, in the present study we

attempt to estimate (i) the degree of the intensity of immunor-

eactivity both to PV and CB in neurons of each center of the

turtle auditory system; (ii) the quantitative ratio between

‘‘weakly’’- and ‘‘strongly’’-labeled either PV- or CB-immunoreac-

tive neurons; (iii) the number of PV- and CB-ir neurons; and

(iiii) the number of neurons that co-localize both proteins. With

this aim, we used double immunofluorescence techniques

followed by counting the number of mono- and double-immu-

nolabeled neurons as well as by measuring the optical density of

neuronal immunofluorostaining.

2. Results

The terminology we use in describing the turtle auditory

nuclei is based on the accepted nomenclature of the

Fig. 1 – Schematic drawings of a caudorostral series of transverse sections through the rhombencephalon (A)–(C),

mesencephalon (D) and forebrain (E) and (F) showing the localization of turtle auditory centers at the level of their

maximum development. Grey areas represent auditory centers. D and L indicate dorsal and lateral axes. Abbreviations:

ADVRdl—dorsolateral area of the anterior dorsal ventricular ridge; ADVRm—medial area of the anterior dorsal ventricular

ridge; ADVRvm—ventromedial area of the anterior dorsal ventricular ridge; cb—cerebellum; Ce—nucleus centralis of the

torus semicircularis; CoA—nucleus cochlearis angularis; CoM—nucleus cochlearis magnocellularis; CoL—nucleus cochlearis

laminaris; cxd—cortex dorsalis; cxl—cortex lateralis; Dma—nucleus dorsomedialis anterior; GLd—nucleus geniculatus

lateralis, pars dorsalis; GLv—nucleus geniculatus lateralis, pars ventralis; GS—ganglion spiralis; Hab—habenula; Ic—nucleus

intercollicularis; Ist—nucleus isthmi magnocellularis; L—nucleus laminaris of the torus semicircularis; nLl—nucleus

lemnisci lateralis; ntrs—nucleus of the solitary tract; OSd—nucleus dorsalis of the oliva superior; OSv—nucleus ventralis

of the oliva superior; Path—pallial thickening; Pe—periventricular area; pedd—pedunculus dorsalis of the lateral forebrain

bundle; Ra—nucleus raphe; Re—nucleus reuniens; Rot—nucleus rotundus; SGP—stratum griseum periventriculare of the

optic tectum; Str—striatum; tro—tractus opticus; TO—optic tectum; V—nucleus ventralis; Vds—nucleus descendens nervi

trigemini; Ve—ventriculus; Ves—vestibular nuclei; VIII—nervus statoacusticus.

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 3 89

brainstem auditory nuclei (Huber and Crosby, 1926; Miller and

Kasahara, 1979), mesencephalic auditory center (Huber and

Crosby, 1926; Browner et al., 1981), thalamic auditory center

(Papez, 1935) and telencephalic auditory center (Johnston,

1915). In all investigated centers, only cells immunoreactive

to PV and/or CB have been counted and discussed, whereas

unlabeled cells have not been considered.

2.1. Auditory ganglion–spiral ganglion

The auditory spiral ganglion (GS) of both turtle species

(Fig. 1A) was composed of three populations of immunoreac-

tive neurons. The first, predominant, population consisted of

medium-sized neurons occupying most of the ganglion and

constituted 76% of all immunoreactive cells. All these neu-

rons were both PV-ir (Fig. 2A, arrows I) and CB-ir (Fig. 2B,

arrows I), thus possessing 100% co-localization (Fig. 2C, med-

ium-sized yellow–green neurons, arrows I), although the

fluorescence intensity (or optical density, OD, measured in

arbitrary units; see below) of the PV-ir neurons was high

(mean OD¼9579, n¼403) and that of the CB-ir neurons was

low (mean OD¼4875, n¼403). The second population of

neurons was situated along the dorsal edge of the GS and

consisted of a few (6%) large-sized cells immunoreactive, as

the members of the first population, to both proteins (Fig. 2A

and B, arrows II). But in contrast to the first population, the

fluorescence intensities of PV-ir (mean OD¼106711, n¼102)

and CB-ir (mean OD¼194718, n¼102) neurons of the second

one were both high. The optical density of the fluorescence of

CB-ir neurons in the second neuronal population was 4 times

higher than that in the first one. As in neurons of the first

population, PV and CB were also co-localized in all neurons of

the second population (Fig. 2C, large yellow–orange cells,

arrows II). The third population consisted of a few (18%)

small-sized neurons, highly (mean OD¼105711, n¼225)

immunoreactive only to PV (Fig. 2A and C, small green

neurons, arrow III).

2.2. First-order and second-order brainstem auditorynuclei—the cochlear complex (Fig. 1A and B)

The axons of GS cells form the auditory root of the VIIIth

nerve and enter the first-order cochlear nuclei, classically

described as the nucleus cochlearis magnocellularis (CoM)

and nucleus cochlearis angularis (CoA), which along with the

second-order nucleus cochlearis laminaris (CoL) form the

cochlear complex (Fig. 1A and B). In the turtle, the cochlear

complex is an elongated monolithic structure within which

the distinction between the CoM and CoA is far from

apparent. The CoM was described as the nucleus cochlearis

posterior (Beccari, 1911) which begins much more caudally

than the entrance of the VIIIth nerve, whereas the CoA or the

nucleus cochlearis anterior of Beccari (1911) forms the rostral

part of this cellular mass. In our material, the CoM and CoA

could be tentatively distinguished by their more caudal (CoM;

Fig. 1A) or more rostral (CoA; Fig. 1B) position in the brain-

stem. Since in our material we could not delineate the border

between these nuclei, we analyzed them together with the

exception of the rostral CoA or caudal CoM. It is worth to note

that the CoL appears as a small separate nucleus ventral to

the CoM (Fig. 1A).

The distribution of PV and CB in the auditory nerve and

neurons of the cochlear complex was similar in both turtle

species. Fibers of the auditory nerve entering CoA and CoM

from the dorsomedial side were PV-ir, others entering CoA

and CoM dorsolaterally were immunoreactive to both PV and

CB (Fig. 4A and D, arrowheads). Together with terminals and

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 390

cellular processes, they formed a dense neuropil surrounding

densely packed cell bodies.

All immunoreactive neurons in the CoA were PV-ir (Figs. 2D,

4A) and 99% of them were also CB-ir (Figs. 2E, 4D; Table 1).

The same was true for the CoM (Fig. 4J). The majority (87%) of

PV-ir neurons were strongly immunoreactive (mean

OD¼131714, n¼245). The intensity of fluorescence in the

rest of PV-ir neurons was up to 1.5 times lower. In general, the

mean fluorescence intensity of all PV-ir neurons (mean

OD¼125712, n¼282) was the highest in the first-order

cochlear nuclei as compared to other auditory centers

(Fig. 3). The fluorescence intensity of almost all CB-ir neurons

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 3 91

was low (mean OD¼4475, n¼245), varying negligibly

(Figs. 2E, 3). PV and CB were co-localized almost in all neurons

of CoA and CoM (Figs. 2F, 4J; Table 1). Among densely-packed

immunoreactive projection neurons, PV-ir and CB-ir elements

of the neuropil including punctate terminals were observed.

In contrast to CoA and CoM, the second-order cochlear

nucleus CoL, which receives auditory afferents from the first-

order cochlear nuclei, contained only a few small-sized CB-ir

neurons (Fig. 4J, arrows).

Most of the arcuate fibers, leaving the cochlear nuclei, were

strongly PV-ir (Fig. 4A, arrows), though some of them also

displayed a weak CB immunoreactivity (Fig. 4D, arrows).

2.3. Second-order brainstem auditory center—thesuperior olivary complex (Fig. 1A and B)

The distribution of PV and CB immunoreactivity in the

superior olivary complex (OS), composed of the ventral

(OSv) and dorsal (OSd) nuclei (Fig. 1A and B), differed slightly

between the two chelonian species. The OSv of Emys con-

tained an about equal portion of PV-ir (53%; Fig. 4B and C;

Table 1) and CB-ir (59%; not demonstrated; Table 1) medium-

sized cells scattered throughout the nucleus. The fluores-

cence intensity was high in both the majority of PV-ir

neurons (73%; Table 1) and CB-ir neurons (67%; Table 1).

At the same time, only 12% of immunoreactive neurons

contained both peptides. The neuropil of the OSv contained

a dense accumulation of strongly PV-ir terminals and fibers

(Fig. 4C) and only few terminals and fibers were CB-ir. Similar

distribution of PV and CB immunoreactivity was observed in

the OSd of Emys. Fibers of the bundle binding OSv and OSd

were predominantly PV-ir. Some PV-ir and CB-ir neurons were

scattered along these fibers.

The OSv of Testudo consisted mainly of CB-ir neurons (77%;

Fig. 4E and F). Some cells (28%) were PV-ir and only 6% of

immunoreactive neurons co-localized PV and CB. Approxi-

mately 60% of CB-ir cells were strongly fluorescent (Fig. 4F)

while the majority (90%) of PV-ir neurons had low

Fig. 2 – Immunoreactivity to PV (green, A,D,G,J), CB (red, B

photomicrographs of transverse sections of the spiral ganglion

(D)–(F), the central (Ce) and laminar (L) nuclei of the torus semic

taken at the level of the maximum development of each center

neurons: arrows I indicate medium-sized strongly PV-ir (A) and

(yellow–green neurons in C); arrows II indicate large-sized neu

colored in yellow–orange co-localizing both peptides (C); arrow

only to PV (A) and (C). (D)–(F): Strongly PV-ir neurons (D) and we

all co-localizing both peptides (yellow–green neurons in F). (G)–

strongly CB-ir neurons in the L (H) of the torus semicircularis th

(G) and weakly CB-ir (H) neuropil in the Ce, fine fibers and termin

(J)–(L): Different intensity of fluorescence in PV-ir (J) and CB-ir (K

(Re). According to the intensity levels double-immunoreactive ne

PV-ir (intensely green in J) and weakly CB-ir (lightly red in K) ne

PV-ir (lightly green in J) and strongly CB-ir (intensely red in K) n

immunoreactive to both PV (intensely green in J) and CB (inten

that many neurons are not double-labeled (green PV-ir and red

Rot–nucleus rotundus. Scale bars, 100 lm in (A)–(C); 30 lm in (D

fluorescence intensity. Along with the small number of PV-

ir terminals and fibers, the neuropil of the OSv contained

numerous strongly CB-ir dendrites (Fig. 4F). The OSd of

Testudo consisted of a few strongly CB-ir neurons with long

interweaving dendrites (Fig. 4E), few PV-ir neurons and

terminals. The majority of fibers connecting these two nuclei

were CB-ir, and some CB-ir neurons were situated among

them (Fig. 4E).

2.4. Second-order brainstem auditory center—the nucleusof the lateral lemniscus (Fig. 1C)

The nucleus of the lateral lemniscus (nLl; Fig. 1C) includes

two parts—a dorsal and a ventral one. In the dorsal nLl of

both turtle species, only CB-ir neurons among moderately

CB-ir neuropil containing terminals and fibers were observed

(Fig. 4H). The ventral nLl of Emys included both strongly CB-ir

and PV-ir neuropil. The majority (91%) of cells were CB-ir, 24%

of which also contained PV (Fig. 4(G)–(I); Table 1). The

fluorescence intensity of CB-ir neurons varied from weak

(37% of CB-ir cells) to strong (63%), whereas the intensity of

the majority (90%) of PV-ir neurons was high (Table 1).

In contrast to Emys, PV immunoreactivity in the ventral part

of the nLl in Testudo was very low.

2.5. Mesencephalic auditory center—the torussemicircularis (Fig. 1D)

The chelonian mesencephalic auditory center- torus semicircu-

laris (TS), which includes the central (Ce) and laminar (L) nuclei

(Fig. 1D)—receives signals from first- and second-order

brainstem auditory nuclei and then relays information to the

thalamic auditory center–nucleus reuniens. The Ce might be

considered as a ‘‘core’’ or lemniscal subdivision of the torus with

the L as its ‘‘belt’’ or an extralemniscal subdivision. In our

previous studies (see Belekhova et al., 2010), using standard

immunohistochemical methods, we have already shown that

,E,H,K) and to both peptides (C,F,I,L) shown on confocal

(GS; (A)–(C) in Testudo, the angular cochlear nucleus (CoA;

ircularis (G)–(I) and the nucleus reuniens (Re; (J)–(L)) in Emys

(see Fig. 1). (A)–(C): Three different populations of ganglion

weakly CB-ir (B) neurons co-localizing both peptides

rons strongly immunoreactive to both PV (A) and CB (B),

III indicates small-sized neurons strongly immunoreactive

akly CB-ir neurons (E) in the angular cochlear nucleus (CoA),

(I): Large-sized strongly PV-ir neurons in the Ce (G) and

at do not co-localize these peptides (I). Note a strongly PV-ir

al-like structures, which do not co-localize two peptides (H).

) neurons in the dorsocentral area of the nucleus reuniens

urons are differently colored (L). Arrow 1 indicates a strongly

uron, which is yellow–green in L; arrow 2 indicates weakly

euron having orange color in L; arrow 3 indicates strongly

sely red in K) neuron having yellow–orange color in L. Note

CB-ir neurons in L). D and L indicate dorsal and lateral axes;

)–(F); 50 lm in (G)–(L).

Table 1 – Percentage of PV-ir and CB-ir neurons and neurons co-localizing both proteins and percentage of neurons ofdifferent fluorescent intensities in the auditory ganglion, brainstem, mesencephalic and thalamic centers of Emys’ centralauditory system.

Mono-PV-

ir neurons

(%)

Mono-CB-

ir neurons

(%)

PV&CB co-

localizing

neurons (%)

Total amount

of PV-ir

neurons (%)

Total amount

of CB-ir

neurons (%)

Stronga

PV-ir

neurons

(%)

Stronga

CB-ir

neurons

(%)

Ganglion spiralis 20 0 80 100 80 100 8

First-order

cochlear nuclei

(CoMþCoA)

1 0 99 100 99 87 0

Nucleus ventralis

of oliva superior

41 47 12 53 59 73 67

Nucleus lemnisci

lateralis, pars

ventralis

9 67 24 33 91 90 63

Nucleus lemnisci

lateralis, pars

dorsalis

0 100 0 0 100 0 73

Torus semicircularis:

Nucleus

centralis, central

area

100 0 0 100 0 100 0

Nucleus

centralis,

peripheral area

39 46 15 54 61 36 45

Nucleus

laminaris

0 100 0 0 100 0 98

Nucleus reuniens

dorsocentralis

33 37 30 63 67 42 59

a The estimation of the level of intensity (strong/weak) of immunoreactive neurons was carried out according to their optical densities

(see experimental procedure), maximal value of which differed in each center but varied from high to low within the range.

Fig. 3 – Mean optical density (arbitrary units) of the PV (grey columns) and CB (black columns) immunofluorescence labeling

in neurons of the dominating population of the spiral ganglion, first-order cochlear nuclei, core (nucleus centralis) and belt

(nucleus laminaris) regions of the torus semicircularis, and of the dorsocentral part (core) of the nucleus reuniens in Emys.

Error bars indicate standard deviations. Note that neurons of the first-order cochlear nuclei possess the maximal intensity of

PV immunofluorescence, and that significant difference (po0.01) of the optical density (stars) exists between PV-containing

neurons of ganglion, cochlear nuclei and central area of the nucleus centralis, from one hand, and peripheral area of the

nucleus centralis and nucleus reuniens, from another hand. On the contrary, neurons of the nucleus reuniens and nucleus

laminaris possess the maximal intensity of CB immunofluorescence with a significant difference (po0.01) of the optical

density (two stars) between these nuclei and more caudal auditory centers.

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 392

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 3 93

both in Emys and Testudo PV and CB immunoreactivities are

complementarily distributed in the nuclei of the TS, and that the

central and peripheral areas of the Ce could be delineated on the

basis of the heterogeneous distribution of PV-ir and CB-ir

elements.

In the present study, double immunofluorescence findings

confirmed our previous observations. Both in Emys and Testudo,

the central area of the Ce was delineated by very intensive PV-ir

neuropil composed of the dense accumulation of immunoreac-

tive terminals and fibers. This area contained mono-PV-ir large-

sized neurons (Fig. 2G and I, 4 K; Table 1). The intensity of their

fluorescence was high (mean OD¼8879, n¼45), comparable to

that of PV-ir neurons in the spiral ganglion (Fig. 3). In the central

area a loosely organized weakly CB-ir neuropil was also present

(Fig. 2H and I). On the contrary, the L contained numerous mono-

CB-ir small- and medium-sized neurons (Fig. 2H and I; Table 1)

fluorescence intensity of which was very high (mean

OD¼127713, n¼177; Fig. 2). Neuropil in the L was weakly PV-ir

(Fig. 2G and I) and moderately intensive CB-ir (Fig. 2H and I).

We did not observe co-localization of PV and CB either in

the central area of the Ce or in the L (Fig. 2I; Table 1). The

peripheral area of the Ce (Fig. 4K), encircling the central core,

contained approximately an equal number of PV-ir (54%) and

CB-ir (61%) neurons (Table 1). Small-sized CB-ir neurons had

different fluorescence intensity, 45% of which were strongly

immunoreactive (mean OD¼6176, n¼128; Fig. 3). Most of

small- and medium-sized PV-ir neurons had moderate fluor-

escence intensity (mean OD¼5776, n¼117; Fig. 3) that was

about 1.6 times lower than that of the PV-ir cells in the central

area of the Ce. Only 36% (Table 1) of PV-ir neurons in the

peripheral area were strongly immunoreactive and their

intensity was comparable to that of PV-ir neurons in the

central area. CB and PV were co-localized in 15% of neurons

in the Ce peripheral area. Its neuropil contained both mod-

erately PV-ir and CB-ir elements.

2.6. Thalamic auditory center—the nucleus reuniens(Fig. 1E)

The thalamic relay auditory nucleus (nucleus reuniens, Re)

consists of two subdivisions—the dorsocentral (core) lemniscal

subdivision and the ventral (belt) extralemniscal one. It receives

projections from both nuclei of the auditory torus via the

tectoreunial tract. In our previous study (Belekhova et al., 2010),

we observed that the core subdivision as well as the tectoreunial

tract displayed both PV and CB immunoreactivity.

In the present study we revealed that the Re dorsocentral

subdivision in Emys contained about equal numbers of PV-ir

(63%) and CB-ir (67%) neurons (Fig. 2J and K; Table 1). The

fluorescence intensity of the CB-ir neurons in the Re (mean

OD¼129711, n¼894) was much higher as compared with that in

lower auditory centers: it was 2.3 times higher than in the spiral

ganglion neurons, 2.5 times higher than in CoA and CoM

neurons, and comparable to that of L neurons of the auditory

torus (Fig. 3). More strongly (59%; OD¼162717, n¼527) and less

strongly (41%; OD¼9679, n¼367) labeled neurons were distin-

guished among Re CB-ir neurons (Table 1). Their ODs varied up to

1.7 times. The fluorescence intensity of PV-ir neurons (mean

OD¼5676, n¼763) was lower than that in lower auditory

centers: it was 1.6 lower than in the neurons of the spiral

ganglion, 2.3 times lower than in neurons of CoA and CoM,

and 1.5 times lower than in Ce neurons of the TS (Fig. 3). Among

PV-ir neurons, there were also strong (42%; mean OD¼6477,

n¼320) and weak (58%; mean OD¼4875, n¼443) neurons,

discriminated by the OD in 1.3 times. The two peptides were

co-localized in 30% of neurons in the Re (Fig. 2L; Table 1). The

neuropil of the Re dorsocentral part included CB-ir and PV-ir

terminals and fibers. The ventral subdivision of the nucleus was

characterized by very weak immunoreactivity to both PV and CB.

Besides dorsocentral and ventral parts, the periventricular band

(the area X of Belekhova et al., 2010), bound by dendrites of the

CB-ir neurons to the dorsocentral part, was also distinguished in

the Re. It contained weakly stained CB-ir neuropil with small-

sized CB-ir neurons.

Similar distribution of PV and CB immunoreactivity was

observed in the Re of Testudo, although with some differences.

First, there were approximately half as many PV-ir neurons in

the dorsocentral part of Testudo’s Re; and second, all PV-ir

neurons were weakly fluorescent (mean OD¼2974, n¼195).

As in Emys, CB-ir neurons were numerous and strongly

labeled, and in spite of smaller number of PV-ir neurons,

the same 30% of neurons co-localized both peptides.

2.7. Telencephalic auditory center (Fig. 1F)

The dorsal thalamic nucleus reuniens projects further to the

telencephalic auditory center—the ventromedial part of the

anterior dorsal ventricular ridge (ADVRvm; Fig. 1F). Using

immunofluorescence, we confirm our previous results

(Belekhova et al., 2010). In both chelonian species, in the

ADVRvm immunoreactivity to PV and CB strongly varied.

Usually we observed a high intensity of immunofluorescence

to CB in the neuropil that occupied the most part of the area

with a dense immunoreactive terminal field concentrated in

its central part. Strongly CB-ir neurons were also found there

(Fig. 4L). On the contrary, immunoreactivity to PV in neuropil

was weak and mostly occupied the ventral part of the area

though thin fibers and terminals could be observed in the

central part of the ADVRvm. Many but weakly PV-ir neurons

were scattered throughout the area.

3. Discussion

Our present findings not only confirm the general pattern of

the distribution of PV and CB in centers of the auditory

system and its transformation along the neuroaxis in two

chelonian species Emys orbicularis and Testudo horsfieldi

(Belekhova et al., 2003, 2004, 2008, 2010), but, on the basis of

the ratio of PV/CB-ir neurons and neurons containing both

peptides and the difference in fluorescence intensity mea-

sured in PV-ir and CB-ir neurons, they provide the possibility

to determine the predominant calcium-binding protein func-

tioning in each auditory center and its lemniscal and extra-

lemniscal subdivisions.

Only projection immunoreactive neurons of the ascending

auditory system will be discussed here, though several studies in

mammals (Fisher and Davies, 1976; Vater et al., 1992; Winer and

Larue, 1996; Cant and Benson, 2003; Jones, 2007; Bender and

Trussell, 2011) and birds (Muller, 1988; von Bartheld et al., 1989;

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 394

Carr et al., 1989; Veenman and Reiner, 1994; Carr and Code, 2000;

Burger et al., 2005; Pinaud and Mello, 2007) have reported that

auditory centers contain as well local circuit neurons which can

also be PV- and/or CB-immunoreactive. Few GABAergic and/or

glycinergic neurons have been observed in the cochlear nuclei;

much more of such neurons have been found in the second-

order nuclei, as well as in the mesencephalic auditory center.

However, it is worth to note that these neurons are not only

inhibitory interneurons, but also projection neurons belonging to

the descending inhibitory cochleopetal system. In the mamma-

lian (Fisher and Davies, 1976; Winer and Larue, 1996; Jones, 2007)

and avian (Domenici et al., 1988; Muller, 1988; Veenman and

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 3 95

Reiner, 1994; Sun et al., 2005; Pinaud and Mello, 2007) thalamic

auditory center, a great diversity in number of interneurons

(including their absence) with various morphological types have

been observed. In the literature, data are extremely scarce with

regard to interneurons in reptilian auditory centers. In lizards,

few GAD-positive cells have been found in the cochlear nuclei,

many in the superior olivary complex and nuclei of the lateral

lemniscus, and less in the central nucleus of the auditory torus

(Yan et al., 2010). In the turtle brainstem and mesencephalic

auditory centers, no interneurons have been reported. However,

in the nucleus reuniens/nucleus medialis in turtles (Belekhova

et al., 1991)/lizards (Yan et al., 2010), GABA/GAD-positive cells

have not been observed. Hence, care should be taken when

comparing the distribution of PV and CB among auditory centers

of amniotes.

3.1. Calcium-binding proteins in the turtle auditorycenters

3.1.1. Spiral ganglion and first-order cochlear nucleiTo begin our discussion, we point out that both PV-ir and CB-

ir hair cells have been revealed in the turtle inner ear and that

almost all such cells contain both peptides (Hackney et al.,

2003). As compared to turtles, a higher concentration of PV in

cochlear hair cells has been demonstrated in birds (Heller

et al., 2002) and mammals (Hackney et al., 2005).

In the spiral ganglion and first-order cochlear nuclei (CoA

and CoM), most immunoreactive neurons are strongly immu-

noreactive to PV and weakly immunoreactive to CB. Only a

small neuronal population is strongly immunoreactive to

both PV and CB. Although the auditory ganglion and first-

order cochlear nuclei contain an approximately equal num-

ber of PV-ir and CB-ir neurons, a great difference in the

intensity of immunolabeling (which reflects the concentra-

tion of peptides in neurons) shows that PV immunoreactivity

prevails in these auditory centers. Moreover, the concentra-

tion of PV in neurons of the first-order cochlear nuclei is the

highest and that of CB is the lowest in comparison to more

rostral auditory centers.

Recently, in the lizard Gekko gecko, the distribution of

calcium-binding proteins in the auditory centers has been

Fig. 4 – Distribution of PV (green) and CB (red) immunoreactivity

(A) and weakly CB-ir (D) arcuate fibers (arrows) arising from the

magnification in the insets (star). Note PV-ir (A) and CB-ir (D) fi

cochlear angular nucleus (CoA). (B), (E): Strong PV immunoreact

superior olivary complex of Emys (B) and strong CB immunorea

between two nuclei of the superior olivary complex. (C) and (F)

Emys (C) and strongly CB-ir neurons, fibers but less terminals i

lemniscus (nLl) of Emys, strongly PV-ir neurons are present only

both in the dorsal and ventral parts of the nucleus (H). Mono-PV

both peptides (arrows) in the nLl ventral part are shown on the

cochlear complex of Emys. Total co-localization of PV and CB in

(CoM; yellow–green neurons) and mono-CB-ir small-sized neur

neurons; arrows). (K): Double immunolabeling in the central and

semicircularis. The central area contains only large-sized PV-ir

either PV-ir or CB-ir neurons. (L): CB-ir fibers, terminals and ne

D and L indicate dorsal and lateral axes. Scale bars, 100 lm in

and (D).

studied (Yan et al., 2010). As in turtles, both PV-ir and CB-ir

neurons (in addition to neurons immunoreactive to another

calcium-binding protein, calretinin, CR) have been described

in its auditory ganglion but without indication of a predomi-

nant specificity of ganglion neurons. Judging by the immu-

noreactivity of cochlear nuclei innervations in birds (Jande

et al., 1981; Takahashi et al., 1987; Braun, 1990; Kubke et al.,

1999) and mammals (Zettel et al., 1991; Braun and

Piepenstock, 1993; Frisina et al., 1995; Caicedo et al., 1996;

Spatz, 2003; Por et al., 2005; Bazwinsky et al., 2008), neurons

of the spiral ganglion likely contain all three calcium-binding

proteins, and in addition to that, in the mammalian spiral

ganglion, PV-ir neurons clearly prevail over CB-ir and CR-ir

neurons though this conclusion does not take into considera-

tion that co-localization of these peptides may occur in

ganglion neurons.

The first-order cochlear nuclei of Gekko are notable for the

predominance of CR-containing neurons. In addition to CR-ir

neurons, similar to the turtle CoM, neurons weakly immu-

noreactive to CB have been observed in the magnocellular

nucleus of lizards, and as opposed to turtles, this nucleus of

lizards does not contain PV-ir neurons. Moreover, in the

angular nucleus of lizards, neither CB-ir nor PV-ir neurons

have been found (Yan et al., 2010).

Chelonian first-order cochlear nuclei also differ from these

nuclei of birds. The data concerning calcium-binding proteins

in the brainstem of birds are scarce and restricted to the

study of the distribution of individual protein (CR, CB or PV).

Similar to CoM and CoA of turtles, the magnocellular nucleus

of the zebra finch (Braun, 1990) contains PV-ir neurons

whereas this nucleus of the chicken and owls contains a

dense population of CR-ir neurons (Rogers, 1987) and possibly

a small population of CB-ir neurons (Jande et al., 1981;

Takahashi et al., 1987; Braun, 1990; Kubke et al., 1999). As

far as we know, no conclusion on the predominance of

particular protein in different auditory centers of birds has

been made. In the mammalian cochlear complex, PV immu-

noreactivity has been shown almost in all neurons whereas

CR and CB immunoreactivity is represented in several specific

neuronal morphotypes predominantly co-localizing with PV

(Zettel et al., 1991; Braun and Piepenstock, 1993; Frisina et al.,

in turtles’ auditory centers. (A), (D): In Emys, strongly PV-ir

cochlear complex, which are also shown at a higher

bers of the auditory nerve (arrowheads) entering first-order

ivity in the ventral (OSv) and dorsal (OSd) nuclei of the

ctivity in these nuclei of Testudo (E). Note a fiber bundle

: Strongly PV-ir neurons, fibers and terminals in the OSv of

n the OSv of Testudo (F). (G)–(I): In the nucleus of the lateral

in its ventral part (G); strongly CB-ir neurons are numerous

-ir (stars), mono-CB-ir (arrowheads) and neurons containing

double-labeled image (I). (J): Double immunolabeling in the

neurons of the first-order cochlear magnocellular nucleus

ons in the second-order cochlear laminar nucleus (CoL; red

peripheral areas of the central nucleus (Ce) of the Emys torus

neurons while the peripheral area contains small-sized

urons in the telencephalic auditory area (ADVRvm) of Emys.

A, B, D, E, K, L; 50 lm in (C), (F), (G)–(J), in the inset in (A)

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 396

1995; Caicedo et al., 1996; Spatz, 2003; Por et al., 2005;

Bazwinsky et al., 2008). The above discussed diversity in the

distribution of various types of calcium-binding proteins in

different neuronal populations both in mammalian and avian

species can be explained by the functional specialization of

the parallel separate circuits in the brainstem auditory

centers processing different aspects of auditory information

(such as intensity, time and frequency of acoustic signals;

Caicedo et al., 1996; Grothe et al., 2004; Bazwinsky et al.,

2008).

3.1.2. Brainstem second-order auditory nucleiIn contrast to the first-order cochlear nuclei, the second-order

auditory nuclei (laminar cochlear nucleus, superior olivary

complex and nucleus of the lateral lemniscus) of turtles, the

portion of CB-ir neurons is generally greater, and only few of

these neurons also contain PV. However, a great diversity of

the CB and PV immunolabeling pattern in these nuclei has

been observed. Such heterogeneity in the distribution of

these peptides can be explained by the presence in them of

several functionally different neuronal populations that give

rise to numerous ascending and descending connections of

these nuclei with both auditory and non-auditory centers.

The difference in the distribution of CB and PV immunor-

eactivity between Emys and Testudo reflects the difference in

their functioning caused by the specific way of life of these

species.

The distribution of calcium-binding proteins in the super-

ior olivary complex and nuclei of the lateral lemniscus of the

lizard (Gekko gecko) is similar to that of turtles: both CB- and

PV-ir neurons have been observed in these nuclei of Gekko

(Yan et al., 2010). The few data concerning the distribution of

calcium-binding proteins in the second-order brainstem

auditory nuclei of birds, only indicate that neurons in the

cochlear laminar nucleus, which is considered as a homo-

logue of the mammalian medial superior olivary nucleus

(Boord, 1969), express either CB (Jande et al., 1981;

Takahashi et al., 1987) or CR (Rogers, 1987; Braun, 1990)

immunoreactivity. A weak CB immunoreactivity has been

found in neurons of the avian superior olivary nuclei and

ventral lateral lemniscal nucleus (Takahashi et al., 1987;

Braun, 1990). In contrast to turtles, lizards and birds, PV

immunoreactivity prevails in the second-order brainstem

auditory nuclei of mammals: strongly PV-ir neurons have

been found in all these centers, whereas CB-ir neurons have

been observed only in the trapezoid body and ventral lateral

lemniscal nucleus. Nevertheless, similar to turtles, a number

of CB-ir neurons increases in the mammalian second-order

auditory nuclei as compared to the first-order auditory nuclei

(Zettel et al., 1991; Caicedo et al., 1996).

3.1.3. Mesencephalic auditory centerThe mesencephalic auditory center is the primary integrative

link in the processing of the auditory information, the basic

organization of which is highly conservative. In reptiles, it is

represented as the torus semicircularis, which is considered

to be the homologue of the nucleus mesencephalicus later-

alis, pars dorsalis (MLd) in birds and inferior colliculus in

mammals (Grothe et al., 2004).

We confirm our previous data showing the complementary

or mutually exclusive localizations of PV versus CB in the

auditory torus of turtles. The auditory torus is the most

illustrative example demonstrating that CB and PV mark

two distinct functional channels. PV predominates in the

central (core) nucleus of the lemniscal auditory channel,

whereas laminar (belt) nucleus of the extralemniscal channel

contains only CB. The intensity of the fluorescence of CB

labeling in neurons of the laminar nucleus as well as that in

neurons of the thalamic auditory center is the highest among

all auditory centers of turtles.

The functional compartmentalization has been also dis-

covered within the proper Ce. Two areas can be described in

the nucleus: the central area that contains large-sized highly

PV-ir neurons and terminals, and the peripheral area that

contains small-sized PV-ir and CB-ir neurons of different

fluorescence intensity, sometimes co-localizing both pro-

teins. An existence of a dense highly PV-ir terminal field in

the central area suggests that the main input from the

brainstem auditory nuclei to the mesencephalic auditory

center is PV-immunoreactive. In lizards, as in turtles, several

subdivisions emerge within the central nucleus of the audi-

tory torus, marked by the different calcium-binding proteins.

But in contrast to turtles, in neurons of this nucleus of

lizards, the predominant type of calcium-binding proteins is

not PV but CB (Yan et al., 2010).

The complementary distribution of CB versus PV is also

typical for the mammalian inferior colliculus. The compart-

mentalization within inferior colliculus central nucleus

increases due to multiple inputs from the auditory and

non-auditory structures manifesting a heterogeneous distri-

bution of calcium-binding proteins in it (Zettel et al., 1991;

Braun and Piepenstock, 1993; Vater and Braun, 1994; Tardif

et al., 2003; Sharma et al., 2009; Zeng et al., 2009). Although in

birds the distinction between the core and the regions in the

MLd has not been finally resolved (Zeng et al., 2007, 2008;

Logerot et al., 2011), most investigators have demonstrated

that PV-ir neurons prevail in its central nucleus whereas

structures surrounding it contain CB-ir cells (Braun, 1990;

Wagner et al., 2003; Logerot et al., 2011).

3.1.4. Forebrain auditory centersThe nucleus reuniens is the relay thalamic center in the

auditory system of turtles, which projects to the ventrome-

dial area of the anterior dorsal ventricular ridge. The homol-

ogy of the thalamic (nucleus reuniens in turtles and

crocodiles, nucleus medialis in lizards, nucleus ovoidalis in

birds) and telencephalic (ADVRvm in reptiles and field L in

birds) auditory centers is widely accepted in all sauropsids

(Papez, 1935; Karten, 1967, 1968; Reiner et al., 2005; Butler

et al., 2011). At the same time, the homology of the forebrain

auditory centers of sauropsids and mammals is still debated.

According to the classical point of view (Karten, 1967, 1968;

Pritz, 1974; Wild et al., 1993; Reiner et al., 2005) nucleus

reuniens/nucleus ovoidalis and ADVRvm/field L are homo-

logous with their mammalian counterparts—the medial gen-

iculate body (MGB) and auditory cortex (AI), correspondingly.

The alternative hypothesis (Davila et al., 2000; Bruce et al.,

2002) proposes that the lizard medial nucleus and avian

nucleus ovoidalis are homologous with some part of the

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 3 97

mammalian complex of the posterior thalamic and intrala-

minar nuclei, while their telencephalic projection areas are

homologous to the claustroamygdaloid complex. Arguments

either in favor or against this hypothesis have been exten-

sively discussed in our previous paper (Belekhova et al., 2010)

and some related papers (see Reiner et al., 2005; Butler et al.,

2011).

According to several hodological, neurochemical, neuro-

physiological and metabolic characteristics (Belekhova et al.,

1985; Khachunts and Belekhova, 1986; Belekhova et al., 2007,

2010), two main regions have been distinguished in the Re:

the lemniscal (core) dorsocentral and extralemniscal (belt)

ventrolateral regions. In the present study, we have been

mainly interested in discussing the core region of the Re. In

this region of Testudo, CB immunoreactivity prevails both in

neurons and neuropil, and the intensity of fluorescence

labeling in neurons is very high, rare PV-ir neurons having a

very weak labeling. In the core of the Re in Emys, an

approximately equal number of PV-ir and CB-ir neurons has

been found, and only a numerical value of the immunofluor-

escence intensity of labeling has allowed us to estimate

whether PV or CB predominates. The intensity of CB fluores-

cence labeling is several times higher than that of PV labeling

and is as high as CB labeling in the laminar nucleus (belt) of

the auditory torus. The fluorescence intensity of PV-ir neu-

rons is the lowest as compared to all lower auditory centers.

These results give evidence that CB predominates in the core

of the Re. The presence of a dense CB-ir terminal field in the

reunial projection area in the telencephalon (ADVRvm) and

weak PV-ir projections there (present study, Belekhova et al.,

2010) supports these evidences.

As we discussed above, since in neurons of auditory gang-

lion and first-order cochlear nuclei that co-localize PV and CB

the intensity of immunofluorescence labeling of PV is extre-

mely high while that of CB is very low, we have suggested

that PV prevails in these auditory centers of turtles. More

rostrally, in the core of mesencephalic auditory torus, only

highly immunolabeled PV-containing neurons are present.

And, owing to the fact that CB-containing neurons predomi-

nate in the core of the thalamic auditory center, we can

conclude that the change of chemospecificity substituting PV

with CB occurs at the thalamotelencephalic level of the turtle

lemniscal auditory channel.

A similar pattern of calcium-binding proteins immunor-

eactivity also exists in lizards. In the thalamic auditory center

of lizards (nucleus medialis), CB-containing cells prevails,

whereas PV-ir cells are either few as in Gekko (Yan et al.,

2010) or completely absent as in Psammodromus (Davila et al.,

2000). However, in contrast to turtles, all lower auditory

centers of lizards contain predominantly CB-ir projection

neurons as do the thalamic ones (Yan et al., 2010). The scarce

data on birds show that the core of the nucleus ovoidalis is

characterized by PV-containing neurons, while the peripheral

belt nuclei contain CB-ir cells (Braun et al., 1985, 1991;

Heizmann and Braun, 1990).

As for the mammalian thalamic auditory center (MGB),

only primates or humans possess strictly alternative distri-

bution of PV and CB in core and belt regions correspondingly

(Jones and Hendry, 1989; Hashikawa et al., 1991; Molinari

et al., 1995; Jones, 1998, 2003; Munkle et al., 2000), whereas in

non-primates, a great interspecies variation in quantity of PV-

and CB-containing neurons in its ventral lemniscal (core)

nucleus have been observed (Celio, 1990; Covenas et al., 1991;

Braun and Piepenstock, 1993; Vater and Braun, 1994; de

Venecia et al., 1995, 1998; Cruikshank et al., 2001; Ashwell

and Paxinos, 2005; Zeng et al., 2009). A more constant feature

is the prevalence of CB-ir neurons in the extralemniscal

nuclei of the MGB (Cruikshank et al., 2001; de Venecia et al.,

1995, 1998; Lu et al., 2009). In spite of variability of calcium-

binding proteins in the core nucleus of MGB in different

mammalian species, its telencephalic projection area (3rd

and 4th layers of the auditory cortex) displays PV-ir terminal

field in most mammalian species studied (Hashikawa et al.,

1991; Molinari et al., 1995; de Venecia et al., 1998; Jones, 1998,

2003; Munkle et al., 2000; Cruikshank et al., 2001; Chiry et al.,

2003; Rubio-Garrido et al., 2007). Apparently, in contrast to the

turtle thalamotelencephalic pathway, the auditory geniculo-

cortical lemniscal pathway of mammals keeps the prevalence

of PV-specificity.

Thus, more conservative feature in the organization of

amniotes mesencephalic and thalamic auditory centers is

the domination of CB-containing projection neurons in their

extralemniscal regions. The lemniscal channel of amniotes

auditory system, and especially thalamotelencephalic link,

possesses greater plasticity. Its organization strongly varies

under the pressure of different adaptive specializations. In

particular, the proportion of projection neurons containing

different calcium-binding proteins significantly varies in

different species of amniotes.

3.2. Functional implications

It is well-known that both investigated proteins (PV and CB)

belong to a group of Ca2þ buffers. However, in light of the

recent data concerning the mechanisms of their functioning,

it was shown that in contrast to PV, which is freely diffusing

‘‘pure’’ Ca2þ buffer, CB can play the role of the activity-

dependent Ca2þ sensor. Binding various enzyme systems in

neurons CB becomes partially immobile and participates in

different cell signaling cascades and other neuronal pro-

cesses (Schmidt et al., 2003, 2005; Schwaller, 2007, 2009). PV

and CB, as Ca2þ buffers, possess different biophysical char-

acteristics. PV has been shown to be a so-called ‘‘slow’’ buffer,

expressing in the highly active neurons with fast, highly

synchronized firing patterns (Kawaguchi et al., 1987; Braun

and Piepenstock, 1993; Chard et al., 1993), whereas ‘‘fast’’ CB

buffer has been found in neurons with different firing

patterns, often in projection neurons possessing a tonic

response. Functional and biophysical properties of PV and

CB make them effective neuronal markers, labeling both

morpho-functionally defined neuronal populations within

the same nucleus and whole functional pathways (Celio,

1990; Andressen et al., 1993; Morona and Gonzalez, 2009).

Thus, PV and CB mark two functionally different pathways in

the mammalian auditory sensory system. PV is inherent to

fast-firing projection neurons in centers of the lemniscal

pathway, whereas CB is predominantly expressed in projec-

tion neurons belonging to the extralemniscal pathway (Celio,

1990; Baimbridge et al., 1992; Braun and Piepenstock, 1993;

Jones, 1998).

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 398

In the present study, using PV and CB as markers, we

have clearly distinguished the core (lemniscal) and belt

(extralemniscal) subdivisions of the turtle mesencephalic

auditory center, where the central area of the central nucleus

(core subdivision) contains large multipolar exclusively PV-ir

neurons, the peripheral multimodal area of the central

nucleus contains smaller either PV-ir or CB-ir neurons while

the laminar nucleus (belt subdivision) is characterized by

several populations of moderate fusiform or oval CB-ir

neurons. These observations correlate with the description

of the cytoarchitecture and neuronal morphology of the torus

semicircularis in the red-eared turtle (Browner et al., 1981)

according to which large-sized neurons delineate a particular

subdivision in the central nucleus, while relatively smaller

neurons of different shape and dendritic arbors compose the

periphery of the central nucleus and laminar nucleus.

Exploring the distribution of calcium-binding proteins in

neural centers, many investigators have pointed out different

levels of immunolabeling in neurons. Naturally, this phenom-

enon needs an explanation. Evidently, the intensity of label-

ing reflects the concentration of different calcium-binding

proteins in neurons. In turn, as a rule, the concentration of

peptide determines its functioning. For example, low and

moderate concentrations of CB enhance the process of

paired-pulse facilitation in neurons whereas a complete

absence or a very high concentration of CB reduces this

effect. Consequently, the concentration of CB in neurons

appears to be crucial for the definition of its function

(Schwaller et al., 2002; Schwaller, 2007, 2009).

In morphological and immunohistochemical investiga-

tions, as in present study, a relative concentration in neurons

of each calcium-binding proteins can be estimated by mea-

suring an intensity of immunofluorescence. As we have

shown, all auditory centers of turtles have neurons that,

according to the intensity level of immunofluorescence,

contain different concentrations of either PV or CB. This is

also true when these peptides are co-localized in the same

neuron. Differences in the intensity of PV/CB immunolabel-

ing allow us to describe three different neuronal populations

in the turtle spiral ganglion. In the first-order cochlear nuclei

of different vertebrates, including amphibians, it has been

shown that highly specialized morpho-functional types of

neurons that receive afferent inputs from distinct cell popu-

lations of the auditory ganglion and that project to different

subdivisions of higher-order auditory centers are selectively

labeled by different calcium-binding proteins (Szpir et al.,

1990, 1995; Caicedo et al., 1996; Grothe et al., 2004; Por et al.,

2005; Fredrich et al., 2009; Yan et al., 2010). The turtle first-

order cochlear nuclei also likely contain several neuronal

types (Miller and Kasahara, 1979) but, on the basis of PV/CB

immunolabeling, we could not delineate different neuronal

populations since the majority of neurons there were equally

highly immunoreactive to PV and weakly immunoreactive to

CB. Probably in addition to PV/CB, neurons of the turtle

cochlear nuclei contain other calcium-binding proteins that

might help to distinguish different neuronal populations

there. Previously in our monoimmunolabeling study, along

with PV- and CB-ir neurons, we have found CR-containing

neurons in the turtle CoM (Belekhova et al., 2008). Although

in the core of the turtle thalamic nucleus reuniens, we have

observed a great variety both of PV- and CB-ir neurons, the

measuring of the fluorescence intensity gave us the possibi-

lity to determine there the predominance of CB. Such a

change of the leading role of PV to CB in the core (lemniscal)

subdivisions of the turtle auditory pathway may be explained

by its less specialized functioning in comparison both to the

core of the turtle mesencephalic auditory center and to the

core of the thalamic auditory center in higher amniotes

reminding rather that of the belt (extralemniscal) subdivi-

sions of auditory centers. Lack of neurophysiological and

behavioral data concerning the function of the turtle auditory

centers impedes an interpretation of the present results.

In the majority of auditory centers, the co-localization of

different calcium-binding proteins in neurons has been

found in all vertebrates studied, including turtles, with a

great diversity of a combination of co-localized peptides and

their concentrations in neurons, and as well of the ratio of

neurons containing several proteins in comparable centers of

different species. All of these depend on the functional load

of the concrete auditory center and not on the stage of

phylogenetic development. For example, in turtles, the max-

imal ratio of PV and CB co-localizing neurons has been

observed in the first-order cochlear nuclei and spiral gang-

lion. At that, the majority of neurons containing both pep-

tides express PV in a high concentration and CB in a low

concentration. The minimal ratio of co-localizing neurons

has been demonstrated in the mesencephalic auditory center

where the highly specialized central area of the toral core

contains large-sized monomodal exclusively PV-ir neurons

that project to the core of the thalamic auditory center. Such

specialization decreases in the thalamic auditory center with

its core containing moderate ratio of PV/CB co-localizing

neurons most of which in contrast to the first-order cochlear

nuclei possess a high immunoreactivity to CB and a low

immunoreactivity to PV.

Nothing is known concerning the interplay of PV and other

calcium-binding proteins contained in the same neuron. In

light of recent data on biophysical properties and mechan-

isms of the intracellular functioning of different calcium-

binding proteins (Schwaller et al., 2002; Schwaller, 2007;

2009), it can be only supposed that, depending on their

concentrations in neurons, PV and CB can differently interact.

For example, possessing different characteristics as Ca2þ

buffers, the combination of PV with low concentration of

CB that takes place in the turtle first-order cochlear nuclei

can facilitate neuronal functioning, while a CB at a high

concentration added to PV, which is the case for the nucleus

reuniens, can reduce neuronal activity. In addition, CB play-

ing completely different role of Ca2þ sensor may be involved

in other cellular processes while PV stays the only buffering

protein. However, in many cases, PV being expressed even at

high levels cannot compensate for a CB-deficiency (Schwaller

et al., 2002; Schwaller, 2007; 2009).

3.3. Evolutionary comments

In turtles, the change of PV-chemospecificity to CB-chemos-

pecificity while passing from the core of the mesencephalic

auditory center to that of thalamic and telencephalic auditory

centers appears along with the progressive decrease of the

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 3 99

clear distinction between the core and belt subdivisions in

the thalamic and especially telencephalic auditory centers.

Particularly, although according to differences in morpholo-

gical, hodological, neurochemical and metabolic properties

and as well in neurogenesis time (Ueda et al., 1983; Belekhova

et al., 1985, 2002, 2010; Zeng et al., 2007), the chelonian

thalamic nucleus reuniens might be subdivided into the core

and belt areas, neurophysiological data evidence that the

core of the nucleus reuniens possesses both lemniscal and

extralemniscal features. The reunial core in addition to

monomodal auditory units that are characteristic for lemnis-

cal system contains also bisensory neurons responding to

both auditory and somatosensory stimuli that are inherent in

the extralemniscal system and that have been found in the

reunial belt (Belekhova et al., 1985; Khachunts and Belekhova,

1986). According to the structural, hodological and metabolic

characteristics, more clear division of the nucleus reuniens

into the core (pars centralis) and belt (pars diffusa) takes

place in crocodiles (Pritz and Stritzel, 1992).

In the telencephalic auditory area (ADVRvm) of turtles,

there is a faint indication of the core-belt organization: its

ventral subdivision is a target area of the reunial projections

whereas its medial subdivision receives overlapping projec-

tions from the auditory nucleus reuniens and somatic

nucleus caudalis (Balaban and Ulinski, 1981). These data give

a reason to assume that the ventral part of ADVRvm is

comparable to the core whereas the medial part is compar-

able to the belt. The developmental delay of the putative core

subdivision of ADVRvm as compared to the belt (Zeng et al.,

2007) supports this assumption. However, these areas could

not be distinguished either on the basis of different distribu-

tion of some neuropeptides (Zeng et al., 2007) or difference in

levels of the metabolic activity (Belekhova et al., 2010), which

are differentially distributed in the core and belt subdivisions

of the subcortical auditory centers in turtles and higher

amniotes. Moreover, the ADVRvm contains even less lemnis-

cal-like neurons than the core of nucleus reuniens, though in

the ventral (core) part of ADVRvm polymodal neurons

respond stronger to acoustic stimuli than to somatic ones

(Khachunts and Belekhova, 1986; see Belekhova et al., 2010).

Further investigations are necessary for determining the

degree of the core-belt differentiation in the turtle telence-

phalic auditory center though even at present it is evident

that it is less pronounced than in the subcortical auditory

centers.

The data obtained in different groups of lower vertebrates

support an idea on the progressive evolutionary development

of the core-belt organization of the brain auditory centers. In

amphibians (frogs), these two subdivisions are clearly distin-

guished only in the torus semicircularis according to the

following characteristics: (i) the presence of PV immunoreac-

tivity and absence of CB immunoreactivity in the core area

(nucleus principalis); (ii) the differential distribution of some

neuropeptides between the core (nucleus principalis) and belt

(nuclei laminaris and magnocellularis); and (iii) the difference

in neurogenesis time between the core and the belt nuclei

(Endepols et al., 2000; Zeng et al., 2008; Morona and Gonzalez,

2009). In the thalamic auditory center, which consists of the

posterior and central nuclei, the core and belt nuclei could

not be distinguished on the basis of these characteristics

(Zeng et al., 2008), although the posterior nucleus receives

projections from the toral core whereas toral belt projects

widely including to the polisensory central nucleus (Feng and

Lin, 1991). In fishes, the core-belt organization has not been

revealed even in the mesencephalic torus where only CB- and

CR-ir neurons have been observed (Castro et al., 2006; see

Huang et al., 2011; Grana et al., 2012).

Although at present there is no possibility to estimate

specific functional features of different CB- and PV-contain-

ing projection neurons in core subdivisions of the rostral

auditory centers in vertebrates, the tendency of the increas-

ing of PV-containing projection neurons in the core subdivi-

sions of auditory centers in parallel with the progressive

differentiation of these centers into the core and belt in the

vertebrate phylogeny is evident. This tendency reaches its

peak in higher mammals (primates) whereas in a great

variety of non-primate mammals the enormous variability

of the ratio of CB and PV-ir projection neurons inheres in the

core of auditory centers likely related to the functional

specificity of the auditory system in different species and

its role in the behavioral organization (Jones, 1998, 2003).

4. Experimental procedure

Brains of three adult Testudo horsfieldi and three adult Emys

orbicularis were used. Care and use of the animals in this

study were approved by the Sechenov Institute (Russian

Academy of Sciences) bioethics committee. Animals were

deeply anesthetized with Nembutal (70 mg/kg body weight)

and then transcardially perfused with 0.7–0.9% heparinized

sodium chloride, followed by ice-cold 4% paraformaldehyde

in 0.1 M phosphate buffer at pH 7.4. Brains were removed

from the skull and subsequently postfixed overnight in the

fresh fixative. After cryoprotecting by storage for 24 h in

20–30% sucrose at 4 1C they were freezed in the Tissue-Tek

O.C.T. Compound (Sakura, NL) at �41 1C.

4.1. Immunofluorescence

Immunohistochemical reactions were performed both on

free floating 40 mm sections and mounted on the object-plate

15 mm sections, cut in the frontal plane on a cryostat (Leica

CE, Germany) and collected in 0.02 M PBS (pH 7.4). Immunor-

eactivity to calcium-binding proteins was revealed using

double immunofluorescence labeling. All immunofluores-

cence procedures were applied at room temperature. For

blocking of non-specific binding sites, brain sections were

pre-incubated for 40 min in the blocking solution containing

0.25% Triton X-100 and 5% normal goat serum in PBS. The

sections were then incubated in the mixture of primary

antibodies consisted of 1:800–1:1000 monoclonal mouse

anti-parvalbumin (Sigma, USA) and 1:5000 polyclonal rabbit

anti-calbindin-D28k (Swant, Switzerland) diluted in the PBS

solution with 0.02% Triton X-100 and 2% normal goat serum

for 24 h under continuous agitation. According to the manu-

facturer, these two antibodies do not cross-react with the

other members of the EF-hand family including calretinin.

After rinsing three times for 15 min in the cold PBS, they were

put for 2 h in the mixture of the secondary antibodies labeled

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 3100

with Alexa Fluor 488 (Invitrogen, USA) for revealing PV and

Alexa Fluor 555 (Invitrogen, USA) for revealing CB both

diluted 1:400 in the PBS solution containing 0.25% Triton

X-100 and 2% normal goat serum. After immunolabeling,

sections were routinely rinsed three times in PBS, then free

floating sections were mounted on the SuperFrost Plus slides

(Menzel, Germany) in PBS, air-dried and coverslipped with the

mounting medium for fluorescence with DAPI H-1200 (Vector

laboratories, USA). For control, both primary antibodies were

replaced by the blocking solution. No labeling was seen under

these conditions. The immunofluorescence labeling of the

cerebellar Purkinje cells served as positive control in the

experiments.

4.2. Image analysis and quantitative methods

Sections were observed and analyzed using the fluorescence

microscope Zeiss AxioImager. A1 (Zeiss, Germany) equipped

with the appropriate filter sets for green Alexa Fluor 488 and

orange Alexa Fluor 555 fluorescence. Images were taken from

representative sections with the digital camera mounted on

the fluorescence microscope. In addition, the analysis of the

sections was made with the help of the confocal microscope

Leica SP5 MF (Leica Microsystems, Germany). In this case,

images were obtained using 20x/40x oil immersion objective

lens with additional twofold–fourfold optical magnification

and saved in tiff format. Only images captured from the

confocal microscope are presented in the paper.

The levels of PV and CB immunoreactivity in neurons in the

auditory centers were estimated in the form of optical

density (OD) by measuring the fluorescence labeling intensity

with the help of the ImageJ 1.37v software. Stacks of virtual

sections 0.5 mm thick from 5 representative 40 mm sections of

each investigated auditory center selected on the mid level

(Fig. 1(A)–(E)) from each turtle were captured from the con-

focal microscope and then converted to 8-bit grayscale

images. Pixel grey values (0–255 Gy levels where 0 is the

absolute black indicating the absence of fluorescence and 255

is the absolute white indicating the maximum of fluores-

cence) of all neurons with a clear visible nucleus in the

selected auditory center were measured in the ImageJ. The

ODs of these neurons were then computed as arbitrary units

by the formula:

OD¼ 10=log10 255=pixel grey value� �

Counting of strongly and weakly immunoreactive neurons

was made according to their ODs. PV/CB-ir neurons were

considered as strongly labeled when their ODs varied from 60

till 200 arbitrary units and weakly labeled when their ODs

was less than 60 arbitrary units. Additionally, the fluores-

cence intensity of neurons in each investigated center varied

from high to low within the general OD range.

Data are presented in the paper as mean values 7standard

deviation in arbitrary units, with OD counting on definite

numbers of neurons (n), or as ratios of ODs of strongly and

weakly immunoreactive cells in the centers examined and

their percentages.

For the rostrocaudal comparison between levels of immu-

noreactivity of the brainstem, mesencephalic, thalamic and

telencephalic auditory centers, OD was measured on the

grayscale images of the sections taken from the fluorescence

microscope under the same conditions (same intensity of the

light and magnification). The data obtained were presented

as a graph of mean optical densities of cells in the auditory

centers (in arbitrary units) 7standard deviation. Since the

two calcium-binding proteins were labeled by different fluor-

ochromes, we made no attempt to compare the levels of

immunoreactivity to the two proteins. All statistical analyses

were carried out by using one-way ANOVA method (SPSS

software package), statistical significance being determined

at po0.01 and po0.001.

The numbers of PV-ir and CB-ir cells, together with the

numbers of double-labeled cells, were determined in the

color images of PV immunostaining, CB immunostaining

and superimposed images of double PV and CB immunos-

taining captured from the confocal microscope. According to

the level of immunoreactivity neurons containing both cal-

cium-binding proteins were differently stained: the

yellow–orange ones-neurons, highly immunoreactive both

to PV and CB; the yellow–green ones-neurons, highly immu-

noreactive to PV and weakly to CB; and the orange ones-

neurons, weakly immunoreactive to PV and highly to CB.

Acknowledgments

We gratefully acknowledge Pr. Sergei M. Antonov for help

with the confocal microscopy. This work was supported by

the following Grant sponsors: (1) Museum National d’Histoire

Naturelle, USM 0501 and CNRS UMR 7221, France; (2) Depart-

ment of Biological Sciences, Russian Academy of Sciences

and Russian Foundation of Basic Research ‘‘Mechanisms of

Physiological Functions: from molecule to behavior’’, Russia;

(3) FQRNT and NSERC, Canada.

r e f e r e n c e s

Anderson, L.A., Wallace, M.N., Palmer, A.R., 2007. Identification ofsubdivisions in the medial geniculate body of the guinea pig.Hear. Res. 228, 156–167.

Andressen, C., Blumcke, I., Celio, M.R., 1993. Calcium-bindingproteins: selective markers of nerve cells. Cell Tissue Res. 271,181–208.

Ashwell, K.W., Paxinos, G., 2005. Cyto- and chemoarchitecture ofthe dorsal thalamus of the monotreme Tachyglossus aculeatus,the short beaked echidna. J. Chem. Neuroanat. 30, 161–183.

Baimbridge, K.G., Celio, M.R., Rogers, J.H., 1992. Calcium-bindingproteins in the nervous system. Trends Neurosci. 15, 303–308.

Balaban, C.D., Ulinski, P.S., 1981. Organization of thalamic affer-ents to anterior dorsal ventricular ridge in turtles. I. Projec-tions of thalamic nuclei. J. Comp. Neurol. 200, 95–129.

von Bartheld, C.S., Code, R.A., Rubel, E.W., 1989. GABAergicneurons in brainstem auditory nuclei of the chick: distribu-tion, morphology, and connectivity. J. Comp. Neurol. 287,470–483.

Bazwinsky, I., Hartig, W., Rubsamen, R., 2008. Characterization ofcochlear nucleus principal cells of Meriones unguiculatus andMonodelphis domestica by use of calcium-binding proteinimmunolabeling. J. Chem. Neuroanat. 35, 158–174.

Beccari, N., 1911. La constituzione, i nuclei terminali e le vie diconnessione del nervo acustico nella Lacerta muralis. Merr.Arch. Ital. Embriol. 10, 646–698.

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 3 101

Belekhova, M.G., Chudinova, T.V., Kenigfest, N.B., Krasnoshchekova,E.I., 2008. Distribution of metabolic activity (cytochromeoxidase)and immunoreactivity to calcium-binding proteins in turtlebrainstem auditory nuclei. Zh. Evol. Biokhim. Fiziol. 44, 302–310.

Belekhova, M.G., Chudinova, T.V., Kenigfest, N.B., Vesselkin, N.P.,2007. Metabolic activity of the thalamic and telencephalicauditory centers of reptiles. Dokl. Biol. Sci. 416, 329–332.

Belekhova, M.G., Chudinova, T.V., Reperant, J., Ward, R., Jay, B.,Vesselkin, N.P., Kenigfest, N.B., 2010. Core-and-belt organiza-tion of the mesencephalic and forebrain auditory centres inturtles: expression of calcium-binding proteins and metabolicactivity. Brain Res. 1345, 84–102.

Belekhova, M.G., Kenigfest, N.B., Karamian, O.A., Vesselkin, N.P.,2004. Distribution of calcium-binding proteins in the centraland peripheral regions of the turtle mesencephalic center,torus semicircularis. Dokl. Biol. Sci. 399, 451–454.

Belekhova, M.G., Kenigfest, N.B., Minakova, M.N., Rio, J.-P.,Reperant, J., 2003. Calcium-binding proteins in the turtlethalamus. Analysis in the light of hypothesis of the ‘‘core-matrix’’ thalamic organization in relation to the problem ofhomology of thalamic nuclei among amniotes. Zh. Evol.Biokhim. Fiziol. 39, 504–523.

Belekhova, M.G., Kenigfest-Rio, N.B., Vesselkin, N.P., Rio, J.-P.,Reperant, J., Ward, R., 2002. Evolutionary significance ofdifferent neurochemical organization of the internal andexternal regions of auditory centers in the reptilian brain:an immunocytochemical and reduced NADPH-diaphorasehistochemical study in turtles. Brain Res. 925, 101–106.

Belekhova, M.G., Kratskin, I.L., Reperant, J., Pierre, J., Veselkin, N.P.,Kenigfest, N.B., Tumanova, N.L., Chkheidze, D.D., 1991. Localiza-tion of GABA-immunoreactive elements in the thalamus of thetortoise Emys orbicularis. Zh. Evol. Biokhim. Fiziol. 27, 676–685.

Belekhova, M.G., Zharskaja, V.D., Khachunts, A.S., Gaidaenko,G.V., Tumanova, N.L., 1985. Connections of the mesencephalic,thalamic and telencephalic auditory centers in turtles. Somestructural bases for audiosomatic interrelations. J. Hirnforsch.26, 127–152.

Bender, K.J., Trussell, L.O., 2011. Synaptic plasticity in inhibitoryneurons of the auditory brainstem. Neuropharmacology 60,774–779.

Boord, R.L., 1969. The anatomy of the avian auditory system. Ann.N.Y. Acad. Sci. 167, 186–198.

Braun, K., 1990. Calcium-binding proteins in avian and mammaliancentral nervous system. Progr. Histochem. Cytochem. 21, 1–64.

Braun, K., Piepenstock, A., 1993. Parvalbumin-immunoreactive neu-rons in the subcortical auditory pathway of the Mongolian gerbil(Meriones unguiculatus). Acta Histochem. Cytochem. 26, 543–554.

Braun, K., Scheich, H., Heizmann, C.W., Hunziker, W., 1991.Parvalbumin and calbindin-D28k immunoreactivity as devel-opmental markers of auditory and vocal motor nuclei of thezebra finch. Neuroscience 20, 853–869.

Braun, K., Scheich, H., Schachner, M., Heizmann, C.W., 1985.Distribution of parvalbumin, cytochrome oxidase activityand 14C-2-deoxyglucose uptake of the zebra finch. I. Auditoryand vocal motor systems. Cell Tissue Res. 240, 101–115.

Browner, R.H., Kennedy, M.C., Facelle, T., 1981. The cytoarchitec-ture of the torus semicircularis in the red-eared turtle.J. Morphol. 169, 207–223.

Bruce, L.L., Kornblum, H.J., Seroogy, K.B., 2002. Comparison ofthalamic populations in mammals and birds: expression ofErbB4n mRNA. Brain Res. Bull. 57, 439–442.

Burger, R.M., Pfeiffer, J.D., Westrum, L.E., Bernard, A., Rubel, E.W.,2005. Expression of GABA(B) receptor in the avian auditorybrainstem: ontogeny, afferent deprivation, and ultrastructure.J. Comp. Neurol. 489, 11–22.

Butler, A.B., Reiner, A., Karten, H.J., 2011. Evolution of the amniotepallium and the origins of mammalian neocortex. Ann. N.Y.Acad. Sci. 1225, 14–27.

Caicedo, A., d’Aldin, C., Puell, J.L., Eybalin, M., 1996. Distributionof calcium-binding protein immunoreactivities in the guineapig auditory brain stem. Anat. Embriol. (Berl) 194, 465–487.

Camp, A.J., Wijesinghe, R., 2009. Calretinin: modulator of neuro-nal excitability. Int. J. Biochem. Cell. Biol. 41, 2118–2121.

Cant, N.B., Benson, C.G., 2003. Parallel auditory pathways: projec-tion patterns of the different neuronal populations in thedorsal and ventral cochlear nuclei. Brain Res. Bull. 60, 457–474.

Carr, C.E., Code, R.A., 2000. The central auditory system of reptilesand birds. In: Dooling, R.J., Fay, R.R., Popper, A.N. (Eds.), Com-parative Hearing: Birds and Reptiles. Springer Verlag, New York,pp. 197–248.

Carr, C.E., Fujita, I., Konishi, M., 1989. Distribution of GABAergicneurons and terminals in the auditory system of the barn owl.J. Comp. Neurol. 286, 190–207.

Castro, A., Becerra, M., Manso, M.J., Anadon, R., 2006. Calretininimmunoreactivity in the brain of the zebrafish, Danio rerio:distribution and comparison with some neuropeptides andneurotransmitter-synthesizing enzymes. II. Midbrain, hind-brain, and rostral spinal cord. J. Comp. Neurol. 494, 792–814.

Celio, M.R., 1990. Calbindin and parvalbumin in the rat nervoussystem. Neuroscience 35, 375–475.

Chard, P.S., Bleakman, D., Christakos, S., Fullmer, C.S., Miller, R.J.,1993. Calcium buffering properties of calbindin D28k andparvalbumin in rat sensory neurons. J. Physiol. 472, 331–357.

Chiry, O., Tardif, E., Magistretti, P.J., Clarke, S., 2003. Patterns ofcalcium-binding proteins support parallel and hierarchicalorganization of human auditory areas. Eur. J. Neurosci. 217,397–410.

Covenas, R., De Leon, M., Alonso, J.R., Arevalo, R., Lara, J., Aijon, J.,1991. Distribution of parvalbumin-immunoreactivity in the ratthalamus using a monoclonal antibody. Arch. Ital. Biol. 129,199–210.

Cruikshank, S.J., Killackey, H.P., Metherate, R., 2001. Parvalbuminand calbindin are differentially distributed within primaryand secondary subregions of the mouse auditory forebrain.Neuroscience 105, 553–569.

Davila, J.C., Guirado, S., Puelles, L., 2000. Expression of calcium-binding proteins in the diencephalon of the lizard Psammo-dromus algirus. J. Comp. Neurol. 427, 67–92.

DiFiglia, M., Christakos, S., Aronin, N., 1989. Ultrastructurallocalization of immunoreactive calbindin-D28k in the ratand monkey basal ganglia, including subcellular distributionwith colloidal gold labeling. J. Comp. Neurol. 279, 653–665.

Domenici, L., Waldvogel, H.J., Matute, C., Streit, P., 1988. Distribu-tion of GABA-like immunoreactivity in the pigeon brain.Neuroscience 25, 931–950.

Endepols, H., Walkowiak, W., Luksh, H., 2000. Chemoarchitectureof the anuran auditory midbrain. Brain Res. Rev. 33, 179–198.

Feng, A.S., Lin, W.Y., 1991. Differential innervation patterns ofthree divisions of frog auditory midbrain (torus semicircu-laris). J. Comp. Neurol. 306, 613–630.

Fisher, S.K., Davies, W.E., 1976. GABA and its related enzymes inthe lower auditory system of the guinea pig. J. Neurochem. 27,1145–1155.

Fredrich, M., Reisch, A., Illing, R.B., 2009. Neuronal subtypeidentity in the rat auditory brainstem as defined by molecularprofile and axonal projection. Exp. Brain Res. 195, 241–260.

Frisina, R.D., Zettel, M.L., Kelley, P.F., Walton, J.P., 1995. Distribu-tion of calbindin D-28k immunoreactivity in the cochlearnucleus of the young adult chinchilla. Hear. Res. 85, 53–68.

Grana, P., Huesa, G., Anadon, R., Yanez, J., 2012. Immunohisto-chemical study of the distribution of calcium binding proteinsin the brain of a chondrostean (Acipenser baeri). J. Comp.Neurol. 520, 2086–2122.

Grothe, B., Carr, C.E., Casseday, J.H., Fritsch, B., Koppl, C., 2004.The evolution of central pathways and their neural processingpatterns. In: Manley, G.A., Popper, A.N., Fay, R.R. (Eds.),

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 3102

Evolution of Vertebrate Auditory System. Springer Verlag, NewYork, pp. 289–359.

Hackney, C.M., Mahendrasingam, S., Jones, E.M.C., Fettiplace, R.,2003. The distribution of calcium buffering proteins in theturtle cochlea. J. Neurosci. 23, 4577–4589.

Hackney, C.M., Mahendrasingam, S., Penn, A., Fettiplace, R., 2005.The concentration of calcium buffering proteins in mamma-lian cochlear hair cells. J. Neurosci. 25, 7867–7875.

Hashikawa, T., Rausell, E., Molinary, M., Jones, E.G., 1991. Parval-bumin- and calbindin-containing neurons in the monkeymedial geniculate complex: differential distribution and cor-tical layer specific projections. Brain Res. 544, 335–341.

Heizmann, C.W., Braun, K., 1990. Calcium-binding proteins:molecular and functional aspects. In: The Role of Calcium inBiological systems. CRC Press Inc., Boca Raton, pp. 21–65.

Heller, S., Bell, A.M., Denis, C.S., Choe, Y., Hudspeth, A.J., 2002.Parvalbumin 3 is an abundant Ca2þ buffer in hair cells.J. Assoc. Res. Otolaryngol. 3, 488–498.

Hevner, R.F., Liu, S., Wong-Riley, M.T., 1995. A metabolic map ofcytochrome oxidase in the rat brain: histochemical, densito-metric and biochemical studies. Neuroscience 65, 313–342.

Huang, Y.F., Zhang, J.Y., Xi, C., Zeng, S.J., Zhang, X.W., Zuo, M.X.,2011. Germinal sites and migrating routes of cells in themesencephalic and diencephalic auditory areas in the Africanclawed frog (Xenopus laevis). Brain Res. 1373, 67–78.

Huber, G.C., Crosby, E.C., 1926. On thalamic and tectal nuclei andfiber paths in the brain of the American alligator. J. Comp.Neurol. 40, 97–227.

Jande, S.S., Maler, L., Lawson, D.E.M., 1981. Immunohistochemicalmapping of vitamin D-dependent calcium-binding protein inthe brain. Nature (Lond.) 294, 765–767.

Johnston, J.B., 1915. The cell masses in the forebrain of turtle,Cistudo carolina. J. Comp. Neurol. 25, 393–468.

Jones, E.G., 1998. Viewpoint: the core and matrix of thalamicorganization. Neuroscience 85, 331–345.

Jones, E.G., 2003. Chemically defined parallel pathways in themonkey auditory system. Ann. N.Y. Acad. Sci. 999, 218–233.

Jones, E.G., 2007. The Thalamus, 2nd ed. Cambridge Univ, Press,Cambridge 1679 p.

Jones, E.G., Hendry, S.H., 1989. Differential calcium-binding pro-tein immunoreactivity distinguishes classes of relay neuronsin monkey thalamic nuclei. Eur. J. Neurosci. 1, 222–246.

Karten, H.J., 1967. The organization of the ascending auditorypathway in the pigeon (Columba livia). I. Diencephalic projec-tions of the inferior colliculus (nucleus mesencephali lateralis.pars dorsalis). Brain Res. 6, 409–427.

Karten., H.J., 1968. The ascending auditory pathway in the pigeon(Columba livia). II. Telencephalic projections of the nucleusovoidalis thalami. Brain Res. 11, 134–153.

Kawaguchi, Y., Katsumaru, H., Kosaka, T., Heizmann, C.W., Hama,K., 1987. Fast spiking cells in rat hippocampus (CA1 region)contain the calcium-binding protein parvalbumin. Brain Res.416, 369–374.

Khachunts, A.S., Belekhova, M.G., 1986. Characteristics of repre-sentations of the auditory and somatosensory systems in thethalamus of the turtle: electrophysiological study. Neurofizio-logia 18, 443–453.

Kubke, M.F., Gauger, B., Basu, L., Wagner, H., Carr, C.E., 1999.Development of calretinin immunoreactivity in the brainstemauditory nuclei of the barn owl (Tyto alba). J. Comp. Neurol.415, 189–203.

Logerot, P., Krutzfeldt, N.O., Wild, J.M., Kubke, M.F., 2011. Subdivi-sions of the auditory midbrain (n. mesencephalicus lateralis,pars dorsalis) in zebra finches using calcium-binding proteinimmunocytochemistry. PLoS One 6, e20686.

Lu, E., Llano, D.A., Sherman, S.M., 2009. Different distributions ofcalbindin and calretinin immunostaining across the medial

and dorsal divisions of the mouse medial geniculate body.Hear. Res. 257, 16–23.

Miller, M.R., Kasahara, M., 1979. The cochlear nuclei of someturtles. J. Comp. Neurol. 185, 221–236.

Molinari, M., Dell’Anna, M.E., Rausell, E., Leggio, M.G., Hashikawa,T., Jones, E.G., 1995. Auditory thalamocortical pathwaysdefined in monkeys by calcium-binding protein immunoreac-tivity. J. Comp. Neurol. 362, 171–194.

Morona, R., Gonzalez, A., 2009. Immunohistochemical localiza-tion of Calbindin-D28k and Calretinin in the brainstem ofanuran and urodele amphibians. J. Comp. Neurol. 515,503–537.

Muller, C.M., 1988. Distribution of GABAergic perikarya andterminals in the centers of the higher auditory pathway ofthe chicken. Cell Tissue Res. 252, 99–106.

Munkle, M.G., Waldvogel, H.J., Faull, R.L., 2000. The distribution ofcalbindin, calretinin and parvalbumin immunoreactivity inthe human thalamus. J. Chem. Neuroanat. 19, 155–173.

Papez, J.W., 1935. Thalamus of turtles and thalamic evolution. J.Comp. Neurol. 61, 433–475.

Partata, W.A., Krepski, A.M.R., Xavier, L.L., Marques, M., Achaval,M., 1999. Distribution of glycogen phosphorylase and cyto-chrome oxidase in the central nervous system of the turtleTrachemys dorbigni. Comp. Biochem. Physiol. A. Mol. Integr.Physiol. 124, 113–122.

Parvizi, J., Damasio, A.R., 2003. Differential distribution of calbin-din D28k and parvalbumin among functionally distinctive setsof structures in the macaque brainstem. J. Comp. Neurol. 462,153–167.

Pinaud, R., Mello, C.V., 2007. GABA immunoreactivity in auditoryand song control brain areas of zebra finches. J. Chem.Neuroanat. 34, 1–21.

Por, A., Pocsai, K., Rusznak, Z., Szucs, G., 2005. Presence anddistribution of three calcium-binding proteins in projectionneurons of adult rat cochlear nucleus. Brain Res. 1039, 63–74.

Pritz, M.B., 1974. Ascending connections of a thalamicauditoryarea in a crocodile Caiman crocodilus. J. Comp. Neurol. 153,199–214.

Pritz, M.B., Stritzel, M.E., 1992. A second auditory area in the non-cortical telencephalon of a reptile. Brain Res. 569, 146–151.

Puelles, L., Robles, C., Martınez-de-la-Torre, M., Martınez, S., 1994.New subdivision scheme for the avian torus semicircularis:neurochemical maps in the chick. J. Comp. Neurol. 340,98–125.

Reiner, A., Yamamoto, K., Karten, H.J., 2005. Organization andevolution of the avian forebrain. Anat. Rec. 287, 1080–1102.

Rogers, J.H., 1987. Calretinin: a gene for novel calcium-bindingprotein expressed principally in neurons. J. Cell Biol. 105,1343–1353.

Rogers, J., Khan, M., Ellis, J., 1990. Calretinin and other CaBPs inthe nervous system. Adv. Exp. Med. Biol. 269, 195–203.

Rubio-Garrido, P., Perez-de-Manzo, F., Clasca, F., 2007. Calcium-binding proteins as markers of layer-I projecting vs deep layer-projecting thalamocortical neurons: a double-labeling analy-sis in the rat. Neuroscience 149, 242–250.

Schmidt, H., Schwaller, B., Eilers, J., 2005. Calbindin D-28k targetsmyo-inositol monophosphatase in spines and dendrites ofcerebellar Purkinje neurons. Proc. Nat. Acad. Sci. U.S.A. 102,5850–5855.

Schmidt, H., Stiefel, K., Racay, P., Schwaller, B., Eilers, J., 2003.Mutational analysis of dendritic Ca2þ kinetics in rodentPurkinje cells: role of parvalbumin and calbindin D28k. J.Physiol. 551, 13–32.

Schwaller, B., 2007. Emerging functions of the ‘‘Ca2þ buffers’’parvalbumin, calbindin D-28k and calretinin in the brain. In:Handbook of Neurochemistry and Molecular Neurobiology.Berlin Heidelberg: Springer-Verlag, p. 197–221.

b r a i n r e s e a r c h 1 4 7 3 ( 2 0 1 2 ) 8 7 – 1 0 3 103

Schwaller, B., 2009. The continuing disappearance of ‘‘pure’’ Ca2þ

buffers. Cell. Mol. Life Sci. 66, 275–300.Schwaller, B., Meyer, M., Schiffmann, S., 2002. ‘New’ functions for

‘old’ proteins: the role of the calcium-binding proteins calbin-din D-28k, calretinin and parvalbumin, in cerebellar physiol-ogy. Studies with knockout mice. Cerebellum 1, 241–258.

Sharma, V., Nag, T.C., Wadhwa, S., Roy, T.S., 2009. Stereologicalinvestigation and expression of calcium-binding proteins indeveloping human inferior colliculus. J. Chem. Neuroanat. 37,78–86.

Spatz, W.B., 2003. Purkinje-like cells in the cochlear nucleus ofthe common tree shrew (Tupaia glis) identified by calbindinimmunohistochemistry. Brain Res. 983, 230–232.

Sun, Z., Wang, H.B., Laverghetta, A., Yamamoto, K., Reiner, A.,2005. The distribution and cellular localization of glutamicacid decarboxylase-65 (GAD65) mRNA in the forebrain andmidbrain of domestic chick. J. Chem. Neuroanat. 29, 265–281.

Szpir, M.R., Sento, S., Ryugo, D.K., 1990. Central projections ofcochlear nerve fibers in the alligator lizard. J. Comp. Neurol.295, 530–547.

Szpir, M.R., Wright, D.D., Ryugo, D.K., 1995. Neuronal organizationof the cochlear nuclei in alligator lizard: a light and electronmicroscopic investigation. J. Comp. Neurol. 357, 217–241.

Takahashi, T.T., Carr, C.E., Brecha, N., Konishi, M., 1987. Calciumbinding protein-like immunoreactivity labels the teminal field ofnucleus laminaris of the barn owl. Neuroscience 7, 1843–1856.

Tardif, E., Chiry, O., Probst, A., Magistretti, P.J., Clarke, S., 2003.Patterns of calcium-binding proteins in human inferior colli-culus: identification of subdivisions and evidence for putativeparallel systems. Neuroscience 116, 1111–1121.

Ueda, S., Takeuchi, Y., Sano, Y., 1983. Immunohistochemical demon-stration of serotonin neurons in the central nervous system ofthe turtle Clemmys japonica. Anat. Embryol. 168, 1–19.

Vater, M., Braun, K., 1994. Parvalbumin, calbindin d-28k, andcalretinin immunoreactivity in the ascending auditory path-way of horseshoe bats. J. Comp. Neurol. 341, 534–558.

Vater, M., Kossl, M., Horn, A.K., 1992. GAD- and GABA-immunor-eactivity in the ascending auditory pathway of horseshoe andmustached bats. J. Comp. Neurol. 325, 183–206.

Veenman, C.L., Reiner, A., 1994. The distribution of GABA-con-taining perikarya, fibers, and terminals in the forebrain and

midbrain of pigeons, with particular reference to the basalganglia and its projection targets. J. Comp. Neurol. 339,209–250.

de Venecia, R.K., Smelser, C.B., Lossman, S.D., McMullen, N.T.,1995. Complementary expression of parvalbumin and calbin-din D-28k delineates subdivisions of the rabbit medial geni-culate body. J. Comp. Neurol. 359, 595–612.

de Venecia, R.K., Smelser, C.B., McMullen, N.T., 1998. Parvalbuminis expressed in a reciprocal circuit linking the medial genicu-late body and auditory neocortex in the rabbit. J. Comp.Neurol. 400, 349–362.

Wagner, H., Gunturkun, O., Nieder, B., 2003. Anatomic markers forthe subdivisions of the barn owl’s inferior-collicular complexand adjacent peri- and subventricular structures. J. Comp.Neurol. 465, 145–159.

Wild, J.M., Karten, H.J., Frost, B.J., 1993. Connections of theauditory forebrain in the pigeon (Columba livia). J. Comp.Neurol. 337, 32–62.

Winer, J.A., Larue, D.T., 1996. Evolution of GABAergic circuitry inthe mammalian medial geniculate body. Proc. Nat. Acad. Sci.U.S.A 93, 3083–3087.

Wong-Riley, M.T., 1989. Cytochrome oxidase: an endogenousmetabolic marker for neuronal activity. Trends Neurosci. 12,94–101.

Yan, K., Tang, Y.Z., Carr, C.E., 2010. Calcium-binding proteinimmunoreactivity characterizes the auditory system of Gekkogecko. J. Comp. Neurol. 518, 3409–3426.

Zeng, S., Li, J., Zhang, X., Zuo, M., 2007. Distinction of neuro-chemistry between the cores and their shells of auditorynuclei in tetrapod species. Brain Behav. Evol. 70, 1–20.

Zeng, S., Lin, Y., Yang, L., Zhang, X., Zuo, M., 2008. Comparativeanalysis of neurogenesis between the core and shell regions ofauditory areas in the chick (Gallus gallus domesticus). Brain Res.1216, 24–37.

Zeng, S.J., Lin, Y.T., Tian, C.P., Song, K.J., Zhang, X.W., Zuo, M.X.,2009. Evolutionary significance of delayed neurogenesis in thecore versus shell auditory areas of Mus musculus. J. Comp.Neurol. 515, 600–613.

Zettel, M.L., Carr, C.E., O’Neill, W.E., 1991. Calbindin-like immu-noreactivity in the central auditory system of the mustachedbat Pteronotus parnelli. J. Comp. Neurol. 313, 1–16.