Dabetologa (1988) 31 : 435-442 Dabetologa 9 Sprngcr-Verlag 1988 Tme-dependent potentaton of nsuln release nduced by alpha-ketosocaproate and leucne n rats: possble nvolvement of phosphonostde hydrolyss W. S. Zawalch Yale Unversty School of Nursng, New Haven, Connectcut, USA Summary. The ablty of the amno acd leucne and ts keto acd, alpha-ketosocaproate, to nduce nsuln release, to ntate phosphonostde hydrolyss, and to amplfy the subsequent nsuln secretory response to glucose was assessed. n slets whose nostol-contanng lpds were prelabelled wth myo[2-3h]nostol, the addton of ether compound resulted n an ncrease n nsuln output, an ncrease n 3H effux, rapd and sgnfcant ncreases n labelled nostol phosphate accumulaton and a sustaned ncrease n 3H efflux after removal of the stmulant. Drect measurements of labelled nostol phosphate accumulaton n slets prevously stmulated wth alpha-ketosocaproate demonstrate that ths sustaned ncrease n ~H effux was the result of a persstent ncrease n phosphonostde hydrolyss and was not smply a consequence of the hydrolyss of preformed nostol phosphates nto more membrane permeable speces. Pror exposure of slets to alpha-ketosocaproate or leucne also resulted n an amplfed secretory response to a subsequent glucose (10 mmol/1) stmulus. Whle peak frst phase nsuln release averaged 66+_4 (mean +_ SEM, n=18) pg.slet -1. mn -1 from control slets, ths value ncreased to 204+_14 and 246_+ 11 pg. slet- 1. mn- 1 n the leucne or alpha-ketosocaproate pretreated slets respectvely. The duraton of ths amplfed response paralleled the duraton of the persstent ncrease n 3H effux. Pror alpha-ketosocaproate exposure also amplfed the subsequent nsuln secretory response to tolbutamde and glyceraldehyde. Whle control (non-pretreated) slets n response to tolbutamde (200.tmol/l) released nsuln at a rate of 50+6 pg.slet-l.mn -1 (n=3), ths frst phase response ncreased to 506+_38 pg.slet -lran -1 n pror alpha-ketosocaproate treated slets. Peak frst and second phase nsuln responses to glyceraldehyde were ncreased 5-fold and 2-fold, respectvely, by pror alpha-ketosocaproate. These results suggest that events coupled to the hydrolyss of membrane nostol-contanng phospholpds nduced by leucne and alpha-ketosocaproate partcpate not only n ther acute nsuln stmulatory acton, but also n ther ablty to nduce tme-dependent potentaton (memory) n solated slets. Key words: slets, phosphonostdes, memory, nsuln secreton, leucne, alpha-ketosocaproate, 3H efflux. t s well establshed that pror glucose stmulaton of pancreatc B cells results n an amplfed nsuln secretory response to a second glucose challenge [1-4]. Some characterstcs of ths phenomenon, termed tmedependent potentaton (TDP) by Grll and assocates [2, 3] have been establshed: (a) glucose must be metabolsed to nduce TDP; (b)tdp develops wthn mnutes and perssts long after removal of the ntal stmulant; (c)ntracellular accumulaton of camp does not seem to play a major role n the process; and (d) glyceraldehyde and the gut hormone cholecystoknn [5] mmc to a large extent ths senstsng effect of glucose. We recently reported that TDP noted wth glucose [6], glyceraldehyde [6], or cholecystoknn [7], appeared to depend on agonst-nduced ncreases n phosphonostde (P) hydrolyss, a response that perssts despte stmulant removal. The duraton of TDP parallels to a remarkable extent the duraton of the sustaned ncrease n P hydrolyss and, moreover, condtons whch abolsh P hydrolyss abolsh TDE n the present report, the capacty of leucne and alphaketosocaproate (KC) to promote P hydrolyss and to nduce TDP was assessed. These partcular stmulants were chosen because they both augment nsuln output from the B cell, an effect that seems to be related, at least n part, to ther ablty to ncrease P metabolsm [8-10]. The present results further emphasze the nvolvement of bochemcal events n the nostol cycle n the nducton of TDP by varous stmulants. Materal and methods Male Sprague-Dawley rats purchased from Charles Rver were used n all studes. The anmals were fed ad lbtum and weghed between
436 W. Zawalch: Memory n rat slets nvolves phosphonostde hydrolyss G2.r5 500 Gz.75 Table 1. nfluence of leucne (20 mmol/l) and alpha-ketosocaproate (15 mmol/l) on nostol phosphate accumulaton n solated perfused rat slets -=,7 o0t- t ~= 1'5 g X h ~ 0.5 7 0 "- " 0 1001- _ / / --CL>r"+", 0 L,,.9o'~"" Gz.75 Gz.;,5 30 40 50 60 70 80 90 DURATON OF PERFUSON (mn Fg.la and b. nsuln release and phosphonostde hydrolyss n response to alpha-ketosocaproate (KC) and leucne, a Alter a 2 h labellng perod wth myo [2-3-] nostol, batches of slets were perfused wth 2.75 mmol/l glucose for 30 mn to establsh stable basal rates of nsuln secreton. They were then stmulated for 60 mn wth KC, 15 mmol/l, (O) or leucne, 20 mmol/l, (l) wth glucose stll present. Ths and all subsequent fgures have been corrected for the dead space (3 ml, equvalent to 3 mn wth a flow rate of 1 ml/mn) n the perfuson system. Mean values are gven + selected SEM's of at least 4 experments, b Fractonal efflux rates of 3H, used to montor P hydrolyss are depcted here n response to KC (O) and leucne (ll). Efflux rates n response to 2.75 mmol/l glucose alone (D) are also shown. These values were calculated accordng to the procedure of Bode et al. [25] and were obtaned from the same slets whose nsuln release profles are gven above 300-400 g. After nembutal-nduced (50 mg/kg) anaesthesa, slets were solated by collagenase dgeston [11]. n some experments, slets were drectly perfused to assess secretory responsveness. n other experments batches of 40-80 slets were loaded onto nylon flters and placed n small glass vals. They were ncubated for 2 h n 200 xl of a myo{2-3h] nostol contanng soluton prepared by addng 10!xC myo[2-3h] nostol (ntal specfc actvty 16.6C/ mmol) to 250 l.tl of ncubaton medum. The medum used for ths ncubaton procedure was smlar to that employed durng the slet perfuson and conssted of 115mmol/1 NaC, 5mmol/ KC, 2.2 mmol/1 CaC2, 1 mmol/l MgC2, 24 mmol/ NaHCO3, and 0.17 g % bovne serum albumn. The soluton was gassed wth 95% 02/5% COz. Glucose (2.75 mmol/l) was also present durng the ncubaton. After termnaton of the ncubaton the slets, stll attached to the nylon flters, were washed wth 5 ml nonradoactve medum and perfused. The ph of the perfuson medum was mantaned at 7.4, the temperature at 37 ~ and the flow rate at 1 ml/mn. slets were usually perfused for 30 mn to establsh stable nsuln secretory rates and then exposed to varous agonsts ndcated n the fgure legends. Perfusate samples were collected at tme ntervals ndcated n the fgures and 200 xl alquots analysed for 3H content when appropr- Protocol P1 P2 P3 (ran) (cpm/40 slets, mean + SE) 1. G2.75 413_+ 30 111+ 12 98 + 9 (50 or 60 mn 2. GzTs--+GzTs+Leucne 621_+ 62 168+ 13 129_+ 7 (30) (20) 3. G2.75 406_+ 23 103_+ 13 88+ 7 (32) 4. Gz75--+Gz75+KC 660+_ 38 179_+ 11 167+_12 (30) (2) 5. Gz75"--+G2.75 + KC 1082_+139 527_+116 283+61 (30) (20) 6. GzT~..+Gz75+LC1 504_+ 41 95+ 14 98_+18 (60) (20) 7. Gz7.~---+Gz75+KC--+Gz7.s 637 29 182+_ 21 108+17 (30) (20) (10) 8. G2.75--+Gz75+KC--+ 1540_.+181 280+_ 28 197+18 (30) (20) G2.75--~G2.75 + LC1 (10) (20) After labellng for 2 h wth myo[2-~h]nostol, groups of slets (n = at least 4 for each condton) were perfused as ndcated. Alter the perfuson, nostol phosphates were extracted wth 10% perchlorc acd and separated as descrbed [13, 14]. Statstcal analyss Ls as follows: Protocol 1 vs Protocol 2 - p < 0.05 for all nostol phosphates, Protocol vs Protocol 5 - p < 0.05 for all nostol phosphates, Protocol 1 vs Protocol 6 - no sgnfcant dfferences, Protocol 3 vs Protocol 4 - p < 0.05 for all nostol phosphates, Protocol 5 vs Protocol 7 - p<0.05 for P; P2, and P3, Protocol 6 vs Protocol 8 - p<0.05 for all nostol phosphates Abbrevatons: P1 - nostol/-phosphate; P2 - nostol 1,4-bsphosphate: P3 -nostol 1,4,5-trsphosphate; +l,3,4-trsphosphate; LCl-lthum chlorde, 10 mmol/ ate as well as nsuln [12] usng rat nsuln (Llly and Co., ndanapols, ndana, USA # 615-1963-12-3) as standard. n some experments the nature of the 3H-contanng molecules was assessed n pooled perfusate samples collected durng the fnal 10 mn of the perfuson. n other experments, the levels of free 3H-nostol, glycerophosphonostol (GP) and labelled nostol phosphates were determned after extracton wth 10% perchlorc acd by methods prevously detaled [13, 14]. Brefly, after neutralsaton wth 0.25-0.28 ml 6 N KOH, the further addton of 5 ml water, and centrfugaton, the supernatant was appled to columns. These columns were prepared by addng (to acheve a length of 3 cm) anon exchange resn (AG # 1-8X Bo-Rad Labs, Rchmond, Calf, USA) to Pasteur ppettes. Further addtons to the column ncluded 10 ml water (to elute free 3H-nostol) and 5 ml 5 mmol/ Borax/60 mmol/l sodum formate (to elute GP). Eluton of the nostol phosphates was accomplshed by the sequental addton of 10 ml 0.1 mol/ formc acd/0.2 mol/l ammonum formate (P0, 0.1 mol/l formc acd/ 0.4mol/l ammonum formate (P2) and 0.1 mol/1 formc acd/ mol/1 ammonum formate (P3). Ths procedure does not dfferen- tate between varous nostol trsphosphate somers. Alquots (0.4 ml) of the eluate were then analysed for radoactve contents. The radosotope used to measure nsuln release (1251-nsuln) was purchased from New England Nuclear (Boston, Mass, USA) and the myo[2-3hlnostol from Amersham (Arlngton Hts., 11, USA). Leu- cne and alpha-ketosocaproate (sodum salt), borax, formc acd and D-glyceraldehyde were purchased from Sgma Chemcal Co.
W. Zawalch: Memory n rat slets nvolves phosphonostde hydrolyss 437 (St.Lous, Mo, USA). Ammonum formate was purchased from Fsher Scentfc Co. (Far Lawn, N J, USA). Tolbutamde (sodum salt) was the generous gft of Upjohn Co. (Kalamazoo, Mch, USA). Statstcal analyss Where approprate, statstcal sgnfcance was determned usng the Student's t-test for unpared data and a p value less than 0.05 was taken as sgnfcant. Values presented n the fgures represent means + SEM of the specfed number of observatons. Results After a 2 h ncubaton perod wth myo[2-3h]nostol to label ther nostol-contanng phospholpds, batches of slets were perfused wth 2.75 mmol/l glucose for 30 ran to establsh basal nsuln secretory rates. Subsequent exposure to 15 mmo/1 KC or 20 mmol/1 leucne (n the contnued presence of 2.75 mmol/1 glucose) was accompaned by an ncreased output of nsuln (Fg.l, top). The response to KC (15 mmol/) was sgnfcantly (p < 0.05) greater at most tme ponts than that noted to leucne (20 mmol/l). nsuln release rates n response to both agonsts tended to fall somewhat durng the fnal mnutes of the 60 mn stmulaton perod. A stuaton smlar to that noted wth nsuln secreton, at least n quanttatve terms, was observed when 3H efflux rates were montored n these same slets (Fg., bottom). Whle prestmulatory effux rates were comparable n the presence of 2.75 mmol/l glucose C.O, J LJ ry" Z _.. CO Z 1 Gz.7~ Gz.75 + KC 800~-, ~,.~,_ 400J 2oo - L f 4.0[- E ' ~ 3.0 2.oL " J Gz75 t ~, o F- ' ~ 10- < X; t.l_ r Gz.75+ KC J / L. 0 L, ~- ----L.~ 0 '3b 40 50 co 30 40 5'o co DURATON OF PERFUSON (ran) Fg.2. Nonradoactve nostol ncreases ~H efflux n response to KC. Groups of 60-80 slets were ncubated for 2 h wth myo[2-3hl nostol, washed wth fresh medum and then perfused. nsuln secretory rates (left panel) and fractonal rates of 3H efflux (rght panel) were measured n response to 15 mmol/l KC alone (9 or KC plus 1 mmol/ nonradoactve nostol (0). Fractonal efflux rates of 3H n response to 2.75 mmol/l glucose plus 1 mmol/ nostol are also depcted (A) T, Table 2. Effect of alpha-ketosocaproate (KC) stmulaton on slet and effluent levels of 3H-nostol (NS), glycerophosphonostol (GP), nostol monophosphate (P1), nostol bsphosphate (P2) and nostol trsphosphate (P3) slets NS GP P1 P2 P3 treatment (cpm/40 slets) A. G27s---+Gz75 B. Gzvs~G2.75+nostoll C. Gz75----~G2.75 + K1C15 D. Gz75---*G2.75 + KC15 + nostoll 1422+241 97+- 8 415+ 23 94+10 69+12 2084+217 110+13 340+_ 25 69+18 57+14 3118 + 193 98 + 17 1143 + 118 512+-38 266+ 21 3978 + 268 69 ::t: 18 676 +_ 53 240 +_ 13 122 + 16 Effluent NS GP Total nostol phosphates P1 lp2 (cpm/40 slets/10 ran) E. G2.75 147+ 16 39+ 7 28+6 F. Gz75+nostoll 317+ 47 41+ 6 35+8 G. Gz75+KC15 896+ 73 51+_ 6 44+_7 H. G2.75 + KC15 + nostoll 1294 + 139 69 +- 12 46 + 9 P3 After a 2 h ncubaton wth (2-3H)nostol to label the phosphonostdes, groups of 60-80 slets were washed and then perfused for 30 mn wth 2.75 mmol/1 glucose to establsh basal stable 3H efflux and nsuln secreton rates. Some slets (A) were mantaned wth 2.75 mmol/l glucose for an addtonal 30 mn whle other groups were exposed to 2.75 mmol/l glucose plus 1 mmol/ nostol (B), 2.75 mmol/l glucose plus 15 mmol/ KC (C), or 2.75 mmol/l glucose plus 15 mmol/l KC plus 1 mmol/l nostol (D). The numbers gven n subscrpt ndcate the mllmolarty of the compound. At the termnaton of the perfuson, the slets were placed n 10% perchlorc acd (PCA) and the varous compounds measured as descrbed n the methods secton. Effluent samples were collected durng the fnal 10 mn of the perfuson wth the substances ndcated and pooled 5 ml samples acdfed wth 500 ~1 PCA. After neutralsaton wth 6 N KOH and centrfugaton, samples were analysed for content of the varous substances. At least 4 experments were performed under each condton. Mean values+ SEM are gven. Because the levels of the ndvdual labelled nostol phosphates were so low n the effluent, they were extracted togehter by the addton of 10 ml 0.1 mol/1 formc acd + 1 mol/1 ammonum formate. Statstcal analyss: lne A vs lne B - p < 0.05 for nostol and P1 ; lne C vs lne D - p <0.05 for nostol, P1, P2 and P3; lne E vs lne F, p <0.05 for nostol; lne G vs lne H, p <0.05 for nostol
438 W. Zawalch: Memory n rat slets nvolves phosphonostde hydrolyss G2.75 G2.75 G2.75 Gto r E G2.75 300 G2.75 Gto Glo 120 /./, / -~ 300 1 o 1 2o01 / C, E O_/fl s O s / o.'''~ 1 t 60 70 80 90 30 40 50 DURATON OF PERFUSON (mn) Fg.& Tme-dependent potentaton (memory) nduced by KC and leucne. Freshly solated slets were perfused for 30 mn wth 2.75 mmol/l glucose and for an addtonal 20 ran wth KC, 15 mmol/l (O) or leucne, 20 mmol/l (0). After a 10 ran washout perod wth 2.75 mmol/l glucose alone, the slets were stmulated for 30 mn wth 10 mmol/l glucose (n =5 for each condton). The response of slets perfused wth 2.75 mmol/l glucose alone ([]) pror to stmulaton wth 10 mmol/l glucose s also depcted. The responses of control slets, mantaned for 60 or 70mn wth 2.75 mmol/l glucose pror to stmulaton wth 10 mmol/1 glucose were vrtually dentcal and pooled values for these experments (n =18) are presented here. The control responses presented n Fgure 4 (left panel) were smlarly calculated (approxmately 0.25%/mn), the addton of KC or leucne was accompaned by quanttatvely dsparate ncreases n perfusate 3H. KC (15 mmol/l) agan proved to be sgnfcantly (p<0.05) more effectve than eucne (20 mmol/l). Also, and smlar to nsuln release rates, 3H efflux rates fell somewhat durng the fnal phase of stmulaton wth both compounds. A 20 mn exposure to KC or leucne also resulted n a sgnfcant ncrease (p < 0.05) n the levels of slet nostol phosphates (Table 1, lnes 2 and 5). Consstent wth the nsuln release results and 3H efflux data, the effect of KC was sgnfcantly greater (p <0.05) than eucne. KC sgnfcantly (p <0.05) ncreased slet levels of labelled nostol phosphates wthn 2 mn after KC exposure (Table 1, lne 4). n a prevous report [15], we demonstrated that the ncluson of nonradoactve nostol (1 mmol/1) n the perfuson medum mproved our ablty to montor glucose-nduced ncreases n 3H efflux. A smlar stuaton occurs wth KC-stmulated slets (Fg.2). Whle nonradoactve nostol had no effect on KC-stmulated secreton (left panel), ts ncluson wth 15 mmo/1 KC (rght panel) resulted n a sgnfcantly greater 3H efflux response (p<0.05), a response that W W _J W Z _J Z 2~176 100 l F o~ OL~/ 10 20 30 10 20 30 DURATON OF STMULATON (mn) Fg.4a and b. Duraton of memory nduced by KC or leucne. a Smlar to the protocol outlned n Fgure 3, slets were stmulated wth KC (9 or leucne (0) for 20 mn. After a 20 ran washout wth 2.75 mmol/l glucose, the slets were stmulated wth 10 mmol/ glucose. Only the duraton of ths stmulatory perod s shown. The control response of slets perfused for 60 or 70mn wth 2.75 mmol/l glucose ([]) pror to stmulaton wth the hgher hexose level s also shown, h A stmulatory protocol smlar to that descrbed prevously was employed except that the nterval between KC(O ) or leucne(q) exposure and 10 mmol/l glucose stmulaton was extended to 60 ran. The response of slets mantaned wth low glucose for 110 mn pror to stmulaton s also shown ([]) was nearly as rapd n onset as nsuln secreton. Furthermore, the ncluson of cold nostol sgnfcantly ncreased (p <0.05) slet content of free 3H-nostol but sgnfcantly reduced (p<0.05) the levels of labelled nostol phosphates n response to KC stmulaton (Table 2). Analyss of effluent radoactvty demonstrated that free 3H-nostol was the major radoactve moety n the effluent of control and stmulated slets (Table 2). Approxmately 70% of the label n the perfusate from control slets (perfused wth 2.75 mmol/l glucose) was free 3H-nostol, a value that ncreased to over 90% when slets were stmulated wth KC alone or n the presence of 1 mmol/1 nonradoactve nostol. Our prevous results wth other B-cell stmulants [5-7, 16] support the concept that events n the nostol cycle play a crucal role n the nducton and mantenance of TDP. Snce both KC and leucne actvate P hydrolyss (Fg.l, Table 1) [8-10], we next nvestgated whether both compounds heghten the secretory response to a subsequent glucose stmulus. The results are shown n Fgure 3. n these experments, freshly solated slets were perfused for 30mn wth 2.75 mmol/l glucose before beng provoked for 20 mn wth KC (15 mmol/l) or leucne (20 mmol/1). n response to ether compound, nsuln output rose. Followng a 10 mn perod wth 2.75 mmol/l glucose
W. Zawalch: Memory. n rat slets nvolves phosphonostde hydrolyss oo-.,c (z.?,~. LEUCNE; ~ G z.r~._. 500 Gz75 G2.75 +TOL G2.75 G2.75 + GLY 625 F, " 439 o.; JJ g G-, LEUCNE~, G275 10 z.75 KC ]r -,:ao t..),.y (3 4/1 / L L,- 30 40 50 60 70 80 90 110 120 DURATON OF PERFUSON (mn} Fg.5. Sustaned effect of KC or leucne on 3H efflux. After a 2 h prelabellng perod, slets were perfused for 30 mn wth low glucose and then stmulated for 20 mn wth KC, 15 mmol/ (O) or leucne, 20 mmol/l(o). The perfuson was contnued for an addtonal 50 mn wth 2.75 mmol/l glucose alone (n =4 for each condton). nsuln release (top) and 3H effux rates (bottom) were measured alone, durng whch tme nsuln release rates rapdly fell to prestmulatory values, the slets were restmulated, ths tme wth 10 mmol/l glucose. The response of control slets mantaned fro 60 or 70 mn (Fg.3 legend) wth 2.75 mmol/l glucose alone before beng provoked wth 10 mmol/l glucose s also shown. When compared to the nsuln response from these control slets, pror exposure to leucne or KC was accompaned by a greater nsuln secretory response to 10 mmol/ glucose. Partcularly dramatc s the frst phase response. n control slets, peak frst phase release averaged 66+4 (n = 18) pg.slet -1- mn -1. Ths value ncreased to 204+14 and 246+ 11 pg.slet -t. mn -t n the leucne-and KC-pretreated groups respectvely. Release rates measured 25-30 mn after the onset of stmulaton were also sgnfcantly greater (p <0.05) n the KC-and leucne-treated slets. f the nterval between the exposure to KC or leucne and 10 mmol/l glucose was extended to 20 or 60 mn, TDP was observed only n the 20 mn nterval group (Fg. 4, left panel). n the case of KC, both frst and second phase responses were amplfed. Wth leucne preexpo- ~" 400 " "~ 200-- " t 500 t - 375./ ;" -./ 125[- z - j / OOj ] oljl 0 10 20 30 0 10 20 30 DURATON OF STMULATON (mn) Fg.6. Tme-dependent potentaton nduced by KC to other agonsts. slets were stmulated for 20 mn wth KC, 15 mmol/ (O), perfused for an addtonal 10 mn wth 2.75 mmol/l glucose alone, and then stmulated wth 2001.tmol/l tolbutamde (left panel) or 5 mmol/1 glyceraldehyde (fght panel). Control responses of slets not prevously stmulated wth KC are also depcted (O). Only the 30 mn stmulator3' perod wth tolbutamde or glyceraldehyde s depcted here. Selected tme ponts were analysed for statstcal sgnfcanoe, *p <0.05. At least 3 experments were performed under each condton sure, only the frst phase nsuln secretory response was sgnfcantly ncreased above control release rates. Extendng the nterval between stmulant presentatons to 60 mn resulted n the abolton of TDP (Fg.4, rght panel). We next examned the mpact of a 20 mn stmulatory perod wth ether K1C or leucne on the efflux of 3H. The results are gven n Fgure 5. Whle acute exposure to ether compound was accompaned by nsuln output (Fg. 5, top), nsuln secreton rates rapdly subsded wthn 10 mn after stmulant removal. n these same slets, 3H efflux rates slowly ncreased wth agonst addton (Fg. 5, bottom) but, unlke nsuln secretory rates, remaned elevated long after KC or leucne removal from the medum. n the case of KC, efflux rates remaned sgnfcantly elevated above prestmulatory values for 40 mn. n the case of leucne, efflux rates remaned elevated above prestmulatory rates for 28 mn after stmulant removal. Of partcular relevance to the noton that the persstent ncrease n P hydrolyss regulates TDP, the duraton of the ncrease n 3H efflux paralleled to a large extent the duraton of TDP (see Fg. 4). The concept that pror KC exposure results n a persstent ncrease n P hydrolyss was substantated further by drect measurements of labelled nostol phosphates n pror KC-stmulated slets (Table 1). Frst, KC rapdly and sgnfcantly ncreased (p < 0.05) the levels of all nostol phosphates measured (Table 1,
440 compare lnes 1 and 3 wth lnes 4 and 5). These levels fell after a 10 mn perfuson wth 2.75 mmol/l glucose alone (Table 1, lne 7), but, n the case of P1 and P2, remaned sgnfcantly hgher than control values (p < 0.05) (lne 1). n other experments control (Table 1, lne 6) or pror KC-treated slets (Table 1, lne 8) were perfused for an addtonal 20 mn wth 2.75 mmol/l glucose plus 10 mmol/ lthum chlorde. Lthum chlorde, by blockng phosphatase acton on nostol phosphates [17], allows these compounds to accumulate ntracellularly. n control slets treated wth 2.75 mmol/1 glucose alone (Table 1, lne 6), ths manpulaton has no sgnfcant effect on levels of these compounds, ndcatng that wth low glucose P hydrolyss s low. However, n the slets prevously exposed to KC, dramatc elevatons of nostol phosphates, partcularly n P1, were noted (Table 1, lne 8). Snce we had already establshed that these nostol phosphate levels decrease after removal of KC from the perfuson medum (Table 1, lne 7), a reasonable explanaton for ths dramatc ncrease n these nostol phosphates s a sustaned actvaton of P hydrolyss nduced by pror KC exposure. Despte these dramatc ncreases n slet nostol phosphate levels under ths condton (pror KC and subsequent exposure to LC), there was no parallel ncrease n nsuln output from these slets (data not shown). We attrbute ths observaton, at least n part, to the nhbtory effect of lthum chlorde on nsuln output [6]. Fnally, experments were desgned to test the specfcty of TDP nduced by KC. The results from these studes (Fg.6) demonstrate that a pror 20 mn exposure to KC results n an amplfed secretory response to ether 200 ~mol/1 tolbutamde (wth 2.75 mmol/l glucose) or 5 mmol/1 D-glyceraldehyde. Dscusson n a seres of precedng publcatons [5-7, 16], we have explored the possble nvolvement of bochemcal events set nto moton by an ncrease n P hydrolyss n the nducton of B-cell memory to glucose. The phenomenon of memory has been known for some tme but ts bochemcal bass has yet to be frmly establshed. Our prevous results wth cholecystoknn [5], glucose [6], glyceraldehyde [6], tolbutamde [16] and a report by Sorenson [18] wth the phorbol ester TPA, all support the concept that events n the nostol cycle are somehow ntmately nvolved n ths process. n the present study, we have expanded the lst of compounds that nduce memory to glucose and, agan, P hydrolyss seems to be an ntegral component of ths phenomenon. n the present experments, we employed KC and leucne to actvate the B cell secretory apparatus. Prevous studes [8, 9] have establshed that both com- W. Zawalch: Memory n rat slets nvolves phosphonostde hydrolyss pounds ncrease the ncorporaton of 32p nto slet nostol contanng phospholpds. KC (10 mmol/1) has also been shown to elewte the levels of labelled nostol phosphates from slets prelabelled wth [2-3H] nostol [10]. A prevous report by Clements et al. [19], however, faled to document a stmulatory effect of leucne or KC on recovery of 3H nostol prelabelled P. Ths observaton may be attrbutable, at least n part, to the low rates of 3H ncorporated nto P and the low levels (8-10 mmol/) of leucne or K1C used n these studes. Even at 20 mmol/l, the mpact of leucne on P hydrolyss s modest. The present studes ndcate that the quanttatve mpact of KC (15 mmol/l) on both nsuln release and P hydrolyss s sgnfcantly greater than leucne (20 mmol/l). t should be emphaszed that the ratonale for usng 3H efflux rates as the ndex of P hydrolyss s based on several pertnent observatons prevously reported. Frst, 3H-nosto s exclusvely ncorporated nto nostol-contanng phospholpds [20, 21]. Second, an ncrease n perfusate 3H s paralleled by a decrease n labelled nostol-contanng phospholpds [20]. Thrd, t s possble to smultaneously montor the knetcs of P hydrolyss and nsuln output n the same slets wth ths approach [14-16, 20]. Fourth, the functonal ntegrty (nsuln release) of the slets can be convenently establshed [14-16, 20], an ssue not often addressed n many prevous reports on P hydrolyss n slets. Along these lnes, for example, lthum chlorde s often ncluded n the medum to facltate the measurements of nostol phosphates generated durng agonst-nduced P hydrolyss. We recently reported [6], n agreement wth other studes [22, 23] that the level (10 mmol/) of lthum commonly employed n many of these studes actually nhbts nsuln output, thus makng the extrapolaton of results obtaned n the presence of ths caton to the dynamc nsuln secretory response tenuous at best. Fnally, and most mportantly, the 3H efflux data were substantated, both qualtatvely and quanttatvely, by drect measurements of labelled nostol phosphate accumulaton n perfused slets. Ths fndng adds further credence to the concept that 3H efflux from 3H-nostol prelabelled slets s ndeed a vald ndex of an ncrease n cellular P hydrolyss. One possble problem wth usng 3H efflux rates to montor P hydrolyss n slets s the fact that whle nosto phosphates accumulate rapdly n response to KC (Table 1, lne 4), sgnfcant ncreases n 3H efflux lag somewhat behnd both ths response and the ncrease n nsuln output. A smlar stuaton occurs wth glucose-stmulated slets [15, 20]. However, we recently reported that the ncluson of nonradoactve nostol mproved our ablty to knetcally montor P hydrolyss usng ths efflux methodology. Smlar studes were conducted wth KC-stmulated slets and the followng major observatons were made: (a) nonradoactve nostol has no apprecable effect on nsuln output
W. Zawalch: Memory n rat slets nvolves phosphonostde hydrolyss 441 (Fg. 2, left panel); (b) however, ts ncluson results n a markedly amplfed 3H effux response to KC stmulaton (Fg.2, rght panel); (c) nonradoactve nostol ncreases the levels of free 3H-nostol n both control and KC-stmulated slets (Table 2); (d) fnally, nonradoactve nostol attenuates the ncrease n slet levels of labelled nostol phosphates noted wth KC stmulaton (Table 2). These fndngs suggest that an ncrease n 3H efflux (prmarly free 3H-nostol, Table 2) most accurately reflects ncreases n P hydrolyss provded cold nostol s ncluded n the medum. We further suggest that cold nostol competes wth lablsed free 3H-nostol (derved from 3H-P) for rencorporaton back nto P. The fact that nonradoactve nostol ncreased slet levels of free 3H-nostol, ncreased 3H efflux but reduced levels of labelled nostol phosphates agrees wth ths suggeston. By competng wth ths rapdly expandng pool of free 3H-nostol for rencorporaton nto P, nonradoactve nostol ncreases the pool of free 3H-nostol allowng more of ths moety to efflux from the slet and appear n the effluent. n the absence of nonradoactve nostol, however, rencorporaton of 3H-nostol back nto P reduces effux of the label from the slet but contrbutes to the slets' ablty to mantan sustaned ncreases n levels of labelled nostol phosphates (and perhaps nsuln secreton as well) after stmulant presentaton (Table 1 and 2). The most cogent data supportng the concept that exposure to KC results n a persstent ncrease n P hydrolyss was obtaned by drectly measurng the accumulaton of labelled nostol phosphates durng and after KC stmulaton (Table 1). The results demonstrate that KC rapdly ncreases Pt, Pz and P3 levels n slets. Removal of KC s paralleled by a reducton (although not to prestmulatory values) n the slet content of these compounds. Part of the persstent 3H efflux response observed from these pror KC stmulated slets may be the result of the degradaton of these preformed nostol phosphates and/or the efflux of ntracellular free nostol (Table 2). However, f these prevously KC-treated slets are subsequently perfused wth lthum chlorde, cellular levels of these nostol phosphates are elevated above those noted n control, non-kc treated slets. The results obtaned n the presence of lthum (Table 1) support the concept that pror KC stmulaton results n a persstent ncrease n P hydrolyss va bochemcal mechansms that reman to be elucdated. When taken together wth our other reports on the same topc, one s left wth the mpresson that P hydrolyss partcpates n agonstnduced nsuln output and prmes the B cell to any subsequent stmulaton. Consderng the nate complexty of events assocated wth P hydrolyss [24], t s not yet possble to defne precsely how nsuln secreton and TDP are nduced by Pl-derved second messenger molecules. Ths promses to reman a fertle area for future studes. Acknowledgements. These studes were supported by the Amercan Heart Assocaton 84-709. The expert techncal assstance of K. C. Zawalch and V. A. Daz, as well as the expert secretaral of Ms. J. Fettes, s gratefully acknowledged. References 1. Grodsky GM, Curry D, Landahl H, Bennett LL (1969) Further studes on the dynamc aspects of nsuln release n vtro wth evdence for a two compartmental storage model. Acta Dabetol tat 6 [Suppl 11:554-579 2. Grll V, Adamson U, Ceras E (1978) mmedate and tme-dependent effects of glucose on nsuln release from rat pancreatc tssue. Evdence for dfferent mechansm of acton. J Cln nvest 61 : 1034-1043 3. Grll V (1981) Tme and dose dependences for prmng effect of glucose on nsuln secreton. Am J Physol 240:E24-E31 4. Ceras E (1975) Potentaton of nsuln release by glucose n man: role of the nsuln response and enhancement of stmul other than glucose. Acta Endocrnol 79:502-510 5. Zawalch WS, Daz VA (1987) Pror cholecystoknn exposure senstzes slets of Langerhans to glucose stmulaton, Dabetes 36:118-122 6. Zawalch WS, Daz VA, Zawalch KC (1988) Role of phosphonostde metabolsm n the nducton of memory n solated rat slets. Am J Physol 254:E609-E616 7. Zawalch WS, Daz VA, Zawalch KC (1987) Cholecystoknn-nduced alteratons n beta cell senstvty: duraton, specfcty and nvolvement of phosphonostde metabolsm. Dabetes 36: 1420-1424 8. Fex G, Lernmark A (1975) Effects of nsuln secretagogues on phospholpd metabolsm n pancreatc B-cells. Bochem Bophys Acta 388:1-4 9. Best L, Malasse WJ (1983) Effects of nutrent secretagogues upon phospholpd metabolsm n rat pancreatc slets. Mol Cell Endocrnol 32:205-214 10. Best L (1986) A role for calcum n the breakdown of nostol phospholpds n ntact and dgtonn-permeablzed pancreatc slets. Bochem J 238:773-779 11. Lacy PE, Kostanovsky M (1967) Method for the solaton of ntact slets of Langerhans from the rat pancreas. Dabetes 16: 35-39 12. Albano JDM, Ekns RP, Martz G, Turner RC (1972) A senstve, precse radommunoassay of serum nsuln relyng on charcoal separaton of bound and free moetes. Acta Endocrnol 70: 487-509 13. Berrdge M J, Dawson RMC, Downes CP, Heslop JP, rvne RF (1983) Changes n the level of nostol phosphates after agonstdependent hydrolyss of membrane phosphonostdes. Bochem J 212:473-482 14. Zawalch WS, Takuwa N, Takuwa Y, Daz VA, Rasmussen H (1987) nteractons of cholecystoknn and glucose n rat pancreatc slets. Dabetes 36:426-433 15. Zawalch WS, Zawalch KC (1988) Phosphonostde hydrolyss and nsuln release from solated perfused rat slets: studes wth glucose. Dabetes (n press) 16. Zawalch WS, Zawalch KC (1988) nducton of memory n rat pancreatc slets by tolbutamde. Dependence on ambent glucose level, calcum, and phosphonostde hydrolyss. Dabetes 37: 816-823 17. Hallcher LM, Sherman WR (1980) The effects of lthum on and other agents on the actvty of myo-nostol-l-phosphatase from bovne bran. J Bol Chem 255:10896-10901 18. 19. Sorenson RLC (1986) slet prmng by phorbol ester. Horm Metab Res 18:353-354 Clements RS Jr, Evans MH, Pace CS (1981) Substrate requrements for the phosphonostde response n rat pancreatc slets. Bochem Bophys Acta 674:1-9
442 W. Zawalch: Memory n rat slets nvolves phosphonostde hydrolyss 20. Axen KV, Schubart UK, Blake AD, Flescher N (1983) Role of Ca 2~ n secretagogue-stmulated breakdown of phosphatdylnostol n rat pancreatc slets. J Cln nvest 72:13-21 21. Clements RS Jr, Rhoten WB (1976) Phosphonostde metabolsm and nsuln secreton from solated rat pancreatc slets. J Cln nvest 57:684-691 22. Anderson JH, Blackard WG (1979) Effect of lthum on pancreatc slet nsuln release. Endocrnology 102:291-295 23. Fontela T, Hermeda OG, Gomez-Acebo J (1987) Dhydroergotamne but not naloxone counteracts lthum as an nhbtor of glucose-nduced nsuln release n solated rat slets n vtro. Dabetologa 30:183-187 24. Zawalch WS (1988) Modulaton of nsuln secreton by phosphonostde-derved second-messenger molecules. Dabetes 37: 137-141 25. Borle A, Uchkawa T, Anderson JH (1982) Computer smulaton and nterpretaton of 45Ca efflux profle patterns. J Membr Bol 68:37-46 Receved: 28 December 1987 and n revsed form: 26 Aprl 1988 Dr. W.S. Zawalch Yale Unversty School of Nursng 855 Howard Avenue P.O. Box 9740 New Haven, CT 06536-0740 USA