J. Physiol. (1967), 189,pp. 247-260 247 With 7 text-figures Printed in Great Britain THE ELECTRICAL PROPERTIES OF RESTING AND SECRETING PANCREAS BY D. G. CLARK, J. R. GREENWELL, A. A. HARPER, ALISON M. SANKEY AND T. SCRATCHERD From the Department ofPhysiology, Medical School, University of Newcastle upon Tyne (Received 15 July 1966) SUMMARY 1. The electrical properties oftheresting and stimulated pancreas have been studied in the anaesthetized cat. There is an inverse relationship betweenthefrequencyofappliedalternatingcurrentandtheresistivityand dielectric constantoftherestingpancreas. Theresistivityisalsoaffectedby thefat contentofthegland. Theimpedancelocusofthepancreasissimilar to that ofother tissues. 2. On intravenous injection of secretin there is a brief increase, followed by a more marked decrease, in conductance and capacitance of the pancreas. The phase ofdecreased conductance is relatedto the flow of pancreatic juice. The decrease in conductance and the flow rate can be characterized by two closely related empirical equations. 3. During secretin stimulation the sodium concentration in pancreatic tissue increases, and the potassium concentration falls. 4. It is tentatively suggested that the decreased conductance across the secreting gland is mostly due to swelling ofthe secretory cells and to a minor degree is the result ofchanges in the composition ofthe secretion in the pancreatic ducts. 5. Inthe anaesthetized cat the mean halflife ofthe secretory action of secretin is 199 sec. INTRODUCTION There is an inverse relationship between the electrical resistivity of tissues and cell suspensions to applied alternating current and the fre- quencyofthe current, dependent onthe highresistance ofcell membranes to current oflow frequency (see Schwan, 1957). Many observations have been made on excised material and on suspensions ofred cells, in which the electrical properties of the cell membrane have been established (Fricke, 1925, 1933). Fewerobservationshavebeenmadeontissuesinsitu. Peserico (1926) andBronk & Gesell (1926) have measuredtheresistance of 248 D. C. CLARK AND OTHERS salivary gland, and both the latter workers and Van Harreveld, Potter & Sloss (1961) observed that the resistance ofthe gland increased whenthe chorda tympani nerve was stimulated. Ofthe other glandular tissues only liver (Schwan & Kay, 1956, 1957) and kidney (Lofgren, 1951) appear to have been studied. As part of an investigation of the cellular changes produced in the exocrinetissue ofthe cat'spancreas byhormonal stimulationtheelectrical properties ofthe gland have been studied, and correlated with alterations in its chemical composition. METHODS Theexperimentswereperformedonfastedcats,anaesthetizedwithchloralose(75mg/kg) and urethane (37.5mg/kg). The splanchnic nerves were cut extraperitoneally, the pyloric sphincter occluded with a tape ligature, and the pancreatic duct cannulated. Electrodes were placed on either side of the tail of the pancreas. In the earlier experiments these consisted oftwo rectangular platinum plates, but to avoid possible artifacts which might result from swelling ofthe stimulated pancreas amore elaborate three terminal electrode system, ofthe 'guardring' type described byGriffiths & Lee (1962), wasused inthelater experiments. The circular electrodes were mounted on a Perspex frame, and the lower electrodewasfixed.Thepositionoftheupperelectrode,0-8cmindiameter,couldbealtered byamicrometer attachmentwhichallowedaccuratemeasurement ofthedistancebetween theelectrodes. Thelowerelectrodeconsistedofaninnerdisk,0-52cmindiameter,surrounded byanouterguardring, 0-13cmwide,withaninsulatingzoneofPerspex,lessthan0*1mm wide, betweenthe two. The guardringwasconnected totheneutralline ofthemeasuring bridgesandtheupperandlowercentralelectrodeswereconnectedtotheirvoltage (E) and current (I) terminalsrespectively. Throughtheseelectrodes alternating currentwasappliedacrossthetailofthepancreas, and the impedance ofthe gland measured at frequencies between 1kc/s and 5Mc/s. The currentwas supplied byaMarconibeat-frequency oscillator, forfrequencies upto 30kc/s, and an Airmec 858 oscillator for higher frequencies up to 5Mc/s. The detectors were a Tektronix 502 oscilloscopeandanAirmecWaveAnalyser853,andthemeasurementswere made withtwoWayne KerrbridgesB221 andB601. Theappliedvoltagevariedbetween 05and 35mVr.m.s., andwithinthisrange the impedancewasconstant. Piecesoftissueremovedfromthetailofthepancreasforchemicalanalysiswereweighed, dried overnight at 1000 C and reweighed. The fat content ofthe drytissue was measured bythemethodofHastings &Eichelberger (1937).Theconcentrationsofsodiumandpotas- sium were determined by flame photometry after digestion ofthe dry tissue in sulphuric andnitricacids(Flink, Hastings& Lowry, 1950).Thechlorideconcentrationwasmeasured bythemethod ofSanderson (1952) afterextraction ofthewettissuewithnitric acid. The secretin used to stimulate the pancreas was prepared by the method ofCrick, Harper & Raper (1949). Resultsare expressedasmeans +s.E. (no. ofobservations). RESULTS The electricalproperties ofthe restingpancreas. Both the resistivity and dielectric constant ofthe gland varied with the frequency ofthe applied current. In six experiments the resistivity was 1031 + 107 Q cm and the dielectric constant 3-17+0x65x 105at 1 kc/s. At 2 Mc/stheyhaddecreased to 225+ 5 Q cm (9) and 6x2+2x4x 102 (9) respectively (Figs. 1 and 2). The ELECTRICAL PROPERTIES OF PANCREAS 249 impedancelocusofthepancreaswasobtainedbymeasuringtheimpedance ofthetissue overarangeoffrequenciesfrom 1 kc/sto 3 Mc/s. Ifthemeans of the specific reactance in nine experiments were graphed against the means of the resistivity the points fell on a semicircle with a centre de- pressed below the resistivity axis. When the curve was extrapolated to zero frequency the resistivity axis was cut at 986+49 Q cm. The cor- responding value forinfinite frequencywas 201 +5 Q cm. Thephase angle betweenlinesfromthecentreofthecircletotheinterceptatzerofrequency andvertical tothe resistivity axis hadameanvalue of64*9+ 1.50 (Fig. 3). The characteristic frequency was found to be 98&9+ 12-6 kc/s (Fig. 1). 1200 1000 0 800 0C) 0 ea 600 .0_ 0 0 400 CC 200 0 10 100 1000 Frequency (kc/s) Fig. 1.Therelationshipofspecificresistanceofpancreatictissuetothefrequencyof the applied alternating current. The curve is drawn through the means of nine experiments, andthebarsindicate S.E.ofthemeans.Thecharacteristicfrequency (CF), at which the specific resistance is the mean between the resistance at zero (ro) andinfinite (r.) frequency, is98-9kc/s. The fat content ofcat pancreas varies considerably in different animals (Clark, Harper & Scratcherd, 1961). Schwan (1957) hasfound that fat has a high specific resistance of 1500-5000 Q cm, andit seemedlikelythatthe total resistivity of the tissue would be influenced by its fat content. In fourteen experiments the specific resistance of the pancreas at 1592 c/s wasmeasured1hraftercompletion oftheoperativeprocedure. Theneutral fat content ofa sample ofpancreas removed at the end ofthe experiment 250 D. G. CLARK AND OTHERS was determined. It was found that there was a highly significant increase (P < 0 002) in the specific resistance ofpancreatic tissue with increase in its fat content (Fig. 4). The electrical properties of the secreting pancreas. In the fasted cat, anaesthetized with chloralose and urethane, there is little or no flow of pancreatic juice unless the gland is stimulatedby secretin to secrete water and bicarbonate. Secretin has no direct effect on enzyme secretion, 106 r- w C 0 104 0 -.0_C 9z 0 10 100 1000 Frequency (kc/s) Fig. 2. The dielectric constant ofpancreatic tissue as afunctionoffrequency. although the first sample of secretin-stimulated juice collected from a previously quiescent pancreas may contain a considerable amount of enzymes which has accumulated inthe pancreatic ducts. With continuous or repeated administration ofsecretin the enzyme content ofthe juice is very low, although the glandular tissue contains a large amount of zymogen material (Harper & Vass, 1941; Harper & Mackay, 1948). Observations were made onthe conductance and capacitance across the pancreas, secreting in response to single i.v. injections of 0 1-0*5 mg secretin. Most ofthe measurements were made at a frequency of 1592 c/s. This frequency, apartfrom its incidental convenience in the calculation of ELECTRICAL PROPERTIES OF PANCREAS 251 400 200 60 400 200 o 10 cCD)a C) N\1 .1 kc/s Cc;2 ° 0 200 Specific resistance Fig. 3. Anexample oftheimpedance locus ofpancreatic tissue. The pointsrelate the specific reactance and specific resistance over the range offrequencies from 1 to 3000kc/s. The line isthebest-fit circular arc, andthe phase angle is 640. 1200 0 0 1100 0 0 1000 a 0 c; OQ , 900 CcC) C4 800 0 700 0 2 4 6 8 10 12 14 16 Fat content ofpancreas (% byweight) Fig. 4. Theeffect offat onthe specific resistance ofthe pancreas. The line, fitted by the method of least squares, has a slope of 25-0 and a y-axis intercept at 762 Elcm, r = 0-776. 252 D. G. CLARK AND OTHERS conductivity, is sufficiently high to eliminate polarization effects, and low enoughto avoid anyappreciable currentflowacross themembranes ofthe pancreatic cells. It seemed likely therefore that electrical changes at this frequency would reflect alterations in the volume or composition of the excracellular fluid of the gland. In resting pancreas at this frequency in twenty-three experiments the range of conductivity was 0-73-1-81 m- mho/cm. After the injection of secretin capacitance and conductance showed parallel alterations, but as the capacitive component is only a 16 15 2, 14 C) 13 94 92 - 90 x 0 88 i 86 U 84 82- 80 - I SN 0 5 10 15 20 25 Minutes Fig.5.Anexampleoftheeffectsonthecapacitanceandconductance ofthepancreas oftheintravenous injection of0.5mgsecretion (SN). small fraction ofthe total impedance at this frequency (Schwan & Kay, 1957) the results will be described in terms of conductance. The rate of secretionwas determinedbyrecordingtheintervalbetweendrops ofjuice. Within 10-20 sec ofa secretin injection there was a slight briefincrease in pancreatic conductance followed by a more marked reduction in con- ductance. The duration of these phases varied with the dose of secretin, but on average the first phase of increased conductance was over in 50-100 sec, after which the conductance decreased rapidly to reach a ELECTRICAL PROPERTIES OF PANCREAS 253 minimum in about 5min, and then slowly returned to the prestimulation level in 15-20 min (Fig. 5). In nine responses to 0-5 mg secretin in seven animals the increase in conductance was 0-8+020% and the decrease 6-6+0-5% of the resting conductance level. Secretion usually began within a minute of the injection, and the rate of flow rapidly reached a maximum, followed by a gradual decline to zero. 0X3 0-2 0 0 K 2 4 6LL Fig. 6. Themeasuredrateofflowofpancreatic juice (steppedline) andchangein conductance, AG, across the pancreas (crosses) after intravenous injection of 0-25mgofsecretinatthedouble-headedarrow.Theshadedareasindicatetherate offlowand conductance change calculated from equations 1 and2 inthetext. There is some evidence (seeDiscussion) that the slight increase in con- ductanceistheresultofatransientincreaseinbloodflowandbloodcontent ofthepancreasfollowinginjectionofsecretin. Ongraphingtheconductance changesandrateofflowofpancreatic juiceitbecame clearthatthesecond phase of decreased conductance followed by a gradual return to resting level was related to the periods of increasing and declining rates of secretion (Fig. 6). The secretory response and conductance changes pro- 254 D. G. CLARK AND OTHERS duced in response to twenty-nine injections of secretin in nine animals have been examined in detail. The amount of secretin injected (0x125 0 50 mg) produced a submaximal secretory response, in the sense that the peak rate of flow lasted for a brief time whereas with larger doses the maximum flow rate persisted for several minutes. It was found that the rate offlow and the second phase ofthe conductance change in these sub- maximal responses couldbe characterized by.two closelyrelated empirical equations. The rate offlow ofjuice (dv/dt) can be represented by the equation dv = K1 In dose e-t/1r (1-e-t/"2) (1) and the change in conductance (AG) during the phase of reduced con- ductance and recovery by the equation AG = K2ln dose e-t'/r (1-et/73), (2) where K1 and K2 are constants, 'dose' = amount of secretin injected in /lg, t = the time from the beginning ofsecretion in seconds, and r1,I2 and T3 are time constants related respectively to the declining phase ofsecre- tion, the rising phase ofsecretion, andthereduction in conductance. They have been found to be: 286+ 6 sec (37), T2, 24+7 sec (6) and T3, l, 220+ 20 sec (7). From equation (1) the halflife ofthe secretory action of secretin (Ti) can be determined, as it equals 0-6932 r1,i.e. 199+4 sec. The volume of pancreatic juice secreted in ml. (V) is proportional to the logarithm ofthe dose ofsecretin injected, a relationship which is defined by integration ofequation 1 in the form T2 V = K1ln dose 1 (3) . 71+r72 In experiments at frequencies other than 1592 c/s it was found that the general pattern ofthe electrical changes after single injections ofsecretin was unaltered, but the extent of the reduction in conductance decreased at higher frequencies. This decrease appeared to bear a linear relationship to the logarithm of the frequency. The smaller capacitance change also decreased as the frequency of the applied current was increased until at the characteristic frequency no change in capacitance could be detected. Above this frequency there was a slight increase in capacitance (Fig. 7). In response to a continuous intravenous infusion ofsecretin, provided the volume oftheinfusionwasinsufficient to alterappreciablythevolumeand composition of the extracellular fluid, the conductance across the gland decreased when secretion was established and remained at this level throughout the period ofinfusion. ELECTRICAL PROPERTIES OF PANCREAS 255 Electrical characteristics ofcat blood. For comparison with the electrical measurements onthepancreas, theresistivityoftensampleseach ofwhole blood and of plasma at 38.50 C was measured. The mean results were 137+2 Q cm for whole blood and 61 +0-65 Q cm for plasma. The electrolyte composition ofthe resting and secretingpancreas. In eight animals small portions of glandular tissue were removed from the tail of the pancreas at rest and during secretion in response to injection of 1 mg secretin. The second sample wasremoved atthetime, usually about 5 min -8 66 e; U 4 - 0~~~~~~~~~~ c;2 0- o 0- +2 1 10 100 1000 Frequency (kc/s) Fig.7. Themaximumpercentagechangeinpancreaticconductance U and capaci- tance * during the second phase of the response to secretin, as a function of frequency. TABLE 1 K Na Na+K Cl H20 Restinggland 94-2+2-2 56-1+3-3 150-4+5-0 46-6+0-7 75-1+0-5 Stimulatedgland 82-1+2-7 70-4+3-2 152-5+4-3 48-1+0-9 76-3+0-7 Meanresults +SE. ofeightexperiments. Electrolytes:m-equiv/kgfat-freetissue. Water percentage offat-free tissue. after the secretin injection, when a simultaneous conductance record had reached its lowest value. The sodium, potassium and chloride concentra- tions in the tissue were determined (Table 1). In all the experiments there werereciprocalchangesinthepotassiumandsodium concentrationsduring secretion. The mean resting potassium concentration of 94-2 m-equiv/kg wet fat-free tissue fell to 82-1 m-equiv/kg on stimulation, and the cor- responding figures for sodium were 56-1 and 70-4 m-equiv/kg. The fall in potassium and rise in sodium concentration on stimulation were both 256 D. G. CLARK AND OTHERS statistically significant (P < 0'01). The slight increases in the concentra- tions of chloride, water and in the sum of the sodium and potassium concentrations on stimulation were not significant. DISCUSSION The electrical properties ofthe resting pancreas are similar to those of other tissues which have been investigated. Schwan & Kay (1557) found thatinliverat 1kc/stherangeofresistivitywasbetween765and1040Qcm andthedielectricconstant 3-2+0-85 x 105. Thecorrespondingmeanfigures for the pancreas are 1031 + 107 Q cm and 3-17+ 0X65x 105. The decrease in pancreatic resistivity and dielectric constant withincreaseinfrequency ofthe applied current is also in agreement with observations on liver and other tissues (Schwan, 1957). The similarity is emphasized by considering theimpedancelocusofthepancreasandothertissues. Theimpedancelocus of the pancreas is characterized by the phase angle of 64 9+ 1 5', which compares with 56-63' for frog sciatic nerve (Cole, 1934), 70-78' for frog gastric mucosa (Teorell & Wersaill, 1945), and figures of 670, 710 and 750 for skeletal muscle (Cole, 1934; Guttman, 1939; Schwan, 1954, respec- tively). The observed relationship between impedance oftissues and frequency ofapplied current is the resultant ofthree groups ofelectrical relaxations, designateda, /3andy. Theacandyrelaxationsinfluence themeasurements atlowandhighfrequencies respectively. Most ofthemeasurements onthe pancreas, which are the basis of the resistivity-frequency relationship (Fig. 1) and the derived impedance locus (Fig. 3), were made over the frequency range dominated by the /3 relaxation. At the lower frequencies thereisevidenceofanacdispersioninthediscrepancybetweentheobserved relationships of specific reactance and specific resistance at these fre- quencies and the impedance locus extrapolated from the main ,3 dis- persion. This accounts for the apparent anomaly of a measured mean resistivity of 1031 Q cm at 1 kc/s and an estimated 961 Q cm at zero frequency. Fromtheresistivity ofpancreatic tissue andtheproportion ofthetissue occupiedbycells approximate valuesfortheresistivityoftheextracellular and intracellular fluids may be obtained by applying the following equations from the theory of electrical resistivity of cell suspensions, developed by Fricke (1924, 1933) and Velik & Gorin (1940) r2+(2 r= ) (4) O)r2 r r (1- )rl+(2+ r-1 (1+25b)rl+2(1- r2'