Secretory responses of intact glomus cells in thin slices of rat carotid body to hypoxia and tetraethylammonium

Sec e o y esponses o in ac glomus cells in hin slices o a ca o id body o hypoxia and e ae hylammonium Rica do Pa dal, Uwe Ludewig, Julia Ga ci ´a-Hi sch eld, and Jose ´Lo ´pez-Ba neo* Depa amen o de Fisiologı´a, Facul ad de Medicina y Hospi al Uni e si a io Vi gen del Rocı´o, Uni e sidad de Se illa, E-41009, Se illa, Spain Communica ed by Clay M. A ms ong, Uni e si y o Pennsyl ania School o Medicine, Philadelphia, PA, Decembe 1, 1999 ( ecei ed o e iew Augus 12, 1999) We ha e de eloped a hin-slice p epa a ion o whole a ca o id body ha allows us o pe o m pa ch-clamp eco ding o mem- b ane ionic cu en s and o moni o ca echolamine sec e ion by ampe ome y in single glomus cells unde di ec isual con ol. In no moxic condi ions (P O 2⬇140 mmHg; 1 mmHg ⴝ133 Pa), mos glomus cells did no ha e measu able sec e o y ac i i y, bu ex- posu e o hypoxia (P O 2⬇20 mmHg) elici ed he appea ance o a la ge numbe o spike-like exocy o ic e en s. This neu osec e o y esponse o hypoxia was ully e e sible and equi ed ex acellula Ca 2ⴙ in lux. The a e age cha ge o single quan al e en s was 46 ⴞ 25 C (nⴝ218), which yields an es ima e o ⬇140,000 ca echol- amine molecules pe esicle. Addi ion o e ae hylammonium (TEA; 2–5 mM) o he ex acellula solu ion induced in mos (>95%) cells es ed (nⴝ32) a sec e o y esponse simila o ha elici ed by low P O 2. Cells non esponsi e o hypoxia bu ac i a ed by exposu e o high ex e nal K ⴙ we e also s imula ed by TEA. A sec e o y esponse simila o he esponses o hypoxia and TEA was also obse ed a e ea men o he cells wi h ibe io oxin o block selec i ely Ca 2ⴙ - and ol age-ac i a ed maxi-K ⴙ channels. Ou da a u he show ha memb ane ion channels a e c i ically in ol ed in senso y ansduc ion in he ca o id body. We also show ha in in ac glomus cells inhibi ion o ol age-dependen K ⴙ channels can con ibu e o ini ia ion o he sec e o y esponse o low P O 2. ca o id-body slice 兩O2sensing 兩glomus cell sec e ion Adecade has passed since pa ch-clamp expe imen s on dispe sed glomus cells om abbi and a ca o id bodies showed ha hey we e elec ically exci able and con ained O 2 -sensi i e ol age-dependen K ⫹ channels (1–6). Al hough some o he elec ophysiological p ope ies and he kind o O 2 -sensi i e K ⫹ channel o glomus cells a ied among animal species, hese obse a ions ga e s ong suppo o a uni ied memb ane model o chemo ansduc ion based on he capabili y o glomus cells o sense educ ions o O 2 ension (P O 2) h ough he closu e o K ⫹ channels, which in u n lead o Ca 2⫹ in lux, ansmi e elease, and ac i a ion o he a e en ibe s o he sinus ne e (see e s. 7 and 8 o e iew). Di ec e idence ha indi idual glomus cells wo ked as O 2 -sensi i e neu osec e o y elemen s was ob ained by moni o ing cy osolic [Ca 2⫹ ] and quan al ca echolamine sec e ion in single Fu a 2-loaded cells (7, 9–11). Nume ous in es iga o s ha e ound O 2 -sensi i e ol age- dependen channels in o he issues (in addi ion o he ca o id body), whe e hese channels pa icipa e in a a ie y o cellula elec ical, mechanical, o sec e o y esponses o acu e hypoxia (see e s. 12 and 13 o e iew). The p ecise ole o ol age-dependen K ⫹ channels in he chemo ecep i e p ope ies o he ca o id body emains, howe e , con o e sial, because he e a e esea che s who belie e ha he O 2 -sensi i e memb ane elec ical e en s in glomus cells a e no di ec ly in ol ed in senso y ansduc ion. A majo a gumen suppo ing his no ion is ha pha macological agen s, such as e ae hylammonium (TEA), 4-aminopy idine (4-AP), o cha ybdo oxin (CTX), which can inhibi he O 2 -sensi i e K ⫹ cu en s in pa ch-clamped glomus cells, do no elici neu al discha ges (o sec e ion) in whole ca o id-body p epa a ions (14–16). Al hough i has been epo ed ha applica ion o CTX can depola ize glomus cells (17), i has also been shown ha TEA and 4-AP ail o depola ize o o inc ease cy osolic [Ca 2⫹ ]in dispe sed a glomus cells (18). This las wo k (18) desc ibed an O 2 -sensi i e, K ⫹ -selec i e leak conduc ance, esis an o TEA and 4-AP, which seemed o con ibu e o he depola iza ion o he cells in esponse o low P O 2. In consequence, i was p oposed ha closu e o he leak K ⫹ channels is an obliga o y ini ial e en in low P O 2-induced exci a ion o a glomus cells. Despi e he conside able amoun o in o ma ion on ca o id- body unc ion al eady a ailable, he unde s anding o he O 2 - sensing mechanisms is p og essing a a ela i ely slow pace. Resea ch on he ca o id body is hampe ed by i s small size, abundance o connec i e issue, and p o use ascula iza ion, all o which make i di icul o ob ain a iable p epa a ion o enzyma ically dispe sed cells wi h consis en physiological p op- e ies. These limi a ions a e pa icula ly impo an when s udy- ing he cellula esponses o hypoxia, because sensi i i y o O 2 is a labile p ocess ha seems o be easily des oyed by uncon olled expe imen al a iables (13). In sea ch o an op imal app oach o in es iga e ca o id-body physiology, we ha e de eloped a hin- slice p epa a ion o he whole o gan ha does no equi e enzyma ic ea men o mechanical dis up ion o he issue. He ein, we show ha he ca o id-body slice p epa a ion used in combina ion wi h single-cell ampe ome y allows he s udy o he elec ophysiology and sec e o y ac i i y o in ac cells in almos op imal condi ions. Ou s udy, which uses his echnique, u he suppo s he iew ha ol age-dependen memb ane ion channels pa icipa e c i ically in chemosenso y ansduc ion. Me hods Ca o id-Body Slicing. Expe imen s we e pe o med on ca o id bodies isola ed om Wis a a s o ages be ween pos na al day 5 and 20. Animals we e anes he ized by injec ion o 0.2 mg 兾 kg sodium pen oba bi al (Eu a-Lende , Mad id). Comple e ca o id bi u ca ions we e quickly emo ed and placed on ice-cooled and O 2 -sa u a ed modi ied Ty ode’s solu ion (in mM: 148 NaCl 兾 2 KCl 兾 3 MgCl 2 兾 10 Hepes 兾 10 glucose, pH 7.4). The ca o id body was dissec ed om he adjacen a e y and hen included in 3% (w 兾 ol) low-mel ing-poin aga ose (FMC). A e quick cooling on ice, he aga ose block was glued wi h cyanoac yla e o he s age o a ib a ome chambe and co e ed by he same cold, O 2 -sa u a ed Ty ode’s solu ion. Slices 100- o 150- ␮ m- hick we e Abb e ia ions: TEA, e ae hylammonium; 4-AP, 4-aminopy idine; CTX, cha ybdo oxin; IbTX, ibe io oxin; TH, y osine hyd oxylase. *To whom ep in eques s should be add essed. E-mail: [email p o ec ed]. The publica ion cos s o his a icle we e de ayed in pa by page cha ge paymen . This a icle mus he e o e be he eby ma ked “ad e isemen ” in acco dance wi h 18 U.S.C. §1734 solely o indica e his ac . A icle published online be o e p in : P oc. Na l. Acad. Sci. USA, 10.1073兾pnas.030522297. A icle and publica ion da e a e a www.pnas.o g兾cgi兾doi兾10.1073兾pnas.030522297 PNAS 兩 Feb ua y 29, 2000 兩 ol. 97 兩 no. 5 兩 2361–2366 PHYSIOLOGY cu wi h s anda d azo blades. The esul ing slices (usually ou ) we e washed wice wi h cold s e ile PBS and placed on 35-mm Pe i dishes wi h DMEM (GIBCO 兾 BRL) supplemen ed wi h 1% ( ol/ ol) penicillin 兾 s ep omycin (GIBCO 兾 BRL) and 10% ( ol 兾 ol) FBS (BioWhi ake ). The slices we e main ained a 37°C in a 5% CO 2 incuba o 48–96 h be o e he expe imen s. A his ime, ma u a ion o he O 2 -sensi i e mechanisms, which occu s a ound pos na al day 10 (19, 20) we e almos comple e. Sensi i i y o he glomus cells o changes in P O 2did no seem o depend on he age o he animals used. Expe imen al Se up, Pa ch-Clamp Reco ding, and Ampe ome ic Mea- su emen o Sec e ion. In mos aspec s, ca o id-body slices we e managed ollowing he p ocedu es desc ibed o hin b ain slices (21, 22). Fo he expe imen s, a slice was ans e ed o a eco ding chambe (⬇200- ␮ l olume) placed on he s age o an up igh Zeiss mic oscope (Axioskop) equipped wi h long dis- ance wa e imme sion objec i es. Once in he chambe , he ca o id-body slice was con inuously pe used by g a i y ( low 1–2 ml 兾 minu e) wi h a solu ion con aining (in mM) 147 NaCl, 2.7 o 4.5 KCl, 23 NaHCO 3 , 1 MgCl 2 , 2.5 CaCl 2 , and 10 glucose. A ly a made o sil e wi e wi h glued ine Nylon h eads was used o hold he slice o he bo om o he chambe . The s anda d (no moxic) solu ion was bubbled wi h a gas mix u e o 5% CO 2 , 20% O 2 , and 75% N 2 (P O 2⬇140 mmHg). Hypoxia was ob ained by con inuously bubbling he solu ion in one o he ese oi s wi h 5% CO 2 and 95% N 2 . The pH o he no moxic and hypoxic solu ions was 7.4. The ime cou se o P O 2changes in he chambe was de e mined wi h a ca bon- ibe elec ode ha was nega i ely pola ized and wo ked as an O 2 elec ode. A e swi ching om no moxia o hypoxia, comple e equilib a ion o he new solu ion in he chambe equi ed be ween 1 and 2 minu es. P O 2in he chambe du ing exposu e o hypoxia eached a alue o ⬇20 mmHg. All pha macological agen s used (ion channel blocke s and ion chela o s) we e ob ained om Sigma and we e added o he ba h solu ion. The osmolali y o he solu ion (⬇300 millios- mol 兾 kg) was main ained by educing he concen a ion o NaCl. In he high K ⫹ solu ions, KCl eplaced NaCl equimola ly. In all he expe imen s, he empe a u e o he solu ions and he chambe was be ween 34 and 37°C. Memb ane cu en s we e eco ded by using he whole-cell con igu a ion o he pa ch- clamp echnique as adap ed in ou labo a o y (23, 24). We used low- esis ance pipe es (1–3 M⍀), capaci y compensa ion, and sub ac ion o linea leakage and capaci i e cu en s. Se ies esis ance compensa ion was be ween 40% and 50%. Pipe e solu ion con ained (in mM) 125 KCl, 4 MgCl 2 ,4MgATP,10 Hepes, and 10 EGTA (pH 7.2 and osmolali y o 285–290 milliosmol 兾 kg). Ionic cu en s we e eco ded wi h an EPC-8 pa ch-clamp ampli ie (HEKA Elec onics, Lamb ech 兾 P alz, Ge many), il e ed a 3 kHz, digi ized, and s o ed on a compu e . Sec e o y e en s we e eco ded wi h a 10- ␮ m ca bon- ibe elec ode connec ed o he inpu o a high-gain cu en o ol age con e e (10, 25). The elec ode was pola ized o ⫹750 mV, a alue mo e posi i e han he oxida ion po en ial o dopamine and he o he ca echolamines in glomus cells (10), and posi- ioned nea a cell unde isual con ol by using a piezoelec ic manipula o . Ampe ome ic cu en s we e also eco ded wi h he EPC-8 pa ch-clamp ampli ie . Cu en s we e il e ed a 100 Hz and digi ized a 250 Hz be o e s o age on compu e . Da a acquisi ion and analysis o ionic and ampe ome ic cu en s we e done wi h an ITC-16 in e ace (Ins u ech, G ea Neck, NY) and PULSE兾PULSEFIT so wa e ( e sion 8.11, HEKA Elec onics). Ty osine Hyd oxylase (TH) Immunocy ochemis y. Ca o id-body slices simila o hose used o he expe imen s desc ibed abo e we e ixed o e nigh a 4°C wi h PBS 兾 4% ( ol/ ol) pa a o mal- dehyde. A e washing in PBS 兾 0.1% T i on X-100 (PBTx), sec ions we e incuba ed wi h PBS 兾 0.3% H 2 O 2 o 2 h and hen Fig. 1. Mo phological appea ance o cells in hin slices o a ca o id body. (A) Low-magni ica ion iew o a slice a ew hou s a e cu ing. (B) Typical cell agg ega e (glome ulus) in a ca o id-body slice main ained o 48 h in a CO2incuba o . Indi idual cells, like he one indica ed by he a ow, can be clea ly di e en ia ed. (C) Ca o id-body slice s ained wi h an ibodies an i- y osine hyd oxylase. No e he ypical appea ance o glomus cells wi h la ge nuclei and he o ganiza ion o he ca o id body in glome uli. 2362 兩 www.pnas.o g Pa dal e al. wi h PBTx 兾 10% ( ol/ ol) FCS 兾 1 mg/ml BSA o ano he 2h o block nonspeci ic si es. A e incuba ion, a polyclonal an i-TH an ibody (1:1,000; Chemicon) was added o e nigh . A e wash- ing in PBTx, sec ions we e incuba ed again wi h bio inyla ed an i- abbi an ibody (1:200; Pie ce) o 12 h. The ea e , sec ions we e incuba ed wi h he ABC ki (Pie ce) o 2 h, and speci ically bound an ibody was e ealed by using 3,3⬘-diaminobenzidine (Sigma) as ch omogen. A e washing in PBS, sec ions we e moun ed on slides and co e slipped. Resul s Fig. 1Aillus a es he ypical appea ance o a a ca o id-body slice o ound shape (⬇400 ␮ m in diame e ) su ounded by he connec i e capsule and a ached issue. Du ing he ew hou s a e slicing, he his ological appea ance o he sec ions was a he uni o m, wi hou clea sepa a ion be ween he glandula , connec i e, and ascula issues. As incuba ion ime p og essed, he slice la ened, and he ex u e became less uni o m, wi h he appea ance o cell agg ega es clea ly sepa a ed om he su - ounding issue and wi h nume ous sphe ical o o oid cells 8–12 ␮ m in diame e (Fig. 1B). These cell agg ega es, simila o he glome uli desc ibed in his ological p epa a ions o he ca o id body (26), we e nicely obse ed in slices immunos ained wi h an i-TH an ibodies showing ypical clus e s o TH-posi i e glomus cells (Fig. 1C). We did mos o ou expe imen s on slices incuba ed o a leas 48 h, because hese slices had be e de ined glome uli ha seemed o con ain a la ge numbe o cells esponding o changes in P O 2(see below). Fo he mea- su emen o whole-cell cu en s o sec e o y ac i i y, he pa ch- clamp and ca bon- ibe elec odes, espec i ely, we e placed adjacen o a well iden i ied cell wi hin a clus e (e.g., he cell nea he a ow in Fig. 1B). We we e able o main ain ca o id- body slices in good condi ion o up o 5 days. The longe i y o he slices was no s udied sys ema ically, al hough i seems ha , wi h special ca e, long-las ing o gano ypic cul u es migh easily be achie ed. A e se ing up he slice p epa a ion, ou main goal was o s udy he sec e o y esponse o hypoxia o in ac cells in he slice (see below); howe e , we also es ed whe he cells in he glome uli could be pa ch-clamped and he e o e we e amenable o elec ophysiological expe imen s. S able whole-cell eco d- ings we e easily ob ained om glomus cells ha had inwa d and ou wa d ol age-dependen cu en s quali a i ely simila o he cu en s seen in enzyma ically dispe sed ca o id-body cells (Fig. 2A Middle and Bo om). Ne peak inwa d cu en a ⫹20 mV was, in mos cells, be ween 50 and 200 pA; howe e , in some cases, Fig. 2. Elec ophysiological and ampe ome ic eco dings om indi idual glomus cells in he slice. (A Top) Mac oscopic cu en s eco ded om a glomus cell by using he whole-cell con igu a ion o he pa ch-clamp echnique. (A Middle) Reco dings o inwa d and ou wa d cu en s om a cell ha had pa icula ly la ge as inwa d cu en . Reco dings elici ed by depola iza ions o ⫹20 and ⫹40 mV om he holding po en ial o ⫺80 mV a e supe imposed. (A Bo om) Supe imposed eco dings om a di e en glomus cell elici ed by depola izing pulses om ⫺80 mV o ⫹20 mV illus a ing he e e sible educ ion o he cu en by TEA. c and a e he con ol and eco e y aces, espec i ely. (B Top) Measu emen o sec e o y ac i i y om glomus cells by ampe ome y. A la ge spike-like exocy o ic e en is shown a an expanded ime base. (B Middle) Sec e o y esponse o a glomus cell o hypoxia (PO2⬇20 mmHg) and o high ex acellula po assium. (B Bo om) Example o a cell ha was insensi i e o changes o PO2(non esponsi e cell) bu ha main ained he ypical sec e o y esponse o high ex acellula po assium. Pa dal e al. PNAS 兩 Feb ua y 29, 2000 兩 ol. 97 兩 no. 5 兩 2363 PHYSIOLOGY as in he example shown in Fig. 2A Middle, we eco ded unusually la ge inwa d cu en s. De ailed elec ophysiological cha ac e iza ion o glomus cells in he slices was no a emp ed in his wo k, bu he la ge inwa d cu en obse ed in his s udy is simila o he Na ⫹ cu en desc ibed in a small popula ion o dispe sed cells (27). In ou expe imen al condi ions, inpu esis ance was be ween 1 and 3 G⍀, wi hin he ange o hose epo ed in isola ed a and abbi glomus cells (18, 24). This inding sugges s ha , unless whole-cell eco ding induced he pa ch-clamped cell o uncouple, he deg ee o elec ical coupling be ween neighbo ing glomus cells is no high. As shown be o e in dispe sed cells (4, 27), he mac oscopic K ⫹ cu en was e e sibly educed by exposu e o ela i ely low concen a ions o TEA (Fig. 2A Bo om). The sec e o y ac i i y o indi idual cells was s udied by measu ing he cu en esul ing om he oxida ion o eleased ca echolamine molecules (Fig. 2B Top). Single exocy o ic e en s appea ed as spike-like signals ep esen ing he usion o indi- idual ca echolamine gic esicles. An example o a la ge sec e- o y spike is shown in Fig. 2B Top Inse . Unde no moxic condi ions, mos cells did no show any measu able sec e o y ac i i y o had some occasional exocy o ic e en s a a equency o less han one pe minu e. Some cells we e, howe e , mo e ac i e wi h a spon aneous sec e o y e en equency o 10 pe minu e o highe . A e swi ching o he hypoxic solu ion, cells esponded wi h a p og essi e inc ease in he equency and ampli ude o he spikes ha pa ially used in o a b oad con- cen a ion en elope (Fig. 2B Middle). Rep oducible esponses o hypoxia we e obse ed h ee o ou imes in he cells i se e al minu es o es we e allowed o eco e y be ween successi e s imuli. Howe e , we ha e no s udied in de ail o he ac o s egula ing he deple ion o eplenishmen o he eleasable esicle pool in ou p epa a ion. Spike equency a he peak o he esponse o hypoxia (measu ed in he minu e a e he ini ial 90 s o exposu e o hypoxia) was 51.8 ⫾22 spikes pe minu e (n⫽ 6; mean ⫾SD). Gi en he p oximi y o he ampe ome ic elec odes o he cell unde s udy, we hink ha mos o he spikes eco ded we e due o exocy o ic e en s om he same cell; howe e , we canno exclude he possibili y ha some small spikes appea ing in he eco dings we e due o sec e o y e en s occu - ing in neighbo ing cells. The p og essi e inc ease in spike ampli ude du ing exposu e o hypoxia was a phenomenon ob- se ed epe i i ely in di e en expe imen s, which p obably esul ed om he in acy oplasmic usion o esicles occu ing be o e elease (compound exocy osis; e . 11). A e swi ching back o he no moxic solu ion, eco e y was as , and cells no mally e u ned o con ol condi ions in less han 30 s. As expec ed om elec ically exci able cells ha ing ol age- dependen Ca 2⫹ channels, all glomus cells made a igo ous sec e o y esponse when exposed o high ex acellula K ⫹ (see Fig. 2B Middle). A e age quan al cha ge o sec e o y e en s elici ed du ing exposu e o hypoxia was 46 ⫾25 C (mean ⫾SD; n⫽218 e en s in six cells; see also Fig. 4A). This alue was ob ained om he ime in eg al o selec ed spikes wi h he as ising phase and slowe decay (Fig. 2B Top Inse ) ypical o Fig. 3. The sec e o y esponse o hypoxia o glomus cells depends on ex acellula Ca2⫹in lux. (Uppe ) Re e sible abolishmen o he esponse o hypoxia du ing he blockade o Ca2⫹channels by addi ion o 0.2 mM CdCl2 o he ex acellula solu ion. (Lowe ) Re e sible abolishmen o he esponse o hypoxia by he emo al and chela ion o Ca2⫹ om he ex acellula solu ion. Fig. 4. Sec e o y esponse o in ac glomus cells o TEA. (A Uppe ) In a cell ha esponded o hypoxia, applica ion o TEA o he ba h elici ed a simila e e sible sec e o y esponse. (A Lowe ) F equency dis ibu ion o he cha ge o sec e o y e en s elici ed by hypoxia and 5 mM TEA in six di e en cells. No e ha he pa ame e s o he dis ibu ion (means ⫾SD) a e he same in he wo expe imen al condi ions. (B) Po en ia ion o he e ec o hypoxia by TEA. (C) Sec e o y esponse o TEA in a cell ha was no esponsi e o hypoxia. 2364 兩 www.pnas.o g Pa dal e al. sec e o y e en s occu ing in he memb ane acing he ampe o- me ic elec ode (10, 25). Assuming ha wo elec onic cha ges a e ans e ed in he oxida ion o a ca echolamine molecule, he a e age numbe o molecules pe esicle is ⬇140,000 ⫾75,000 (mean ⫾SD), a alue simila o ou es ima e in dispe sed abbi glomus cells (10). Besides he cells wi h cha ac e is ic sec e o y esponse o hypoxia desc ibed abo e, we also obse ed non e- sponsi e cells. Some o hese eco dings could ha e been om ype II cells, which a e nonexci able (24) and appea in e min- gled wi hin he glome ulus wi h glomus ( ype I) cells. Howe e , mos o he non esponsi e cells had he ypical appea ance o glomus cells and showed a powe ul sec e o y esponse o high ex acellula K ⫹ (Fig. 2B Bo om). In he O 2 -sensi i e glomus cells, he neu osec e o y esponse o hypoxia was almos com- ple ely and e e sibly abolished by he addi ion o Cd 2⫹ o he ex acellula solu ion (Fig. 3 Uppe ) o a e emo al o ex a- cellula Ca 2⫹ (Fig. 3 Lowe ). These obse a ions a e simila o hose epo ed in abbi glomus cells (10, 11) and u he show ha ansmi e elease elici ed by exposu e o physiological le els o hypoxia (⬇20 mmHg o highe ) depends on he in lux o ex acellula Ca 2⫹ . In neona al and adul cells, he majo con ibu o s o he O 2 -sensi i e mac oscopic K ⫹ cu en a e ol age- and Ca 2⫹ - dependen K ⫹ channels, which a e blocked by ex e nal TEA (4, 17, 20, 27). Fo his eason, we es ed o see whe he applica ion o TEA elici ed a sec e o y esponse simila o ha igge ed by hypoxia. In mos cells s udied (31 o 32), TEA a concen a ions o 2 o 5 mM, bo h o which p oduce blockade o a la ge p opo ion o K ⫹ channels ( e s. 4, 17, and 27; Fig. 2A), induced a clea sec e o y esponse (Fig. 4A Uppe ). This e ec o TEA was obse ed e en in quiescen cells (wi hou measu able spon- aneous quan al elease) and ega dless o whe he ex acellula K ⫹ was 2.7 o 4.5 mM. Spike equency a he peak o he esponse o 5 mM TEA (measu ed in he minu e a e he ini ial 90 s o exposu e o he blocke ) was 42 ⫾17 spikes pe minu e (mean ⫾SD; n⫽6), and a e age quan al cha ge was 43 ⫾30 C (mean ⫾SD; n⫽275 e en s in six cells). These alues, as well as he dis ibu ion o quan al e en s (Fig. 4A Lowe ), a e simila o hose es ima ed wi h e en s elici ed by hypoxia, sugges ing ha bo h TEA and exposu e o low P O 2can igge he elease o he same esicle pool. As shown in Fig. 4B, TEA also po en ia ed he sec e o y esponse induced by hypoxia. We ha e no es ed whe he his addi i e e ec is s ill p esen when high (sa u a ing) concen a ions o TEA a e used; howe e , we ha e eco ded one cell ha esponded o hypoxia bu was insensi i e o TEA alone. As men ioned be o e, TEA could elici ca echol- amine elease e en om cells ha we e no esponsi e o changes in P O 2(Fig. 4C). Thus, hese da a indica e ha di ec blockade o he O 2 -sensi i e K ⫹ channels wi h TEA can elici a igo ous sec e o y esponse om in ac glomus cells. Fu he suppo o his no ion came om expe imen s wi h ibe io oxin (IbTX), a selec i e blocke o he Ca 2⫹ - and ol age-ac i a ed maxi K ⫹ channels ha can induce ca echolamine sec e ion in clus e s o ca o id-body cells (28, 29). Fig. 5 illus a es ha IbTX elici ed a sec e o y esponse om indi idual glomus cells simila o ha e oked by hypoxia o TEA; washou o IbTX was, howe e , slowe han eco e y om he o he s imuli. Discussion In his epo , we desc ibe he ca o id-body slice p epa a ion and show ha his echnique allows us o s udy he elec o- physiology and cellula esponses o single glomus cells o hypoxia wi hou he need o enzyma ic ea men o mechan- ical dis up ion o he issue. We ha e consis en ly eco ded sec e o y ac i i y elici ed by hypoxia in single a glomus cells by measu ing ca echolamine elease wi h a ca bon- ibe am- pe ome ic elec ode. The neu osec e o y esponse o hypoxia in single a glomus cells is quali a i ely simila o he esponse ha we desc ibed be o e in dispe sed abbi glomus cells (10, 11). In bo h cases, ca echolamine elease e oked by physio- logic low P O 2le els (⬇20 mmHg o highe ) depends absolu ely on ex acellula Ca 2⫹ in lux. Thus, ou da a gi e u he suppo o he iew ha glomus cells beha e as O 2 -sensi i e p esynap ic-like elemen s, in which Ca 2⫹ in lux igge ed by hypoxia-induced depola iza ion is he undamen al e en lead- ing o ansmi e elease and ac i a ion o he a e en senso y ibe s wi h which he glomus cells o m synapses (7–11). In e es ingly, we obse ed ha some cells ha had an appea - ance ypical o glomus cells and ha esponded wi h exocy o ic ca echolamine sec e ion o high ex acellula K ⫹ we e insensi i e o changes in P O 2. The insensi i i y o hese cells o hypoxia was no ela ed o animal age and could be explained ei he by he need o lowe P O 2le els o become depola ized o by he ac ha , in his subse o glomus cells, he O 2 -sensi i e machine y is inac i a ed o uncoupled om he K ⫹ channels. We belie e ha he las possibili y is mo e likely o wo easons. Fi s , like he hypoxia- esponsi e cells, he non esponsi e ones we e s imu- la ed by TEA. Second and mo e impo an ly, when we obse ed a la ge numbe o non esponsi e cells in a pa icula expe imen , he lack o O 2 -sensi i i y ended o be main ained in all he slices o he p epa a ion. The e o e, i seems ha insensi i i y o hypoxia depends on uncon olled a iables, some o hem in- oduced by he expe imen al p o ocol, which may modi y he chemical s a us o he O 2 -sensing molecule o al e i s associa- ion wi h he ion channels. In ac , we ealized long ago ha he O 2 -sensi i i y o ion channels is a labile phenomenon ha can be easily al e ed by cell dissocia ion p ocedu es (13). Ano he impo an obse a ion in his s udy is ha ex e nal applica ion o TEA and IbTX consis en ly elici ed a sec e o y esponse ha in mos aspec s was simila o he esponse e oked by hypoxia. This simila i y con as s wi h he lack o TEA e ec on dispe sed a glomus cells (18) and indica es ha di ec blockade o O 2 -sensi i e ol age-dependen K ⫹ channels, which in hese cells a e hose blocked by TEA o IbTX (4, 17, 27, 28), can lead o sec e ion. A co olla y o hese da a is ha inhibi ion Fig. 5. Sec e o y esponse o in ac glomus cells o IbTX. In a cell ha esponded o hypoxia (Uppe ), applica ion o IbTX (200 nM) o he ba h elici ed a simila e e sible sec e o y esponse (Lowe ). No e ha eco e y om IbTX is ela i ely slow p obably because o he high a ini y o oxin binding o he channel. Pa dal e al. PNAS 兩 Feb ua y 29, 2000 兩 ol. 97 兩 no. 5 兩 2365 PHYSIOLOGY o ol age-dependen K ⫹ channels by hypoxia can con ibu e o ini ia ion o he sec e o y esponse elici ed by low P O 2.A TEA- esis an leak K ⫹ cu en inhibi ed p e e en ially by anoxia o ex eme low P O 2le els has been desc ibed in dispe sed a glomus cells (18). Whe he he leak and he TEA-sensi i e channels a e p esen in he same a glomus cell ype and whe he hei unc ions o e lap he same ange o P O 2 alues a e ques ions ha should be add essed in u u e expe imen al wo k. Based on ou p esen da a, i is di icul o unde s and he insensi i i y o TEA (o o CTX and 4-AP) o whole ca o id- body p epa a ions epo ed by some au ho s (14–16). Al hough he molecula iden i y and pha macology o he K ⫹ channels in he glomus cell-a e en ibe synapse a e unknown (see, how- e e , e s. 4, 5, 17, 27, and 30), i is logical o pos ula e ha , ega dless o he mechanisms in ol ed in he ansduc ion o he hypoxic s imulus, he e mus be, a he synapse i sel and along he a e en ne e ibe , se e al ypes o ol age-dependen K ⫹ channels whose blockade should al e ansmi e elease as well as he equency and shape o he p opaga ed ac ion po en ials. Because TEA as well as CTX and 4-AP a e ela i ely la ge, poo ly lipid-soluble molecules, a possible explana ion could be ha he blocke s do no di use a he app op ia e concen a ion in o he ex acellula space o ca o id bodies ei he supe used (wi h he su ounding connec i e capsule in ac ) o pe used h ough he ca o id a e y. We wish o hank D . Juanjo Toledo-A al o help in he TH immunos aining o ca o id-body slices. R.P. is a p edoc o al ellow o he Spanish Fo macio´n del Pe sonal In es iga o p og am. U.L. was ecipien o a pos doc o al ellowship o he Eu opean Molecula Biology O ganiza ion. Resea ch was suppo ed by he Spanish Di ec- cio´n Gene al de Ensen˜auza Supe io , Fundacio´n La Caixa, and he Andalusian go e nmen . 1. Duchen, M. R., Caddy, K. W. T., Ki by, G. C., Pa e son, D. L., Pon e, J. & Biscoe, T. J. (1988) Neu oscience 26, 291–311. 2. Lo´pez-Ba neo, J., Lo´pez-Lo´pez, J. R., U en˜a, J. & Gonza´lez, C. (1988) Science 242, 580–582. 3. Delpiano, M. A. & Heschele , J. (1989) FEBS Le . 249, 195–198. 4. Pee s, C. (1990) Neu osci. Le . 119, 253–256. 5. Gan o nina, M. D. & Lo´pez-Ba neo, J. (1991) P oc. Na l. Acad. Sci. USA 88, 2927–2930. 6. S ea, A. & Nu se, C. A. (1991) P lu¨ge s A ch. 418, 93–101. 7. Lo´pez-Ba neo, J., Beno , A. R. & U en˜a, J. (1993) News Physiol. Sci. 8, 191–195. 8. Gonza´lez, C., Alma az, L. & Obeso, A. (1994) Physiol. Re . 74, 829–898. 9. Buckle , K. J. & Vaughan-Jones, R. D. (1994) J. Physiol. (London) 476, 423–428. 10. U en˜a, J., Fe na´ndez-Chaco´n, R., Beno , A. R., Al a ez de Toledo, G. & Lo´pez-Ba neo, J. (1994) P oc. Na l. Acad. Sci. USA 91, 10208–10211. 11. Mon o o, R. J., U en˜a, J., Fe na´ndez-Chaco´n, R., Al a ez de Toledo, G. & Lo´pez-Ba neo, J. (1996) J. Gen. Physiol. 107, 133–143. 12. Lo´pez-Ba neo, J. (1996) T ends Neu osci. 19, 435–440. 13. Lo´pez-Ba neo, J., Mon o o, R. J., O ega-Sa´enz, P. & U en˜a, J. (1998) in Oxygen Regula ion o Ion Channels and Gene Exp ession, eds. Lo´pez-Ba neo, J. & Wei , K. S. (Fu u a, New Yo k), pp. 127–144. 14. Doyle, T. P. & Donnelly, D. F. (1994) J. Appl. Physiol. 77, 2606–2611. 15. Osanai, S., Bue k, D. G., Mokashi, A., Chugh, D. K. & Lahi i, S. (1997) B ain Res. 747, 324–327. 16. Lahi i, S., Roy, A., Rozano , C. & Mokashi, A. (1998) B ain Res. 794, 162–165. 17. Wya , C. N. & Pee s, C. (1995) J. Physiol. (London) 483, 559–565. 18. Buckle , K. J. (1997) J. Physiol. (London) 498, 649–662. 19. Donnelly, D. F. & Doyle, T. P. (1994) J. Physiol. (London) 475, 267–275. 20. Ha on, C. J., Ca pen e , E., Peppe , D. R., Kuma , P. & Pee s, C. (1997) J. Physiol. (London) 501, 49–58. 21. Edwa ds, F. A., Konne h, A., Sakmann, B. & Takahashi, T. (1989) P lu¨ge s A ch. 414, 600–612. 22. Schneggenbu ge , R., Lo´pez-Ba neo, J. & Konne h, A. (1992) J. Physiol. (London) 445, 261–276. 23. Hamill, O. P., Ma y, A., Nehe , E., Sakmann, B. & Sigwo h, F. (1981) P lu¨ge s A ch. 391, 85–100. 24. U en˜a, J., Lo´pez-Lo´pez, J. R., Gonza´lez, C. & Lo´pez-Ba neo, J. (1989) J. Gen. Physiol. 93, 979–999. 25. Chow, R. H. & on Ruden, L. (1995) in Single Channel Reco ding, eds. Sakmann, B. & Nehe , E. (Plenum, New Yo k), pp. 245–275. 26. Fidone, S. J. & Gonza´lez, C. (1986) in The Respi a o y Sys em II: Handbook o Physiology, ed. Fishman, A. P. (Am. Physiol. Soc., Be hesda), pp. 557–621. 27. Lo´pez-Lo´pez, J. R., Gonza´lez, C. & Pe´ ez-Ga cı´a, M. T. (1997) J. Physiol. (London) 499, 429–441. 28. Gal ez, A., Gimenez-Gallego, G., Reuben, J. P., Roy-Con ancin, L., Feigen- baum, P., Kaczo o ski, G. J. & Ga cia, M. L. (1990) J. Biol. Chem. 264, 11083–11090. 29. Jackson, A. & Nu se, C. (1997) J. Neu ochem. 69, 645–654. 30. Gan o nina, M. D. & Lo´pez-Ba neo, J. (1992) J. Gen. Physiol. 100, 401–426. 2366 兩 www.pnas.o g Pa dal e al.