B R A I N R E S E A R C H 1 3 8 1 ( 2 0 1 1 ) 9 5 – 1 0 5
ava i l ab l e a t www.sc i enced i r ec t . com
www.e l sev i e r . com/ loca te /b ra i n res
Research Report
The dorsal pallium in zebrafish, Danio rerio(Cyprinidae, Teleostei)
Thomas Muellera,b,⁎, Zhiqiang Donga, Michael A. Berberoglua, Su Guoa
aDepartment of Bioengineering and Therapeutic Sciences and Programs in Human Genetics, University of California, San Francisco,CA 94143-2811, USAbInstitute of Biology I, Department of Developmental Biology, University of Freiburg, Hauptstrasse 1, D-79104 Freiburg, Germany
A R T I C L E I N F O
⁎ Corresponding author. Institute of BiologyFreiburg, Germany. Fax: 49 761 203 2597.
Article history:Accepted 31 December 2010Available online 8 January 2011
Zebrafish as a neurogenetic model system depends on the correct neuroanatomicalunderstanding of its brain organization. Here, we address the unresolved question regardinga possible zebrafish homologue of the dorsal pallial division, the region that in mammalsgives rise to the isocortex. Analyzing the distributions of nicotine adenine dinucleotidephosphate diphorase (NADPHd) activity and parvalbumin in the anterior zebrafishtelencephalon, we show that against previous assumptions the central (Dc) zonepossesses its own germinative region in the dorsal proliferative zone. We define thecentral (Dc) zone as topologically corresponding to the dorsal pallial division of othervertebrates (mammalian isocortex). In addition, we confirm through BrdU-labelingexperiments that the posterior (Dp) zone is formed by radial migration and homologousto the mammalian piriform cortex. Based on our results, we propose a new developmentaland organizational model of the zebrafish pallium—one which is the result of a complexoutward–inward folding.
The mammalian isocortex is considered the crowning achieve-ment of evolution because it forms the neurological substratefor cognitive and emotive human mental processes (Rakic,2009). It develops from what is called the dorsal pallial
I, Department of Develop
.uni-freiburg.de (T. MuellCtx, cortex; CP, caudate pdorsal telencephalon; Dlposterior zone of the dorsminence; lot, lateral olfaNT, nucleus taeniae; OB,pallium; Sep, septum; TV,l, lateral nucleus of the veis
er B.V. All rights reserved
division. Searching for the evolutionary origin of this structurehas been one of the most challenging questions in compar-ative neurology (Medina and Abellan, 2009). A dorsal pallialdivision homologous to the mammalian isocortex evolvedwith jawed vertebrates (gnathostomes) and is present indiverse anamniotes like sharks, lungfish, and frogs (Gonzalez
mental Biology, University of Freiburg, Hauptstrasse 1, D-79104
er).utamen; D, dorsal telencephalon (pallium); Dc, central zone of the, lateral zone of the dorsal telencephalon; Dm, medial zone of theal telencephalon; EN, entopeduncular nucleus; GP, globus pallidus;ctory tract; LP, lateral pallium; LV, lateral ventricle; MGE, medialolfactory bulb; P, pallium; pirCtx, piriform cortex; PSB, pallial–telencephalic ventricle; V, ventral telencephalon (subpallium); Vd,ntral telencephalon; VP, ventral pallium; Vv, ventral nucleus of the
96 B R A I N R E S E A R C H 1 3 8 1 ( 2 0 1 1 ) 9 5 – 1 0 5
and Northcutt, 2009; Northcutt, 1981; Northcutt, 2009; Pombalet al., 2009; Rodriguez-Moldes, 2009; Wicht and Northcutt,1998). Ray-finned fish (actinopterygians) like zebrafish havebeen denied this privilege. Comparative studies have notestablished a distinct cortex homologue (Northcutt, 2008). Wealso lack specific markers that could help identify the cortexregion. Molecular markers (pax6 and reelin) which label themammalian cortex in a characteristic, stage dependentmanner are not expressed in regions qualifying for a cortexhomologue in zebrafish (Costagli et al., 2002; Wullimann andRink, 2001). Also, none of the extensive molecular and geneexpression studies on embryonic and larval stages of zebrafishhave indicated a cortex homologue (Mueller and Wullimann,2005; Mueller and Wullimann, 2009).
The main obstacle for identifying pallial divisions inzebrafish is the unusual development of the teleosteantelencephalon (Fig. 1). The telencephala of zebrafish andother ray-finned fish develop through a unique process ofoutward folding called eversion. The exact nature of thiseversion process has been a subject of debate for the past130 years. A number of eversion models have been proposed,ranging from very simple to highly elaborate (Braford, 1995;Braford, 2009; Butler, 2000; Gage, 1883; Nieuwenhuys, 2009;Northcutt and Davis, 1986; Northcutt, 2008; Studnička, 1894;
Fig. 1 – Development of the telencephalon in teleosts and mammgiving rise to the telencephalon. (B and D) The teleostean telencetelencephalon (pallium) where proliferative zone and ventricularlocation and orientation of radial glia). The development of the mdivisions which in mammals give rise to the hippocampus, corteEversionmodels simplified after (Nieuwenhuys andMeek, 1990) a(mouse) telencephalon develops through evagination. Proliferatidivisions in mouse simplified (Puelles et al., 2000).
Wullimann and Mueller, 2004; Yamamoto et al., 2007).However, little developmental evidence has validated any ofthese models. As a result, there is no consensus on the exactanatomical delineation of even well established pallialhomologies such as the teleostean pallial amygdala, thehippocampus, and the piriform cortex (Nieuwenhuys, 2009;Northcutt, 2008). Yet, all of the participants in the currentdebate agree that the exact anatomical delineation of thesehomologies and the identification of the dorsal palliumdepend on a complete topological analysis of the teleosteaneversion (Nieuwenhuys, 1962; Nieuwenhuys, 2009).
We chose a comparatively simple yet effective method fordeciphering the zebrafish pallium. To determine andmap truepallial histogenetic units, we studied consecutive sections ofadult zebrafish that were stained against nicotine adeninedinucleotide phosphate diphorase (NADPHd) activity andparvalbumin. The differential staining patterns of both ofthese markers visualized pallial zones and their topologicalsite of origin. For the first time, we show that the central (Dc)zone reaches the dorsal proliferative zone at the rostral pole ofthe telencephalon. Dc comprises its own germinative zone oforigin and, thus, is a true pallial histogenetic unit. In asubsequent BrdU long-term labeling experiment, we provideadditional evidence that the posterior (Dp)-zone, is the result
als. (A) Coronal view of the vertebrate anterior neural tubephalic outward folding (eversion) leads to a dorsalsurface are located on its dorsal most site (indicated byedial (MP), dorsal (DP), lateral (LP), and ventral (VP) pallial
x, piriform cortex, and pallial amygdala is poorly understood.nd (Mueller andWullimann, 2009). (C andD) Themammalianve zones are inwardly oriented towards the ventricle. Pallial
97B R A I N R E S E A R C H 1 3 8 1 ( 2 0 1 1 ) 9 5 – 1 0 5
of radial migration as proposed earlier (Wullimann andMueller, 2004). We also defined the topological origin of bothof these divisions. In sum, we here propose a modification ofthe partial eversion model (Wullimann and Mueller, 2004)—one that recognizes the central (Dc) zone as a true pallialdivision topologically corresponding to the dorsal pallium.
Fig. 2 – Coronal sections of a zebrafish telencephalon stainedagainst NADPHd-activity. (A) The differential distribution ofnicotine adenine dinucleotide phosphate diphorase(NADPHd)-activity reveals distinct pallial divisions in therostral pole of the zebrafish telencephalon. The lateral (Dl)zone of the pallium is strongly NADPHd-positive and caneasily be discerned from the weakly stained central (Dc) zoneand the unstained medial (Dm) and posterior (Dp) zones.(B) For comparison. GAD67 mRNA is a marker for GABAergiccells. The pallium is defined by scarce GAD67-expressingcells and, thus, can be distinguished from the subpallium,which is defined by dense GAD67-expressing cellpopulations in the subpallial parts of the telencephalon,namely the dorsal (Vd) and ventral (Vv) zones of the ventraltelencephalon (Vd) and the entopeduncular nucleus (EN).(C) Schematic coronal sections with topographicalnomenclature. Note that the dorsal (Dd) zone of the dorsaltelencephalon is not found in Fig. 1A and B.
2. Results
2.1. NADPHd-activiy as a marker for pallial units
The contrasting coloration of the NADPHd-activity stainallowed us to distinguish the individual zones of the zebrafishpallium (Fig. 2A). The lateral (Dl) zone located at thedorsolateral part was marked most distinctly in dark blue. Incontrast, the central (Dc) zone at the core of the pallium onlyshowed sparse NADPHd-positive cells. The medial (Dm) zonefacing Dl as well as the posterior (Dp) zone ventral to Dl werefree of NADPHd-activity (Fig. 2A). The locations of the pallialzones are the basis of the common topographical nomencla-ture (Fig. 2A–C) (Nieuwenhuys, 1963; Nieuwenhuys and Meek,1990). Our results, however, deliver the foundation for atopological terminology insofar as we define the four truehistogenetic pallial units and their topological sites of origin.
2.2. NADPHd-activity at the rostral pole of thetelencephalon
We traced these topological sites of origins analyzing consec-utive sections of the rostral pole of adult zebrafish. Here, wecapitalized on the fact that, at its rostral pole, all major zonesof the pallium are easily discernable with a stain againstNADPHd-activity (Fig. 3A–E). We found that the position of thepallial zones at the rostral pole of the telencephalon (describedin Fig. 2A–C) differed significantly from those seen atmidlevels (Fig. 3A–E). At most anterior sections (Fig. 3A andB), the bluish NADPHd-positive Dl zone did not face themedial(Dm) zone but sat laterally to the sparsely NADPHd-positivecentral (Dc) zone. Dc was in these anterior sections verticallysandwiched between the medial (Dm) and the lateral (Dl)zones. Most importantly, the central (Dc) zone—our candidatefor a dorsal pallial division—extended to the dorsal prolifer-ative matrix (marked by arrows in Fig. 3A and B). Dc, thus,comprises its own germinative field in the dorsal proliferationzone.
Atmore posterior levels (Fig. 3C), the lateral (Dl) zone beganto cover the central (Dc) zone including its dorsal mostproliferative matrix (marked by arrows). Also, a large inden-tation started to become visible between Dl and Dm. Morecaudally, this indentation enlarged (Fig. 3D) and formed thesulcus ypsiloniformis (Y) (Fig. 3E). On more caudal sections,the dorsal proliferation zone of Dc—the germinative sheet oforigin—was visible as a distinctive line of red cells (arrows inFig. 3B and C) delineating the border to neighboring medial(Dm) and lateral (Dl) zones. At more caudal sections, thestrongly NADPHd-positive lateral (Dl) zone fully covered thisgerminative sheet (arrows pointing on red cells in Fig. 3C andD) and the entire sparsely labeled central (Dc) zone (Fig. 3C–E).Together, these findings indicate that during development the
central (Dc) zone sinks ventrally into the center of the pallium.We call this process invagination in our outward–inwardfolding hypothesis. The major force of this process may bethe differential growth processes of the medial (Dm) andlateral (Dl) zones during development. We hypothesize thatduring development the medial (Dm) and lateral (Dl) zonegrow laterally and medially, respectively. This processes wecall protrusions.
This displacement of Dc by Dl alters our understanding ofthe sulcus ypsiloniformis (Y) as well (Fig. 3B). The sulcusypsiloniformis (Y) has been interpreted as an independent andstable anatomical landmark between Dm and Dl. Our results,however, indicated that this is true only for posterior levels ofthe telencephalon. At most anterior levels, the sulcus ypsilo-niformis (Y) marked the border between the medial (Dm) zone
Fig. 3 – NADPHd-activity in consecutive sections of the zebrafish pallium. (A and B) In these anterior most sections, the central(Dc) zone is vertically sandwiched between the medial (Dm) and the lateral (Dl) zones. Here, the central (Dc) zone includes itsgerminative field in the proliferative matrix (arrows in Fig. 1A and B). The sulcus ypsiloniformis (Y) appears as a smallindentation (Fig. 1B) between the medial (Dm) and central (Dc) zone. (C and D) The dark blue-stained lateral (Dl) zoneincreasingly covers the central (Dc) zone on more posterior sections. At this level, the central (Dc) zone appears ventrallydisplaced and the sulcus ypsiloniformis is visible as a large gap (Y) between the medial (Dm) and lateral (Dl) zones. Note, Dc'sgerminative zone is visible on top of the ventrally displaced central (Dc) zone (arrows in Fig. 2C, D). (E) The lateral (Dl) zonecompletely covers the central (Dc) zone at subsequent posterior sections. We hypothesize, that the sulcus ypsiloniformis (Y inpanels B–E) and the central position of Dc are results of a complex outward–inward folding (compare Fig. 7A). Thick black arrowsindicate the protrusions of the medial (Dm) and lateral (Dl) zones during this hypothesized folding process.
98 B R A I N R E S E A R C H 1 3 8 1 ( 2 0 1 1 ) 9 5 – 1 0 5
and the central (Dc) zone (Fig. 3B). This is because here thecentral (Dc) zone, including its germinative proliferative layer,reaches the dorsal top of the telencephalon. A dorsal (Dd) zoneis absent in zebrafish.
2.3. Parvalbumin immunoreactivity confirms four pallialunits
To confirm that the zebrafish pallium comprises only fourhistogenetic units, we examined the distribution of parval-bumin. And, indeed, parvalbumin differentially marked thesame four pallial zones in the adult zebrafish as NADPHd(Fig. 4A–D). For example, the lateral (Dl) zone displayed many
parvalbumin-positive cells ventrally adjacent to the dorsalproliferative layer, parvalbumin-positive cells in the periph-ery of Dl, as well as strongly labeled parvalbumin-positivefibers (nicely displayed in Fig. 5E). The central (Dc) zone, incontrast, markedly differed from the overlying lateral (Dl)zone through the absence of parvalbumin-positive cells inboth, its proliferative layer and in its periphery. The presenceof many parvalbumin fibers in Dc set this pallial unit apartfrom the adjacent medial (Dm) and the posterior (Dp) zones,which both lacked expression of parvalbumin. Note, also thelateral olfactory tract (lot), which is known to send olfactoryprojections to the posterior (Dp) zone as well as to the nucleustaeniae (Levine and Dethier, 1985; Northcutt, 2008), was
Fig. 4 – Coronal sections of a zebrafish telencephalon stainedagainst parvalbumin. (A–D) The differential distribution ofparvalbumin-positive cells and fibers reveals all four pallialdivisions in the anterior zebrafish telencephalon. The lateral(Dl) zone of the pallium shows many parvalbumin-positivecells ventrally adjacent to the dorsal proliferation zone andits periphery (nicely seen in panels C and D) and displaysmany parvalbumin-positive fibers (A–D). In contrast, thecentral (Dc) zone shows also many parvalbumin-positivefibers but is defined by the absence of parvalbumin-positivecells. The medial (Dm) zone is largely free ofparvalbumin-positive structures at all of these levels. Thelateral olfactory tract (lot; A) and its principal projection areaand the posterior (Dp) zone (C and D) are both free ofparvalbumin. Sparse parvalbumin-positive neurons arefound in the ventral (Vv) and dorsal (Vd) nuclei of thesubpallium (B–D). Parvalbumin-positive fiber bundles at thepallial–subpallial boundary can be followed from theseanterior sections (arrow in B) that send projections to thecentral (Dc) and lateral (Dl) zones. Note, the sulcusypsiloniformis (Y) is a large indentation at these anteriorlevels (C andD) andDc reaches the ventricular surface formedby it. A dorsal (Dd) zone described in other teleosts is missingaccording to our results.
99B R A I N R E S E A R C H 1 3 8 1 ( 2 0 1 1 ) 9 5 – 1 0 5
defined by the absence of parvalbumin-positive fibers(Fig. 4A).
The majority of parvalbumin-positive fibers in the central(Dc) and lateral (Dl) zone likely originated from two mainsources: an anterior and a posterior one. Anteriorly, parvalbu-min-positive fibers formed fiber bundles at the pallial–subpallial boundary (black arrows, Fig. 4C and D) that weretraceable towards the rostral pole of the telencephalon; these
fiber bundles stemmed from the dorsal (Vd) and ventral (Vv)nuclei of the subpallium. Here, many parvalbumin-positivecells displayed the morphology of projecting neurons (blackarrowheads, Fig. 4C and D). At posterior levels, Dc and Dlappeared to receive parvalbumin-positive projections via thepartially labeled lateral forebrain bundle. These projectionslikely stemmed from nuclei of the preglomerular complex,which is known to project to both the lateral (Dl) was well asthe central (Dc) zone in goldfish (Northcutt, 2006).
2.4. Parvalbumin-immunoreactivity confirms thehistogenetic unit Dc
To confirm that the central (Dc) zone is a true histogeneticunit, we examined the distribution of parvalbumin-reactivityat the most rostral pole of the pallium. Indeed, the central (Dc)zone reached its own germinative zone within the periven-tricular site of proliferation (white arrows in Fig. 5A) whensandwiched between the medial (Dm) and the lateral (Dl)zones. The lateral (Dl) zone, defined by strongly labeledparvalbumin-positive cells ventrally lining the dorsal prolif-erative zone, appeared to protrude medially. It increasinglycovered the central (Dc) regions in subsequent sections(Fig. 5B–D). Note, that the neuroepithelium anteriorly coveringthe central (Dc) zone (white Fig. 5A) is horizontally sand-wiched between Dc and Dl on more posterior sections (arrowsin Fig. 5B–D). At midlevels of the pallium, this neuroepithelialsheet was no longer visible between the central (Dc) andlateral (Dl) zones (Figs. 4D and 5E). Here, the lateral (Dl) zonemarked by many parvalbumin-positive cells completelycovered Dc and was attached to the medially facing medial(Dm) zone. Likewise, an open sulcus ypsiloniformis was onlyvisible at more anterior sections (Y in Fig. 4C and D). Again, nodorsal (Dd) zone was found in these stainings, confirming theabsence of this zone in zebrafish.
In sum, our results show that the zebrafish pallium, like themammalian one, possesses four true histogenetic divisions. Inzebrafish these four are the medial (Dm), the lateral (Dl), theposterior (Dp), and the central (Dc) zones. The central (Dc)zone is a distinct developmental entity (i.e., a histogeneticunit)—one that possesses its own field of origin in the dorsalproliferative matrix. We hypothesize that Dc is ventrallydisplaced during development where its initial dorsal positionshifts to the center of the pallium. The main force of thisinvagination process might be due to the medially orientedprotrusion of the lateral (Dl) zone. A dorsal (Dd) zone is notpresent in zebrafish in contrast to earlier reports (Wullimannet al., 1996). The overgrown part of Dl has beenmisinterpretedas a region corresponding to Dd.
2.5. The posterior zone (Dp) is the result of radialmigration
To identify corresponding pallial divisions between teleosts andmammals—and,most importantly, to defineDc as topologicallycorresponding to the mammalian dorsal pallium—we need tounderstand the topological origin of each of these four domains.Apart from the central (Dc) zone, the posterior (Dp) zone of theteleostean telencephalon was the other key divisions withdisputed place of origin (Nieuwenhuys, 2009). Yet, returning to
Fig. 5 – Distribution of parvalbumin in consecutive sections of the zebrafish pallium. (A–E) Parvalbumin is the second markerthat allows discerning all four true histogenetic units in the zebrafish pallium. The central (Dc) zone is defined by manyparvalbumin-positive fibers, whereas the lateral (Dl) zone also exhibits many parvalbumin-positive cells ventrally adjacent tothe proliferation zone and in its periphery. The medial (Dm) and posterior (Dp) zones are largely free of parvalbumin. (A and B)Anteriorly, Dc extents to the dorsal proliferation and, thus, includes its germanitive field in the proliferative matrix (whitearrows inA andB). At this level, the central (Dc) zone is vertically sandwiched between themedial (Dm) and the lateral (Dl) zonesconfirming the results of the NADPHd-activity stains. Note, in this case, the sulcus ypsiloniformis (Y) does not appear as gap(D) between themedial (Dm) and central (Dc) zone. (B–D) On subsequentmore posterior sections, the strongly labeled lateral (Dl)zone increasingly covers the central (Dc) zone. (E) The lateral (Dl) zone completely covers the central (Dc) zone on midlevels ofthe telencephalon. Thick black arrows indicate the protrusions of the medial (Dm) and lateral (Dl) zones during thehypothesized outward folding process (compare Fig. 7A).
100 B R A I N R E S E A R C H 1 3 8 1 ( 2 0 1 1 ) 9 5 – 1 0 5
our NADPHd-stained sections (in particular Fig. 3C), we get afirst glance at Dp's genesis. In Fig. 3C, the medial (Dm) zone isvertically subdivided. We can differentiate between a verticalfield formed by densely populated cells (red arrowhead inFig. 3C) and a parallel vertical field formed by scattered cells(green arrowhead in Fig. 3C) running alongside the former. Weinterpret this second subdivision as an adult reminiscent traceof an extensive radial migration across the pallium. Thismigration exiles the posterior (Dp) zone to the pial surface
below the lateral (Dl) zone as has been suggested in the partialeversion model (Lillesaar et al., 2009; Mueller and Wullimann,2009; Wullimann and Mueller, 2004).
We determined whether Dp's origin is the result of such aradial migratory activity by performing long-term-BrdUexperiments on larval zebrafish ranging from three to eightand nine days after fertilization. We used an antibody againstHu-proteins, a marker for newborn and differentiating neu-rons, as counterstain. In this way, the pallium (P) was easily
101B R A I N R E S E A R C H 1 3 8 1 ( 2 0 1 1 ) 9 5 – 1 0 5
distinguishable from the subpallium (S) based on itsdiffering proliferation and migration patterns (Fig. 6A–D).The heterogeneous subpallium showed strong proliferativeand migratory activity (Fig. 6A–D) and exhibited many post-mitotic double-labeled BrdU- and Hu-cells (yellow cellsnicely seen in Fig. 6A). In addition, migrated subpallialneurons mono-stained for Hu-proteins (green) appearedbrighter than pallial ones. The pallium displayed a greathomogeneity and dimly lit up as a uniform area of weaklyHu-labeled, green neurons. Here, double-labeled BrdU- andHu-positive cells (orange) remained in the vicinity of theirdorsal proliferative zones of origin marked in red as BrdU-positive. Against this consistent green background of BrdU-negative cells, we identified individual red oval BrdU-positive cells migrating mediolaterally across the pallium(arrowheads in Fig. 6A–D).
Fig. 6 – BrdU long-term pulse–chase experiments. (A–D) A pulse–(dpf) revealed distinct pallial (P) versus subpallial (S) proliferation aby BrdU and Hu-antibodies compared to the pallium and alloweradially migrating BrdU-positive and Hu-negative cells (arrowhe(arrowhead) originating in the proliferation zone of the pallium incells from a similar area towards the pial surfacewhere the primosection further caudally. The primordial posterior (Dp) zone conssurface of the telencephalon. In contrast, the pallial proliferativeorange BrdU-/Hu-positive cells. (E) Schematic cartoon that describof radial migration of cells across the entire pallium as detectedexperiments.
This chain of migrating cells originated in a lateral field ofthemedial (Dm) zone andmoved towards the cell poor area ofthe primordial posterior (Dp) zone (Fig. 6A–D). Once these ovalshaped, red migrating cells arrived at their destination, theytransformed into round-shaped, red neurons forming theposterior (Dp) zone (Fig. 6D). In addition, the post-mitoticBrdU-positive cells of Dp were distinctly different fromadjacent BrdU-positive cells of the lateral (Dl) zone. Cells ofDl were double-labeled for Hu and lit up yellow while themigrated cell population of Dp appeared red only; thus, Dplacked these yellow Hu-positive neurons. This difference incell population reveals that Dp is not a derivative of theproliferative matrix of the lateral (Dl) zone nor of any otherproliferative zone close by, as similarly suggested by Lillesaaret al. (2009). Rather, the developmental foundation of Dp is theabove described radial migration of neurons across the
chase experiment from three to eight days postfertilizationndmigration patterns. The subpallium (S) is stronger stained
d to delineate the pallial–subpallial boundary (PSB). (A) Twoads) forming a chain. (B) A migrating BrdU-positive cella medial area. (C) A chain of radially migrating BrdU-positiverdial posterior (Dp) zone forms. (D) Same brain as panel C, oneisting of dimly red, round BrdU-positive cells close to the pialzone above is formed by shiny red, BrdU-positive cells andes the genesis of the primordial posterior (Dp) zone as a resultwith the above described BrdU-/Hu double labeling
102 B R A I N R E S E A R C H 1 3 8 1 ( 2 0 1 1 ) 9 5 – 1 0 5
pallium (Mueller et al., 2008; Wullimann and Mueller, 2004).The domain of origin of those migrating neurons lies laterallyadjacent to the primordial medial (Dm) zone.
Now thatwehave identified all true pallial divisions and theirpossible mode of development, we can perform a topologicalanalysis of the zebrafish pallium. We propose a new develop-mental model of the teleostean pallial eversion (Fig. 7A) charac-terized by three main morphogenetic processes. First, theposterior (Dp) zone is the result of a radial migration across theentire pallium (Fig. 7A1) aspreviously suggested (WullimannandMueller, 2004). Second, the central (Dc) zone including itsproliferative matrix invaginates during development (Fig. 7A2).Third, the lateral (Dl) zone subsequently overgrows the central(Dc) zone in a medially oriented protrusion (Fig. 7A4). Accordingto ourmodel, the dorsal (Dd) zone is not present in zebrafish andmight not be a histogenetic unit in other teleosts. The dorsal (Dd)zone in other teleostsmay be, as the sulcus ypsiloniformis (Y) inzebrafish, a corollary of the morphogenetic movements duringthe complex eversion process.
Fig. 7 –Complex eversionmodel and topological organization of thsame four pallial divisions present in mammals. Adult location adevelopmental events. First, the posterior (Dp) zone (the lateral ppallium as suggested earlier (Wullimann andMueller, 2004). Secopallium (DP) that invaginates from a dorsal most position. Third,protrudes during development. Note, this model explains the orinvagination. (B) Schematic section with topological nomenclatucompared to mouse. The zebrafish telencephalon possesses a vemammalian basolateral amygdala, a lateral pallium (LP) correspocorresponding to the (iso-) cortex and, a medial pallium (MP) corrtract (lot) of zebrafish is homologous to the tract with the same n(Medina et al., 2004; Puelles et al., 2000).
If we now topologically compare the teleostean telenceph-alon with the mouse model (Figs 1 and 7A–C), we concludethat the medial (Dm) zone corresponds to the ventral pallium(VP), the posterior (Dp) zone to the lateral pallium (LP), thecentral (Dc) zone to the dorsal (pallium), and the lateral (Dl)zone to the medial pallium (MP).
3. Discussion
In this study, we identify in zebrafish the dorsal pallialdivision, which topologically corresponds to the dorsalpallium of other jawed vertebrates, including the isocortex ofmammals. The teleostean dorsal pallial division has beenoverlooked because of its obscured development. For the firsttime, we establish the central (Dc) zone as a histogenetic unit,one that includes its own germinative field of origin in thedorsal proliferative matrix. We hypothesize, that Dc's locationat the center is the result of an invagination process caused by
e zebrafish pallium. (A) The teleostean pallium consists of thend topological origin are obscured by three majorallium, LP) is the result of radial migration across the entirend, the central (Dc) zone at the core of the pallium is the dorsalthe lateral (Dl) zone is the medial pallium (MP) that mediallyigin of the sulcus ypsiloniformis as a corollary of there. The zebrafish pallium and its four pallial divisionsntral pallium (VP) that topologically corresponds to thending to themammalian piriform cortex, a dorsal pallium (DP)esponding to the hippocampus. Note that the lateral olfactoryame in mouse. (C) Mouse telencephalon simplified after
103B R A I N R E S E A R C H 1 3 8 1 ( 2 0 1 1 ) 9 5 – 1 0 5
the protruding medial (Dm) and lateral (Dl) zones duringdevelopment (Fig. 7A).
Until now, the central (Dc) zone of the zebrafish palliumhas not been considered a true histogenetic unit. Instead, Dchas been treated as a deeper zone of the periventricular zoneslike the medial (Dm), the dorsal (Dd), and the lateral (Dl) zones(Northcutt, 2006; Northcutt, 2008: Nieuwenhuys, 2009). Simi-larly, Dc together with periventricular parts of Dd and Dl hasbeen discussed as a possible homologue of the dorsal pallium(Wullimann and Mueller, 2004). These assumptions, however,must be rejected as we clearly show that the central (Dc) zoneis a true histogenetic unit and, thus, a distinct entity of theadult zebrafish brain.
Our interpretation of pallial divisions—apart from Dc and Dd—fundamentally agrees with current research. The medial Dm,lateral Dl, and the posteriorDp zones of the zebrafish palliumareestablished homologues of the pallial amygdala, hippocampus,and piriform cortex, respectively (Braford, 1995; Northcutt, 2006;Portavella et al., 2002). In regard to Dl, we agree with thesuggestion that the dorsal (Dld) and ventral (Dld) zones in otherteleosts than zebrafish are subdivisions of the samehistogeneticunit, i.e., the medial pallium (Northcutt, 2006).
Our anatomical delineation of Dm, Dl, and Dp is further-more supported by molecular marker distributions in theadult and developing teleostean telencephalon. For example,the interpretation of the medial (Dm) zone as the pallialamygdala is consistent with behavioral studies (Portavellaet al., 2002; Portavella et al., 2004), the presence of ascendingfibers from basal forebrain cholinergic cell groups, and thedistribution of calretinin-positive cells (Castro et al., 2006;Mueller et al., 2004). Also, the interpretation of the lateral (Dl)zone as the teleostean hippocampus is supported by behav-ioral experiments and molecular marker distribution (Castroet al., 2006; Portavella et al., 2002). The posterior (Dp) zone ofthe teleostean pallium has been previously established as theregion homologous to the piriform cortex of amphibians andother tetrapods (Braford, 1995). Dp is well distinguishable fromthe overlaying lateral (Dl) zone, which expresses NADPHd andparvalbumin (this study) as well as neuropeptide Y (Castroet al., 2006). Overall, what we describe as a conserved pallialorganization in zebrafish is consistent with developmentalgene expression patterns and mirrors the conserved molecu-lar organization of subpallial divisions. Like in mammals, theteleostean pallium—including its dorsal pallial division—isinvaded by tangentially migrating GABAergic cells thatoriginate in a region homologous to the medial ganglioniceminence (Martyniuk et al., 2007; Mueller et al., 2006; Muelleret al., 2008; Mueller and Guo, 2009; Mueller and Wullimann,2009; Retaux et al., 2008).
Our results have significant implications for the under-standing of the pallial evolution in ray-finned fish andvertebrates in general. The identification of a histogeneticunit in zebrafish that topologically corresponds to the dorsalpallium of other vertebrates is the first evidence for a dorsalpallium in ray-finned fish. A possible dorsal pallial division inteleosts has been discussed as being convergent because therehad been no signs of a dorsal pallium in non-teleost ray-finned fish (Northcutt, 2006; Northcutt, 2008). For example, adorsal pallial division has not been identified in Polypterus, themost basal group of ray-finned fish (Holmes and Northcutt,
2003; Northcutt et al., 2004; Northcutt, 2008). The identificationof Dc as the possible homologue of the dorsal pallium in othervertebrates changes this situation. For example, sturgeons—asecond clade of basal ray-finned fish and closely related toPolypterus—do possess a central (Dc) zone homologous to oneof teleosts (Adrio et al., 2008; Northcutt, 2008; Pinuela andNorthcutt, 2007). This latter finding supports the hypothesisthat a dorsal pallium is an ancestral character of ray-finnedfish, homologous to those of other vertebrates.
The dorsal (Dd) zone, however, has to be excluded as apossible candidate for representing either a convergent orhomologous dorsal pallium in zebrafish.We believe that Dd ofother teleosts than zebrafish is a composite area formedduring the invagination process described in this study.
The identification of a distinct dorsal pallial division inzebrafish and its development through eversion has implica-tions beyond evolutionary theory. For once, we provide astable geography of the zebrafish pallium for succeedingphysiological, anatomical, and behavioral studies. Our de-velopmental paradigm of a dorsal pallial division overgrownby the medial (Dm) and the lateral (Dl) zones and our findingof tangentially migrating cells that form the posterior (Dp)zone need to be further analyzed in future studies regardingthe genetic regulation of these processes in zebrafish.
4. Experimental procedures
4.1. Animal treatments
We used 20 adult zebrafish (age, 6–12 months) from our localbreeding colony at UCSF. Our zebrafish were anesthetizedwith tricaine methanosulfonate (MS222, Sigma) perfused withSörensen phosphate buffer (PB, pH 7.4), and perfusion fixedwith 4% paraformaldehyde (PFA, in PB). We immediatelyremoved the brains after perfusion and postfixed them in 4%PFA for 24–48 h. Then, we washed the fixed brains three timesfor 15 minwith PB and then transferred them into a solution of30% sucrose in PB (wt./vol.) for 16 h (overnight). On the nextday, we transferred the cryoprotected brains to Tissue-Tek®O.C.T.™ and froze them at −20 °C. We analyzed three adultbrains stained against NADPHd-activity and five stainedagainst parvalbumin. Larval zebrafish that we first treatedwith BrdU (see BrdU-labeling) were anesthetized with tricainemethanosulfonate (MS222) for a couple of minutes and thenfixed overnight with 4% PFA. For this study, we analyzed sevenlarval brains stained against BrdU and Hu-proteins.
4.2. NADPHd-activity histochemistry
We performed our histochemical detection of NADPHd-activity in adult zebrafish brains according to Giraldez-Perezet al. (2008). We incubated the brain sections in a solution of0.1 M Tris buffer (pH 8.0), 0.05% Triton-X-100 (Sigma-Aldrich),1 mM beta-NADPH (Sigma-Aldrich), and 0.8 mM nitrobluetetrazolium (Sigma-Aldrich) at 37 °C for 1–2 h. Afterwards,we rinsed the slides with the brain sections three times inphosphate buffer (PB) and fixed them for 1 h in 4% PFA. Laterwe dehydrated them in a graded series of ethanol andcoverslipped them with Entellan (Merck).
104 B R A I N R E S E A R C H 1 3 8 1 ( 2 0 1 1 ) 9 5 – 1 0 5
4.3. Parvalbumin immunostaining
We applied a monoclonal mouse antibody against parvalbu-min (MAB1572, 1:2000, Millipore) on cryosectioned brainsections (thickness, 18–35 μm). We used a secondary antibody(rabbit IgG) coupled with horseradish peroxidase (Elite ABCKit, Vectastain PK-6102) and diaminobenzidine (DAB) aschromogen as described elsewhere (Mueller et al., 2004). TheDAB incubation of slides with brain sections involved heavymetal intensification (50 mg DAB plus 3 ml 1% nickel-sulfateplus 3 ml 1% cobalt chloride in 200 ml PBS). After 20 min ofpreincubation, we added 600 μm of 0.3% H2O2. DAB wasallowed to react with H2O2 for 20 min. Later, we rinsed theslides three times in PB, dehydrated them in a graded series ofethanol, and coverslipped them with Entellan (Merck).
4.4. BrdU-labeling
We incubated 12 larval zebrafish in a 10 mM solution of BrdU(dissolved in system water) for 3–8 days post fertilization asdescribed elsewhere (Mueller and Wullimann, 2002). In orderto cryoprotect our sections, we left our fixed larvae in 30%sucrose in PB overnight and then transferred them to Tissue-Tek® O.C.T.™ and frozen them at minus 20 °C. Sections witha thickness of 16 μm were prepared at a cryotome. Immuno-histochemistry was performed as follows: we washed ourslides with sections in PBS (3×, 10 min) and then incubatedthem in 4 N HCl for 20 min at room temperature. Afterwards,wewashed the sections first with PBS (3×, 10 min) and later inPBS+0.5% Triton (2×, 5 min) and PBS (3×, 10 min). Then weblocked our sections in 3% bovine serumalbumin (BSA) in PBSfor 30 min and incubated them with primary antibodies(diluted in 3% BSA-PBS) overnight at 4 °C. The following day,we washed our sections in PBS (3×, 10 min) and in PBS+0.1%Triton (2×, 5 min). Later, we incubated our sections withsecondary antibodies (diluted in PBS+0.1% Triton) for 2 h,followed by washes with PBS (6×, 10 min). We coverslippedour slides using Dako fluorescent mounting medium. Ourprimary antibodies were anti-BrdU (rat, 1:2000, Abcam) andanti-HuC/D (mouse, 1:1500, Invitrogen). Our secondary anti-bodies were anti-rat 568 and anti-mouse 488 (both Alexa,1:200 dilution). We took our images using a Zeiss compoundmicroscope and we used Adobe Photoshop CS for our imageprocessing.
Acknowledgments
We thank Steffi Dippold and Mimi Zeiger for their editorialwork and Glenn Northcutt for the comments on the manu-script. This work is supported by the NIH AA016021, NS042626,UCSF Byers Award, and Sandler Family Foundation.
R E F E R E N C E S
Adrio, F., Anadon, R., Rodriguez-Moldes, I., 2008. Distribution ofsomatostatin immunoreactive neurons and fibres in thecentral nervous system of a chondrostean, the Siberiansturgeon (Acipenser baeri). Brain Res. 1209, 92–104.
Braford Jr., M.R., 1995. Comparative aspects of forebrainorganization in the ray-finned fishes: touchstones or not?Brain Behav. Evol. 46, 259–274.
Braford Jr., M.R., 2009. Stalking the everted telencephalon:comparisons of forebrain organization in basal ray-finnedfishes and teleosts. Brain Behav. Evol. 74, 56–76.
Butler, A.B., 2000. Topography and topology of the teleosttelencephalon: a paradox resolved. Neurosci. Lett. 293, 95–98.
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. I. Olfactory organand forebrain. J. Comp. Neurol. 494, 435–459.
Costagli, A., Kapsimali, M., Wilson, S.W., Mione, M., 2002.Conserved and divergent patterns of Reelin expression in thezebrafish central nervous system. J. Comp. Neurol. 450, 73–93.
Gage, S.P., 1883. The brain of Diemyctylus viridescens, from larval toadult life and comparisons with the brain of Amia andPetromyzon. The Wilder Quarter-Century Book. Comstock PublCo, Ithaca, NY, pp. 259–313.
Giraldez-Perez, R.M., Gaytan, S.P., Ruano, D., Torres, B., Pasaro, R.,2008. Distribution of NADPH-diaphorase and nitric oxidesynthase reactivity in the central nervous system of thegoldfish (Carassius auratus). J. Chem. Neuroanat. 35, 12–32.
Gonzalez, A., Northcutt, R.G., 2009. An immunohistochemicalapproach to lungfish telencephalic organization. Brain Behav.Evol. 74, 43–55.
Holmes, P.H., Northcutt, R.G., 2003. Connections of the pallialtelencephalon in the Senegal bichir, Polypterus. Brain Behav.Evol. 61, 113–147.
Levine, R.L., Dethier, S., 1985. The connections between theolfactory bulb and the brain in the goldfish. J. Comp. Neurol.237, 427–444.
Lillesaar, C., Stigloher, C., Tannhauser, B., Wullimann, M.F.,Bally-Cuif, L., 2009. Axonal projections originating from rapheserotonergic neurons in the developing and adult zebrafish,Danio rerio, using transgenics to visualize raphe-specific pet1expression. J. Comp. Neurol. 512, 158–182.
Martyniuk, C.J., Awad, R., Hurley, R., Finger, T.E., Trudeau, V.L., 2007.Glutamic acid decarboxylase 65, 67, and GABA-transaminasemRNA expression and total enzyme activity in the goldfish(Carassius auratus) brain. Brain Res. 1147, 154–166.
Medina, L., Legaz, I., Gonzalez, G., De Castro, F., Rubenstein, J.L.,Puelles, L., 2004. Expression of Dbx1, Neurogenin 2, Semaphorin5A, Cadherin 8, and Emx1 distinguish ventral and lateral pallialhistogenetic divisions in the developing mouseclaustroamygdaloid complex. J. Comp. Neurol. 474, 504–523.
Medina, L., Abellan, A., 2009. Development and evolution of thepallium. Semin. Cell Dev. Biol. 20, 698–711.
Mueller, T., Vernier, P., Wullimann, M.F., 2004. The adult centralnervous cholinergic system of a neurogenetic model animal,the zebrafish Danio rerio. Brain Res. 1011, 156–169.
Mueller, T., Wullimann, M.F., 2005. Atlas of Early Zebrafish BrainDevelopment: A Tool for Molecular Neurogenetics. Elsevier.
Mueller, T., Vernier, P., Wullimann, M.F., 2006. A phylotypic stagein vertebrate brain development: GABA cell patterns inzebrafish compared with mouse. J. Comp. Neurol. 494,620–634.
Mueller, T., Wullimann, M.F., Guo, S., 2008. Early teleostean basalganglia development visualized by zebrafish Dlx2a, Lhx6, Lhx7,Tbr2 (eomesa), and GAD67 gene expression. J. Comp. Neurol.507, 1245–1257.
Mueller, T., Guo, S., 2009. The distribution of GAD67-mRNA in theadult zebrafish (teleost) forebrain reveals a prosomeric patternand suggests previously unidentified homologies to tetrapods.J. Comp. Neurol. 516, 553–568.
105B R A I N R E S E A R C H 1 3 8 1 ( 2 0 1 1 ) 9 5 – 1 0 5
Nieuwenhuys, R., 1962. The morphogenesis and the generalstructure of the actinopterygian forebrain. ActaMorphol. Neerl.Scand. 5, 65–78.
Nieuwenhuys, R., 2009. The forebrain of actinopterygiansrevisited. Brain Behav. Evol. 73, 229–252.
Nieuwenhuys, R., 1963. The comparative anatomy of theactinopterygian forebrain. J. Hirnforsch. 13, 171–192.
Nieuwenhuys, R., Meek, J., 1990. The telencephalon ofactinopterygian fishes. In: Jones EG, P.A. (Ed.),Comparative Aspects of Cortical Structure. Plenum Press,New York, pp. 31–73.
Northcutt, R.G., 1981. Evolution of the telencephalon innonmammals. Annu. Rev. Neurosci. 4, 301–350.
Northcutt, R.G., Davis, R.G., 1986. Telencephalic organization inray-finned fishes. Fish Neurobiology Vol. 2. Higher BrainArea and Functions. University of Michigan Press, Ann Arbor,MI, pp. 41–98.
Northcutt, R.G., Plassmann, W., Holmes, P.H., Saidel, W.M., 2004. Apallial visual area in the telencephalon of the bony fishPolypterus. Brain Behav. Evol. 64, 1–10.
Northcutt, R.G., 2006. Connections of the lateral and medialdivisions of the goldfish telencephalic pallium. J. Comp.Neurol. 494, 903–943.
Northcutt, R.G., 2009. Telencephalic organization in the spottedAfrican Lungfish, Protopterus dolloi: a new cytological model.Brain Behav. Evol. 73, 59–80.
Pinuela, C., Northcutt, R.G., 2007. Immunohistochemicalorganization of the forebrain in the white sturgeon, Acipensertransmontanus. Brain Behav. Evol. 69, 229–253.
Pombal, M.A., Megias, M., Bardet, S.M., Puelles, L., 2009. New andold thoughts on the segmental organization of the forebrain inlampreys. Brain Behav. Evol. 74, 7–19.
Portavella, M., Vargas, J.P., Torres, B., Salas, C., 2002. The effects oftelencephalic pallial lesions on spatial, temporal, andemotional learning in goldfish. Brain Res. Bull. 57, 397–399.
Portavella, M., Torres, B., Salas, C., 2004. Avoidance response ingoldfish: emotional and temporal involvement of medial andlateral telencephalic pallium. J. Neurosci. 24, 2335–2342.
Puelles, L., Kuwana, E., Puelles, E., Bulfone, A., Shimamura, K.,Keleher, J., Smiga, S., Rubenstein, J.L., 2000. Pallial andsubpallial derivatives in the embryonic chick and mousetelencephalon, traced by the expression of the genes Dlx-2,Emx-1, Nkx-2.1, Pax-6, and Tbr-1. J. Comp. Neurol. 424, 409–438.
Rakic, P., 2009. Evolution of the neocortex: a perspective fromdevelopmental biology. Nat. Rev. Neurosci. 10, 724–735.
Retaux, S., Pottin, K., Alunni, A., 2008. Shh and forebrain evolutionin the blind cavefish Astyanax mexicanus. Biol. Cell 100, 139–147.
Rodriguez-Moldes, I., 2009. A developmental approach to forebrainorganization in elasmobranchs: new perspectives on theregionalizationof the telencephalon.BrainBehav.Evol. 74, 20–29.
Studnička, F.K., 1894. Zur Lösungeiniger FragenausderMorphologiedes Vorderhirnes der Cranioten. Anat. Anz. 307–320.
Wicht, H., Northcutt, R.G., 1998. Telencephalic connections in thePacific hagfish (Eptatretus stouti), with special reference to thethalamopallial system. J. Comp. Neurol. 395, 245–260.
Wullimann, M.F., Rink, E., 2001. Detailed immunohistology of Pax6protein and tyrosine hydroxylase in the early zebrafish brainsuggests role of Pax6 gene in development of dopaminergicdiencephalic neurons. Brain Res. Dev. Brain Res. 131, 173–191.
Wullimann, M.F., Mueller, T., 2004. Teleostean and mammalianforebrains contrasted: evidence from genes to behavior. J.Comp. Neurol. 475, 143–162.
Yamamoto, N., Ishikawa, Y., Yoshimoto, M., Xue, H.G., Bahaxar, N.,Sawai, N., Yang, C.Y., Ozawa, H., Ito, H., 2007. A newinterpretation on the homology of the teleosteantelencephalon based on hodology and a new eversion model.Brain Behav. Evol. 69, 96–104.
