Diabetologia 1, 28-32 (1965)

A Method for the Study of the Metabolism of Isolated Mammalian Islets of Langerhans and Some Preliminary Results

B y

H A R ~ r KEEN, ROBERT SELLS a n d R . J o ~ N JARRETT From the Depar tment of Medicine; Guy's Hospi tal Medical School, London,

Received February 1, 1965

Summary. Isolated islets of Langerhans, micro-dissec- ted from the duct-l igated pancreas in the rat , survived for at least four hours of incubation, their rate of glucose oxidation continuing linearly during this period. Oxida- t ion of 1 - ltC-glueose was considerably higher when the glucose concentration in the medium was raised and com- parison with other tissue suggested tha t glucose oxidation of the islets, in part icular t ha t by the hexose monophos- phate pathway, was specifically s t imulated by the higher glucose concentrations. This may well be directly associa- ted with the glucose stimulus to insulin release. Act iv i ty of the hexose monophosphate pa thway has previously been demonstrated in both fish and mammal ian islets and in human islet cell tumours, both by enzyme estimations and by studies with labelled glucose, and the suggestion has been made tha t act ivi ty of the pa thway may be l inked with insulin synthesis (FIELD, 1964).

This work was supported by Grant No. AM 6135 MET of the Nat ional Inst i tutes of Health, Bethesda, Maryland. Our thanks are due to Dr. JA~ES B. FIELD, in whose la- boratories the earliest studies were init iated.

Rdsumd. Des ilots de Langerhans isol~s par micro- dissection chez le rat , apr@s ligature des canaux pancr@ati- ques, survivent pendant au moins quatre heures d'ineu- barton (dn vitro~), leur ry thme d 'oxydat ion du glucose se poursuivant d 'une mani~re constante pendant cette p6- riode. L 'oxydat ion du 1--~4C-glucose 6tait consid6rable- merit plus 61ev6e lorsque la concentration en glucose du milieu @fair augment6e, et la comparaison avec le com- por tement d 'autres tissus a sugg@r6 que l 'oxydat ion du glucose par les ilots, en particulier celle qui s'effeetue par la vote de l 'hexose monophosphate, 6tait stimul@e spgci- fiquement par les concentrations les plus halites en glucose.

Ce phgnomgne pourrai t bien 6tre directement assoei@ avec le stimulus que le glucose exeree pour libgrer l 'insuline. L 'aet ivi t6 de la voie de l 'hexose monophosphate a gt6 antgrieurement dgmontr@e k la lois sur les ilots de poisson et de mammif6re ainsi que sur los eellules de tumeurs insulaires humaines, k la fois par des estimations enzy- matiques et par des 6tudes effectu@es avec du glucose marqu6, et la suggestion a gt6 faite que l 'aetivit6 de eette voie mgtabolique pouvai t 6tre li@e k eelle de la synthgse de l ' insuline (FIELD, 1964).

Zusammes,fassung. Nach Ligatur der Pankreasaus- fiihrungsg/~nge bei der Ra t t e durch Mikrodissektion g e - wonnene Langerhans'sehe Inseln k6nnen dureh Inkuba- t ion in vitro am Leben erhals werden and bleiben fiir mindestens 4 Stunden in der Lage, Glukose mit konstanter Geschwindigkeit zu oxydieren. Eine Erh6hung der Glu- kosekonzentrat ion des Mediums fiihrt zu einer betr/ieht- lieh gesteigerten Oxydat ion der Glukose-l-14C. Diese Sti- mulierung der Glukoseoxydation dutch die Inselzellen scheint selektiv zu sein, wie der Vergleieh mit der Stimu- lation, die man unter den gleiehen Bedingungen durch andere Gewebe erhglt, annehmen laI]t. Besonders die di- rekte Oxydat ion des Glukose-6-phosphats scheint gestei- gert zu sein. Es ist m6glich, dal? dieser Stoffwechseleffekt yon einer dutch Glukose gesteigerten Insulinsekretion be- gleitet wird. Das Vorhandensein des Pentosezyklus in den Langerhans 'schen Inseln von Fischen and S/iugetieren, so- wie in den Zellen yon Tumoren mit Insulinaktivit~it ist be- reits nachgewiesen worden. Diese Beobachtungen stii tzen sich auf enzymatische Messungen und vergleichende Stoff- wechseluntersuehungen mit maxkierter Glukose, undes ist mhglich, dal~ die Akt iv i ta t des Pentosephosphatzyklus mit derj enigen der Insulinsynthese verkniipft ist. (FrELD, 1964)

Based on a communica t ion given to the Meet ing of the Medical and Scientific Sect ion of the Br i t i sh Diabe- t ic Associat ion, September , 1963.

The s t u d y of is let funct ion in the m a m m a h a n pan- creas is h indered b y the dispersion of the is let t issue th rough the organ. ScOT~ (1912) f irst used the met l iod of t y ing the pancrea t ic ducts to produce a t r o p h y of the ac inar t issue, and i t is well known t h a t the first clear demons t r a t i on of the hypoglyeemic ac t ion of saline ex t rac t s of canine pancreas b y BANTING a n d BEST (1922) involved the use of th is technique. W e have a p p h e d the same procedure in the r a t to make a pre- pa ra t i on of surv iv ing m a m m a l i a n islets for in vitro metabohc studies.

Methods

a) Duct ligation and its effects. The gross a n a t o m y of the pancrea t ic ducts of the W i s t a r R a t is shown in Fig. 1, a d rawing made af ter in jec t ing I n d i a n ink

t h r o u g h the common bile duct . I n over a dozen such

prepara t ions , the a r r angemen t of the ducts was r emark- ab ly cons tant . There are two ma in ducts , b o t h drain- ing in to the bile duct . The upper dra ins t h e gastr ic and splenic por t ions of the pancreas and the lower dra ins t h a t pa r t of t h e pancreas ly ing below and be- h ind the t ransverse colon. I n p re l imina ry exper iments i t was found t h a t the lower duc t was technica l ly more diff icul t to locate and tie. Subsequent ly , therefore , the upper duc t only was t ied.

Male ra t s were anaes the t i sed wi th e ther and the pancreas exposed, t h rough a med ian ven t ra l incision, b y d rawing out the s tomach and duodenum and tu rn ing t h e m back on to the chest wall. U n d e r the dissect ing microscope (magnif ica t ion x 25) the uppe r duc t was ident i f ied and two or three l iga tures of fine silk t i ed in a p p r o x i m a t e l y the posi t ions shown in Fig. 1. The or- gans were then rep laced and the wound closed. The ope ra t ed ra t s were ki l led a t two or th ree d a y in te rva l s and the pancreas f ixed in 10% forraol-sal ine. Para f f in sections were s t a ined wi th h a e m a t o x y l i n and eosin and examined sys t ema t i ca l ly under the l ight microscope.

Vol. 1, No. i H. KEEl'; et al. : A Method for the Study of the Metabolism 29

b) Metabolic Studies. These were carried out using islets removed from rats operated upon four to six weeks previously. The atrophic portion of the pan- creas was readily identified, the normal, pink, fleshy aeinar tissue being replaced by glistening, t ransparent

~rosplenic omentum

~Spleen

Pancreas

Fig. 1. Pancrea t i c ducts of the Wis t a r ra t . Drawing of a specimen injected in r i v e wi th Ind i an ink th rough the conmlon bile duct. Arrows indicate

sites of l igat ion

They have a peculiarly speckled appearance, due to the many small capillaries coursing over the surface and penetrating between the cells. An individual islet was removed by grasping with the forceps the fine, duct-like cord of connective tissue to which it is at- tached (Fig. 2). Care was taken not to grasp the fragile islets between the forceps. After removal, the cleaned islets (Fig. 3) were collected in a small, black-based container, filled with chilled buffer. As each batch was completed, it was transferred to the incubation medium contained in the centre well of the micro-incubation apparatus previously described by K~]~_~ et al. (1963).

Batches of ten islets were incubated in 0.15 ml of KREBS-RIZ~G~g bicarbonate buffer together with 14C- labelled glucose and unlabelled glucose. The reaction was stopped and 14C02 expelled from the medium by the injection of 0.1 ml N/10 IIC1 into the centre well. The 14C02 was collected, during one hour's diffusion at room temperature, in 1 ml of phenylethylamine solu- tion (Woeller, 1961), injected into the vial base at the end of incubation. The phenylcthylamine solution with its content of 14C02 was then added to 10 ml of phosphor (0.3~o P.P.O. in xylene) and counted in a liquid scintillation counter (Nuclear Enterprises, Ltd.,). After removing the islets, the residual incubation medium was neutralised with 0.t ml N/10 NaOH, and stored at --20~ for the subsequent determination of

Fig. 2. Aeinar a t rophy . P h o t o g r a p h of r a t pancreas four weeks a f te r due t l igation. The islets appear as solid, ovoid bodies aga ins t the t rans lucent fa t . The forceps grasp a connect ive tis-

sue cord to which an islet is a t t ached

Fig. 3. I so la ted r a t islet ( • 500). Pho tomic rograph of mi- erodissected islet. Below, r ight , is the connect ive t issue

skeleton, by which the islet is handled

fat. After the atrophic zone was detached from the rest of the pancreas it was transferred to a black-based Petri dish containing chilled Krebs-l%inger bicarbonate buffer and teased apar t under the dissecting microscope (magnification x 25), using fine watehmaker 's forceps. The islets are ovoid, sometimes lobed bodies, often clustered in groups, and pale and solid in appearance against the translucent, f a t ty background (Fig. 2).

its insulin content, which was carried out by a modi- fication of the ra t epididymal fat pad assay of lVIART~ et al. (1958).

Results

a) Histology. Our early experience with duet - - tied rats showed that only about 20~o of rats killed six or more weeks after operation showed the necessary fa t ty

r i g . 4. Ra t pancreas : 7 days af ter l igat ion ( • 200). I n the lower pa r t of the photograph, the pancreas is normal ; above i t is par t ia l ly atrophied, w i th

cystic areas

r i g . 6. R a t pancreas : 13 days af ter I igat ion ( • 200). An area of regenera t ion (centre) of both islet and acinar t issue in a region of a t rophy and fibrosis

Fig. 5. R a t pancreas : 10 days a f te r l igat ion ( • 720). Abnormal islet (A) in region of cystic a t rophy (B). Note the wide capil lary spaces

Fig. 7. R a t pancreas : 24 days a f te r l igat ion ( x 200). Regenera t ion of a ac inar t issue has occurred. The hyper t rophied islet (centre) is demarca t ed by

capsule of connective t issue

Vol. 1, No. i I-I. KEEN et al. : A Method for the Study of the Metabolism 31

replacement of aeinar tissue. Consequently, systematic studies of pancreatic histology at varying intervals after ligation were undertaken:

1--d days: The lobe behind the ligature showed several changes. At its centre there was necrosis of the aeini and the islets, though islet cell 'ghosts' persisted. In the surviving cells, zymogen granules were scanty and the nuclei were poorly defined. At the periphery, both acinar and islet cells were swollen and capillaries were partially obliterated. Infil tration of leucocytes and exudate was seen around and at the periphery of the lobe.

4--10 days: Necrosis of the central aeini was more advanced, with fibrous infiltration but persistence of islet cell ghosts. Peripherally, there was well-marked cystic acinar dilatation (Fig. 4), with few granules visible, loss of nuclear profile, cloudy swelling of the cytoplasm and pale staining. In general, there was enlargement of the islets, but the capillaries appeared to be wide open (Fig. 5). Towards the tenth day the appearance of the peripheral islets was normal.

10--17 days: Amid the fa t ty replacement of the necrotic acinar tissue there were loci of regenerating acinar and islet cells. The peripheral aeinar tissue was still cloudy, but granules were reappearing. Cystic duets were plentiful. (Fig. 6).

17--24 days: The regenerating foei of aeinar tissue were now proliferating. In later sections, normal acinar tissue was seen interspersed with scarred glands and cystic ducts. Islets were sparse in the regenerating areas. In some later sections, however, individual islets were seen regenerating, embedded in fat or fibrous tissue and apparently with an increased blood supply (Fig. 7).

I t was concluded that, in many of the apparently unsuccessful ligation experiments, drainage of the affected lobe had been re-established via collateral ductules, with repopulation of the lobe by aeinar tissue. Examinat ion of several specimens under the dissecting microscope supported this view, for in many cases, the ligatures, which had most certainly invested the main duets, were now to be found lying free in the tissue, with a few strands of fibrous tissue within the knot. I t would appear, from the above studies, tha t the ideal t ime to use the duet-tied animals is 2 - -4 weeks after operation.

b) Metabolic Studies. Batches of ten islets were incubated with 1-14C-glucose for varying periods of up to four hours at 37~ in a shaking incubator. Fig. 8 demonstrates tha t the oxidation of glucose to C02 continued in a linear fashion for four hours of incuba- tion. Assay of the insulin content of the incubation medium (Fig. 9) showed tha t considerable amounts of insulin had been released. Allowing for dilution, the actual insulin concentration in the incubation media was estimated as 5--50 mu/ml.

In further experiments, groups of islets were incu- bated in glucose concentrations of 40 and 400 mg ~o respectively. The specific activity of the 1J4C-glucose

was the same in each case (10 ~c/mg). In one experi- ment for comparison, 20 mg pieces of rat epididymal fat were also incubated at these two glucose concen- trations. The amounts of 1~C02 produced are shown in Fig. 10. In the high glucose medium, the islets produced

2 r Th~ue e._) . . . . . . . . . 2 Control

I

0 0 I Z 3 4 z

ttOUPS of/17cuSclll'on

Fig. 8. The effect of dm'at ion of incubat ion upon ~4C0~ product ion f rom 1-~C-glueose by isolated islets. The h i s tog ram labelled zero t ime represents

the value for m e d i u m incubated wi thout islets

~ I0 "lO 3

8 , l o z

2

o Bu/~ep +

Znsukh #o Zdef

7-lT~Ue lYedi2im Medium /giTzlted II10 Wl/k

Bul~er

s pm;7t = m e o n o f '

3 uplnke~

Fig. 9. Assay of the insulin-like ac t iv i ty of the incubat ion med ia f rom the exper iment depicted in Fig. 8. The assay is based upon the glucose up take

of r a t ep id idymal fa t

%

O "

i

.;+260 % / [ ~o/o/ed, , /~§ fs/e/ (~)

/ , , / I - - - -

, ,/// / ~, +67~ Epidi~igrncl fat (s)

~o

G/ucose Concez/rg//bz rag%

Fig. 10. The effect of glucose concentrat ion upon 14CO~ product ion f rom t-zdC-glucose by isolated islets of Lange rhans and r a t ep id idymal fa t , respect ively. The solid lines represent an exper iment in which islets and fa t were direct ly compared. The broken line represents a fur ther experi- men t , in which islets alone were incubated. Three separa te pieces of fa t and two batches of islets were incuba ted a t bo th h igh and low glucose concen-

t ra t ions for three hours

32 I t . KEEN el) al. : A Method for the Study of the Metabolism Diabetologia

four t imes as much laC02 as in the low glucose medium, a much grea ter increase t han t h a t seen with the epid idy- real fat , suggest ing t h a t high glucose concentra t ions m a y have had a specific s t imu la to ry effect on ox ida t ive mechanisms wi th in the islet. A fur ther exper iment compared the ra te of ox ida t ion of 1JaC-glucose b y islets i ncuba ted a t glucose concentra t ions of 20 and 200 mg % respect ively , wi th t h a t of small f ragments of r a t pancrea t i c fa t of s imilar dimensions to the islets, t a k e n f rom the a t roph ic pancreas ad jacen t to t hem and incuba ted under the same condit ions. Fig. 11 shows the s t r ik ing difference be tween the two tissues.

~/e[-

~ 8 ~ ~ fcncreal/c

111 ~ "r~/"

, o ~ r "- o,,z zo zoo

G/ucose Concenlfo//on T3,g %

Fig. i l . A compal%on of the effect of glucose concentration upon the pro- duetion of ~4C0~ h'om l-~4C-glucose by isolated islets and by fragments of pancreatic fat, respectively. Two batches of islets and two of fat fragments were incubated at each of three concentrations of glucose for three hours

Table. Glucose concentration and 1-C and 6-C oxidation

20mg % medium 200mg % medium

200/20

cpm from i--14C02 6--~CO2 1/6

1256 948 1,3 9288 3697 2,5

7,4 3,9

F ina l ly , the ~4C02 produc t ion b y isolate dislets f rom 1-14C-glucose was compared with t h a t f rom 6-14C-glucose in low and high glucose concentra t ions (20 and 200 mg %). Table 1 shows t h a t 14C02 produc- t ion was grea ter in the high glucose med ium in bo th cases, b u t t h a t the re la t ive increase was grea ter in the case of the glucose label led in the 1 posit ion.

Discussion

Several approaches have been used in the s t u d y of ~-cell me tabo l i sm and its re la t ionship to insulin syn- thesis and release (FIELD, 1964). A grea t deal of infor- ma t ion has been ga ined f rom the s t u d y of the "pr inci - pa l i s le t" of te leost fish, b u t i t appears t h a t there m a y be i m p o r t a n t differences be tween the piseine and the

m a m m a l i a n islet. Studies of insulin p roduc t ion using perfused pancreas (GI~oI)sI~Y et al., 1963), or pieces or slices of pancreas (CooRE and RA~])LE, 1962; R-CA~DE- LA et al., 1963) f rom ma mma l s have been repor ted ; MiALtIE and MEYER (1961) used pieces of duc t - t i ed r a t pancreas . However , because of the overwhelming excess of ac inar t issue present , i t is no t possible wi th these techniques to s t u d y the re la ted i n t e rmed ia ry metab- olism of the islets. The technique descr ibed in th is paper offers a m e t h o d b y which bo th insulin p roduc t ion and metabol ic processes can be s tud ied s imul taneously .

Since this work was begun, HET,LERSTRO~ (1964) has descr ibed a m e t h o d for the micro-dissect ion of islets in the mouse, r a t or guineapig wi thou t pr ior duc t l igat ion. Hav ing since t r i ed bo th methods , we feel t h a t r emova l of the islets is easier and quicker if the ducts are first t ied. This is of some impor tance , in t h a t 80- -100 islets are requiI.ed for the average exper iment . I n addi t ion , there is the fur ther advan t age of work ing wi th t issue cer ta in ly u n c o n t a m i n a t e d b y pro teo ly t ic enzymes f rom the exocrine pancreas .

References. BATTING, F. G. and C. H. BEST : Pancreatic extracts. J. Lab. Clin. Med. 7. 464--472 (1922). -- COORE H.G. and P . J . t~ANDLE: Secretion of insulin by rabbi t pancreas in vitro. Biochem J. 84, 78P (1962). -- FIELD, J .B . : Factors concerned with insulin synthesis and re- lease. Metabolism, 13. 407--421 (1964). - - G R O D S K Y , G. M., A.A. BAT~S, L.L. BENNETT, C. VCELLA, N.B. Me- WILLIAMS, and D.F . SMIT~, Effects of carbohydrates on secretion of insulin from isolated ra t pancreas. Amer. J. Physiol. 205, 638 644 (1963). -- HELLE~ST~O~, C.: A method for the micro-dissection of intact pancreatic islets of mammals. Acta Endocrinologica. 45, 122--132 (1964). -- KEE~, H., J .B . FIELD and I. PASTAN: A simple method for in vitro metabolic studies using small volumes of tissue and medium. Metabolism. 12, 143-- 147 (1963). -- MARTIN, D.B., A.E. RENOWn and Y.M. DAGENAIS: An assay for insulin-like act ivi ty using ra t adipose tissue. Lancet. 2, 76--77 (1958). -- MIALHE, P. and V. M E , R : Secretion d'insuline par le pancreas du ra t in vitro. C.R. Acad. Sci. (Paris). 253, 1861--1863 (1961). -- R-CANDELA, J .L . , D. MAI~TIN-HERNANDEZ and T. CASTILLA-CORTA- ZAR : Stimulation of insulin secretion in vitro by adenosine tr iphosphate. Nature, 197, 1304 (1963). -- Seo~T, E .L . : On the influence of intravenous injections of an ext rac t of the pancreas on experimental pancreatic diabetes. Amer. J. Physiol. 29, 306-- 310 (1912). -- WOELLEIr F.I-I. : Liquid scintillation counting of C1~O2 with phenylethyl- amine. Analyt . Bioehem. 2, 508--511 (1961).

~Z[ARR Y KEEN, Depar tment of Medicine Guy's Hospital Medical School London, S.E. 1. England