zfnlrr, a novel leucine-rich repeat protein is preferentially expressed during regeneration in...
TRANSCRIPT
zIi
PMD
zpbdmdszpc1Iiavdstmd
I
aiFtpimaio
d
MCNMolecular and Cellular Neuroscience 13, 167–179 (1999)
Article ID mcne.1999.0742, available online at http://www.idealibrary.com on
1CA
fNLRR, a Novel Leucine-Rich Repeat Proteins Preferentially Expressed during Regenerationn Zebrafish
eter Bormann, Lukas W. A. Roth, David Andel,
anuel Ackermann, and Eva Reinhardepartment of Pharmacology, Biozentrum, University of Basel, 4056 Basel, Switzerland
pac(ictttowhtcp1amof1hafcZsftwiat
fNLRR is a novel transmembrane protein that is mostrominently expressed during regeneration of the ze-rafish central nervous system. Retinal ganglion cells andescending spinal cord neurons strongly increase zfNLRRRNA levels after axotomy in the adult. In contrast, duringevelopment expression is hardly detectable and is re-tricted to a few sensory systems. In the adult brain,fNLRR mRNA is found at low levels in several motor andremotor systems. Sequence analysis reveals that zfNLRRontains in its extracellular region 12 leucine-rich repeats,
immunoglobulin-like domain and 1 fibronectin typeII-like domain. The same protein binding motifs weredentified in transmembrane proteins from frog, mouse,nd human. Together, they constitute a novel family ofertebrate neuronal leucine-rich repeat proteins. Threeistinct isoforms are identified so far. On the basis of itstructural features and expression pattern, we proposehat zfNLRR functions as a neuronal-specific adhesionolecule or soluble ligand binding receptor, primarilyuring restoration of the nervous system after injury.
NTRODUCTION
Axon-extending neurons in the embryo and in thedult during restoration of the nervous system afternjury express a similar set of genes (Skene, 1989;awcett, 1992; Aubert et al., 1995; Caroni, 1997). One ofhe few exceptions is the growth-associated cytoskeletalrotein MAP1x that is expressed in the embryo but not
n the adult (Woodhams et al., 1989). Furthermore, twoembrane-associated proteins, reggie-1 and reggie-2,
re expressed in regrowing retinal ganglion cells (RGCs)n the adult, while no expression is detected in embry-nic retina (Schulte et al., 1997).
In a search for genes that are specifically upregulateduring regeneration of the adult nervous system, wei
044-7431/99 $30.00opyright r 1999 by Academic Pressll rights of reproduction in any form reserved.
erformed differential display of RNA from lesionednd control zebrafish retina (Liang and Pardee, 1992). Inontrast to adult mammalian central nervous systemCNS)-neurons, fish CNS-neurons are capable of regrow-ng their axons after lesion and reestablish functionalircuits (Sperry, 1948; Becker et al., 1997). Here we reporthe isolation of a novel zebrafish transmembrane pro-ein, zfNLRR (zebrafish neuronal leucine-rich repeat),hat contains in its extracellular domain a combinationf distinct motifs, including leucine-rich repeats (LRRs)ith characteristic flanking regions (Kobe and Deisen-
ofer, 1994). The LRR-motif provides an ideal conforma-ion for binding to other proteins. Therefore, all LRR-ontaining proteins are thought to be involved inrotein–protein interactions (Kobe and Deisenhofer,994). LRRs are found in proteins with diverse functionsnd cellular locations, including extracellular, adhesiveolecules, such as small proteoglycans that bind vari-
us components of the extracellular matrix and growthactors (Kobe and Deisenhofer, 1994; Hocking et al.,998). The function of only a few of the LRR-proteinsas so far been determined. In Drosophila, connectinnd capricious are involved in neuromuscular junctionormation (Nose et al., 1992; Shishido et al., 1998),haoptin in photoreceptor cell development (Krantz andipursky, 1990), and slit in the pathfinding of commis-ural axons (Rothberg et al., 1990). Other members of theamily constitute signal transducing receptors, such ashe neurotrophin receptors trk (Schneider and Sch-
eiger, 1991). In addition to the LRR-motif, the newlysolated zfNLRR contains an immunoglobulin-like and
fibronectin type III-like domain, modules characteris-ic for many cell adhesion molecules of the Ig superfam-
ly (Fields and Itoh, 1996).We show that zfNLRR mRNA is low abundant in the
167
efmRnTwma
R
zw
r4slazrbaepcmzcTrCbleDt(ttdHvzsfaHNt
b1c((ztatmxmoolcepstctir1isCn
ptlw1oAasm
zN
gsttt
168 Bormann et al.
mbryonic nervous system, where it is restricted to aew sensory systems. In response to axonal lesion its
RNA level is drastically increased in regeneratingGCs and in descending spinal cord neurons that haveot expressed zfNLRR during embryonic development.his demonstrates that zfNLRR is one of the few geneshose expression pattern reflects the different require-ents for neuronal growth in the embryo and in the
dult after lesion.
ESULTS
fNLRR cDNA Codes for a Transmembrane Proteinith Distinct Motifs
By differential display of RNA populations fromegenerating and control zebrafish retina, we isolated a00-bp-long fragment from the 38 end of a cDNA that istrongly transcribed in RGCs in response to optic nerveesion. The full-length cDNA was obtained by screening
cDNA library from adult control and regeneratingebrafish retina. It comprises 6578 bp, including an openeading frame of 2232 bp, a 58 untranslated region of 300p, and a 38 untranslated region of 4048 bp. Sequencenalysis reveals the presence of a 350-bp-long DANAlement within the 38 untranslated region, starting atosition 5278. DANA elements constitute a family ofomposite, tRNA-derived, short interspersed DNA ele-ents that are associated with mutational activities in
ebrafish (Izsvak et al., 1996). The deduced proteinontains 744 aa and includes two hydrophobic stretches.he one in the most N-terminal region is likely toepresent a signal peptide while the other, close to the-terminal region, is highly reminiscent of a transmem-rane domain (Fig. 1A). Analysis of the 660-aa extracel-
ular domain reveals the presence of 12 LRRs that arencompassed by flanking cystein clusters (Kobe andeisenhofer, 1994), a single Ig-like domain, and adjacent
o the membrane, a fibronectin type III-like domainFigs. 1A and 1B; Carnemolla et al., 1996). According tohe division of the immunoglobulin superfamily intohree sets by Williams and Barclay (1988), the Ig-likeomain found in zfNLRR belongs to the C2 type.owever, it also contains structural features found in
ariable domains. Therefore, the Ig-like domain offNLRR may belong to the distinct structural set called Iet that includes several of the immunoglobulin super-amily domains forming the cell adhesion moleculesnd surface receptors (Harpaz and Chothia, 1994).omology searches based on the alignment method of
eedleman and Wunsch (1970) showed overall similari-ies of zf NLRRR to two leucine-rich repeat transmem-bi
rane proteins from mouse, mNLRR-1 (Taguchi et al.,996) and mNLRR-3 (Taniguchi et al., 1996), and toDNAs from xenopus (xNLRR-1; AB014462) and humanhNLRR-3; AC004142) and an EST from humanAI015776). Therefore, we named the isolated proteinfNLRR, for zebrafish neuronal leucine-rich repeat pro-ein. zfNLRR shows a similarity of 64 and 63% on themino acid level to mNLRR-1 and xNLRR-1, respec-ively (Fig. 1C). Its extracellular domain contains 28 aa
ore than the extracellular domain of mNLRR-1 andNLRR-1. Its similarity to mNLRR-3 and hNLRR is 52%.NLRR-1 and xNLRR-1 show a 86% similarity to each
ther and hNLRR and mNLRR-3 are 90% similar to eachther. This indicates that zfNLRR is not a true homo-
ogue of the so far identified NLRR proteins, butonstitutes a novel member of the family. This hypoth-sis is strengthened by analysis of the 62-aa-long cyto-lasmic tail. The intracellular portion of zfNLRR is moreimilar to mNLRR-1, xNLRR-1, and the human ESThan to hNLRR and mNLRR-3. However, there is lessonservation among zfNLRR, mNLRR-1, and xNLRR-1han between mNLRR-1 and xNLRR-1 (Fig. 1C). Interest-ngly, a box of 11 aa is contained within the intracellularegion that is identical in all members of the family (Fig.B). Two putative mammalian endocytosis motifs arencluded within this conserved box, a tyrosine-basedignal conforming to the YxxØ motif (Chen et al., 1990;ollawn et al., 1990) and a dileucine-type motif (Letour-eur and Klausner, 1992; Pieters et al., 1993).Pulse-labeling experiments reveal that zfNLRR is
osttranslationally modified. COS cells were transientlyransfected with a zfNLRR/myc fusion cDNA, pulse-abeled with [35S]methionine, and immunoprecipitated
ith a c-myc-specific mAb (9E10; Evan et al., 1985; Fig.D). One dominant band is visible at a molecular weightf approximately 120 kDa at the beginning of the chase.fter 1 h, a higher band at approximately 140 kDa
ppears. This strongly indicates that zfNLRR is a glyco-ylated protein and that glycosylation increases itsolecular weight by approximately 20 kDa.
fNLRR Is Expressed in the Embryonic and Adultervous System
zfNLRR gene expression in zebrafish was investi-ated by Northern blot and in situ hybridization analy-is. Northern blot analysis reveals a single zfNLRRranscript with a length of approximately 6700 nt in allissues examined (Fig. 2). Long exposure times result inhe appearance of additional bands, due to unspecific
inding of the probe to ribosomal RNA contaminationsn the poly(A)1 RNA preparations (Fig. 2; embryo,
FIG
.1.
zfN
LR
Ris
apo
sttr
ansl
atio
nally
mod
ified
tran
smem
bran
epr
otei
nw
ith
dis
tinc
tmot
ifs
inth
eex
trac
ellu
lar
and
intr
acel
lula
rpa
rts.
(A)A
min
oac
idse
quen
ceof
zfN
LR
R.1
2le
ucin
e-ri
chre
peat
s(u
nder
lined
)wit
ha
char
acte
rist
icam
ino
and
carb
oxy
flan
king
regi
on(d
otte
d),
anIg
-lik
ed
omai
n(b
oxed
),an
da
fibro
nect
inty
peII
I-lik
ed
omai
n(d
ashe
d)
are
foun
din
its
extr
acel
lula
rpa
rt.T
hetr
ansm
embr
ane
dom
ain
isd
oubl
eun
der
lined
.The
intr
acel
lula
rpa
rtco
ntai
nsa
stre
tch
ofam
ino
acid
s(s
had
edbo
x)th
atsh
ows
high
sim
ilari
tyto
othe
rm
embe
rsof
the
NL
RR
fam
ily.
(B)
The
cons
erve
dbo
xw
ithi
nth
ein
trac
ellu
lar
dom
ain
defi
nes
diff
eren
tiso
form
sw
ithi
nth
eso
far
char
acte
rize
dm
embe
rsof
the
fam
ily.(
C)P
erce
ntag
eof
sim
ilari
tyon
the
amin
oac
idle
vel
amon
gd
iffer
ent
mem
bers
ofth
efa
mily
.mN
LR
R-1
and
xNL
RR
-1(d
ark
shad
e)an
dm
NL
RR
-3an
dhN
LR
R(l
ight
shad
e),r
espe
ctiv
ely,
repr
esen
td
isti
nct
isof
orm
s.(D
)Pu
lse-
chas
eex
peri
men
tw
ith
zfN
LR
R/
myc
-tra
nsfe
cted
CO
Sce
lls.z
fNL
RR
was
imm
unop
reci
pita
ted
usin
ga
mA
bsto
c-m
yc,f
ollo
wed
byan
alys
ison
a7.
5%SD
S–po
lyac
ryla
mid
ege
l.C
hase
peri
ods
are
ind
icat
edon
the
top
ofea
chlin
e.In
the
last
line,
cells
wer
ela
bele
dw
ith
[35S]
met
hion
ine
for3
h.A
rrow
son
the
righ
tsid
epo
intt
opo
sttr
ansl
atio
nally
mod
ified
and
nonm
odifi
edzf
NL
RR
.Mol
ecul
arw
eigh
tsar
ein
dic
ated
inkD
a.
Regenerating CNS Neurons Express zfNLRR 169
bapt(ahdzet
zsftEtgnatob
szbdltmtoss
pmhnmslmstFti(rab(oi(tgsccoetntatc(4pi
FeferewletT
170 Bormann et al.
rain). By far the strongest signal is obtained in regener-ting retina whose optic nerve has been crushed 5 daysreviously (Fig. 2; retina 1). An extended exposure of
he blot is needed to detect a band in control retinaretina 2), in 48-h embryos, and in the adult brain. Inddition, the amount of mRNA loaded was muchigher for adult brain and embryos than for retina (foretails see ‘‘Experimental Methods’’). This shows thatfNLRR mRNA is very low abundant in zebrafishmbryos at times when extensive neurite outgrowthakes place.
We analyzed by in situ hybridization the site offNLRR expression in the embryo. The most prominentignal, starting at approximately 24 h of development, isound in bilateral structures in the developing forebrain,he presumptive olfactory bulbs (Figs. 3A and 3B).xpression of zfNLRR mRNA corresponds with the
ime when axons from the olfactory placode arise androw toward the olfactory bulb that has not yet evagi-ated at this stage (Wilson et al., 1990). Visualizingxon-extending neurons with an antibody to acetylatedubulin (Black and Keyser, 1987) shows that axons
IG. 2. Northern blot analysis of zfNLRR expression during regen-ration in the adult and embryonic development. Poly(A)1 RNAsrom regenerating (retina 1) and control (retina 2) retina, 48-hmbryos, and adult brain were analyzed using a probe to the codingegion of zfNLRR. A single transcript was detected in all tissuesxamined (arrow). The lane holding the RNA from regenerating retinaas shorter exposed than the rest of the blot. The amount of mRNA
oaded was 4.5 times higher for adult brain and 12 times higher formbryos than for retina. Stars indicate unspecific binding of the probeo ribosomal RNA contamination in the poly(A)1 RNA preparations.he size marker is in nucleotides.
riginating in the olfactory epithelium connect to therain in regions corresponding to the zfNLRR mRNA
et
ignal (Fig. 3C). This reinforces the assumption thatfNLRR is expressed in the differentiating olfactoryulb. A weak hybridization signal is seen in RGCs of theeveloping retina (Fig. 3D) and in neuromasts of the
ateral line system (Fig. 3E). To detect zfNLRR mRNA inhese places, the color reaction had to be proceeded for a
uch longer time than in the olfactory bulbs, suggestinghat the expression level is much lower than in thelfactory bulbs. Taken together, zfNLRR mRNA expres-ion in the embryo is restricted to a few developingensory systems.
Northern blot analysis revealed that zfNLRR is ex-ressed in the adult zebrafish brain (Fig. 2). To deter-ine the site of expression, we performed in situ
ybridization studies. Brain areas were identified andamed according to the zebrafish brain atlas (Wulli-ann et al., 1996). In the adult, the olfactory bulbs are
till zfNLRR mRNA positive (Fig. 4A). Positive cells areocated around glomeruli and may represent a subset of
itral cells (Fig. 4B). Several brain regions receivingecondary olfactory projections, such as the ventralelencephalic area (Fig. 4D) and the nucleus taeniae (NT;ig. 4A8) are zfNLRR mRNA positive. The tegmentumhat includes many motor structures expresses zfNLRRn restricted areas, presumably in the oculomotor nucleusNIII) and in the most rostral portion of the superioreticular formation (SRF; Fig. 4C). Figure 4C also showsccumulation of zfNLRR mRNA in cells that mayelong to the nucleus of the medial longitudinal fascicleNMLF), where the major descending spinal projectionsriginate. Several structures in the cerebellum, includ-
ng the granular eminence are zfNLRR mRNA positiveFig. 4E). The cerebellar crest that is considered part ofhe medulla oblongata and receives input from theranular eminence shows a strong zfNLRR mRNAignal in whole mount brains (Fig. 4E). Large, definedells, located in the layer of the Purkinje cells in theerebellum express zfNLRR (Figs. 4F and 4F8). In tele-sts, this layer contains, in addition to Purkinje cells,urydendroid cells that form the main efferent connec-ions from the cerebellum, including outputs to theucleus of the medial longitudinal fascicle, the oculomo-
or nucleus, and the reticular formation (Wullimann etl., 1996). It remains to be elucidated which cell-type inhe Purkinje cell layer expresses zfNLRR. In the spinalord, distinct motoneurons of the W and RI typeWesterfield et al., 1986) show zfNLRR expression (Fig.G). In summary, in the adult brain, zfNLRR is ex-ressed in sensory systems and in motor, premotor, and
ntegrative centers. The staining pattern indicates that
xpression is neuronal. However, we cannot excludehat neuron-associated glial cells also express zfNLRR.
z
zrhoa(aisV(b5Rs
msnatsl(i(wtmii
gammwg6icz
D
z
waszImLitsvahfp
FwetgAµFb(sttlsaIoE
Regenerating CNS Neurons Express zfNLRR 171
fNLRR Is Expressed in Regenerating CNS Neurons
We originally isolated zfNLRR from regeneratingebrafish retina. To determine the site and temporalegulation of its expression during regeneration, in situybridization studies were performed on retinas whoseptic nerves had been crushed at different times prior tonalysis. No zfNLRR mRNA is detected in control retinaFig. 5A). Figures 5B–5F show that regenerating RGCsre the only cells in the retina responding to the crush byncreasing the zfNLRR mRNA level. A subset of RGCstart to raise RNA levels at 24 h postaxotomy (Fig. 5B).irtually all RGCs express zfNLRR at 3 days after crush
Fig. 5C). The relative mRNA level reaches a maximumetween days 3 and 5 and decreases thereafter (Figs.D–5F). At 9 days after lesion, only a small number ofGCs are still zfNLRR mRNA positive (Fig. 5F). Noignal is obtained with a sense probe (not shown).
To examine whether the upregulation of zfNLRRRNA after lesion is restricted to the primary visual
ystem or whether other CNS neurons also respond toerve injury by increasing zfNLRR mRNA levels, wenalyzed different brain regions following spinal cordransection (Fig. 6). Several nuclei, including the somato-ensory medial funicular nucleus (MFN) and the reticu-ar formation, that are affected by spinal cord lesionBecker et al., 1997) show low levels of zfNLRR mRNAn control fish (Fig. 6B). Also, the vagal motor nucleusNxm) that does not project to the spinal cord shows a
eak hybridization signal (Figs. 6A and 6B). Lesion ofhe spinal cord leads to a drastic increase of zfNLRR
RNA. Most brain nuclei with descending fibers, includ-ng the medial longitudinal fascicle (MLF) and thenferior reticular formation (IRF) in the medulla oblon-
IG. 3. zfNLRR expression during zebrafish embryonic developmehole-mount brains. Bilateral structures in the telencephalon, presu
mbryos. (C) Extending axons are visualized with an antibody to acetyo regions in the forebrain that are zfNLRR mRNA positive. (D) Coronanglion cell layer express zfNLRR. (E) Horizontal section through abbreviations: d, diencephalon; ey, eye; h, hindbrain; l, lens; m, musclem in D.IG. 4. zfNLRR expression in the adult zebrafish brain, analyzed by iulbs, the nucleus taeniae (enlarged in A’), the hypothalamus (arrowhB) Saggital section through the olfactory bulb. (C) Saggital section thection through the medial part of the telencephalon. Positive cellselencephalic area. (E) Dorsal view of the cerebellum in a whole-mounhe arrowheads to the crista cerebellaris. (F) Saggital section through thayer and may be Purkinje cells or eurydendroid cells. (F8) Enlargemeubset of the cells in the Purkinje cell layer express zfNLRR (arrowheadre zfNLRR mRNA positive. Abbreviations: C, central canal; CCe, cere
L, inferior lobe of hypothalamus; Mo, medulla oblongata; NMLF, nuclculomotor nucleus; SRF, superior reticular formation; TeO, optic tectum; V; 100 µm in B, C, D, F; 50 µm in G.ata (Fig. 6A), the medial octavolateralis nucleus (MON),nd the intermediate reticular formation (IMRF) in theetencephalon (Fig. 6C), drastically increase zfNLRRRNA levels after spinal cord lesion. Lesioned neuronsere visualized by applying fluorescently labeled, retro-
radely transported dextran beads to the cut site. FigureD shows that regions of strong zfNLRR mRNA stain-ng contain nuclei with major spinal projections. Inontrast to RGCs, several of these brain nuclei expressfNLRR at detectable levels without lesion.
ISCUSSION
fNLRR Contains Several Protein-Binding Motifs
We have cloned a novel cDNA of a gene, zfNLRR,hose mRNA is most abundant in regenerating RGCs
nd in brain nuclei with descending projections afterpinal cord transection. Sequence analyses revealed thatfNLRR cDNA codes for a transmembrane protein typea that contains a distinct combination of adhesion
otifs in its extracellular domain. The largest motif is aRR. All members of the LRR family analyzed so far, are
nvolved in protein-protein interactions. The residueshat determine the structure of the LRR-motif are con-erved within this family, while the intervening residuesary considerably and determine the specificity andffinity of the interaction with ligands (Kobe and Deisen-ofer, 1995). The trk family of neurotrophin receptors,or example, contains an LRR domain with three re-eats. The intermediate residues of the second repeat
nalyzed by in situ hybridization. (A, B) Saggital and dorsal view ofy the olfactory bulbs (arrowheads) express zfNLRR mRNA in 24-htubulin. Fibers originating in the olfactory epithelium (arrows) projecttion through the eye of a 48-h embryo. Cells in the developing retinalembryo. Neuromasts of the lateral line system are zfNLRR positive.
tina; sp, spinal cord; t, telencephalon. Scale bars, 50 µm in A, B, C, E; 25
hybridization. (A) Ventral view of a whole-mount brain. The olfactory), and the eminentia granularis (arrows) are zfNLRR mRNA positive.h the mesencephalon. (Plane of section indicated in A). (D) Saggitalcated by arrows, may be part of the dorsal nucleus of the ventralin. Arrows point to the zfNLRR expressing eminentia granularis andebellum. zfNLRR mRNA-positive cells are located in the Purkinje celld double labeling with the nuclear dye ‘‘Hoechst’’ shows that only a) A coronal section through the spinal cord reveals that motoneuronscorpus; DH, dorsal horn; Dl, lateral zone of dorsal telencephalic area;
nt, amabllatedal sec48-h; r, re
n situeadsroug
, indit brae cernt ans). (G
bellar
eus of the medial longitudinal fascicle; NT, nucleus taeniae; NIII,, ventral telencephalic area; VH, ventral horn. Scale bars: 500 µm in A,
172 Bormann et al.
aizflckacpaaets
sk
DiHsemo
Ietggep
FmnNlz AbbrI , medv .
Regenerating CNS Neurons Express zfNLRR 173
re responsible for the specificity of neurotrophin bind-ng in trkA and trkB (Windisch et al., 1995). The LRRs offNLRR are surrounded by an amino- and a carboxy-anking region that are characterized by similarly spacedysteins (Rothberg et al., 1990; Fisher et al., 1991). Little isnown about the role of these flanking cystein clustersnd only a small number of LRR containing proteinsarry both of them. It has been proposed that theseroteins commonly function in neural differentiationnd/or developmental processes as adhesive proteinsnd/or receptors (Kobe and Deisenhofer, 1994; Suzukit al., 1996). Mutations in the carboxy-flanking region ofhe Drosophila toll-protein has been shown to influenceignal transduction (Hashimoto et al., 1991).
Several LRR proteins are expressed in the nervous
IG. 6. Expression of zfNLRR mRNA in the adult, regenerating CNSedulla oblongata. Neurons of the MFN and IRF are affected by the
uclear dye ‘‘Hoechst’’ shows the arrangement of the cells in this areaxm are not affected by the lesion. They express zfNLRR constitutivel
esion by increasing the zfNLRR mRNA level. (D) Retrogradely tranfNLRR positive regions in C contain neurons with descending axons.RF, inferior reticular formation; MFN, medial funicular nucleus; MLFagal motor nucleus; RV, rhombencephalic ventricle. Scale bar, 100 µm
ystem. Their binding partners and function is onlynown for a few of them and has been studied best in
pr
rosophila. Connectin, toll, and capricious are involvedn neuromuscular junction formation (Nose et al., 1992;
alfon et al., 1995; Rose et al., 1997; Shishido et al., 1998),lit in the pathfinding of commissural axons (Rothbergt al., 1990), and chaoptin in photoreceptor cell develop-ent (Krantz and Zipursky, 1990). The only homology
f zfNLRR to these proteins resides in the LRR domain.In addition to the LRR domain, zfNLRR contains an
g-like domain and a FN type III like repeat in itsxtracellular region. Only a few LRR-proteins are knownhat contain additional Ig-like domains. Included in thisroup are LIG-1 that is expressed in a small subset oflial cells (Suzuki et al., 1996), ISLR, an abundantlyxpressed gene cloned in human (Nagasawa et al., 1997),eroxidasin, a Drosophila extracellular protein with
ys after spinal cord transection. (A) Coronal section at the level of then and express high levels of zfNLRR mRNA. (A8) Staining with theNonlesioned control at the same level as shown in A. Neurons of theNeurons in the reticular formation and the MON react to spinal cord
ted dextran beads (applied to the lesion site) demonstrate that theeviations: CC, cerebellar crest; IMRF, intermediate reticular formation;ial longitudinal fascicle; MON, medial octavolateralis nucleus; Nxm,
, 5 dalesio. (B)y. (C)spor
eroxidase activity (Nelson et al., 1994), and the trk-eceptor family (reviewed in Meakin and Shooter, 1992).
TbtIataRmdsnhicFtcfcfntRer
z
eqzFsfmhasxoptdl
1ttYd
Fa ted it inneo
174 Bormann et al.
he FN type III-like repeat is adjacent to the transmem-rane region of zfNLRR and shows similarity to the FNype III motif of Tenascin-R (Carnemolla et al., 1996).g-like domains in combination with FN type III repeatsre found in several neuronal cell-adhesion molecules ofhe immunoglobulin superfamily. Often, these proteinsre involved in homophilic binding (Brummendorf andathjen, 1996). Homophilic binding of NCAM and L1ay play a role in appropriate glia:axon interactions
uring development and regeneration of the nervousystem (Martini, 1994; Bernhardt et al., 1996). Prelimi-ary experimental data obtained for zfNLRR suggesteterophilic, rather than homophilic binding. Often,
ndividual zfNLRR-positive cells are surrounded byells that do not express the gene (compare Figs. 4–6).urthermore, transient expression of zfNLRR in cul-ured cells did not result in a preference of transfectedells to clump or in a more intense labeling of theusionprotein at contact sites (not shown). However, weannot exclude that zfNLRR is involved in neuriteasciculation by homophilic binding. Members of theeural immunoglobulin superfamily have been shown
o undergo complex interactions (Brummendorf andathjen, 1996) and it may be that zfNLRR exerts differ-
IG. 5. Changes of zfNLRR mRNA levels during regeneration of lenalyzed by in situ hybridization at different times thereafter (indicaransiently express zfNLRR. Abbreviations: gcl, ganglion cell layer; inl,uter plexiform layer. Scale bar, 50 µm.
nt functions in the embryo, in the adult, and duringegeneration that are based on distinct binding partners.
Pr
fNLRR Is a Member of the NLRR Family
zfNLRR shows homology to LRR-proteins from differ-nt vertebrate species (Fig. 1). We analyzed their se-uence in more detail and found in all of them, as infNLRR, an additional Ig-like domain and a putativeN type III repeat adjacent to the membrane. Thistrongly supports the hypothesis that they constitute aamily of neuronal adhesion or receptor proteins.
NLRR-1, xNLRR-1, and a human EST seem to beomologues, based on a similarity of 86% on the aminocid level. Likewise, mNLRR-3 and hNLRR are 90%imilar, whereas their similarity to mNLRR-1 andNLRR-1 is only 60–64% (Fig. 1C). zfNLRR shows annly 52–64% similarity to the so far identified NLRRroteins indicating that it constitutes a novel member of
he family. It remains to be elucidated whether theifferent members of the NLRR family bind to distinct
igands.Within the intracellular region, we found a stretch of
1 aa that is present in all NLRR proteins. Contained inhis box are two mammalian clathrin mediated endocy-osis motifs, a tyrosine-based signal conforming to thexxØ motif (Chen et al., 1990; Collawn et al., 1990), and aileucine-type motif (Letourneur and Klausner, 1992;
d RGCs. (A–F) Optic nerves were crushed at 0 h, 0 days and retinasn the upper right corner). RGCs are the only cells in the retina thatr nuclear layer; ipl, inner plexiform layer; onl, outer nuclear layer; opl,
sione
ieters et al., 1993; Bremnes et al., 1994). Preliminaryesults indicate that ectopically expressed zfNLRR is
erms1ra1se
o(1pzesa1emaami1ttpq
zR
RuazwlhcomrrScgT
old
tsitzrtbiasfa(rtconmmtrsRtoediwimCn
E
D
dfuSt
Regenerating CNS Neurons Express zfNLRR 175
ndocytosed in COS cells (not shown). Endocytosis andecycling mechanisms are relevant for cell adhesion
olecules, such as integrins during cell migration (Law-on and Maxfield, 1995; Lauffenburger and Horwitz,996). The L1 subfamily of cell adhesion molecules hasecently been reported to be endocytosed preferentiallyt the rear of axonal growth cones (Kamiguchi et al.,998). This may be a means to regulate the adhesivetrength during neuronal growth in response to thextracellular environment.
mNLRR-1 and mNLRR-3 are the only other membersf the family that have been studied in more detailTaguchi et al., 1996; Taniguchi et al., 1996; Ishii et al.,996). They are distinctly regulated and the expressionatterns of both genes differ from the one for zfNLRR.fNLRR mRNA is found at low levels in developingmbryos, where expression is restricted to a few sensoryystems (Fig. 3). In contrast, mNLRR-1 and mNLRR-3re widely expressed in embryonic mice (Taguchi et al.,996; Taniguchi et al., 1996). Some overlapping areas ofxpression can be found. In the adult brain, zfNLRR andNLRR-1 are expressed in the cerebellum and zfNLRR
nd mNLRR-3 mRNA are both found in the embryonicnd adult olfactory system. Northern blot analysis ofNLRR-1 in adult brain revealed two strong bands,
ndicating alternative splicing events (Taguchi et al.,996). A single transcript was found for zfNLRR in allissues examined (Fig. 2). Once the binding partners ofhe NLRR proteins have been determined it will becomeossible to better understand the functional conse-uences of the individual expression patterns.
fNLRR mRNA Levels Are Increased inegenerating Neurons
In the adult zebrafish, lesioning the optic nerve causesGCs to regenerate. Several genes are transiently upreg-lated during this period (Perry et al., 1987; Stuermer etl., 1992; Bernhardt et al., 1996; Bormann et al., 1998).fNLRR mRNA levels start to increase coordinatelyith these well known growth-associated genes in
esioned RGCs (Fig. 5). However, virtually all RGCsave decreased their zfNLRR mRNA level back toontrol levels at 10 days after lesion, at a time whenther growth-associated genes are still elevated (Bor-ann et al., 1998). With reference to the time course of
egeneration in goldfish, where regenerating RGC axonseach the tectum 10–13 days after lesion (Murray, 1976;tuermer and Easter, 1984), it seems that lesioned RGCsease to express zfNLRR at, or shortly prior to, synapto-
enesis, suggesting a role for zfNLRR during this event.o what extent the difference in the temporal regulationap
f the different neuronal growth-associated genes afteresion relies on distinct functions remains to be eluci-ated.The association of zfNLRR expression and regenera-
ion in the adult is further substantiated by the fact thateveral brain nuclei with descending projections stronglyncrease zfNLRR mRNA levels after spinal cord transec-ion. Interestingly, the same neurons did not expressfNLRR at detectable levels in the embryo. Similarly,eggie-1 and reggie-2, two membrane-associated pro-eins, have been reported to be present in regenerating,ut not embryonic RGCs (Schulte et al., 1997). This
ndicates that regenerating neurons in the adult expresst least two set of genes. One set includes the well-tudied neuronal growth-associated proteins that areound in virtually all neurons during neurite outgrowthnd axon elongation, in the embryo and in the adultSkene, 1989). A second set contains genes that are eitherequired in only a subset of axon-extending neurons orhat allow growing neurons to reach their targets specifi-ally in the adult environment after lesion. The findingf genes that are selectively expressed in regeneratingeurons indicates that regeneration in the adult is not aere recapitulation of embryonic neuronal develop-ent (Aubert et al., 1995). Interestingly, several genes
hat were found to be coordinately regulated duringetinal regeneration in zebrafish, showed distinct expres-ion patterns in the embryo (L. Roth, P. Bormann, and E.einhard, unpublished observations). Therefore, it is
empting to assume that nerve lesion leads to activationf common regulatory elements in their promoter/nhancer region. It remains to be elucidated whetherifferent regulatory elements induce gene transcription
n the embryo and during regeneration in the adult andhether these are conserved. This is of particular
nterest in respect to the fact that in contrast to fish,ammals are not able to functionally regenerate afterNS-lesions and concomitantly do not reexpress neuro-al growth-associated genes.
XPERIMENTAL METHODS
ifferential Display
Differential display was performed as previouslyescribed (Liang and Pardee, 1992). RNA was isolated
rom regenerating and control retinas 3 days after lesionsing the RNeasy Total RNA kit (Qiagen AG, Basel,witzerland). Different primer combinations were usedo perform the PCR reaction. The primers 58-T GC-38
14nd 58-AGCCCTCAGA-38 resulted in an amplificationroduct from the 38 end of the zfNLRR cDNA. The
ecccTt
AGmcsPMm(
Ta
sipvsrt
z(1Iss1tndaNa3foi3crsa
2crmsmciwbetmspb(
Z
mAmnawwwacitemtDknEoas
I
t
176 Bormann et al.
xcised band was reamplified by PCR and subcloned. Tolone a full-length zfNLRR cDNA, a Lambda ZAPIIDNA library was generated from regenerating andontrol adult zebrafish retinas (Stratagene, La Jolla, CA).wo overlapping partial clones were linked by PCR andhe full-length cDNA subcloned.
The sequencing reactions were performed using anBI Prism dye terminator sequencing kit (No. 402080), aeneAmp 9600 thermal cycler and an ABI 373A Auto-ated sequencer (all Perkin–Elmer Corp., CA). All
lones were sequenced on both strands. Sequence analy-is was done with the Wisconsin Sequence Analysisackage VMS version 8.0 (Genetics Computer Group,adison, WI), including the Gap program for align-ents and the BLAST network service of the NCBI
National Center for Biotechnology Information, U.S.A.).
ransfection of COS Cells, Metabolic Labeling,nd Immunoprecipitation
zfNLRR cDNA, without the sequence coding for theignal peptide (starting at amino acid 25), was clonednto the expression vector 5myc-pcDNA1 (generouslyrovided by Dr. Graham Jones, Dept. Physiology, Uni-ersity of Basel, Basel, Switzerland), which contains aignal sequence and five myc epitopes 58 to the uniqueestriction site ClaI. The resulting fusion protein con-ains five extracellular myc epitopes.
To determine posttranslational modifications offNLRR, COS-7 cells (Gluzman, 1981) cultured in DMEMGibco BRL, Gaithersburg, MD), supplemented with0% fetal calf serum (FCS), 10 mM sodium pyruvate, 10U/l penicillin, and 100 mg/l streptomycin, were tran-iently transfected with a zfNLRR/myc fusion con-truct, using the DEAE-dextran based method (Cullen,987) at subconfluency. Forty-eight hours after transfec-ion, the cells were incubated in labeling media (methio-ine-free, glutamine-free MEM, supplemented with 10%ialyzed FCS; Gibco BRL, Gaithersburg, MD) for 15 mint 37°C. Then 50 µCi [35S]methionine (New Englanduclear, Boston, MA) per 35-mm tissue culture dish was
dded and the metabolic labeling carried out for 3 h at7°C. For pulse–chase experiments, the cells were pulsedor 15 min at 37°C with 50 µCi [35S]methionine, washednce with phosphate-buffered saline (PBS), and chased
n COS culture medium for different time-periods at7°C. All subsequent steps were carried out on ice. Theells were washed twice with PBS, scraped off with aubber policeman in 1 ml solubilization buffer (100 mM
odium phosphate, pH 8.0; 1% Triton X-100; 0.1% NaN3;nd 40 µg/ml PMSF), and passed five times through aPf
5-gauge needle. After incubation for 1 h the lysate wasentrifuged at 105,000g for 1 h, at 4°C. Protein A–Sepha-ose CL-4B beads (Pharmacia, Uppsala, Sweden; approxi-
ately 12 mg/35-mm dish) were washed with 100 mModium phosphate, pH 8.0, 0.1% NaN3, then 2.5 µl of aonoclonal mouse anti-myc antibody (9E10; 103 con-
entrated; Evan et al., 1985) was added to the beads andncubated for 2 h at room temperature (RT). The beads
ere rinsed twice with solubilization buffer and incu-ated with the resulting cell supernatant for 1.5 h on annd-over-end shaker at 4°C. The beads were washed 4imes with solubilization buffer, rinsed once with 100
M sodium phosphate, pH 8.0, and once with 10 mModium phosphate, pH 8.0. Protein was eluted from therotein A beads by boiling in electrophoresis sampleuffer and separated on a 7.5% SDS–polyacrylamide gelLaemmli, 1970).
ebrafish Maintenance and Surgery
Zebrafish were obtained from the Oregon AB line andaintained as described previously (Westerfield, 1995).dult zebrafish were anesthetized by immersion in 150g/liter MS222 (Sigma, Buchs, Switzerland). The optic
erve was exposed by cutting the skin over the left eyend the eye was slightly pushed to the side. The nerveas crushed near the posterior wall of the orbit withatchmaker’s forceps. The completeness of the axotomyas assured by the translucent appearance of the nerve
t the crush site. Fish that bled were not used. Theontralateral optic nerve of each fish was left intact withts retina serving as a control. For spinal cord transec-ions, a longitudinal incision was made dorsally toxpose the vertebral column that was then cut approxi-ately 3–4 mm caudal to the brainstem/spinal cord
ransition zone. The wounds were sealed with Tissucoluo S (Immuno AG, Vienna, Austria) and the fish wereept at 28°C. To visualize the cell bodies of lesionedeurons CellTracker Orange CMTMR (Molecular Probes,ugene, OR) was applied to the lesion site of someperated fish. Brains were removed for analysis 5 daysfter lesion. Embryos were obtained as previously de-cribed (Westerfield, 1995).
n Situ Hybridization
Digoxigenin-labeled riboprobes for in situ hybridiza-ion were generated within the coding region of zfNLRR.
robes were prepared according to the instructionsrom the DIG labeling kit (Boehringer-Mannheim,
Mtiratss
wb(iwo0tSPaMlmihta5aPWassgcdconMmy
isstPt
WS
abS1ie(ni
N
iSMabTQs14fNUJa(f0pfcfa
A
cW
Regenerating CNS Neurons Express zfNLRR 177
annheim, Germany). Template DNA was removed byreatment with DNase (Promega, Catalys AG, Wall-sellen, Switzerland). Unincorporated nucleotides wereemoved by using purification columns from the RNe-sy Total RNA kit (Qiagen AG, Basel, Switzerland). Theranscripts were eluted in 30 µl H2O; hybridizationolution was added to 100 µl (Westerfield, 1995) andtored at 220°C until use.
In situ hybridization studies were performed onhole mount zebrafish embryos, retinas and brains, and
rain sections following the protocol of Westerfield1995). Briefly, eyes were removed and the retinassolated in PBS under a dissecting microscope. Embryos
ere dechorionated. All tissues were fixed in 4% PFAvernight at 4°C. After several washes in PBS containing.1% Tween 20 (PBST), the specimens were transferredo methanol and left at 220°C for at least 30 min.amples were rehydrated in a graded series of methanol/BST, washed in PBST, and refixed in 4% PFA. Protein-se K-treatment (10 µg/ml; Boehringer-Mannheim,annheim, Germany) was performed for 10 min or
onger (up to 50 min for older embryos) at RT. Speci-ens were fixed again in 4% PFA, followed by prehybrid-
zation for 2–4 h at 55°C in hybridization solution andybridized overnight at 55°C using a 1:100 dilution of
he digoxigenin-labeled probe (see above). Washes weret a maximal stringency of 0.23 SSC; 0.1% Tween 20 at5°C. The hybridized probes were visualized usingnti-digoxigenin-AP Fab fragments (diluted 1:2000 inBST; Boehringer-Mannheim), following the protocol ofesterfield (1995). Times for color development, chosen
ccording to the probes used and the developmentaltage of the embryos, ranged from 4 to 8 h. Thepecimens were fixed in 4% PFA and either cleared in araded series of glycerol/PBS for whole mount micros-opy or cryoprotected in 20% Sucrose/PBS and embed-ed in OCT Tissue-tek (Miles Inc., Elkhart, IN) forryosectioning. Then 12-µm sections were cut, mountedn slides, and examined by microscopy. To visualizeuclei, Hoechst dye (bisbenzimide H33258, Boehringer-annheim; 1 mg/ml in PBS) was included in theounting media. Cleared embryos were freed from the
olk and mounted in glycerol.This protocol was slightly changed for in situ hybrid-
zation on brain section. Fresh brain were frozen, cryo-ectioned (12–14 µm), and mounted on precoated coverlides (SuperFrost/Plus; Menzel-Glaser; Germany). Sec-ions were fixed at RT in 4% PFA for 20 min, washed in
BST, and prehybridized. Subsequent steps followedhe protocol described above.
wtF
hole Mount Immunohistochemicaltaining of Embryos
Embryos were pretreated for immunohistochemistrys described in Westerfield (1995). The primary anti-ody (mouse anti-acetylated tubulin, Sigma, Buchs,witzerland) was diluted 1:1000 in BDP 1 N (1% BSA;% DMSO in PBSTx 1 10% normal sheep serum) andncubated for 3 h at RT. After several washes, thembryos were incubated with the secondary antibodysheep anti-mouse IgAP; Boehringer-Mannheim) over-ight at 4°C and the alkaline phosphatase was visual-
zed as described in Westerfield (1995).
orthern Blot Hybridization
Northern blot hybridization was performed follow-ng the protocols of Sambrook et al. (1989) and the ‘‘DIGystem User Guide for Filter Hybridization’’ (Boehringer-annheim). mRNA was isolated from retinas, embryos
t different developmental stages, and adult zebrafishrain, according to the instructions given in the RNeasyotal RNA kit together with the Oligotex mRNA kit (alliagen AG, Basel, Switzerland). The transcripts were
eparated in a formaldehyde gel. mRNA derived from0 µg of total RNA from retina, 125 µg from embryo, and5 µg from brain was loaded. The mRNA was trans-erred to a nylon membrane (Schleicher Schuell, NY; 13, Keene, NH), using a pressure blotter (Posiblot) andV crosslinked (Stratalinker 2400; both Stratagene, La
olla, CA). Hybridization with DIG-labeled probes (seebove) was performed at 68°C overnight in a HybritubeGibco BRL, Life Technologies AG, Basel, Switzerland)ollowed by several washes at a maximal stringency of.13 SSC; 0.5% SDS at the same temperature. Therobes were detected using anti-digoxigenin-AP Fab
ragments, diluted 1:108000 in Boehringer buffer and thehemoluminescence reagent CDP-Star (TROPIX, Bed-ord, MA). Exposure was 2 min for regenerating retinand 20 min for adult brain, control retina, and embryo.
CKNOWLEDGMENTS
We are grateful to Dr. H. P. Hauri and Mrs. K. Bucher for suggestionsoncerning the metabolic labeling and immunoprecipitation assays.e thank Dr. M. A. Ruegg for many valuable comments. This work
as supported by a grant from the Swiss National Science Founda-ion, the Theodore OTT Foundation, and the ‘‘Sandoz-Stiftung zurorderung der Medizinisch-biologischen Wissenschaften.’’
R
A
B
B
B
B
B
B
C
C
C
C
C
E
F
F
F
G
H
H
H
H
I
I
K
K
K
K
L
L
L
L
L
M
M
M
N
N
N
178 Bormann et al.
EFERENCES
ubert, I., Ridet, J. L., and Gage, F. H. (1995). Regeneration in the adultmammalian CNS: Guided by development. Curr. Opin. Neurobiol. 5:625–635.
ecker, T., Wullimann, M. F., Becker, C. G., Bernhardt, R. R., andSchachner, M. (1997). Axonal regrowth after spinal cord transectionin adult zebrafish. J. Comp. Neurol. 377: 577–595.
ernhardt, R. R., Tongiorgi, E., Anzini, P., Schachner, M., Bernhardt,R. R., Tongiorgi, E., Anzini, P., and Schachner, M. (1996). Increasedexpression of specific recognition molecules by retinal ganglion cellsand by optic pathway glia accompanies the successful regenerationof retinal axons in adult zebrafish. J. Comp. Neurol. 376: 253–264.
lack, M. M., and Keyser, P. (1987). Acetylation of alpha-tubulin incultured neurons and the induction of alpha-tubulin acetylation inPC12 cells by treatment with nerve growth factor. J. Neurosci. 7:1833–1842.
ormann, P., Zumsteg, V. M., Roth, L. W., and Reinhard, E. (1998).Target contact regulates GAP-43 and alpha-tubulin mRNA levels inregenerating retinal ganglion cells. J. Neurosci. Res. 52: 405–419.
remnes, B., Madsen, T., Gedde-Dahl, M., and Bakke, O. (1994). An LIand ML motif in the cytoplasmic tail of the MHC-associatedinvariant chain mediate rapid internalization. J. Cell Sci. 107:2021–2032.
rummendorf, T., and Rathjen, F. G. (1996). Structure/function relation-ships of axon-associated adhesion receptors of the immunoglobulinsuperfamily. Curr. Opin. Neurobiol. 6: 584–593.
arnemolla, B., Leprini, A., Borsi, L., Querze, G., Urbini, S., and Zardi,L. (1996). Human tenascin-R. Complete primary structure, pre-mRNA alternative splicing and gene localization on chromosome1q23-q24. J. Biol. Chem. 271: 8157–8160.
aroni, P. (1997). Intrinsic neuronal determinants that promote axonalsprouting and elongation. Bioessays 19: 767–775.
hen, W. J., Goldstein, J. L., and Brown, M. S. (1990). NPXY, a sequenceoften found in cytoplasmic tails, is required for coated pit-mediatedinternalization of the low density lipoprotein receptor. J. Biol. Chem.265: 3116–3123.
ollawn, J. F., Stangel, M., Kuhn, L. A., Esekogwu, V., Jing, S. Q.,Trowbridge, I. S., and Tainer, J. A. (1990). Transferrin receptorinternalization sequence YXRF implicates a tight turn as the struc-tural recognition motif for endocytosis. Cell 63: 1061–1072.
ullen, B. R. (1987). Use of eukaryotic expression technology in thefunctional analysis of cloned genes. Methods Enzymol. 152: 684–704.
van, G. I., Lewis, G. K., Ramsay, G., and Bishop, J. M. (1985). Isolationof monoclonal antibodies specific for human c-myc proto-oncogeneproduct. Mol. Cell. Biol. 5: 3610–3616.
awcett, J. W. (1992). Intrinsic neuronal determinants of regeneration.Trends Neurosci. 15: 5–8.
ields, R. D., and Itoh, K. (1996). Neural cell adhesion molecules inactivity-dependent development and synaptic plasticity. TrendsNeurosci. 19: 473–480.
isher, L. W., Heegaard, A. M., Vetter, U., Vogel, W., Just, W., Termine,J. D., and Young, M. F. (1991). Human biglycan gene. Putativepromoter, intron-exon junctions, and chromosomal localization. J.Biol. Chem. 266: 14371–14377.luzman, Y. (1981). SV40-transformed simian cells support the replica-tion of early SV40 mutants. Cell 23: 175–182.alfon, M. S., Hashimoto, C., and Keshishian, H. (1995). The Dro-sophila toll gene functions zygotically and is necessary for proper
motoneuron and muscle development. Dev. Biol. 169: 151–167.arpaz, Y., and Chotia, C. (1994). Many of the immunoglobulinsuperfamily domains in cell adhesion molecules and surface recep- Ntors belong to a new structural set which is close to that containingvariable domains. J. Mol. Biol. 238: 528–539.ashimoto, C., Gerttula, S., and Anderson, K. V. (1991). Plasmamembrane localization of the Toll protein in the syncytial Dro-sophila embryo: importance of transmembrane signaling for dorsal-ventral pattern formation. Development 111: 1021–1028.ocking, A. M., Shinomura, T., and McQuillan, D. J. (1998). Leucine-rich repeat glycoproteins of the extracellular matrix. Matrix Biol. 17:1–19.
shii, N., Wanaka, A., and Tohyama, M. (1996). Increased expression ofNLRR-3 mRNA after cortical brain injury in mouse. Brain Res. Mol.Brain Res. 40: 148–152.
zsvak, Z., Ivics, Z., Garcia-Estefania, D., Fahrenkrug, S. C., andHackett, P. B. (1996). DANA elements: A family of composite,tRNA-derived short interspersed DNA elements associated withmutational activities in zebrafish (Danio rerio). Proc. Natl. Acad. Sci.USA 93: 1077–1081.
amiguchi, H., Long, K. E., Pendergast, M., Schaefer, A. W., Rapoport,I., Kirchhausen, T., and Lemmon, V. (1998). The neural cell adhesionmolecule L1 interacts with the AP-2 adaptor and is endocytosed viathe clathrin-mediated pathway. J. Neurosci. 18: 5311–5321.
obe, B., and Deisenhofer, J. (1994). The leucine-rich repeat: A versatilebinding motif. Trends Biochem. Sci. 19: 415–421.
obe, B., and Deisenhofer, J. (1995). Proteins with leucine-rich repeats.Curr. Opin. Struct. Biol. 5: 409–416.
rantz, D. E., and Zipursky, S. L. (1990). Drosophila chaoptin, amember of the leucine-rich repeat family, is a photoreceptor cell-specific adhesion molecule. Embo J. 9: 1969–1977.
aemmli, U. K. (1970). Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 227: 680–685.
auffenburger, D. A., and Horwitz, A. F. (1996). Cell migration: Aphysically integrated molecular process. Cell 84: 359–369.
awson, M. A., and Maxfield, F. R. (1995). Ca(21)- and calcineurin-dependent recycling of an integrin to the front of migratingneutrophils. Nature 377: 75–79.
etourneur, F., and Klausner, R. D. (1992). A novel di-leucine motif anda tyrosine-based motif independently mediate lysosomal targetingand endocytosis of CD3 chains. Cell 69: 1143–1157.
iang, P., and Pardee, A. B. (1992). Differential display of eukaryoticmessenger RNA by means of the polymerase chain reaction. Science257: 967–971.artini, R. (1994). Expression and functional roles of neural cellsurface molecules and extracellular matrix components duringdevelopment and regeneration of peripheral nerves. J. Neurocytol.23: 1–28.eakin, S. O., and Shooter, E. M. (1992). The nerve growth factorfamily of receptors. Trends Neurosci. 15: 323–331.urray, M. (1976). Regeneration of retinal axons into the goldfishoptic tectum. J. Comp. Neurol. 168: 175–195.agasawa, A., Kubota, R., Imamura, Y., Nagamine, K., Wang, Y.,Asakawa, S., Kudoh, J., Minoshima, S., Mashima, Y., Oguchi, Y., andShimizu, N. (1997). Cloning of the cDNA for a new member of theimmunoglobulin superfamily (ISLR) containing leucine-rich repeat(LRR). Genomics 44: 273–279.eedleman, S. B., and Wunsch, C. D. (1970). A general methodapplicable to the search for similarities in the amino acid sequenceof two proteins. J. Mol. Biol. 48: 443–453.elson, R. E., Fessler, L. I., Takagi, Y., Blumberg, B., Keene, D. R.,Olson, P. F., Parker, C. G., and Fessler, J. H. (1994). Peroxidasin: A
novel enzyme-matrix protein of Drosophila development. Embo J.13: 3438–3447.ose, A., Mahajan, V. B., and Goodman, C. S. (1992). Connectin: a
P
P
R
R
S
S
S
S
S
S
S
S
S
T
T
W
W
W
W
W
W
Regenerating CNS Neurons Express zfNLRR 179
homophilic cell adhesion molecule expressed on a subset of musclesand the motoneurons that innervate them in Drosophila. Cell 70:553–567.
erry, G. W., Burmeister, D. W., and Grafstein, B. (1987). Fast axonallytransported proteins in regenerating goldfish optic axons. J. Neuro-sci. 7: 792–806.
ieters, J., Bakke, O., and Dobberstein, B. (1993). The MHC classII-associated invariant chain contains two endosomal targetingsignals within its cytoplasmic tail. J. Cell Sci. 106: 831–846.
ose, D., Zhu, X., Kose, H., Hoang, B., Cho, J., and Chiba, A. (1997).Toll, a muscle cell surface molecule, locally inhibits synapticinitiation of the RP3 motoneuron growth cone in Drosophila.Development 124: 1561–1571.
othberg, J. M., Jacobs, J. R., Goodman, C. S., and Artavanis-Tsakonas,S. (1990). slit: an extracellular protein necessary for development ofmidline glia and commissural axon pathways contains both Egf andLrr domains. Genes Dev. 4: 2169–2187.
ambrook, J., Fritsch, E. F., and Maniatis, T. (1989). Molecular Cloning: ALaboratory Manual, 2nd ed. Cold Spring Harbor Laboratory Press,Cold Spring Harbor, New York.
chneider, R., and Schweiger, M. (1991). A novel modular mosaic ofcell adhesion motifs in the extracellular domains of the neurogenictrk and trkB tyrosine kinase receptors. Oncogene 6: 1807–1811.
chulte, T., Paschke, K. A., Laessing, U., Lottspeich, F., and Stuermer,C. A. (1997). Reggie-1 and reggie-2, two cell surface proteinsexpressed by retinal ganglion cells during axon regeneration.Development 124: 577–587.
hishido, E., Takeichi, M., and Nose, A. (1998). Drosophila synapseformation: Regulation by transmembrane protein with leu-richrepeats, CAPRICIOUS. Science 280: 2118–2121.
kene, J. H. (1989). Axonal growth-associated proteins. Annu. Rev.Neurosci. 12: 127–156.
perry, R. W. (1948). Patterning of central synapses in regeneration ofthe optic nerve in teleosts. Physiol. Zool. 21: 351–361.
tuermer, C. A., Bastmeyer, M., Bahr, M., Strobel, G., and Paschke, K.(1992). Trying to understand axonal regeneration in the CNS of fish. W
tuermer, C. A., and Easter, S. S., Jr. (1984). A comparison of the normaland regenerated retinotectal pathways of goldfish. J. Comp. Neurol.223: 57–76.
uzuki, Y., Sato, N., Tohyama, M., Wanaka, A., and Takagi, T. (1996).cDna cloning of a novel membrane glycoprotein that is expressedspecifically in glial cells in the mouse brain. Lig-1, a protein withleucine-rich repeats and immunoglobulin-like domains. J. Biol.Chem. 271: 22522–22527.
aguchi, A., Wanaka, A., Mori, T., Matsumoto, K., Imai, Y., Tagaki, T.,and Tohyama, M. (1996). Molecular cloning of novel leucine-richrepeat proteins and their expression in the developing mousenervous system. Brain Res. Mol. Brain. Res. 35: 31–40.
aniguchi, H., Tohyama, M., and Takagi, T. (1996). Cloning andexpression of a novel gene for a protein with leucine-rich repeats inthe developing mouse nervous system. Brain Res. Mol. Brain Res. 36:45–52.esterfield, M., McMurray, J. V., and Eisen, J. S. (1986). Identifiedmotoneurons and their innervation of axial muscles in the zebrafish.J. Neurosci. 6: 2267–2277.esterfield, M. (1995). The Zebrafish Book. A Guide for the Laboratory Useof Zebrafish (Danio rerio), 3rd ed., University of Oregon Press,Eugene.illiams, A. F., and Barclay, A. N. (1988). The immunoglobulinsuperfamily—Domains for cell surface recognition. Annu. Rev.Immunol. 6: 381–405.ilson, S. W., Ross, L. S., Parrett, T., and Easter, S. S., Jr. (1990). Thedevelopment of a simple scaffold of axon tracts in the brain of theembryonic zebrafish, Brachydanio rerio. Development 108: 121–145.indisch, J. M., Auer, B., Marksteiner, R., Lang, M. E., and Schneider,R. (1995). Specific neurotrophin binding to leucine-rich motif pep-tides of TrkA and TrkB. FEBS Lett. 374: 125–129.oodhams, P. L., Calvert, R., and Dunnett, S. B. (1989). Monoclonalantibody G10 against microtubule-associated protein 13 distin-guishes between growing and regenerating axons. Neuroscience 28:
49–59.ullimann, M. F., Rupp, B., and Reichert, H. (1996). Neuroanatomy ofJ. Neurobiol. 23: 537–550. the Zebrafish Brain. A Topological Atlas. Birkhauser Verlag, Basel.
Received December 14, 1998Revised February 11, 1999
Accepted February 18, 1999