De elopmen /Plas ici y/Repai
Regula ion o Gephy in Clus e Size and Inhibi o y Synap ic
Cu en s on Renshaw Cells by Mo o Axon Exci a o y Inpu s
Da id Gonzalez-Fo e o,
1,2
Angel M. Pas o ,
2
E ic J. Geiman,
1
Bea iz Benı´ ez-Temin˜o,
2
and F ancisco J. Al a ez
1
1
Depa men o Ana omy and Physiology, W igh S a e Uni e si y, Day on, Ohio 45435, and
2
Depa amen o de Fisiologı´a y Zoologı´a, Facul ad de Biologı´a,
Uni e sidad de Se illa, 41012-Se illa, Spain
Renshaw cells ecei e a high densi y o inhibi o y synapses cha ac e ized by la ge pos synap ic gephy in clus e s and mixed glycine gic/
GABAe gic inhibi o y cu en s wi h la ge peak ampli udes and long decays. These p ope ies appea adap ed o inc ease inhibi o y
e icacy o e Renshaw cells and ma u e pos na ally by mechanisms ha a e unknown. We es ed he hypo hesis ha he e osynap ic
in luences om exci a o y mo o axon inpu s modula e he de elopmen o inhibi o y synapses on Renshaw cells. Thus, e anus (TeNT)
and bo ulinum neu o oxin A (BoNT-A) we e injec ed in amuscula ly a pos na al day 5 (P5) o, espec i ely, ele a e o educe mo o
axon i ing ac i i y o ⬃2 weeks. A e TeNT injec ions, he a e age gephy in clus e a eas on Renshaw cells inc eased by 18.4% a P15
and 28.4% a P20 and dec eased a e BoNT-A injec ions by 17.7% a P15 and 19.9% a P20. The a e age size di e ences esul ed om
changes in he p opo ions o small and la ge gephy in clus e s. Whole-cell eco dings in P9–P15 Renshaw cells a e P5 TeNT injec ions
showed inc eases in he peak ampli ude o glycine gic minia u e pos synap ic cu en s (mPSCs) and he as componen o mixed
(glycine gic/GABAe gic) mPSCs compa ed wi h con ols (60.9% and 78.9%, espec i ely). GABAe gic mPSCs inc eased in peak ampli-
ude o a smalle ex en (45.8%). Howe e , because o he compa a i ely longe decays o synap ic GABAe gic cu en s, o al cu en
ans e changes a e TeNT we e simila o synap ic glycine and GABA
A
ecep o s (56 s 48.9% inc eases, espec i ely). We concluded
ha mo o axon exci a o y synap ic ac i i y modula es he de elopmen o inhibi o y synapse p ope ies on Renshaw cells, in luencing
ec ui men o pos synap ic gephy in and glycine ecep o s and, o lesse ex en , GABA
A
ecep o s.
Key wo ds: mo oneu ons; de elopmen ; spinal co d; bo ulinum oxin; e anus oxin; GABA
A
ecep o ; glycine ecep o ; ecu en
inhibi ion
In oduc ion
Inhibi o y synapses modula e exci a o y inpu s and shape neu-
onal i ing. The e o e, inhibi o y synap ic p ope ies mus be se
acco ding o exci a o y inpu s eng hs and he i ing ac i i y o
indi idual neu ons. Howe e , he inhibi o y synapse p ope ies
ha a e in luenced by local exci a o y ac i i y a e poo ly known.
One impo an ea u e ha a ec s inhibi o y synap ic s eng h is
ecep o ec ui men , and his in u n depends on pos synap ic
densi y (PSD) s uc u e (Nusse e al., 1997, 1998; Lim e al.,
1999; Oleske ich e al., 1999). Indeed, we p oposed ha he size
o pos synap ic clus e s o gephy in, a key sca old p o ein o he
inhibi o y PSD (T ille e al., 1985; Ki sch e al., 1993; Ki sch and
Be z, 1998), co ela es wi h pos synap ic glycine gic cu en am-
pli udes (Lim e al., 1999; Van Zunde e al., 2004). In mos
neu ons, la ge gephy in clus e s occu in dis al dend i es; how-
e e , Renshaw cells exhibi la ge p oximal gephy in clus e s ha
colocalize glycine and GABA
A
ecep o s (Al a ez e al., 1997;
Geiman e al., 2002a). Acco dingly, glycine and GABAe gic syn-
ap ic cu en s in Renshaw cells a e la ge han in o he spinal
in e neu ons (Gonzalez-Fo e o and Al a ez, 2005). The mecha-
nisms ha ma u e la ge gephy in clus e s and glycine/GABAe gic
cu en s in Renshaw cells a e unknown.
Renshaw cells p o ide ecu en inhibi ion o mo oneu ons
and modula e mo oneu on ec ui men , i ing a e, and p op i-
ocep i e e lexes. The main exci a o y inpu o Renshaw cells
a ises om choline gic in aspinal ecu en colla e als o mo o
axons (Renshaw, 1946; Eccles e al., 1954; Al a ez e al., 1999) and
esul s in igo ous long-las ing pos synap ic esponses ha e oke
sus ained i ing wi h ini ial ins an aneous equencies o ⬃1500
Hz (Eccles e al., 1954). E en single ac ion po en ials p opaga ing
in indi idual mo o axons induce high- equency discha ges in
Renshaw cells in he anes he ized ca spinal co d ( an Keulen,
1981). No o he spinal in e neu on shows simila synap ic e-
sponses o mo o axons, and hus he speci ic p esence o powe -
ul inhibi o y synapses on Renshaw cells sugges s a ela ionship
wi h he unique p ope ies o hei exci a o y ac i i y. Inhibi o y
synap ic p ope ies on Renshaw cells de elop du ing he i s
h ee pos na al weeks, in conjunc ion wi h he onse o locomo-
ion and he ma u a ion o en al ho n mo o ci cui s (Geiman
e al., 2000; Gonzalez-Fo e o and Al a ez, 2005). These obse a-
ions led o he sugges ion ha inhibi o y synapse de elopmen
Recei ed July 21, 2003; e ised No . 19, 2004; accep ed No . 20, 2004.
This wo k was suppo ed by Na ional Science Founda ion G an 9984441 o F.J.A. and Minis e io de Ciencia y
Tecnolog´ia–Fondo Eu opeo de Desa ollo Regional G an BFI2003-01024 o A.M.P. D.G.-F. was suppo ed by a
pos doc o al ellowship om he Minis e io de Educacio´n, Cul u a y Depo e (Spain). We hank D s. R. E. W. Fy e, J.
S e n, J. Dean, and S. P. Schneide o help ul sugges ions du ing he pe o mance o his wo k and D s. T. Cope and
B. Walmsley o help ul sugges ions on p e ious e sions o his manusc ip .
Co espondenceshouldbeadd essed oD .F anciscoJ.Al a ez,Depa men o Ana omyandPhysiology,W igh
S a e Uni e si y, 3640 Colonel Glenn Highway, Day on, OH 45435. E-mail: [email p o ec ed].
DOI:10.1523/JNEUROSCI.3725-04.2005
Copy igh © 2005 Socie y o Neu oscience 0270-6474/05/250417-13$15.00/0
The Jou nal o Neu oscience, Janua y 12, 2005 •25(2):417–429 • 417
on Renshaw cells could be modula ed by ac i i y in hei mo o
axon synap ic inpu s (Geiman e al., 2000). To es his hypo he-
sis, we de eloped an in i o expe imen al model ha pe mi ed
ch onic al e a ion o mo oneu on i ing in neona es. We in-
c eased and educed mo o ac i i y, espec i ely, wi h e anus
neu o oxin (TeNT) and bo ulinum neu o oxin se o ype A
(BoNT-A) du ing he pe iod o ma u a ion o inhibi o y syn-
apses on Renshaw cells [ om pos na al day 5 (P5) o P20]. Then,
we analyzed he esul ing gephy in clus e s and inhibi o y synap-
ic cu en s. We show ha changes in mo o axon ac i i y signi -
ican ly al e ed gephy in clus e ing and glycine/GABA
A
synap ic
cu en s. Thus, we concluded ha ac i i y a neighbo ing exci a-
o y synapses in luence he ec ui men o inhibi o y pos synap-
ic sca olding p o eins and ecep o s, sugges ing a mechanism
ha helps ma ch inhibi o y synapses o local exci a o y inpu s.
P elimina y obse a ions we e p esen ed in abs ac o m
(Geiman e al., 2002b; Al a ez e al., 2004).
Ma e ials and Me hods
Toxin injec ions. Wis a a pups o ei he sex ecei ed injec ions in he
gas ocnemius wi h TeNT, BoNT-A, o ehicle (0.9% s e ile saline)
(bo h oxins we e gi s om J. O. Dolly, Impe ial College, London, UK).
A second injec ion in he same muscle con ained 1
␮
l o 1% Fas Blue and
1% Diamidino Yellow (EMS-Polyloy, G ossUms ad , Ge many) and was
used o e og adely label spinal co d sec ions con aining he a ge mo-
o pool.
Injec ions we e made a P5 in he le gas ocnemius muscle unde
e he anes hesia a e exposing he muscle ia a small incision in he skin.
The app op ia e oxin doses o ehicle we e deli e ed using 2
␮
l olume
injec ions in o he muscle belly. Then, he incisions we e closed wi h
Ve -seal (B. B aun Su gical, Emmenb u¨cke, Swi ze land), and he ani-
mals we e e u ned o hei mo he s a e hey ully eco e ed om an-
es hesia. Toxin e ec s, ankle e anic ex ension a e TeNT o ankle laccid
pa alysis a e BoNT-A injec ion, we e moni o ed daily. To quan i a-
i ely analyze he ime cou se o clos idial oxin ac ion on mo o ou pu
and he neu omuscula junc ion (NMJ), some animals we e p epa ed o
eco ding ne e ac i i y o muscle ension a 6 h a e he injec ion o a
1, 2, 4, 6, 8, 10, 12, 14, and 16 d a e he injec ions. Fo mo phological
analysis o gephy in clus e s in calbindin-immuno eac i e (IR) Renshaw
cells, ea ed animals we e killed a P15 o P20. To analyze inhibi o y
cu en s (glycine gic and GABAe gic) in Renshaw cells, we used a spinal
co d slice p epa a ion as desc ibed p e iously (Gonzalez-Fo e o and Al-
a ez, 2005). I is echnically di icul o p ese e in ac mo oneu ons in
his p epa a ion beyond P15, and mo oneu on s abili y seems impo an
o ob aining good quali y Renshaw cell eco dings; hus, we a ge ed
Renshaw cells o he ages P9–P15 o analyze inhibi o y cu en s a e
oxin injec ions. Only animals showing clea oxin e ec s ha s a ed
wi hin he i s 48 h a e injec ion we e used in subsequen mo pholog-
ical o elec ophysiological analyses.
Neona al animals show dec eased sensi i i y o clos idial oxins (Pas-
o e al., 2003), he e o e we i s injec ed di e en doses o BoNT-A
( om 5 o 100 ng/kg) in he gas ocnemius muscle o P5 neona al a s
and obse ed hei e iciency a elici ing speci ic ipsila e al ankle pa alysis
wi h minimal e ec s on o e all animal de elopmen . Su i al a es,
g ow h o he li e s, and pa alysis a he ankle o o he join s ipsila e al
and con ala e al o he injec ion we e no ed o each dose. The inal dose
chosen was 30 ng/kg. A his dose, mos animals su i ed he oxin injec-
ions and showed ob ious laccid pa alysis es ic ed o he injec ed leg.
Pa alysis was no ed by he lack o oo ex ension when ouching he
oo pad and immobili y o he ipsila e al oo when pinching he ail ip
o g abbing he animal by he ail. In addi ion, pa alyzed animals ended
o d ag he a ec ed oo du ing locomo ion. Animals ha ecei ed injec-
ions o lowe doses o oxin showed dec eased le els o pa alysis and
as e eco e ies. Highe doses led o inc eased mo ali y and he appea -
ance o con ala e al and sys emic e ec s.
Fo TeNT, pups ecei ed injec ions o 25 ng/kg a e es ing ini ial
doses anging 15–50 ng/kg o ind su i al a es o ⬎90% o he pups.
Animals ha ecei ed injec ions o 25 ng/kg showed speci ic ipsila e al
e aniza ion o he ankle wi hin 48 h a e injec ion (i.e., he a ec ed oo
is pa alyzed in con inuous ex ension). Te anic ex ension las ed om 8 o
10 d and s a ed o eco e a e wa d.
Animals we e inspec ed daily. Animals showing signs o dis ess we e
emo ed om he s udy. No signi ican di e ences in weigh we e ob-
se ed be ween injec ed animals and li e ma es a he doses used o he
s udy. In addi ion, animals we e disca ded i , a e his ological p ocess-
ing, ew e og adely labeled mo oneu ons we e obse ed.
All animal p ocedu es we e pe o med acco ding o Na ional Ins i-
u es o Heal h guidelines app o ed by he W igh S a e Uni e si y Lab-
o a o y Animal Use Commi ee and in acco dance wi h he Eu opean
Union (86/609/EU) and Spanish legisla ion (BOE 67/8509-12, 1988).
Reco ding o clos idial oxin ac ions on he NMJ. Measu emen s o NMJ
unc ion a e TeNT o BoNT-A injec ions we e ob ained by compa ing
he o ce gene a ed by he gas ocnemius muscle a e di ec s imula ion
o he muscle ibe s o a e ne e s imula ion (indi ec muscle ibe
s imula ion). Ra pups we e decapi a ed, and he le leg was apidly
dissec ed ou in cold K eb’s solu ion (in mM: 113 NaCl, 4.5 KCl, 1
MgSO
4
, 2 CaCl
2
, 11 glucose, 1 Na
2
HPO
4
, and 25 NaHCO
3
). The calca-
neal endon o he gas ocnemius muscle and he calcaneous bone we e
de ached om he ankle. O he ankle ex enso s we e de ached om he
calcaneous bone and dene a ed so ha hey would no in e e e wi h
gas ocnemius con ac ion. The dissec ed leg was ans e ed o he e-
co ding ba h, and he ibia was pinned down o a Sylga d base. The
calcaneal endon was secu ed o a s ain gauge ansduce (Dynamome-
e UF1; Pioden Con ols) unde isome ic condi ions and coupled o an
ampli ie (NL108; Neu oLog Sys em; Digi ime , Welwyn Ga den Ci y,
He o dshi e, UK). A sil e bipola hook elec ode was used o s imu-
la ion o he ibial ne e. One addi ional bipola s imula ing elec ode
was o ked a ound he muscle belly. Fo ces gene a ed by ne e s imula-
ion we e es ima ed by he a e age o h ee o six s imuli (15 V, 0.2 msec
du a ion; deli e ed a 0.5 Hz). The o ce om di ec muscle s imula ion
was a e aged om he same numbe o s imuli (15 V, 2 msec du a ion;
deli e ed a 0.5 Hz). Fo ce ansduce signals we e digi ized wi h a Mi-
c o1401 MK II analog- o-digi al con e e (5 kHz, 16 bi ; CED, Cam-
b idge, UK) and analyzed wi h Signal A e age (CED).
Reco dings o mo o ne e ac i i y. Animals we e anes he ized wi h
e he (dose o e ec ) and su gically p epa ed o elec oneu og aphic
bila e al eco dings o he ibial ne es. The ibial ne e was exposed and
dissec ed ee o connec i e issue, and a bipola hook elec ode was
shaped a ound he ne e o ensu e gen le con ac o bo h leads. Elec-
odes we e made o 200
␮
m Te lon-insula ed sil e wi e. The elec ode
loca ion was insula ed wi h Pa a ilm and ba hed in mine al oil. The
muscle and skin we e su u ed o hold he elec ode in place, and he ne e
was cu dis al o he eco ding si e. Then, he a pups we e allowed o
sligh ly eco e om anes hesia ( he e is li le ne e ac i i y when deeply
anes he ized). The le el o anes hesia kep he animal elaxed, and pups
did no mo e du ing eco dings. The elec oca diog am was audio mon-
i o ed o luc ua ions in hea a e du ing expe imen s. The a pups
we e gen ly es ained in a small Pe spex box gen ly cushioned wi h
co on so ha he animal adop ed a neu al es ing posi ion wi h he
o elimbs ex ended. Ligh s we e dimmed so no o s a le he animal, and
eco dings we e s a ed. Simul aneous bu s ing ne e ac i i y was e oked
bila e ally by ubbing he ail wi h a glass p obe. Elec oneu og aphic
ac i i y was eco ded AC-coupled in common mode ejec ion using a
NL104 di e en ial ampli ie (Digi ime ) and il e ed om 10 Hz o 10
kHz plus a 50 Hz no ch il e . The signal o con ol and ea ed ne es was
hal -wa e ec i ied, and he a ea unde he cu e was measu ed in se-
lec ed 1 sec pe iods o synch onic bu s s o ac i i y in con ol and ea ed
sides o pai wise compa isons. The pups we e eu hanized a he end o
he expe imen .
Tissue p epa a ion and immunocy ochemical p ocedu es. A P15 and
P20, oxin-injec ed a pups we e ixed by pe usion wi h 4% pa a o -
maldehyde in 0.1 Mphospha e bu e (PB), pH 7.4, ollowed by 30 min
pos ixa ion in he same ixa i e, hen washed and s o ed in 15% suc ose
in 0.1 MPB un il use. Fo y-mic ome e - hick sec ions om he L5 spinal
co d segmen we e ob ained on a eezing sliding mic o ome.
The sec ions we e dual-immunolabeled o gephy in (mouse mono-
418 •J. Neu osci., Janua y 12, 2005 •25(2):417–429 Gonzalez-Fo e o e al. •Ma u a ion o Inhibi o y Synapses on Renshaw Cells
clonal an ibody; 1:100 in PBS con aining 0.1% T i on X-100; Boeh inge
Mannheim, Indianapolis, IN) and calbindin ( abbi polyclonal an ibody;
1:4000 dilu ion; Swan , Bellinzona, Swi ze land). Renshaw cells we e
iden i ied by hei in ense calbindin immuno eac i i y (An al e al., 1990;
A d idsson e al., 1992; Sanna e al., 1993; Ca e al., 1998; Fallah and
Clow y, 1999; Geiman e al., 2000; McDonough e al., 2001). Immuno-
eac i e si es we e isualized wi h donkey an i- abbi , an i-mouse, o
an i-goa IgGs coupled o cyanine 3 (Cy3) o FITC (Jackson ImmunoRe-
sea ch, Wes G o e, PA).
Gephy in clus e a ea analysis. Calbindin-IR Renshaw cells we e cho-
sen andomly o analysis a low magni ica ion om sec ions con aining
ⱖ10 e og adely labeled mo oneu ons. Gephy in immuno luo escence
was only imaged a e he Renshaw cells o be used o analysis we e
selec ed. To gua an ee ha all soma ic gephy in clus e s we e imaged,
only calbindin-immunolabeled Renshaw cells wi h he comple e soma a
con ained wi hin he sec ion we e used. Then, gephy in and calbindin
immuno luo escence in Renshaw cells we e simul aneously imaged a
high magni ica ion (60⫻oil imme sion objec i e; nume ical ape u e,
1.4 digi ally zoomed 3⫻; 1024 ⫻1024 pixel images; pixel size, 0.0064
␮
m
2
) using an Olympus Fluo iew FX con ocal mic oscope. To ob ain
he highes quali y clus e mo phology, con ocal se ings we e op imized
o he bes signal/noise a io and o use he ull dynamic ange a ailable
(4095 in ensi y le els). Op ical sec ions we e ob ained e e y 0.5
␮
m
z-s eps. This z-s ep size allowed image “o e sampling” o mos clus e s,
equen ly iden i ied in adjacen op ical sec ions ( ocal op ical dep h o
he objec i e, ⬃0.6–0.7
␮
m), and p e en ed he loss o he dimme and
smalle clus e s, usually imaged in jus one ocal plane. Gephy in clus e s
we e analyzed in con ocal image s acks o he whole-cell body using
Image P o Plus so wa e ( e sion 4.1; Media Cybe ne ics, Sil e Sp ing,
MD). Con ocal images we e analyzed one op ical sec ion a a ime o all
en ace clus e s p esen a he op and bo om su aces (supplemen al Fig.
1, a ailable a www.jneu osci.o g as supplemen al ma e ial). The ocal
dep h o ou objec i e was mo e han enough o image he whole ex en
o en ace clus e s wi hin he op ical sec ion. En ace clus e s we e only
measu ed in op ical sec ions whe e hey displayed he maximum immu-
no luo escence b igh ness. Clus e s imaged h ough many op ical sec-
ions we e equen ly no en ace and lie on egions o high memb ane
cu a u e. These angen ial clus e s we e excluded om he analysis
gi en he di icul y o ge ing accu a e size es ima es om hem. Mea-
su ed clus e s ep esen app oxima ely one- hi d o all clus e s on he cell
bodies o indi idual neu ons. This ela i ely high p opo ion o en ace
clus e s esul s in pa because o la ening o he neu ons in he z-axis
wi hin he his ological p epa a ion. Renshaw cells we e de ined by hei
s ong calbindin immuno luo escence. Indi idual gephy in clus e s in
Renshaw cells we e segmen ed au oma ically om he image by using a
h eshold le el equal o 25–33% o he maximum b igh ness in each
op ical sec ion. This alue was empi ically de ined acco ding o gephy in
immuno luo escence b igh ness in di e en animals such ha h esh-
olded clus e s we e aced jus inside he luo escence di ac ion halo.
This acing pe ime e was p e iously ound o esul in he mos accu-
a e measu emen s o clus e size and was compa able o elec on mic o-
scopic obse a ions (Geiman e al., 2000). Di e ences in clus e immu-
no luo escence b igh ness because o dep h wi hin he issue we e
co ec ed by ela ing he h esholding c i e ia o he maximum b igh -
ness eco ded in each op ical sec ion. A eas we e measu ed o all en ace
gephy in clus e s ound a he op and bo om su aces o he neu on. We
also eco ded maximum and minimum diame e s, in eg a ed op ical
densi y, and immuno luo escence densi y. We did no analyze in ensi y
alues u he because we obse ed a sys ema ic dec ease in b igh ness
wi h issue dep h. Mos o his dec ease should be a ibu ed o a iable
amoun s o ligh dispe sion induced by he issue. These al e a ions in
in ensi y did no a ec a ea measu emen s (i.e., en ace clus e s sampled
a each side o he cell showed no signi ican di e ences in a e age a ea).
We p e iously showed ha small changes in immuno luo escence in en-
si y do no signi ican ly change he appa en sizes o gephy in clus e s
(Geiman e al., 2000). Fo clus e densi y measu emen s, all gephy in
clus e s on Renshaw cell soma a we e coun ed and di ided by he su ace
a ea o Renshaw cell soma a. Su ace a eas we e es ima ed by app oxi-
ma ing he cell body o an ellipsoid o he same maximum and minimum
diame e s.
Whole-cell eco dings o inhibi o y cu en s. We eco ded en al ho n
in e neu ons om spinal co d slices ob ained om P9–P15 animals ha
ecei ed injec ions o TeNT (25 ng/kg) a P5, as desc ibed abo e. Ren-
shaw cell eco dings we e ob ained as desc ibed p e iously (Gonzalez-
Fo e o and Al a ez, 2005). Animals ecei ed injec ions o FITC-
conjuga ed chole a oxin agmen B (CTb) (2.5% in saline) in he
gas ocnemius muscle 2–5 d be o e elec ophysiological eco dings. Re -
og ade anspo o CTb allowed isualiza ion o he gas ocnemius
mo oneu on pool. Sec ions wi h la ge numbe s o e og adely labeled
mo oneu ons we e iden i ied a low magni ica ion (4⫻) using epi luo-
escence (Olympus BX 50WI mic oscope), and hen small en al in e -
neu ons (⬍20
␮
m diame e ) we e sampled in he “Renshaw cell egion”
using in a ed di e en ial in e e ence con as op ics and high magni i-
ca ion (40⫻; wa e imme sion). Reco ded neu ons we e iden i ied a e
eco ding as Renshaw cells using neu oana omical his ochemical c i e ia
as desc ibed below.
Animals we e anes he ized (50 mg/kg pen oba bi al) and decapi a ed,
and hei spinal co ds we e emo ed quickly. Dissec ion was in ice-cold
(⬍4°C) suc ose-modi ied a i icial CSF (S-aCSF) bubbled wi h 95% O
2
and 5% CO
2
. S-aCSF composi ion was as ollows (in mM): 26 NaHCO
3
,
10 glucose, 3 KCl, 1.25 NaH
2
PO
4
, 2 MgCl
2
, 2.4 CaCl
2
, and 218 suc ose.
T ans e se slices om he L4–L5 segmen s (300–400
␮
m hick) we e cu
using an OTS-4000 issue slice (Elec on Mic oscopy Sciences, Fo
Washing on, PA). Slices we e ans e ed o no mal oxygena ed aCSF (in
which suc ose was emo ed and 130 mMNaCl and 2 mMCaCl
2
we e
added) and incuba ed o 1 h a 36°C and hen s abilized a oom em-
pe a u e (22–25°C) o a leas 30 min be o e elec ophysiological e-
co dings. Finally, he slices we e ans e ed in o he eco ding chambe
and pe used (a a a e o 4 ml/min) wi h oxygena ed no mal aCSF a
oom empe a u e. Pa ch elec odes (2–5 M⍀ esis ance) con ained he
ollowing (in mM): 120 CsCl, 4 NaCl, 4 MgCl
2
,1Cl
2
Ca, 10 HEPES, 0.2
EGTA, 3 Mg-ATP, and 0.3 GTP-T is. In all o he expe imen s, 0.4%
neu obio in (Vec o Labo a o ies, Bu lingame, CA) was added o he
in e nal solu ion. Only eco dings wi h access esis ance be ween 5 and
20 M⍀we e accep ed o analysis. Access esis ance was moni o ed con-
inuously, and he eco ding was abandoned i i changed ⬎15%. Junc-
ion po en ials we e co ec ed a e inse ing he pipe e in o he ba h.
Cells we e ol age clamped a ⫺75 mV. The e e sal po en ial o chlo-
ide cu en s was close o 0 mV; hus, GABAe gic and glycine gic cu -
en s we e eadily de ec ed as inwa d cu en s well sepa a ed om base-
line noise. Synap ic cu en s we e eco ded and low-pass bessel il e ed a
5 kHz wi h a Axopa ch 200B ampli ie (Axon Ins umen s, Union Ci y,
CA). Da a we e digi ized a 10 kHz and acqui ed using Axog aph ( e -
sion 4.6; Axon Ins umen s). Fo each cell, we ob ained wo o h ee
segmen s o 5 min con inuous eco dings o pha macologically isola ed
spon aneous synap ic cu en s o in e es .
Minia u e spon aneous synap ic cu en s o GABAe gic and/o glycin-
e gic o igin (mPSCs) we e isola ed wi h e odo oxin (TTX) (1
␮
M;
Alomone, Je usalem, Is ael), 6-cyano-7-ni oquinoxaline-2,3-dione
(CNQX) (10
␮
M; Toc is, B is ol, UK), and D- ubocu a ine chlo ide (30
o 10
␮
M; Sigma, S . Louis, MO) applied by ba h pe usion. Glycine gic
o GABAe gic mPSCs we e subsequen ly isola ed by adding ei he bicu-
culline me hiodide (10
␮
M; Sigma) o s ychnine hyd ochlo ide (0.25
␮
M; Sigma).
O -line da a analysis was pe o med using pClamp 9.0 and Mini-
Analysis so wa e (Synap oso ). E en s we e de ec ed by se ing a de ec-
ion h eshold alue o h ee imes he oo mean squa e noise. The
de ec ion h eshold was ⬃10 pA. We ou inely sc olled h ough de ec ed
e en s and isually ejec ed supe imposed o spu iously de ec ed e en s.
Peak mPSC ampli udes we e measu ed a he absolu e maximum o he
cu en s. Time o decay was calcula ed as he ime om peak o 33%
decay o he peak ampli ude. Rise imes we e measu ed as he ime
elapsed om 10 o 90% o he peak mPSC ampli ude. The cha ge ans-
e ed by each ype o synap ic e en (picocoulombs) was measu ed as
he in eg a ed a ea unde a e aged mPSCs. Cu e i ing o mPSC decays
( om peak o end) o single- o double-exponen ial equa ions was pe -
Gonzalez-Fo e o e al. •Ma u a ion o Inhibi o y Synapses on Renshaw Cells J. Neu osci., Janua y 12, 2005 •25(2):417–429 • 419
o med on a e aged aces wi h he simplex algo i hm leas squa es ex-
ponen ial i ing me hod p o ided by MiniAnalysis so wa e.
Pos eco ding iden i ica ion o Renshaw cells. We iden i ied Renshaw
cells by he high densi y o choline gic con ac s immuno eac i e o he
esicula ace ylcholine anspo e (VAChT) on hei dend i es (Al a ez
e al., 1999; Gonzalez-Fo e o and Al a ez, 2005). Reco ded neu ons we e
illed wi h neu obio in (0.4% in pipe e solu ion), and he slices we e
ixed in 4% pa a o maldehyde o 30 min and hen s o ed a 4°C in PBS.
Immunohis ochemis y was pe o med in he hick slices used o e-
co ding. To acili a e an ibody pene a ion, he slices we e dehyd a ed
and ehyd a ed h ough g aded alcohols and xylene, hen washed in PBS,
blocked wi h no mal ho se se um [1:10 in PBS and 0.3% T i on X-100
(TX)], and incuba ed ee- loa ing o 24–48 h a oom empe a u e in
p ima y an ise a agains VAChT (dilu ed 1:1000 in PBS and 0.3% TX).
The ea e , he slices we e washed and incuba ed o 2 h in s ep a idin–
Cy3 and FITC-coupled donkey an i-goa an ibodies dilu ed in PBS and
TX (1:50 dilu ions; Jackson ImmunoResea ch). Finally, he slices we e
moun ed on glass slides and co e slipped wi h Vec ashield (Vec o Lab-
o a o ies). Neu obio in-labeled cells we e i s analyzed a low magni i-
ca ion o con i m hei posi ion in he Renshaw cell a ea (Gonzalez-
Fo e o and Al a ez, 2005) and e ealed he ex ension and o ien a ions o
dend i ic ees and axonal a bo iza ions. Then, he labeled neu ons
(Cy3) and VAChT immuno luo escence (FITC) we e scanned a high
magni ica ion (60⫻2) wi h an Olympus Fluo iew FX con ocal mic o-
scope. Cell mo phology and VAChT-IR bou on con ac s we e econ-
s uc ed om s acks o op ical sec ions (1
␮
m s ep size). Reco ded neu-
ons we e classi ied as Renshaw cells i hey we e con ac ed by a high
numbe o VAChT-IR e minals on hei soma odend i ic su aces. Only
synap ic cu en s om iden i ied Renshaw cells we e analyzed.
S a is ical analysis. ANOVA es s we e un o compa e he pe cen age
o di e ences o elec oneu og aphic ac i i y be ween injec ed and non-
injec ed ne es wi h ime a e he injec ion and o con i m he simili ude
o gephy in clus e s in he noninjec ed sides o di e en animals. Pai ed
es s we e pe o med on animal means o compa e a e age Renshaw cell
gephy in clus e a eas in he noninjec ed and injec ed sides o each oxin
and age. The a e age glycine gic and GABAe gic peak ampli udes o
decays we e compa ed be ween con ol and expe imen al Renshaw cells
wi h unpai ed es s (expe imen al and con ol Renshaw cell samples
we e ob ained om slices ob ained in se e al animals). Dis ibu ion his-
og ams and cumula i e cu es o gephy in clus e a eas and inhibi o y
mPSCs eco ded in expe imen al and con ol Renshaw cells we e com-
pa ed using Kolmogo o –Smi no (K–S) es s. Da a a e exp essed as
mean ⫾SE. Signi icance o all s a is ical es s was se a p⬍0.05. S a is-
ics we e un in SigmaS a e sion 2.0 o S a is ica e sion 5.1.
Image composi ion and p esen a ion. Figu e composi ion and labeling
we e done in Co elD aw 8.0 o SigmaPlo 4.0. Fo p esen a ion pu poses,
con ocal images o gephy in clus e s we e enhanced using a high Gauss
il e (Image P o-Plus). This il e imp o es isualiza ion o ine de ail in
small luo escen punc a e s uc u es. Measu emen s we e done in he
aw images wi h no pos cap u e modi ica ions.
Resul s
In amuscula TeNT and BoNT-A, espec i ely, inc eased
and dec eased mo o axon ou pu and muscle ac i i y in
neona al a s
Neona al a s show dec eased sensi i i y o clos idial oxins
compa ed wi h adul s (Pas o e al., 2003). To maximize oxin
e ec s in neona es, we de e mined he la ges dose o each oxin
ha injec ed in amuscula ly in he gas ocnemius muscle o P5
a pups was compa ible wi h adequa e su i al (see Ma e ials
and Me hods). Doses o 25 ng/kg TeNT and 30 ng/kg BoNT-A
esul ed in su i al a es a P20 close o 90 and 80%, espec i ely.
Toxin- ea ed neona es and non ea ed li e ma es showed no
signi ican di e ences in weigh gain, sugges ing no mal o e all
de elopmen wi h hese oxin doses. Animals ha ecei ed injec-
ions a P5 wi h 25 ng/kg TeNT showed ipsila e al e anic ex en-
sion o he ankle wi hin 24–48 h a e injec ion. Te anic ex en-
sion was clea ly isible du ing he nex 8–10 d and hen slowly
eco e ed. In con as , animals ha ecei ed injec ions o 30
ng/kg BoNT-A showed laccid pa alysis o he ankle join ha
s a ed wi hin 24 h a e injec ion. BoNT-A e ec s began o e-
co e ⬃6–8 d a e injec ion (P11–P13). These sho ime
cou ses o oxin-induced pa alysis con as wi h he long-las ing
pa alysis (15–20 d o TeNT and 1–3 mon hs o BoNT-A) in-
duced in adul animals (Sanna e al., 1993; Mo eno-Lo´pez e al.,
1997; Gonzalez-Fo e o e al., 2003).
Nex , we cha ac e ized oxin ac ions on NMJ neu o ansmis-
sion and pe iphe al ne e mo o axon ac i i y. The deg ee and
ime cou se o NMJ block was es ima ed in a pups o 7–21 d o
age as he a io o peak o ce ampli ude e oked in he isola ed
gas ocnemius muscle a e di ec elec ical s imula ion o mus-
cle ibe s and a e ne e s imula ion (i.e., indi ec s imula ion)
(Fig. 1A–C). Maximum o ce peak ampli udes a e di ec muscle
s imula ion we e compa able in magni ude be ween injec ed and
noninjec ed muscles. Ne e-induced muscle con ac ions in
noninjec ed muscles o muscles injec ed wi h TeNT we e 80–
95% o he o ce gene a ed by di ec s imula ion applied o he
muscle (Fig. 1A,B), sugges ing no mal NMJ ansmission. In
con as , muscle o ce p oduced by ne e simula ion in BoNT-
A-injec ed muscles was ⬍5% o he o ce gene a ed by di ec
muscle s imula ion a P7 and eco e ed o ⬃50% by P21 (Fig.
1B,C). In conclusion, 30 ng/kg BoNT-A in neona al a pups
induced NMJ blocks ha pa ially eco e ed du ing he second
pos injec ion week. TeNT exe ed li le o no ac ion on NMJ
ansmission a his dose.
To cha ac e ize he ime cou se o e ec s on en al ho n
mo o ou pu , we compa ed mo o axon ac i i y eco ded in
pe iphe al ne es ipsila e al and con ala e al o oxin injec ions.
B anches o he ibial ne e inne a ing he gas ocnemius mus-
cle in he injec ed and con ala e al con ol sides we e p epa ed
o ex acellula eco dings o bila e al e lex ne e ac i i y elic-
i ed by mechanical s imula ion ( i m ubbing wi h a glass od) o
he ail. The ne es we e cu dis al o eco ding si es o p e en
con amina ion om muscle senso y a e en ac i i y. Examples
o aw ex acellula eco dings in con ol and ea ed ne es a e
shown in Figu e 1, Dand E. An a e age pe cen age o di e ence
o he e oked ac i i y in con ol and expe imen al sides was cal-
cula ed o each animal om he a eas unde he cu e o hal -
wa e- ec i ied elec oneu og aphic eco dings. Th ee animals
we e analyzed wi h each oxin and a each pos injec ion da e (6 h
and 1 d a e a P5 oxin injec ion and hen e e y 2 d un il P19).
TeNT-injec ed animals showed d ama ic inc eases in mo o e-
sponse compa ed wi h he con ol side. In con as , BoNT-A-
ea ed ne es show dep essed mo o ne e esponses (Fig.
1D,E). Signi ican changes in bila e al esponses we e de ec ed
om 1 o 10 d a e injec ion (i.e., P15) a e TeNT and om 6 h
o 12 d a e injec ion (P17) in BoNT-A- ea ed animals (Fig. 1F)
( wo-way ANOVA; LSD es o pos hoc compa isons; p⬍0.05).
Maximal esponse di e ences a e TeNT peaked a 4–6 d a e
injec ion (P9–P11) and a e BoNT-A peaked a 2–4 d a e in-
jec ion (P7–P9). Responses in oxin-injec ed and con ala e al
ne es we e no signi ican ly di e en 12 d a e injec ion (P17–
P19) o TeNT and 14 d (P19) o BoNT-A (Fig. 1F).
In summa y, TeNT induced in neona es a ch onic inc ease in
mo o ou pu ha was iden i ied by e anic ex ension o he ankle
and by inc eases in mo o ne e e lex esponse ac i i y. In con-
as , BoNT-A dec eased mo o ou pu , bes measu ed by d a-
ma ic educ ions in e lex esponses. In addi ion, BoNT-A e ec-
i ely blocked NMJ neu o ansmission. The peak e ec s o
BoNT-A occu ed as e han wi h TeNT, and he e ec s o bo h
420 •J. Neu osci., Janua y 12, 2005 •25(2):417–429 Gonzalez-Fo e o e al. •Ma u a ion o Inhibi o y Synapses on Renshaw Cells
oxins eco e ed du ing he second week a e injec ion, e u n-
ing o con ol le els ⬃P19. I was hus expec ed ha hese changes
in mo o ou pu eco ded pe iphe ally should e lec pa allel cen-
al al e a ions in mo o axon inpu s on o Renshaw cells.
Gephy in clus e size inc ease in Renshaw cells a e TeNT
and dec ease a e BoNT-A injec ions
Renshaw cells de elop dis inc i ely la ge gephy in clus e s on
hei cell bodies and p oximal dend i es, and mos clus e g ow h
occu s be ween P10 and P20 (Geiman e al., 2000). We a gued
ha any accele a ion in gephy in clus e g ow h could be bes
de ec ed a P15, bu i al e a ions we e only exp essed in mo e
ma u e clus e s, hen hey could be mo e no iceable a P20.
Mo eo e , P15 is he oldes age a which
inhibi o y synap ic cu en s in Renshaw
cells we e eliably eco ded in spinal co d
slices (Gonzalez-Fo e o and Al a ez,
2005). We he e o e analyzed gephy in
clus e sizes in Renshaw cells a P15 and
P20 a e unila e al injec ions o TeNT o
BoNT-A a P5.
Renshaw cells we e sampled in he en-
al 200
␮
m o he spinal co d and iden i-
ied by hei calbindin immuno eac i i y
in sec ions con aining a leas 10 mo oneu-
ons e og adely labeled om oxin-
injec ed muscles (Fig. 2A–C). Renshaw
cells we e sampled andomly in con ol
and expe imen al sides a low magni ica-
ion, and hen gephy in clus e s we e im-
aged a high magni ica ion wi h con ocal
mic oscopy. Renshaw cell soma a and
p oximal dend i es exhibi ed a a ie y o
gephy in clus e s om e y small ound
punc a o la ge and b igh clus e s show-
ing pe o a ions o scalloped con ou s.
Renshaw cells ipsila e al o TeNT-injec ed
mo o pools equen ly displayed in-
c eased numbe s o e y la ge and b igh
clus e s compa ed wi h Renshaw cells lo-
ca ed in he con ala e al side (Fig. 2D,E).
In con as , Renshaw cells ipsila e al o
BoNT-A injec ions exhibi ed ewe la ge
clus e s (Fig. 2F).
Calbindin immuno luo escence was
simila in he injec ed and noninjec ed
sides o P15 and P20 TeNT- ea ed ani-
mals; howe e , i was educed in he in-
jec ed side o mos BoNT-A- ea ed ani-
mals a P20 ( ou o six) and a P15 ( h ee
o ou ) (supplemen al Fig. 2, a ailable a
www.jneu osci.o g as supplemen al ma e-
ial) (Sanna e al., 1993). The numbe o
calbindin-IR neu ons in he en al 200
␮
m o laminas VII and IX (Renshaw cells)
was, howe e , simila in bo h spinal co d
sides in all o hese animals. BoNT-A-
injec ed animals analyzed a P15 con-
ained some en al neu ons in he Ren-
shaw cell a ea wi h weak calbindin
immuno luo escence ha was displaced o
he pe iphe y o cell soma a (da a no
shown). Neu ons wi h his calbindin im-
muno luo escence pa e n we e equen ly co e ed by gephy in
clus e s oo dim o image. These clus e s we e unmeasu able wi h
ou echniques, and he e o e hese cells could no be analyzed.
No such cells we e encoun e ed in TeNT-injec ed animals o in
BoNT- ea ed animals analyzed a P20.
Quan i a i e analysis o gephy in clus e a eas showed an in-
c ease in gephy in clus e size on Renshaw cells ipsila e al o
TeNT injec ions and a dec ease in BoNT-A- ea ed sides com-
pa ed wi h hei espec i e con ala e al sides. We i s calcula ed
he a e age gephy in clus e sizes o Renshaw cells sampled om
oxin-injec ed and con ala e al sides and pe o med pai ed
compa isons wi hin animals be ween oxin-injec ed and nonin-
jec ed sides. An a e age pe cen age o change was hen calcula ed
Figu e 1. Pe iphe al ac ions o TeNT and BoNT-A. A–C, Toxin ac ions on he NMJ. D–F, Toxin ac ions on mo oneu onal i ing
eco dedby elec oneu og aphyo he ibial ne e.A, Rep esen a i e ension eco dingso gas ocnemiusmuscle con ac ionsin
P9andP21animals ha ecei edinjec ionso 25ng/kg TeNTa e di ec (applied o hemuscle)andindi ec (applied o he ibial
ne e)s imula ion.Musclecon ac ionsa esimila a e ei he s imula ion,sugges ingunblockedneu omuscula ansmission.B,
Same as in A, bu o P7 and P15 animals ha ecei ed injec ions o 30 ng/kg BoNT-A. P o ound neu omuscula ansmission
blockade occu ed a P7 (muscle does no con ac by s imula ing he ne e) ha pa ially eco e ed a P15. C, Time cou se o
changes in indi ec (ne e) o di ec (muscle) s imula ion ension-e oked a ios a e ea men wi h clos idial neu o oxins.
Con ol da a a e om he con ala e al noninjec ed side. Two-way ANOVA indica es signi ican di e ences be ween BoNT-A
ea men a all ime poin s wi h espec o he con ol side (LSD o pos hoc compa isons a a signi icance le el o p⬍0.05). D,
E, Selec ed epochs o ne e discha ge a P7, P15, and P19 in TeNT- and BoNT-A- ea ed animals in esponse o s imula ion o he
ail ip.Mo o esponseswe einc easedin heTeNT- ea edne eanddep esseda e BoNT-A.F,Timecou seo changesinne e
ac i i y esponsesexp essed as a pe cen ageo he con ol side o hesame ime in e al. Te anuso dep essed i ing de eloped
o e 2–4da e ea men .Signi ican changesoccu omP6 oP15 o TeNTand omP6 oP17 o BoNT-A ea men ( wo-way
ANOVA; LSD es o pos hoc compa isons; p⬍0.05). Sham-ope a ed animals measu ed a P8 show simila esponses in bo h
sides.
Gonzalez-Fo e o e al. •Ma u a ion o Inhibi o y Synapses on Renshaw Cells J. Neu osci., Janua y 12, 2005 •25(2):417–429 • 421

o each oxin and su i al da e (Fig. 3). In
a second analysis, we pooled oge he all
gephy in clus e s sampled o each oxin
and age and cons uc ed his og am dis i-
bu ions and cumula i e p obabili y unc-
ions o hei a eas (Fig. 4).
Renshaw cell gephy in clus e a ea in-
c eased, on a e age, 18.4% in P15 TeNT-
injec ed animals, om 0.24 ⫾0.003
␮
m
2
con ala e al o he TeNT-injec ed side
(con ol) o 0.29 ⫾0.021
␮
m
2
in ipsila -
e al expe imen al Renshaw cells (n⫽5 an-
imals; n⫽42 con ol and 42 expe imen al
Renshaw cells). A P20, he a e age in-
c ease in size was 28.4%. The a e age
gephy in clus e a eas we e 0.32 ⫾0.020
␮
m
2
con ala e al and 0.41 ⫾0.019
␮
m
2
ipsila e al o he TeNT injec ions in P20
animals (n⫽4 animals; n⫽54 and 55
con ol and expe imen al Renshaw cells,
espec i ely). La ge a ea es ima es a P20
e lec he ma u a ion o gephy in clus e
sizes on Renshaw cells wi h age (Geiman e
al., 2000). Pai ed compa isons o gephy in
clus e a eas indica ed ha di e ences in
clus e size es ima ed in con ol Renshaw
cells (con ala e al) and in Renshaw cells
ipsila e al o TeNT injec ions we e signi i-
can a bo h ages (pai ed es ; p⬍0.05)
(Fig. 3A,C,E).
A e age gephy in clus e size dec eased
a e BoNT-A injec ions. A P15, we calcu-
la ed an a e age 17.7% dec ease in gephy-
in clus e size. A e age gephy in clus e
a eas dec eased om 0.31 ⫾0.009
␮
m
2
con-
ol o 0.25 ⫾0.020
␮
m
2
ipsila e al o
BoNT-A injec ions (n⫽4 animals; n⫽45
and 51 con ol and expe imen al Renshaw
cells, espec i ely). Simila o TeNT expe i-
men s, di e ences a e BoNT-A we e la ge
a P20. In hese animals, gephy in clus e a -
eas we e dec eased by 19.9%, on a e age,
om 0.34 ⫾0.010
␮
m
2
con ala e al o
0.28 ⫾0.022
␮
m
2
ipsila e al o BoNT-A in-
jec ions (n⫽6 animals; n⫽50 and 53 con-
ol and expe imen al Renshaw cells, espec-
i ely). Di e ences in gephy in clus e a ea
a e BoNT-A we e also s a is ically signi i-
can a bo h ages (pai ed es ;p⬍0.05) (Fig.
3B,C,E).
The indi idual a eas o Renshaw cell
gephy in clus e s always a ied be ween wo o de s o magni ude
( om 0.07 o ⬎1.5
␮
m
2
). Al hough la ge clus e s a e qui e dis inc
and de ine Renshaw cells, smalle clus e s (i.e., clus e s ⬍0.4
␮
m
2
)
always p edomina e (Geiman e al., 2000). In e es ingly, gephy in
clus e size dis ibu ion his og ams indica ed ha he p opo ion o
small clus e s (i.e., ⬍0.4
␮
m
2
) dec eased ipsila e al o he injec ion
in TeNT- ea ed animals in a o o inc easing he dis ibu ion skew
owa d la ge gephy in clus e sizes (Fig. 4A,B). Thus, cumula i e
p obabili y unc ions o gephy in clus e sizes we e shi ed owa d
la ge alues in he TeNT- ea ed side compa ed wi h he con ala -
e al side. This shi was la ge a P20 han a P15. The dis ibu ion
his og ams o gephy in clus e sizes in Renshaw cells loca ed in he
en al ho ns ipsila e al o BoNT-A injec ions showed opposi e
changes, he dis ibu ion his og ams showed smalle pe cen ages o
gephy in clus e s o la ge size, and cumula i e p obabili y unc ions
we e shi ed o he le , owa d smalle size clus e s. Changes in his-
og am dis ibu ions and cu e shi s a e BoNT-A we e also mo e
e iden a P20. Changes in gephy in clus e size dis ibu ions we e
signi ican in all cases ( p⬍0.05; K–S es ). We concluded ha TeNT
inc eased he a e age gephy in clus e size by speci ically inc easing
he p opo ion o la ge size gephy in clus e s (⬎0.4
␮
m
2
), whe eas
BoNT-A esul ed in opposi e changes.
Each gephy in clus e ep esen s he PSD o an independen
synap ic complex, and equen ly, se e al synap ic complexes a e
Figu e 2. Expe imen al design and al e a ions o gephy in clus e s uc u e on Renshaw cells a e TeNT o BoNT-A injec ions.
A,Diag amillus a ing hebasicci cui be weenmuscleand hespinalco dmo o ne wo k a ge edbype iphe allyadminis e ed
clos idial oxins.Toxinsand e og ade ace swe e injec edin hegas ocnemius muscle.BoNT-A a ec sNMJsa ex a usaland
in a usal muscle ibe s, and some en e s he e og ade anspo owa d he spinal co d. TeNT bypasses he NMJ and is e o-
g adely anspo ed o he cell bodies, whe e i exe s i e ec s o e inhibi o y synapses on he mo oneu on (MN) cell somas (see
Discussion o de ails and e e ences). B, Re og ade ace s (Fas Blue and Diamidino Yellow) we e injec ed wi h each oxin o
ehicle o iden i y spinal co d sec ions con aining la ge numbe s o a ec ed mo oneu ons. The gas ocnemius mo o pool is
e og adelylabeledbyFas Blue(mo oneu onsoma a)andDiamidinoYellow(mo oneu onla genuclei).C,The egionwi hin he
whi e box in Dis magni ied in C, and calbindin immuno eac i i y is added o show Renshaw cells in he icini y o oxin- ea ed
mo oneu ons. Renshaw cells in sec ions con aining ⬎10 labeled mo oneu ons we e sampled andomly o analysis om bo h
injec ed (con aining he labeled mo oneu ons) and con ala e al sides (unlabeled, con ol). D, Con ocal econs uc ion o he
su aceo aRenshawcell(RC) om hecon olsideo aP20animal.Thep oximalsoma odend i icsu aceisco e edwi hgephy in
clus e s o di e en sizes and shapes, usually la ge han clus e s on p ocesses o o he neu ons in he adjacen neu opil. E,F,
Simila su ace econs uc ions o Renshaw cells (RC) om en al ho ns con aining TeNT- ea ed (E) o BoNT-A- ea ed (F)
mo oneu ons. Calbindin immuno eac i i y o hese Renshaw cells is illus a ed in hei espec i e inse s (bo om igh ). Se e al
op ical sec ions we e supe imposed o ende en ace iews o gephy in clus e s co e ing he op su aces o he neu ons. All
Renshaw cells exhibi ed la ge gephy in clus e s ha easily dis inguished hem om o he neu ons; howe e , la ge clus e s we e
mo e equen inTeNT- ea edRenshawcellsandlessa e BoNT-A.Scaleba s:B,500
␮
m;C,100
␮
m;D–F,10
␮
m;inse s,5
␮
m.
422 •J. Neu osci., Janua y 12, 2005 •25(2):417–429 Gonzalez-Fo e o e al. •Ma u a ion o Inhibi o y Synapses on Renshaw Cells
o med by indi idual bou ons (T ille e al., 1985; Al a ez e al.,
1997). Thus, inc easing he numbe o synap ic complexes pe
bou on could compensa e o a dec ease in hei indi idual sizes.
Howe e , he su ace densi ies o gephy in clus e s (i.e., numbe
o clus e s pe 100
␮
m
2
o soma ic su ace memb ane) did no
change a e TeNT ea men s a any su i al da e (pai ed es s;
p⬎0.05) (Fig. 3F). The e was a end owa d dec eased densi ies
in Renshaw cells ipsila e al o BoNT-A injec ions a P15, bu i
did no each s a is ical signi icance ( p⫽0.057; pai ed es ).
This end was no obse ed a P20 ( p⫽0.26; pai ed es s).
TeNT induced a signi ican inc ease in inhibi o y cu en s in
Renshaw cells
The mo phological analyses desc ibed abo e sugges ela i ely
small bu signi ican changes in gephy in clus e ing a e al e ing
mo o ne e ac i i y wi h clos idial oxins. Thus, we in es iga ed
he unc ional consequences o hese s uc u al changes by e-
co ding inhibi o y (glycine gic and GABAe gic) minia u e syn-
ap ic cu en s (mPSCs) in Renshaw cells a e al e ing mo o
axon ac i i y. Mo oneu on iabili y limi s Renshaw cell eco d-
ings in spinal co d slices o he i s wo pos na al weeks
(Gonzalez-Fo e o and Al a ez, 2005). Reco ded neu ons we e
neu obio in illed and iden i ied as Renshaw cells using his olog-
ical c i e ia (Fig. 5) (Gonzalez-Fo e o and Al a ez, 2005). Neu-
ons displaying many VAChT-IR con ac s on hei soma and
dend i es we e classi ied as Renshaw cells. VAChT-IR con ac s
occu ed a highe densi y on dend i ic segmen s a some dis-
ance om he cell body as shown o adul Renshaw cells (Fig.
5B,D) (Al a ez e al., 1999). All o he cells iden i ied as Renshaw
cells we e loca ed in he en almos 200
␮
m o laminas VII and
IX (Geiman e al., 2000; Gonzalez-Fo e o and Al a ez, 2005). All
o he cells had local dend i ic a bo s ha ex ended li le beyond
200
␮
m o he cell bodies, and in some cases, we e ealed a local
axon ha a ge ed mo oneu on pools ( he cell shown in Fig. 5Ais
a good example). Some non-Renshaw cell in e neu ons and mo-
oneu ons we e also eco ded (Fig. 5C), bu hey we e no ana-
lyzed u he .
As epo ed p e iously, pha macologically isola ed spon ane-
ous GABAe gic/glycine gic mPSCs in Renshaw cells o his age
ha e cha ac e is ics di e en om o he neu ons (Gonzalez-
Fo e o and Al a ez, 2005). Renshaw cell GABAe gic/glycine gic
mPSCs a e cha ac e ized by as e en s (glycine gic) wi h la ge
han usual peak ampli udes, he p esence o many slow e en s
(GABAe gic), and a ela i e high equency o biphasic e en s
wi h as and slow componen s in hei decays. In Renshaw cells
sampled om he TeNT- ea ed side, he mos ob ious change
Figu e 3. E ec s o TeNT and BoNT-A on he de elopmen o gephy in-IR clus e s on Ren-
shaw cells. A,B, Plo s illus a ing a e age clus e sizes (A) and densi ies (B) ob ained om i e
P15 (F) and ou P20 (E) TeNT- ea ed animals. Poin s connec ed by lines ep esen he
a e age clus e a eas o densi ies in he con ol (con ala e al) and expe imen al (ipsila e al)
sideo hesameanimal.Clus e sizesanddensi iesinc eased omP15 oP20.Gephy inclus e s
onTeNT- ea edRenshawcellswe e alwaysla ge hanon hecon olside, whe easnochange
wasobse edinclus e densi y.C,D,SameasinAandB,bu o da aob ained om ou P15and
six P20 (only 3 in D) BoNT-A- ea ed animals. BoNT-A ea men had opposi e e ec s on he
clus e size de elopmen . The a e age clus e a ea in all o he animals was smalle on he
BoNT-A-injec ed side, whe eas clus e densi y emained unchanged ela i e o con ol, excep
o one P15 animal ha was clea ly di e en om he es . E,F, Analysis o a e age changes
(exp essed as a pe cen age o a ia ion; mean ⫾SE) in gephy in clus e a ea (E) o densi y (F)
on Renshaw cells induced by TeNT and BoNT-A ea men s a P15 ( ) and P20 (䡺). Gephy in
clus e a eas signi ican ly inc eased o declined on he ea ed side a e TeNT o BoNT-A, e-
spec i ely; clus e densi y was sligh ly educed only in BoNT-A-injec ed animals. Howe e , his
di e encewas below hes a is ical signi icancele el( p⬎0.05). As e isksindica esigni ican
di e ences ela i e o he con ala e al (un ea ed) side ( p⬍0.05; pai ed S uden ’s es ).
Figu e4. Gephy in clus e sizedis ibu ionsa e changedinopposi e di ec ionsbyTeNT and
BoNT-A. A,B, Dis ibu ion his og ams and cumula i e p obabili y unc ions o gephy in clus e
a eas measu ed on Renshaw cells om i e P15 (A) and ou P20 (B) TeNT- ea ed animals.
Da awe eob ained om 42 con ol (g ay lines;n⫽2375clus e s) and 42 expe imen al (black
lines; n⫽2287 clus e s) Renshaw cells (RCs) a P15 (A) and om 55 con ol (g ay lines; n⫽
4870clus e s)and54expe imen al(black lines; n⫽5167 clus e s) RCs a P20. B,C, Same as in
Aand B, bu o RCs om BoNT-A- ea ed animals (n⫽4 a P15 and n⫽6 a P20). Da a a e
pooled om 45 con ol (g ay lines; n⫽2935 clus e s) and 51 expe imen al (black lines; n⫽
2923 clus e s) RCs a P15 (C) and om 55 con ol (g ay lines; n⫽4870 clus e s) and 53
expe imen al (black lines; n⫽5167 clus e s) RCs a P20 (D). Cumula i e p obabili y unc ions
we e p og essi ely shi ed o he igh , and he a ea dis ibu ions we e skewed owa d la ge
sizes a he ages P15 and P20 a e TeNT. In con as , BoNT-A ea men caused opposi e shi s
o he igh (lowe clus e a eas) in gephy in clus e a ea dis ibu ions. All ou dis ibu ions o
gephy inclus e sizes ipsila e al o he injec edsidewe e signi ican lydi e en om con ala -
e al con ol dis ibu ions ( p⬍0.05; K–S es ). Bin wid h, 0.05
␮
m.
Gonzalez-Fo e o e al. •Ma u a ion o Inhibi o y Synapses on Renshaw Cells J. Neu osci., Janua y 12, 2005 •25(2):417–429 • 423
was an inc ease in he peak ampli ude o he as e en s and he
as componen o mixed e en s (Fig. 6A). To quan i a i ely an-
alyze mPSC al e a ions in Renshaw cells sampled con ala e al
(n⫽8) and ipsila e al (n⫽8) o he TeNT injec ion side, we
ob ained a e age aces o all mPSCs in each cell. These a e age
cu en s included e en s wi h as , slow, and mixed decays; hus,
hei decays we e bes i ed by he sum o wo exponen ial unc-
ions (Fig. 6B). We measu ed he a e age cu en peak ampli-
ude, he es ima ed as and slow decay ime cons an s (
␶
and
␶
s
,
espec i ely), he a io o he as and slow componen s o he
absolu e cu en peak (A
/A
s
a io), and he o al cha ge unde he
a e age cu en cu e (Fig. 6C–F). Signi ican inc eases in peak
ampli ude, A
/A
s
a io, and o al cha ge we e de ec ed in Renshaw
cells ipsila e al o TeNT injec ions compa ed wi h con ala e al
con ol Renshaw cells ( p⬍0.05; S uden ’s es ). Peak ampli-
udes inc eased by 78.9% om an a e age o 117.56 ⫾10.87 pA
in con ol compa ed wi h 210.31 ⫾16.83 pA in he TeNT- ea ed
side. Con ol Renshaw cell a e age peak ampli udes we e simila
o hose epo ed p e iously in a di e en sample o noninjec ed
animals a he same age (Gonzalez-Fo e o and Al a ez, 2005). In
addi ion, A
/A
s
a ios almos doubled om a 3.36 a io in con ol
Renshaw cells o a 6.45 a io in he TeNT-injec ed side (91.8%
inc ease). This sugges s ha he as componen o he cu en s
inc eased in size ela i e o he slowe componen . No signi ican
di e ences we e de ec ed in he decay ime cons an s o he as
(
␶
⫽3.27 ⫾0.16 msec in con ol and 3.57 ⫾0.26 msec in he
TeNT-injec ed side; p⫽0.35; S uden ’s es ) o slow (
␶
s
⫽
39.00 ⫾2.31 msec in con ol and 48.21 ⫾14.48 msec in he
Figu e 5. Ana omical cha ac e iza ion o eco ded Renshaw cells. A, Selec ed low-magni ica ion con ocal op ical planes h ough a 300-
␮
m- hick slice o a P10 spinal co d con aining a
neu obio in-labeled Renshaw cell (s ep a idin–Cy3 is shown in whi e o be e display de ails o i s dend i es and axon) and VAChT-IR bou ons (g een; FITC). This image is om he con ol side o
a TeNT-injec ed animal. VAChT immuno eac i i y is weak inside mo oneu on cell somas and e y s ong in C e minals in lamina IX (LIX) and su ounding Renshaw cells (Al a ez e al., 1999). This
Renshawcell hadap o use localaxonala bo iza ion ha a ge ed en al mo o pools.Inse , Va icoseaxoncolla e als om he egion labeledwi han as e iskin A. Theyellow do ed lineindica es
he g ay/whi e ma e bo de . B, High-magni ica ion image showing a pa ial econs uc ion o he cell soma and dend i es o he Renshaw cell shown in A(whi e do ). VAChT-IR e minals appea
ing een.ManyVAChT-IRbou onscon ac hecellsoma(bo omle inse ;a ows)anddend i es.Theboxeddend i ic egion(bo om igh )isshowna highe magni ica ionin hebo omle inse .
C, Low magni ica ion o selec ed op ical planes om he TeNT-injec ed side o a 300-
␮
m- hick P9 spinal co d slice. VAChT-IR bou ons appea as small do s in g een (FITC). In addi ion, mo oneu on
soma acon aining e og adely anspo edFITC–CTba eindica edwi ha owheads.In hissec ion, h eeneu onswe e eco dedandlabeled(openci clesindica ed heloca ionso hecellbodies).
Cell body images o eco ded cells a e shown in he inse s a he le . The neu on a he op is a mo oneu on ha also con ained FITC–CTb. The neu on a he bo om co esponds o a Renshaw cell
showna highe magni ica ioninD. The cellin he middle didno display VAChT-IR con ac so Renshaw cellcha ac e is icsand p obably ep esen sano he kind o en alin e neu on.The yellow
do ed line indica es he g ay/whi e ma e bo de as be o e. D, High magni ica ion o he Renshaw cell in he TeNT- ea ed side o he slice shown in C. VAChT-IR con ac s a e indica ed in he cell
soma and p oximal dend i es (a ows). The densi y o VAChT-IR con ac s inc eases a some dis ance om he cell body (boxed dend i ic egions shown in he indica ed inse s). The dis ibu ion o
VAChT-IR con ac s esembles ha o adul Renshaw cells (Al a ez e al., 1999). LF, La e al uniculs; VF, en al uniculus; LIX, lamina IX. Scale ba s: A,C, 200
␮
m; B,D, inse s, 5
␮
m.
424 •J. Neu osci., Janua y 12, 2005 •25(2):417–429 Gonzalez-Fo e o e al. •Ma u a ion o Inhibi o y Synapses on Renshaw Cells
TeNT-injec ed side; p⫽0.12; S uden ’s es ) componen s, al-
hough decays appea ed somewha longe ipsila e al o he TeNT
injec ions. The amoun o cha ge unde he cu e o a e age
mPSCs inc eased by 42.4%, om 1.88 ⫾0.63 pC in con ol o
2.68 ⫾0.65 pC in he TeNT-injec ed side. These esul s s ongly
sugges ha he unde lying cu en s esponsible o as compo-
nen s (glycine gic) inc eased in size much mo e apidly han
hose esponsible o he slow componen s (GABAe gic). To an-
alyze each componen sepa a ely, we isola ed glycine gic mPSCs
wi h bicuculline (10
␮
M) and GABAe gic mPSCs wi h s ychnine
(0.25
␮
M).
As expec ed bicuculline-isola ed glycine gic mPSCs showed a
la ge inc ease in peak ampli ude wi h no change in decay, ise
imes, o equency (Fig. 7A–D). We quan i a i ely examined
pha macologically isola ed glycine gic cu en s in 8 and 15 Ren-
shaw cells con ala e al and ipsila e al o TeNT injec ions. The
dis ibu ion his og ams o mPSC peak ampli udes (Fig. 7E)
changed in simila ways o gephy in clus e s sizes a e TeNT, and
he p opo ion o small e en s was diminished in a o o a s ong
skew owa d la ge cu en ampli udes. As a esul , he cellula
a e age peak ampli ude was signi ican ly inc eased by 60.87%
(127.14 ⫾10.66 pA in con ol compa ed wi h 204.53 ⫾42.81 pA
in he TeNT- ea ed side; p⬍0.001; S uden ’s es ). The o al
cha ge was signi ican ly inc eased in pa al-
lel (56.0% inc ease; 0.58 pC ⫾0.05 pC in
con ol o 0.90 ⫾0.06 pC in he TeNT-
injec ed side; p⫽0.003; S uden ’s es ).
No signi ican di e ences we e no ed in
decay (
␶
⫽3.93 ⫾0.39 msec in con ol and
4.3 ⫾0.39 msec in he TeNT-injec ed side)
o equency (1.21 ⫾0.29 Hz in con ol;
1.15 ⫾0.25 msec in he TeNT-injec ed
side).
Changes in GABAe gic mPSCs a e
TeNT we e mo e sub le. Quali a i ely,
GABAe gic mPSCs equen ly appea ed
augmen ed in peak ampli ude (Fig. 8A)
and wi h somewha slowe decays (Fig.
8B,D)(
␶
⫽37.88 ⫾3.39 msec in con ol
and 46.86 ⫾3.61 msec in he TeNT-
injec ed side; p⫽0.13; S uden ’s es ) bu
simila ise imes (Fig. 8C). Peak ampli-
ude his og ams and cumula i e p obabil-
i y plo s demons a ed s a is ically signi i-
can shi s in hei dis ibu ion owa d
la ge alues ( p⬍0.05; K–S es ); how-
e e , in ou sample (n⫽5 in bo h con ol
and TeNT- ea ed sides), a e age cellula
di e ences in peak ampli ude did no
each s a is ical signi icance (44.05 ⫾4.71
pA in con ol compa ed wi h 64.22 ⫾
11.15 pA in he TeNT-injec ed side; p⫽
0.13; S uden ’s es ), mos likely because
o la ge cell- o-cell a iabili y, pa icula ly
in he TeNT- ea ed side. Ne e heless,
cha ge ans e was signi ican ly inc eased
by 49.86% in he TeNT- ea ed side
(1.75 ⫾0.16 pC in con ol o 2.63 ⫾0.32
pC in he TeNT-injec ed side; p⫽0.04),
sugges ing a change in GABAe gic mPSC
a e TeNT ea men . I is in e es ing ha
he much longe ime cou se o GABAe -
gic e en s esul ed in cha ge ans e al e -
a ions in GABAe gic mPSCs a e TeNT simila o hose es i-
ma ed o glycine gic mPSCs, despi e much smalle changes in
peak ampli ude.
Discussion
The main indings o his s udy a e ha pos synap ic gephy in
clus e sizes and inhibi o y mPSC peak ampli udes, pa icula ly
he glycine gic componen s, we e modi ied in pa allel in Ren-
shaw cells a e al e a ion o mo oneu on ac i i y. Inhibi o y syn-
apses on Renshaw cells display abundan GABA/glycine co ans-
mission (Geiman e al., 2002a; Gonzalez-Fo e o and Al a ez,
2005). In e es ingly, synap ic GABAe gic cu en s we e a ec ed
o smalle deg ees. Gi en he igh ela ionship be ween gephy in,
glycine ecep o clus e ing (Ki sch e al., 1993; Feng e al., 1998;
Ki sch and Be z, 1998; Kneussel e al., 1999, 2001; Meie e al.,
2000), and pos synap ic glycine gic cu en s (Lim e al., 1999;
Van Zunde e al., 2004), he mos pa simonious explana ion is
ha gephy in/glycine ecep o clus e ing on Renshaw cells is
modula ed by mo o axon exci a o y synap ic inpu s and ha his
mechanism con ibu es o egula e pos synap ic cu en size.
mPSC equencies and gephy in clus e su ace densi ies did no
signi ican ly change, sugges ing ha he numbe o elease si es
(each opposed by a gephy in clus e ) and hei elease p obabil-
Figu e 6. Inhibi o y synap ic cu en s a e up egula ed in Renshaw cells loca ed ipsila e al o TeNT injec ions. A, T aces o
spon aneouslyoccu ingminia u e “inhibi o y” PSCs eco ded omaP11 Renshaw cell in hecon olside ( op aces) and oma
P10 Renshaw cell eco ded in he TeNT-injec ed side 5 d a e injec ion (bo om aces). Neu ons we e ol age clamped a ⫺75
mV,andmPSCswe e isola ed wi h TTX (1
␮
M),CNQX(10
␮
M),andD- ubocu a ine(30
␮
M).Th eedi e en ypes o mPSCs could
beiden i ied: as decaying(F),slowdecaying(E),andmixede en s(as e isks).Peakampli udeso as andmixedmPSCswe e
la ge in TeNT- ea ed Renshaw cells. B, A e age mPSCs o all o he ypes o e en s ( as decaying, slow decaying, and mixed
e en s) om he cells illus a ed in Adu ing 5 min o con inuous eco dings. Decay phases o a e age mPSCs we e always bes
i edby double-exponen ial unc ions(supe imposeddashedlines), inwhich
␶
and
␶
s
a e he decay imecons an so he “ as ”
and “slow” componen s, and A
and A
s
ep esen he ela i e con ibu ion o he as and slow componen s o he absolu e peak
ampli ude.The a e ageso cu en acesinRenshaw cellsin heTeNT- ea edside (bo om ace)showed ala ge inc ease in he
peak ampli ude o he as componen and a sligh slowing o he decay o he slow componen . C–F, His og ams showing
quan i a i e compa isons o peak ampli ude (C), slow (
␶
s
) and as (
␶
) decay ime cons an s (D), A
/A
s
a io (E), and cha ge
ans e ed(F) o a e agedmPSCs ob ainedincon ol (n⫽8)and Renshaw cellsin heTeNT- ea edside (n⫽8).Theas e isks
indica esigni ican di e encesbe weenbo hg oups( p⬍0.05;S uden ’s es ).Theg ow hinampli udeo he as componen
and he small leng hening o he slow componen bo h signi ican ly con ibu ed o an inc ease in he cha ge ans e ed by he
a e aged mPSCs in Renshaw cells eco ded in he TeNT- ea ed side.
Gonzalez-Fo e o e al. •Ma u a ion o Inhibi o y Synapses on Renshaw Cells J. Neu osci., Janua y 12, 2005 •25(2):417–429 • 425