Cytotoxic effect of IFN-γ plus TNF-α on human islet cells

Journal of Autoimmunity (1991) 4,291-306

Cytotoxic Effect of IFN-y plus TNF-a on Human Islet Cells

Gloria Soldevila,*-f Massimo Buscema,? Mala Doshi,? Roger F. L. James,1 Gian Franc0 Bottazzot and

Ricardo Pujol-Borrell*

*Laboratorio the Inmunologia, Hospital Universitario ‘Germans Trias i Pujol’,

P.O. Box 72, Badalona 08916 (Barcelona), Spain, t Immunology Department,

University College and Middlesex School of Medicine, Arthur Stanley House,

40-50 Tottenham Street, London WlP 9PG, and $Department of Surgery, Clinical

Sciences Building, Leicester Royal Infirmary, P.O. Box 65, Leicester LE2 7LX, UK

(Received 12June 1990 and accepted 21 December 1990)

We have previously reported that the combination of IFN-1 plus TNF-a is able to induce the de nova expression of HLA class II on human g cells.

In the present study, we have investigated the effect of these cytokines, alone or in combination, on the function and viability of human islet cells in vitro. Three hour insulin release was markedly reduced in human islet monolayer cultures after 4 days’ exposure to 1000 U/ml of the combination TNF-a plus IFN-y (36.7 I7.7, y0 of the control i SEM) or to TNF-a alone (49.5 + 7% of the control) while IFN-y had little effect. On direct inspection cell damage was clearly detected only in the cultures treated with TNF-a plus IFN-1 in which staining by indirect immunofluorescence (IFL) for insulin revealed that the number of g cells was also significantly reduced, thus suggesting a real cytotoxic effect of this cytokine combination. This effect was not g cell specific since glucagon release and the number of a cells were also reduced in the cultures exposed to 1FN-q plus TNF-a. “Cr release experiments supported the cytoxlcity of these cytoldnes to normal islet cells. There was a time course relationship between class II induction (2 days) and the cytotoxic effect of IFN-1 plus TNF-a (4 days) on the same islet cells. In conclusion, these results indicate that the combination of IFN-y and TNF-a exerts a cytotoxic effect on human islet cells in vitro.

Introduction

During the last two decades evidence has been accumulating which indicates that autoimmunity plays a decisive role in the ultimate destruction of the pancreatic islet p

Correspondence to: Prof. Ricardo Pujol-Borrell, Unidad the Inmunologia, Hospital Universitario ‘Germans Trias i Pujol’, P.O. Box 72, Badalona-08916 (Barcelona), Spain.

291

0896-841 l/91/020291 + 16 $03.00/O 0 1991 Academic Press Limited

292 Gloria Soldevila et al.

cells in Type 1 (insulin-dependent) diabetes [l]. More recently, the recurrence of insulitis in pancreata of normal twins grafted to their diabetic identical co-twins [2] and the clinical improvement of the disease with immunosuppressive therapy in

newly diagnosed cases [3] have further strengthened the evidence supporting a

pathogenetic role of the autoimmune response. The factors responsible for the initiation of this autoimmune process and the precise effector mechanism(s) of fl cell destruction remain, however, elusive.

In the pancreas of newly-diagnosed diabetic patients, islets show a number of immunological alterations which include mild lymphocytic infiltration (insulitis)

[4, 51, immunoglobulin deposition [6, 71, an overall increase in expression of HLA class I molecules and ‘ectopic’ class II expression [7-91. Remarkably, the ‘ectopic’ expression of class II and the recent finding of IFN-a [lo] are the only p-cell specific

phenomena. As in Type 1 diabetes the destruction is confined to l3 cells; there is a suggestion that these latter findings are relevant to pathogenesis.

In an attempt to identify the mediators responsible for the changes in HLA expression observed in vivo, we studied the effect of a variety of cytokines on the p

cells and found that the combination of IFN-)I plus TNF-a is required to induce

HLA class II expression by the /3 cells [ Ill. While performing these studies we noticed a reduction in the number of p cells in

the cultures in which the monolayers had been exposed to the combination of IFN-), plus TNF-a for more than 4 days. This observation was of potential interest because

it suggested a possible role of these cytokines in the destruction of p cells. In fact, there were already several reports of the toxicity of IL-l j3, IFN-y and TNF-a, alone or in combination, to islet cells [ 12-161, but most data were generated from studies

using rat and mouse islets. We decided, therefore, to extend our studies to confirm the toxicity of these cytokines to human islets. We report here experiments on

the cytotoxicity of IFN-?/ and TNF-a to human islets, their cell specificity, the synergism between these two cytokines and the relationship between cytotoxicity and HLA molecule induction in human l3 cells.

Materials and methods

Materials

Pancreatic tissue Eight human pancreata were used for the study. All of them were obtained from HLA typed cadaveric kidney donors provided by the Leicester Royal Infirmary (see Table 1).

Reagents

Cytokines Recombinant human interferon gamma (rIFN-y) and tumor necrosis factor alpha (rTNF-a) were kindly provided by Dr G. R. Adolf, Boehringer Institute, Vienna. Cytokines were purified to high specific activity (3 x lo7 U/ml and 4 x IO6 U/ml, respectively) from an E. coli recombinant DNA source (99% homogeneous; endo- toxin contamination < 0.125 ng/mg protein). IFN-y and TNF-a were used alone or

Cytotoxic effect of IFN-1 plus TNF-a 293

Table 1. Human pancreatic tissue

Pancreas

HP66 HP68 HP70 HP72 HP74 HP75 HP79 HP8 1

A3,- Al,2 A3,9 A2,lO Al,- A2,9 Al,30 Al,28

Haplotype

B7,15 cw3 B8,21 cw7 B7,22 cw7 B16,- cw4 B8,12 - B12,15 - B8,21 Cw7 B8,12 -

DR2,4 DR2,3 DR2,5 DR5,8 DR2,3 DR2,7 DR2,- DR7,7

Table 2. List of monoclonal antibodies used for double indirect immunofluorescence staining

Antibody Specificity Working

Isotype Preparation concentration Source

RFDR- 1 HLA class II IgM supernatant l/3 Dr G. Janossy non-polymorphic

PEP030 mouse anti- IgG ascites I/SO-l/l00 Dr Jorgensen human C peptide

GLUOOl mouse anti- IgG cell culture l/20 Dr Jorgensen human glucagon supernatant

LE61 anti-keratin IgG concentrated neat Dr B Lane supernatant

in combination at a concentration of 1,000 U/ml. This was found in previous studies to be optimal for HLA class II induction.

Antibodies and other reagents The specificity and source of the monoclonal antibodies (mAbs) used are listed in Table 2. mAbs were used as culture supernatants or as purified ascites, adjusting the concentration to 1-5 pg.

TRITC (tetramethyl rhodamine isothiocyanate) labelled aflinity purified goat anti-mouse IgG serum (y-chain specific) and FITC (fluorescein isothiocyanate) labelled affinity purified goat anti-mouse IgM serum (p-chain specific) were obtained from Southern Biotechnology, AL, USA.

Methods

Preparation of human islet cell monolayers Human pancreatic islets were prepared following a method previously described [17]. Briefly, introduction of collagenase (Type XI, 3-7 ml/g; Sigma, Dorset, UK) was performed via the main pancreatic duct and the gland incubated at 37°C. When


digestion was complete, the dispersed pancreas was washed with Minimal Essential Medium (Gibco, London, UK) and the islets were isolated by a discontinuous BSA

gradient formed on an IBM 2991 cell separator. Islets then underwent a second digestion using a trypsin/collagenase solution (trypsin 0.5Y0 Sigma/collagenase

0.02%, Worthington, NJ, USA) in calcium- and magnesium-free Hanks Balanced Salt Solution (HBSS, Flow, Irvine, Scotland). Aliquots of 3 x IO4 cells were plated

on 6 mm diameter glass coverslips placed in 96-well plates (Nunc, Copenhagen,

Denmark) in a final volume of 0.2 ml of Complete Tissue Culture Medium (CTCM)

(RPM1 1640 containing 2 mM glutamine, Gibco, London, UK; 5 pg/ml transferrin, Sigma; and lo-’ M hydrocortisone, Sigma) supplemented with loo/ fetal calf serum

(FCS, Seralab, Crawley, Sussex, UK). Cultures were kept for 2-3 days at 37°C in a 5Oi, CO,/air humidified incubator to allow the cells to attach and spread. After this process, the purity of the islet cell preparation was estimated to be in the range of 6O-9Oo/o, as judged by the indirect immunofluorescence staining of monolayer cul-

tures with anti-insulin antibodies. At days 3 and 7, the medium was changed and the

cultures supplemented with the cytokines.

Cell lines The human pancreatic cell line HP 62 was generated in the London laboratory by transfection of islet monolayer cultures with the plasmid pX-8 which contained SV40 early region [ 181. This cell line retained insulin production during the initial

six passages and has grown in continuous culture for 2 years, maintaining epithelial features [19 and submitted]. The mouse fibroblastoid cell-line L929 [20], the human

pancreatic exocrine cell-line BxPC-3 [21] and the human thyroid cell-line HT 93 (also generated by transfection with pX-8) [22], were used as cell specificity controls

in these experiments.

Assessment of cytotoxicity Four different assays were performed to assess the cytotoxic effect of cytokines to the islet cells: insulin and glucagon release, direct inspection on the inverted microscope,

differential counting of the 8, a and ductal cells in the monolayers, and specific 5’Cr

release assay.

Insulin and glucagon release assay. At days 0, 4, 7 and 11, after the addition of cytokines, cultures were washed and incubated for 3 h at 37°C with 0.2 ml of fresh CTCM without cytokines. At the end of this period, the medium was collected for

the insulin and glucagon assays, and the cultures were stained by double IFL for HLA expression and hormone/cytokeratin content.

Insulin release was measured by insulin radioimmunoassay (RIA) [23] with a detection limit of 6 pU/ml, an intra-assay coefficient of variation (CV) of 3”b and inter-assay CV of 9%.

Glucagon release was measured using an RIA kit (ICN, Biomedicals Inc, CA) with

a detection limit of 25 pg/ml and an intra-assay CV of 2-6%.

Morphological assessment. Cultures were inspected daily by phase contrast on an inverted microscope (Wild Leitz Ltd, UK). Degree of cell spreading, extension of

Cytotoxic effect of IFN-y plus TNF-a 295

the monolayer, presence of floating debris, vacuolization, pycnotic nuclei and other

morphological changes were recorded.

Cell counting in the monolayer cultures. To assess the number of p, a and epithelial/ ductal cells in the monolayers, cultures were stained by IFL using cell specific monoclonal antibodies (Table 2). Staining was performed on days 0, 4, 7 and 11

following a protocol previously described [24]. Briefly, cells were first stained for surface HLA class II molecules using mouse mAb RFDR-1 followed by FITC-

labelled goat anti-mouse IgM serum. After washing, the monolayers were fixed

with chilled acetone/methanol 501 b v/v to expose cytoplasmic antigens and then

stained with mAbs to human C-peptide, glucagon or to cytokeratin followed by TRITC-labelled goat anti-mouse IgM serum.

Non-specific staining was ruled out by labelling parallel cultures with an unrelated mAb of the same immunoglobulin class or normal mouse serum at a dilution of 1: 500. Undesirable cross-reactivities among reagents from different species were assessed by omitting one of the layers in turn. All preparations were examined with a

x 63 oil immersion objective under a UV Zeiss III photomicroscope. Duplicate

cultures were evaluated independently by two observers. Only experiments in which the control cultures contained at least 200 cells were considered valid. The inter-observer intra-assay CV was of 6.2”, .

Specific “Cr release assay. Both primary cultures of islet cells (SO”, pure) and the HP 62 human pancreatic islet cell line (tested Mycoplasma free) were used as

substrate for this assay.

Labelling of the cells: human normal islet cells were used at culture day 2 to 3 and the HP 62 cells when 50v,,, confluent. Cultures in 96-well plates were washed twice

with RPM1 1640 alone (FCS was found to interfere with 51Cr uptake by islet cells), and each well was supplemented with 20 ~1 of ?r (2 &i per well, sodium chromate, Amersham, London, UK) and kept for 10 min at 37”C, after which another 80 ~1 of RPM1 1640 were added. Plates were incubated for another 50 min at 37°C and cultures were washed with 51Cr-free medium. The proportion of initial 51Cr uptake

was determined by measuring the radioactivity after lysis of the cultures with lo”,,

SDS for 5 min, and expressed as the ratio: cpm in the lysate/total cpm added x 100.

After labelling, the cultures were washed four times with CTCM and kept in the incubator for 8-12 h for the cells to recover. Cultures were washed again and supple-

mented with 200 ~1 of the same CTCM alone or containing IFN-y, TNF-a or the combination IFN-y plus TNF-a to a final concentration of 1,000 U/ml. Nine micro- cultures were used for each treatment and for the nonstimulated controls. At day 4 after stimulation, 100 ~1 aliquots of the supernatant were collected from each well and

radioactivity measured in a y-counter. Percent specific cell lysis was calculated as follows:

Specific Yr release = test medium cpm - spontaneous cpm

x 100 total cpm - spontaneous cpm

The initial 5’Cr uptake was 3% in the primary cultures of islet cells and 7?$ in the cell-line HP 62. Percentages of spontaneous release (control medium) in the human


islet cells and the islet cell line were < 20% and 35 y0 respectively (calculated as the q0 of the total 5’Cr release measured by dissolving the cells in lo”/, SDS solution). Nine

cultures were taken for each of the treatments and controls.

IL-ID measurement and neutralization Production of IL-ll3 in the primary islet cultures was measured in the culture

supernatants after 4 days’ incubation with 1,000 U/ml of IFN-y + TNF-a using an ELISA technique (IL-18 detection kit, Cistron Biotechnology) with a lower limit of

sensitivity of 20 pg/ml. To test whether or not the observed cytotoxic effect of IFN-)I plus TNF-a was due to the release of IL- 1 b, blocking experiments were performed

by adding a neutralizing mAb to IL-lp (Cistron Biotechnology) to the cultures. At the concentration used this mAb is able to block up to 10 U/ml of IL-IS.

Statistical analysis

Statistical significance was determined by using the Student t-test and P values

co.05 (two tailed) were considered significant. The program Statworks@ and an Apple Macintosh SE/30 computer were used for the calculations.

Results

Insulin and glucagon release

At day 4, a significant reduction in 3 h insulin release was detected in the cultures treated with 1,000 U/ml of the combination of IFN-y plus TNF-a (36.74, -&7.7”,,

of the insulin in the control; P<O.OOl), and TNF-a alone (49.59, +7O, of the control; P<O.OOl) but not in the cultures treated with IFN-y alone. The same

effect was detected after 7 days of incubation with the combination of cytokines (38.67, f 11.3% of the control; P< 0.001) and TNF-a alone (42.6”,, &- 8.7:, of the

control; P<O.OOl) (see Figure 1).

The effect of cytokines on 3 h insulin release in individual pancreata is shown in

Figure 2. Although there was variability between glands, IFN-)I plus TNF-a, or TNF-a alone produced a clear inhibitory effect on 3 h insulin release in all of them, both at day 4 (Figure 2a) and at day 7 (Figure 2b) after stimulation.

Three-hour glucagon release was also reduced in the cultures treated with

1,000 U/ml of IFN-y plus TNF-a (30:~ f 8.5,% of the control f SEM; P< 0.01) or

with TNF-a alone (40.276 + 9”/b of the control; P < 0.01) but not with IFN-y alone (Figure 3). Again, cultures prepared from some glands were found to be more sensitive to cytokines than others.

Morphological assessment and cell counting

Inspection on inverted microscopy showed clear cell damage in the islet cell cultures treated with 1,000 U/ml of the combination of IFN-7 plus TNF-a (Figure 4). From days 4-7 progressive vacuolization, which finally resulted in the disintegration of the islet cell monolayer, was observed. Viability of cells spontaneously detached from the monolayer culture was assessed by acridine orange/ethidium bromide staining and


IFN-gamma TNF-alpha IFN + TNF

Figure 1. Insulin release at days 4 and 7 after the addition of 1,000 U/ml of IFN-y, TN combination IFN-y +TNF-a (1,000 U/ml each). Data are expressed as mean values obtainec human pancreata (mean& SEM). *PC 0.01; **P<O.OOl.

‘F. -a or the if rom four

(b)

60

40

20

0

I

J m IFN-gamma TNF-alpha TNF+IFN IFN-gamma TNF-alpha TNF+IFN

Figure 2. Effect of cytokines on insulin release from human islet cell monolayer cultures. Results represent 3 h insulin release from four individual pancreata after 4 (a) or 7 (b) days’ stimulation with 1,000 U/ml of IFN-I, TNF-a or the combination IFN-y +TNF-a.

found to be < 1 %, indicating that all viable cells remained in the monolayer cultures. By contrast, cells which remained attached to the coverslip were > 95% viable.

A significant decrease in the number of l3 cells was detected only in the monolayer cultures treated with 1,000 U/ml of the combination of IFN-y plus TNF-a (37.20/,+6.5°/0 of the control* SEM; P<O.OOl, Figure 5). Neither IFN-y nor TNF-a alone was able significantly to reduce the number of insulin positive cells. In the monolayer cultures treated with IFN-y plus TNF-a (but not in those treated with only one of the cytokines) there was a reduction in the number of a cells (Figure


-& 100 L

E T _ : 80- z ? 60- * a, z T *

f 40- T L E ; 20-

75 A7 0 r-3 IFN-gamma TNF-alpha lFN+TNF

Figure 3. Effect of cytokines on 3 h glucagon release on islet monolayer cultures. Results expressed as the percentage of glucagon released in the controls after four days’ incubation with l,OOOU/ml of IFN-I, TNF-a or IFN-y+TNF-a. Data represent mean values obtained from four human pancreata (means* SEM). *P<O.Ol.

Figure 4. Cytotoxicity of the cytokine combination IFN-y +TNF-a visualized on the inverted microscope. (a) Control cultures incubated for 4 days with normal complete tissue culture medium and (b) islet cell monolayers after 4 days’ incubation in the presence of 1,000 U/ml of the combination of IFN-1 + TNF-a.

6). The number of cytokeratin positive cells (epithelial/ductal) was moderately reduced in the cultures treated with IFN-)I plus TNF-a but the ratio /3 cells/ epithelial cells was much lower than in the controls, thus indicating a pronounced selectivity of the cytotoxic action towards the islet cells (Figure 7).

When cultures were treated with 100 U/ml of the combination of IFN-)I and TNF-a, no clear cytotoxic effect was detected in the monolayer cultures. By contrast, this concentration of cytokines was capable of inducing HLA class II, as demonstrated by double IFL staining of the same monolayer cultures (data not shown).


T 1

*

iI_ II - IFN -gamma TNF-alpha IfN + TNF

Figure 5. Reduction in the number of l3 cells present in the islet cell monolayer after stimulation with IFN-I, TNF-a or the combination IFN-y +TNF-a. Mean values obtained from four human pancreata (means+ SEM). *P<O.OOl.

;; IOO- L

z 9 80-

% d 60- In zz E 40-

2 $ 20-

z” OL

-5 T

II- * J IFN-gamma TNF-alpha IFNfTNF

Figure 6. Effect of cytokines IFN-y, TNF-a and the combination IFN-y +TNF-a on the number of glucagon producing cells, identified by mAb Glu-001, as indicated in the text. The results are expressed as percentage of the number of glucagon cells in the non-stimulated control cultures. The data are shown as mean values of three experiments (means + SEM). *P < 0.05.

l HP74

Control TNF-alpha IFN+TNF

Figure 7. Effect of cytokines on the number of epithelial cells in islet cell monolayer cultures obtained from two different human pancreata. Results show a preferential decrease of l3 cells detected at day 4 after the addition of l,OOOU/ml of IFN-y and/or TNF-a. Data are expressed as the ratio No. g cells/No. epithelial cells x 100.

Specific 51 Cr release assay

The percentage of specific 51Cr release showed specific cell lysis after 4 days’ incubation of the islet cell cultures with 1,000 U/ml of either IFNq, TNF-a or with

300 Gloria Soldevila et tall.

Figure 8. Specific Wr release assay in human pancreas HPt31. Islet cell monolayer cultures were used as substrate to assess specific cell killing mediated by cytokines. Cells were labelled with Yr and release was measured at day 4 after the addition of 1,000 U/ml of IFN-I, TNF-a or the combination EN-7 + TNF-a. Results represent percentages of specific Wr release obtained from nine islet cell cultures (means + SD).

40 -

30 -

20 -

IO 20 30 40 50 60 70 Class II positive beta cells (%I

Figure 9. Correlation between class II expression of human l3 cell and the reduction in the number of j3 cells. Results are expressed as percentage of class II positive fl cells and number of g cells in the cultures obtained from four different human pancreata at day 4 after the stimulation with 1,000 U/ml of IFN-7 + TNF-a.

the combination IFN-y plus TNF-a. The specific %r release was only significantly higher in cultures exposed to IFN-y + TNF-a (66.9% f 2 1.6, y0 specific 5’Cr release + SD, PcO.01, Figure 8).

Effect of IFN-y plus TNF-a on the B cells: correlation between class II expression induction and cytotoxicity

An inverse correlation was found between the percentage of class II positive and the total number of l3 cells in the cultures at day 4 after stimulation with 1,000 U/ml of IFN-y plus TNF-a (r = 0.9, Figure 9). A time lapse between the two effects was found, i.e., class II expression was already detectable at day 2, while cytotoxicity was not detected until day 4.

IL-lb detection and neutralization

IL-ll3 was not detected in the supernatant from control or from IFN-y, TNF-a or IFN-y plus TNF-a treated cultures (six cultures, two experiments). The addition of

Cytotoxic effect of IFN-7 plus TNF-a 301

Table 3. Specific “Cr release assay on HP62 cell line

Treatment exp 1 exp 2 exp 3

Control 1780.5+213.1 2132.lk95.2 2300.3 + 594.9

IFNy 1,000 U/ml 1965.4+ 254.2 2253.3+ 194.3 2067.7 + 254.9

TNFa 1,000 U/ml 2001 f 137* 2307.3 k484.6 1970.6 + 209.5 TNF + IFN 1,000 U/ml 2280.3 _+ 190.2* 3124.9+396.1* 2898.3* 948.3*

Chromium release (cpm+ SD) from human pancreatic islet cell line HP62 after 4 days of incubation in the presence of 1,000 U/ml of IFN-y, TNF-a or with the combination of IFN-+y + TNF-a. *Levels of significance (student t-test, F<O.O5). Data obtained from nine replicate cultures.

the IL-lp neutralizing mAb did not block the cytotoxic effect of IFN-y + TNF-a as assessed by inspection of the cultures under the inverted microscope (nine cultures, three experiments).

Cell lines

Cytotoxicity to the cell lines was primarily assessed by direct inspection on the inverted microscope. The addition of TNF-a to the cultures of HP 62, HT 93 and BxPC-3 cell lines did not produce any morphological changes. TNF-a, as expected, was highly cytotoxic to the L929 mouse fibroblastoid cell line. In the IFN-y-treated cultures, the HP 62, HT 93 and BxPC-3 cells adopted anlelongated shape and nu- merous floating dead cells were visible. The synergistic cytotoxic effect of EN-y plus TNF-a was clearly seen in the HP 62, HT 93 and BxPC-3 cells. 51Cr release experiments confirmed the toxicity of IFN-y plus TNF-a to the HP 62 cell line compared to the controls (Table 3).

Discussion

The cytotoxicity of different preparations of lymphokines to islet cells has been extensively evaluated in different in vitro systems using mouse and rat islets [18], demonstrating that it is influenced by many factors, such as using free floating islets or monolayers, the presence of FCS or the level of glucose in the culture medium [25]. In addition, the same cytokine preparations may have opposite effects when used at low or high concentration [26,27], and may be cytotoxic in only some species [ 151. Moreover, the experimental systems used to detect the cytotoxic effect have often been indirect (i.e., measurement of insulin release), and very rarely included cell-specificity control (i.e., assessment of the effect on other cell types, such as glucagon producing cells or other pancreatic epithelial cells). Finally, direct evidence of the toxic effect of cytokines to the islet cells in vivo is, remarkably, missing and in

the nearest model, the perfusion of intact rat pancreas, the results are inconclusive

WI. The present work provides firm evidence of the in vitro toxicity of the combination

IFN-)I plus TNF-a for human islet p cells. Cytotoxicity is demonstrated not only


indirectly by the inhibition of 3 h insulin release but also directly by the great reduction in the number of viable p cells present in the cultures, and is supported by ?r release


a slight cytotoxic effect. However, when TNF-a was used in combination with IFN-y, cell damage was observed which was even clearer with 1,000 U/ml, confirming the synergism and the

toxicity detected on the 51Cr release assay. This phenomenon is similar to that described in some pancreatic tumor cell lines that were resistant to TNF-a but sensitive to the combination TNF-a plus IFN-y [40]. The higher susceptibility of

the HP 62 cell line to the cytotoxic effects of IFN-y and of TNF-a plus IFN-y parallels its higher sensitivity to class II induction and seems to be a feature common

to many transformed cell lines.

Our results using human islet cells are in keeping with those of Campbell et al., who have also shown that the combination IFN-)I +TNF-a is cytotoxic to free floating murine islet cells [ 151. Puke1 et al. [ 141 used a system comparable to ours, with rat islets on monolayers, but cytotoxicity was only assessed by direct inspection and 51Cr release. These authors tested a number of cytokines and found a hierarchy of

cytotoxic effects with different combinations, i.e., IFN-y + IL-l /3 > IFN-y + TNF- a > TNF-a + IL-l@ Rabinovitch et al. [41] have recently reported the cytotoxicity of these cytokines to human islets using 51Cr release assay and measuring insulin and

glucagon production and content. Their data, like ours, show that in humans the

combination IFN-y plus TNF-a is very cytotoxic and that this effect is not b cell

specific. As IL-lb (alone or in combination with IFN-y and/or TNF-a) has been found

cytotoxic for p cells in a number of in vitro experimental systems, we considered it important to rule out the possibility that IL-l was mediating the cytotoxic effect that

we have observed. IL- 10 was not only undetectable in the culture supernatant, but the cytotoxic effect of the cytokine combination was unaffected by the addition of a neutralizing mAb to IL- 1 p. It could thus be concluded that IL-l p was not involved in the islet cytotoxicity caused by IFN-)I plus TNF-a.

The 51Cr release assay has been a parameter commonly used to assess cytotoxicity.

It has confirmed the cytotoxicity of the cytokines IFN-y plus TNF-a to islet cells and had good correlation with the other methods used, i.e., cell counting and hormone secretion. However, this assay by itself does not show if the cells that are killed in

vitro correspond only to 0 cells and therefore it should be supported, as here, by complementary studies.

Although IFN-a has been detected in ‘diabetic’ p cells [lo], TNF-a and IFN-y are

difficult to detect in pathological specimens, and so far there are no reports of their presence in Type 1 diabetic pancreas. The observed effects of IFN-y plus TNF-a on the human l3 cells, i.e., class II induction and subsequent cytotoxicity, could account for both ‘ectopic’ class II expression and for their destruction. However, it remains to be explained how these cytokines, which in vitro are active on the different islet cell

types, could be p cell specific in vim. One possibility is that the cell specificity is conferred by the T cells, which could secrete both cytokines and deliver them


selectively to the p cells. This would be facilitated by the existence of a membrane form of TNF-a [42]. This explanation is, however, in conflict with the relatively small number of lymphocytes infiltrating the islets. Direct demonstration of IFN-)I, TNF-a or IL-lp in the diabetic islets by immunohistochemistry or by in

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

References

Bosi, E., I. Todd, R. Pujol-Borrell, and G. F. Bottazzo. 1987. Mechanisms of autoimmunity: relevance to the pathogenesis of type 1 diabetes. R. S. de Fronzo and J. S. Skyler, eds, Diab. Met. Rev. 3: 893-928 Sutherland, D. E. R., F. C. Goetz, and A. F. Michael. 1989. Recurrence of disease in pancreas transplants. Diabetes 38 Suppl 1: 85-87 Assan, R., G. Feutren, M. Debray-Sachs, et al. 1985. Metabolic and immunological effects of cyclosporin in recently diagnosed type 1 diabetes mellitus. Lancet i: 67-71 Gepts, W. 1984. The pathology of the pancreas in human diabetes. In Immunology in Diabetes. D. Andreani, U. Di Mario, K. F. Federlin and L. G. Hedding, eds, Kimpton, London. Chapter 2, pp. 21-34 Foulis, A. K. and J. A. Stewart. 1984. The pancreas in recent-onset Type I (insulin- dependent) diabetes mellitus: insulin content of islets, insulitis and associated changes in the exocrine acinar tissue. Diabetologia 26: 456-461 Sai, P., M. Kremer, M. F. Nomballais, and G. Dillet. 1984. Antibodies spontaneously bound to islet cells in Type 1 diabetes (Letter to Editor). Lancet ii: 233-234 Bottazzo, G. F., B. M. Dean, J. M. McNally, E. H. Mackay, P. G. F. Swift, and D. R. Gamble. 1985. In situ characterisation of autoimmune phenomena and expression of HLA molecules in the pancreas in diabetic insulitis. N. Engl. J. Med. 313: 353-360 Foulis, A. K. and M. A. Farquharson. 1986. Aberrant expression of HLA DR antigens in insulin containing beta cells in recent onset type 1 (insulin-dependent) diabetes mellitus. Diabetes 35: 1215-1226 Hanafusa, T.,A. Miyazaki, J. Miyagawa, S. Tamura, M. Inada, K. Yamada, Y. Shinji, H. Katsura, K. Yamagata, N. Itoh, H. Asakawa, C. Nakagawa, A. Otsuka, S. Kawata, N. Kono, and S. Tarui. 1990. Examination of islets in the pancreas biopsy specimens from newly diagnosed type 1 (insulin-dependent) diabetic patients. Diabetologia 33: 105-l 11 Foulis, A. K., M. A. Farquharson, and A. Meager. 1987. Immunoreactive alpha- interferon in insulin-secreting beta cells in Type 1 Diabetes. Lancer i: 1423-I 427


11.

12.

13.

14.

15.

16.

17.

18.

19.

20. 21. 22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

Pujol-Borrell, R., I. Todd, M. Doshi, G. F. Bottazzo, R. Sutton, D. Gray, G. R. Adolf, and M. Feldmann. 1987. HLA class II induction in human islet cells by interferon-y plus TNF-a. Nature 326: 304-306 Bendtzen, K., T. Mandrup-Poulsen, J. Nerup, J. H. Nielsen, C. A., Dinarello, and M. Svenson. 1986. Cytotoxicity of human pI7 Interleukin-1 for pancreatic iselt of Langerhans. Science 232: 1545-1547 Mandrup-Poulsen, T., K. Bendtzen, C. A. Dinarello, and J. Nerup. 1987. Human Tumor Necrosis Factor potentiates human interleukin l-mediated rat pancreatic beta-cell cytotoxicity. J. Immunol. 139: 4077-4082 Pukel, C., H. Baquerizo, and A. Rabinovitch. 1988. Destruction of rat islet cell monolayer cultures by cytokines. Synergistic interactions of interferon-y, tumor necrosis factor, lymphotoxin and interleukin 1. Diabetes 37: 133-136 Campbell, I. L., A. Iscaro, and L. C. Harrison. 1988. IFN-), and tumor necrosis factor. Cytotoxicity to murine islet of Langerhans. J. Immunol. 141: 2325-2329 Mandrup-Poulsen, T., S. Helqvist, J. Molvig, L. D. Wogensen, and J. Nerup. 1989. Cytokines as immune effector molecules in autoimmune endocrine diseases with special relevance to insulin-dependent diabetes mellitus. Autoimmunity 4: 198-218 Lake, S. P., P. D. Basset, A. Larkins, J. Revell, K. Walzak, J. Chamberlain, G. M. Rumford, N. J. M. London, I’. S. Veitch, P. R. F. Bell, and R. F. L. James. 1989. Large- scale purification of human islet utilizing discontinuous albumin gradient on IBM 2991 cell separator. Diabetes 38 Suppl 1: 143-145 Fromm, H. and P. J. J. Berg. 1982. Deletion mapping of DNA regions required for SV-40 early region promoter function in vivo. Mol. App. Gen. 1: 457-481 Buscema, M., V. Marini, G. Soldevila, F. Folli, R. Sutton, R. James, R. Pujol-Borrell, and G. F. Bottazzo. 1988. Towards immortalization of human pancreatic endocrine cells: an approach using SV40 viral DNA. Diabetes 37: 80A European Collection of Animal Cell Cultures. Cat. No. 85011425 American Type Culture Collection ATCC, CRL 1687 Belfiore, A., T. Mauerhoff, R. Pujol-Borrell, R. Mirakian, and G. F. Bottazzo. 1986. Spontaneous HLA class II expression in SV-40 transfected human thyroid cells. J. Endocrin. Invest. 9: 61 Morgan, C. R. and A. Lazarow. 1963. Immunoassay of insulin two antibody system. Diabetes 12: 115-126 Pujol-Borrell, R., I. Todd, M. Doshi, D. Gray, M. Feldmann, and G. F. Bottazzo. 1986. Differential expression and regulation of MHC products in the endocrine and exocrine cells of the human pancreas. Clin. Exp. Immunol. 65: 128-139 Mandrup-Poulsen, T., G. A. Spinas, S. J. Prowse, B. S. Hansen, D. W. Jorgensen, K. Bendtzen, J. H. Nielsen, and J. Nerup. 1987. Islet cytotoxicity of interleukin 1. Influence of culture conditions and islet donor characteristics. Diabetes 36: 641-647 Spinas, G. A., T. Mandrup-Poulsen, J. Molvig, L. Baek, K. Bendtzen, C. A. Dinarello, and J. Nerup. 1986. Low concentration of interleukin- 1 stimulate and high concentrations inhibit insulin release from isolated rat islets of Langerhans. Acta Endocrinol. (Copenh.) 113: 551-558 Comens, P. G., B. A. Wolf, E. R. Unanue, P. E. Lacy, and M. L. MC Daniel. 1987. Interleukin 1 is potent modulator of insulin secretion from isolated rat islets of Langerhans. Diabetes 36: 963-970 Wogensen, L. D., V. Kolbbachofen, P. Christensen, C. A. Dinarello, T. Mandrup- Poulsen, S. Martin, and J. Nerup. 1990. Functional and morphological effects of IL-l beta on the perfused rat pancreas. Diubetologiu 33: 15-23 Orci, L., B. Thorens, M. Ravazzola, and H. F. Lodish. 1989. Localization of pancreatic beta cell glucose transporter to specific plasma membrane domains. Science 245: 295-298 Soldevila, G., M. Doshi, R. James, R. Sutton, G. F. Bottazzo, and R. Pujol-Borrell. 1990. Induction of HLA-DR, DP, DQ subregion products on human !3 cells by the cytokine combination IFN-y plus TNF-a. Autoimmunity 6: 307-317 Pujol-Borrell, R. and G. F. Bottazzo. 1988. Puzzling diabetic transgenic mice: a lesson for human type 1 diabetes? Immunol. Today 9: 303-306


32. 33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

Parham, P. 1988. Intolerable secretion in tolerant transgenic mice. Nature 333: 500-503 Rabinovitch, A., H. Baquerizo, C. Pukel, and W. Sumoski. 1989. Effects of cytokines on rat pancreatic islet cell monolayer cultures: distinction between functional and cytotoxic effects on islet l3 cells. Regional Immunol. 2: 77-82 Wong, G. H. W. and D. V. Goeddel. 1986. Tumor necrosis factor a and 8 inhibit virus replication and synergize with interferons. Narure 323: 819-822 Sugarman, B. J., B. B. Aggarwal, P. E. Hass, I. S. Figari, M. A. Palladino, and H. M. Shepard. 1985. Recombinant human tumor necrosis factor-a. Effects on proliferation of normal and transformed cells in vitro. Science 230: 943-945 Ruggiero, V., J. Tavernier, W. Fiers, and C. Baglioni. 1986. Induction of the synthesis of tumor necrosis factor receptors by interferon-gamma. 3. Zmmunol. 136: 2445-2450 Buscema, M., I. Todd, U. Deuss, L. Hammond, R. Mirakian, R. Pujol-Borrell, and G. F. Bottazzo. 1989. Influence of tumor necrosis factor-a on the modulation by interferon-y of HLA class II molecules in human thyroid cells and its effect on interferon-y binding. 3. Clin. Endocrinol. Metab. 69: 433-439 Beresini, M. H., M. J. Lempert, and L. B. Epstein. 1988. Overlapping polypeptide induction in human fibroblasts in response to treatment with IFN-alpha, IFN-gamma, IL-lalpha, IL-lbeta and TNF.3. Immunol. 140: 485-493 Rubin, B. Y., S. L. Anderson, R. M. Lunn, N. K. Richardson, G. R. Hellerman, L. J. Smith, and L. L. Old. 1988. Tumor necrosis factor and IFN induce a common set of proteins. 3. Zmmunol. 141: 1180-l 184 Schmiegel, W. H., J. Caesar, H. Kalthoff, H. Greten, H. W. Schreiber, and H. G. Thiele. 1988. Antiproliferative effects exerted by recombinant human tumor necrosis factor-a (TNF-a) and Interferon-y (IFN-y) on human pancreatic tumor cell lines. Pancreas 3: 180-188 Rabinovitch, A., W. Sumoski, R. V. Rajotte, and G. L. Wrnok. 1990. Cytotoxic effects of cytokines on human pancreatic islet cells in monolayer culture.3. Clin. Endocrinol. Metab. 71: 152-156 Kriegler, M., C. Perez, K. DeFay, I. Albert, and S. D. Lu. 1988. A novel form of TNF/Cachectin is a cell surface cytotoxic transmembrane protein: ramifications for the complex physiology of TNF. Cell 53: 45-53

Cytotoxic effect of IFN-γ plus TNF-α on human islet cells

Documents

Transcript of Cytotoxic effect of IFN-γ plus TNF-α on human islet cells