Pancreatic effects of ethionine: blockade of exocytosis ......nophagy) can be observed, and, 24 h later, autophagocytosis is noted. Finally, shortly before the onset of pancreatic

Pancreatic effects of ethionine: blockade of exocytosis and appearance of crinophagy and autophagy precede cellular necrosis

HEIICHIRO KOIKE, MICHAEL L. STEER, AND JACOPO MELDOLESI Department of Pharmacology and CNR Center of Cytupharmacolugy, University of Milan, 20129 Milan, Italy; and Department of Surgery, Beth Israel Hospital and Harvard Medical School, Boston, Massachusetts 02215

KOIKE, HEIICHIRO, MICHAEL L. STEER, AND JACQPO MEL- DOLESI. Pancreatic effects of ethionine: blockade of exocytosis and appearance of crinophagy and autophagyprecede cellular necrosis. Am. J. Physiol. 242 (Gastrointest. Liver Physiol. 5): G297-G307, 1982.-Young female mice fed a choline-deficient, ethionine-supplemented (CDE) diet for 24 h develop hemorrhagic pancreatic necrosis with a &day mortality rate of -50%. At the end of the diet administration, the in vivo discharge of digestive enzymes is blocked, images of exocytosis and luminal membrane recycling disappear, and zymogen granules accumulate within acinar cells. The general ultrastructure, however, remains well preserved, protein synthesis is normal, and intracellular transport of secretory proteins is only slightly retarded. Thus, the CDE diet does not affect the general phenomenon of membrane fusion-fission but specifically inhibits that associated with exocytosis. Twenty-four hours after withdrawal of the CDE diet, discharge of zymogen granules into lysosomes (crinophagy) can be observed, and, 24 h later, autophagocytosis is noted. Finally, shortly before the onset of pancreatic necrosis, cells of nearly normal appearance are noted to be scattered among cells showing varying degrees of lesion up to complete disruption. Thus, the CDE-induced pancreatic necrosis results from a sequence of events: blockade of exocytosis evolves to crinophagy and autophagy, which might lead to lysosomal activation of zymogens.

secretion; intracellular transport; exocytosis; membrane recycling; lysosomes; pancreatitis; choline deficiency; freeze fracture; immunocytochemistry; cell fractionation; electron microscopy

-.

SECRETORY PROTEINS synthesized by pancreatic acinar cells are vectorially transported through a series of discrete, membrane-enclosed compartments: the ER,l Golgi complex, and secretory granules. In the latter compartment, secretory proteins are concentrated before their eventual discharge into the extracellular space by exocytosis. The process by which functional continuity is established between the various compartments (i.e., fusion-fission of their limiting membranes) is poorly under-

’ Abbreviations used: AHPN, acute hemorrhagic pancreatic necrosis; CCK-PZ, cholecystokinin-pancreozymin; CDE, choline-deficient, ethionine-supplemented (diet); E face, extracellular or endoplasmic fracture face; ER, rough-surfaced endoplasmic reticulum; IMP, intramembrane particle(s); P face, protoplasmic fracture face; RM, rough-surfaced microsomes; SM, smooth-surfaced microsomes; TCA, trichloroacetic acid; ZG, zymogen granules.

stood, and it is not clear how these processes are regu- lated (for review, see Refs. 20, 25, 27).

Lombardi et al. (17, 18) have noted that young female mice fed a choline-deficient diet supplemented with ethionine (the ethyl analogue of methionine) develop fatal AHPN. They reported that the earliest morphological change in the pancreas of mice fed CDE diet was an increase in the acinar cell content of ZG and that this change was evident within 24 h (17). Later studies by Gilliland and Steer (8) revealed that in vivo zymogen synthesis was almost unaffected, whereas in vivo zymogen discharge was almost totally abolished within 24 h of ingesting the CDE diet. Furthermore, removal of the CDE diet and reinstitution of a regular diet free of ethionine, after exposure to the CDE diet for 24 h, did not prevent the development of AHPN with a &day mortality rate of -50% (8). These studies suggested that the initial event triggered by the CDE diet is a blockade of exocytosis and that this is followed by a cascade of events that eventually lead to the appearance of AHPN.

In the present communication, we report the results of studies that have investigated the effects of the CDE diet on the ultrastructure of the pancreatic acinar cell and on the intracellular transport of digestive enzymes. The latter process has been evaluated by cell fractionation 24 h after the start of CDE diet administration. The ultrastructural studies used a number of techniques (conven- tional thin sectioning, freeze fracture, and both immuno- and enzyme cytochemistry) to evaluate the changes present at various times (O-72 h) after withdrawal of the CDE diet. These studies support the suggestion that the CDE diet causes an early and selective blockade of ZG discharge. As a result, crinophagy and autophagocytosis occur.

MATERIALS AND METHODS

Female CD-l mice (weighing 12-14 g and 21-23 days old) were purchased from Charles River Breeding Labo- ratories, Calco, Italy. They were fasted for 24 h and then fed either the CDE or control diet for 24 h, as previously described (8). At the end of this period, the animals were starved for an additional 24 h and then fed regular laboratory chow ad libitum. Groups of mice fed either the CDE or control diet were killed by decapitation at 0, 24,34,48,58, and 72 h after the initial 24-h period of diet

0193~1857/82/0000-0000$01.25 Copyright 0 1982 the American Physiological Society G297

by 10.220.33.1 on April 5, 2017

http://ajpgi.physiology.org/D

ownloaded from

http://ajpgi.physiology.org/

G298

administration. In agreement with previous observations (B), -50% of the mice that were fed the CDE diet for 24 h and were not killed died with APHN by the 5th day, whereas none of the control diet-fed mice died.

Pulse labeling and cell fractionation. Cell fiactiona- tion was performed after in vivo pulse labeling. Z-[4,5-3H] leucine (5 $X/mouse) was injected into the tail vein, and groups of three animals were killed at varying times thereafter. In each group, the labeled pancreatic tissue was homogenized in 10 vols unbuffered 0.3 M sucrose. Homogenates were fractionated by differential and density-gradient centrifugation, using a protocol originally developed for fractionating the guinea pig pancreas (21). Biochemical analyses and electron microscopic exami- nation of mouse pancreas fractions obtained using this procedure indicated that they correspond closely to their guinea pig counterparts.

Electron microscopy. A thorough analysis of the effects of various fixatives on the ultrastructure of pancreatic acinar cells was performed with the aim of obtaining an adequate preservation of the structures induced by CDE feeding. Fixative solutions, maintained at 0°C were in- ftitrated in the gland interstitium with a syringe. Small tissue blocks were trimmed with razor blades, and fixation was continued by immersion for 2-3 h at O*C. Fixative solutions for thin-section electron microscopy were 1% Os04 in either phosphate or cacodylate buffer (0.12 M, pH 7.4); 1% 0~04 in phosphate buffer mixed with potassium ferricyanide (I mg/ml) shortly before use; and 1% glutaraldehyde plus 0.5% formaldehyde (freshly prepared from paraformaldehyde) in phosphate buffer, followed by 1% 0~04 in cacodylate buffer. After 0~0~ fixation, samples were extensively rinsed in 0.12 Veronal buffer (pH 6.5), block stained by uranyl acetate, dehydrated in ethanol, and embedded in Epon 812. Thick sections were stained with toluidine blue 0 and used for orientation; thin sections were doubly stained with uranyl acetate and lead nitrate and examined using a Philips EM 400 electron microscope. Samples for freeze-fracture studies were fixed in 1% glutaraldehyde-0.5% formaldehyde as described above, infiltrated in glycerol, frozen in Freon 22, fractured, and replicated as described in Ref. 33. Samples for enzyme cytochemistry (i.e., localization of acid phosphatase activity) were fixed for 30 min in cold 2.5% glutaraldehyde-2% formaldehyde in cacodylate buffer (90 mM, pH 7.4) containing 2.5 mM CaCl2. After extensive rinsing in this buffer, followed by cold 7.5% sucrose, the tissue blocks were embedded in 10% agar, cut into thick sections using a Smith and Farquhar tissue chopper (Servall TCZ, Du Pont Instruments, Newton, CT), and incubated at 37°C for 30-60 min in medium containing 2.4% P-glycerophosphate, 5% sucrose, 50 mM acetate buffer (pH 5), and 0.12% lead nitrate. At the end of the incubation, the slices were washed extensively with 7.5% sucrose and then with cacodylate buffer, postfixed in 0~0~ in the same buffer, block stained, and embedded in Epon as already described.

Isolated tissue fractions [RM, SM, and ZG] were fixed in situ with 1% 0~04 in phosphate buffer, embedded, sectioned, and examined as described in Ref. 21.

Immunocytochemistry. Immune serum was obtained from rabbits injected with solubilized contents of guinea pin ZG (21) as the antigen. This antiserum produced six

KOIKE, STEER, AND MELDOLESI

positive arcs when analyzed against the antigen by crossed immunoelectrophoresis. Tissue samples were fixed in 0.2% glutaraldehyde-2% formaldehyde in phosphate buffer as described above, washed in the same buffer containing 10% sucrose, and then infiltrated in 2.3 M sucrose in phosphate buffer. After freezing in Freon 22 cooled by liquid nitrogen, the samples were thin sectioned at -100°C in a Servall Porter Blum LTC-2- cryodevice attached to a MT 2B ultramicrotome. Thin cryosections were collected and exposed to antiserum (diluted 1125 to 1:50 in phosphate-buffered saline) and then washed according to the protocol recommended by Tokuyasu (34). The bound immunoglobulin molecules were labeled using colloidal gold particles coated with protein A and purified by sucrose density-gradient centrifugation before use. The labeled sections were stained and embedded as described by Tokuyasu (34).

The immune serum used in these studies was evaluated using cryosections of pancreatic tissue obtained from rats and guinea pigs as well as mice. Results comparable with those reported for antiamylase and antitrypsinogen by Geuze et al. (7) were obtained, i.e., the compartments containing high concentrations of secretory zymogens (ZG, condensing vacuoles, Golgi cisternae, and acinar lumina) were heavily and consistently labeled. In addition, labeling of ER cisternae was significantly above background. In contrast, sections exposed to preimmune serum or immune serums prepared using nonpancreatic antigens (electrophorus cholinesterase and black widow spider venom) revealed labeling that was of low intensity and random distribution (see Table 1) ,

To compare quantitatively the specific immunolabeling of the various cell compartments, printed pictures of acinar cell cryosections obtained from mice fed the CDE diet for 24 h and then starved for a further 24 h were analyzed morphometrically. Specifically, we measured the density (number per unit area) of protein A-colloidal gold particles appearing over individual compartments in

TABLE 1. Distribution of protein A-g&d particles in subcellular compartments

Serum

Compartment Antizyrnogen, Preimmune, particles/ym’ particles/~m’

Large vacuoles 60.3 + 6.5 4.2 +- 1.7 (12.3) 66)

Dense areas of vacuole content 80.0 k 9.0 3.1 k 0.6 (5.3) 0.7)

Zymogen granules 125J t 25.4 4.1 -+ 0.8 (21.9) (13.1)

Condensing vacuoles and Golgi 60.0 k 21.6 cisternae (11.5)

Endoplasmic reticulum 12*1 k 4.0 4.5 & 2.9 cisternae (4.1) (2.U

Mitochondria 4.7 + 0.9 4.2 -1- 1.2 (5.4) (6J)

Nuclei 4.8 k 1,9 3.2 k 0.4 (18.4) (2.U

Cytoplasmic ground substance 5.6 k 1.9 5.1 t 2 (5.0) Gw

Values are means t SE of data obtained from the pancreatic acinar cells of mice killed after 24 h of CDE feeding followed by 24 h of starvation. Values in parentheses indicate the total analyzed areas (pm’). The morphometric analysis was carried out as described in MATERIALS AND METHODS.

by 10.220.33.1 on April 5, 2017


ownloaded from


ETHIONINE EFFECTS ON MOUSE PANCREAS G299

cryosections exposed to either the antizymogen or preimmune serum, both diluted 1:35. Areas were measured using a MOP- 1 Kontron-Zeiss apparatus.

Analytical procedures. Protein was measured according to the method of Lowry et al. (19) using bovine serum albumin as the standard, DNA according to Burton (3), and amylase according to Bernfeld (2) after Triton X-100 solubilization of the latent enzyme. The procedure for measuring TCA- insoluble radioactivity is given in Ref. 33.

Materials. Z-[4,5-3H]leucine (52 Ci/mmol) was purchased from the Radiochemical Centre, Amersham, UK. All other chemicals were of reagent grade and purchased from sources previously specified (33).

were

FIG+

mogena leucine. 0 and

6

1 3 5 hours

1. Time course of TCA-precipitable radioactivity in total ho- .tes prepared from pancreatic tissue pulse labeled with Z-[3H]- Mice were injected intravenously with 5 PCi of tracer at time

o- killed in groups of 3 after indicated time periods. m---e, mice treated with CDE diet for 24 h.

Control

4

B

1 I 1 1 3 5

hours

RESULTS

Intracellular transport of radioactive protein. The time course of 3H-labeled TCA-insoluble material in pancreas homogenates and in subcellular fractions prepared from homogenates at varying times after in vivo pulse labeling of mice fed for 24 h either the control or CDE diet is shown in Figs. 1 and 2. The data for homogenates are expressed on a cell basis (i.e., dpm/mg DNA) and those for cell fractions are expressed as protein specific activity (i.e., dpm/mg protein).

Incorporation of tracer into pancreatic tissue protein is similar for both the control and CDE groups. However, TCA-insoluble radioactivity is rapidly lost in the control group, whereas no decrease is observed in the CDE group (Fig. 1).

Subcellular fractionation of the pancreas indicates, as originally described by Palade and associates (13, 14, 25, 29,30), that newly synthesized secretory proteins migrate through various intracellular compartments. In the control group, RM (corresponding to ER) show the highest level of specific radioactivity at the end of the pulse, and this is followed by a rapid decline. The SM (equaling Golgi complex) show an initial increase followed by a sustained decrease in radioactivity. The ZG show late accumulation followed by a gradual decline in specific radioactivity that is due to discharge of zymogens into the extracellular space.

The process of ER-Golgi transport (i.e., decrease in RM and increase in SM specific radioactivity) is not markedly altered in the CDE group (Fig. 2, A and B). In contrast, the peak level of ZG specific radioactivity is only -30% of that reached in the control group (Fig. 2C). It should be noted, however, that the CDE diet causes ZG accumulation in acinar cells (see Refs. 4 and 6 and data below) and a large increase in the pancreatic content of digestive enzymes (approximately threefold in these experiments) (data not shown, cf. Ref. 8). Thus, the reduced ZG peak specific radioactivity noted in these

C

FIG l 2. Time course of TCA-precipitable radioactivity in subcellular fractions isolated from pancreatic homogenates prepared from mice labeled in vivo by intravenous injection of Z-[3H]leucine and killed at various times thereafter (see Fig. 1). A: rough-surfaced microsomes, B: smooth-surfaced microsomes; C: zymogen granules. o- --0, Control mice; h-4, mice fed with CDE diet for 24 h.

\ \ \ k

\ \

\ \

\ \

by 10.220.33.1 on April 5, 2017


ownloaded from


G300 KOIKE, STEER, AND MELDOLESI

studies does not indicate diminished entry of newly synthesized secretory proteins into ZG but merely reflects the presence of an expanded ZG compartment after exposure to the CDE diet.

The decline in ZG radioactivity in vivo is clearly blocked by prior administration of the CDE diet. As noted in Fig. ZC, the specific radioactivity of the ZG fraction in the CDE group does not decline from the peak level over a 5-h period, whereas it falls to less than 30% of the peak value in the control group during this time. This observation is consistent with that made using pancreatic homogenates (Fig. 1).

Electron microscopy. The ultrastructure of the mouse acinar cell closely resembles that of the rat and guinea pig as previously described (7,lZ 13,25,29, 33), although the mouse acinar cell has fewer ZG and more images of exocytosis (often multiple or compound, Fig. 3, A-C). Such images are relatively infrequent in the other species. The limiting membranes of discharged ZG appear to evolve into large coated pits (Fig. 3B), which later develop into coated vesicles (15). According to recent data reported by Herzog and Reggio (12), these coated vesicles permit recycling to the cytoplasm of excess membrane inserted at the cell surface during ZG-plasmalemma fusion.

No gross changes in acinar cell ultrastructure are observed in samples taken 24 h after CDE diet administration is begun (Fig. 3D). Specifically, the general organization of the nucleus, ER, and Golgi apparatus is not appreciably different from that of mice given the control diet. In contrast, ZG are greatly increased in number and are closely packed in the supranuclear area of the cytoplasm. ZG are also seen very close to, and sometimes strictly opposed to, the lateral as well as the luminal plasmalemma (Fig. 30). In contrast to samples obtained from control diet-fed animals, clear images of exocytosis and large coated pits and vesicles have almost completely disappeared after 24-h exposure to the CDE diet. Al- though the size of acinar lumina are variable and not greatly different from those of the control group, the density of the filamentous substance contained in the lumina is considerably decreased after exposure to the CDE diet (Fig. 3D), and some lumina appear to be completely empty.

The ultrastructural appearance described above is also observed in samples obtained from mice fed the CDE diet for 24 h and then starved for 24 h before being killed (Fig. 4). At this time, however, the acinar cells also contain membrane-bound vacuoles, which are usually of large size and distributed throughout the cytoplasm, and most of which contain a filamentous material. This material is often clumped into large, discrete masses that have a density similar to that of ZG cores (Fig. 4, A-C). In freeze fracture, these vacuoles appear as large, irregularly spherical or ovoidal organelles. Their limiting membranes have IMP that are more concentrated on the P- than on the E-fracture face. The density of these IMP (i.e., number per unit area) in the membranes of the large vacuoles is much lower than in the plasmalemma, ER, or Golgi membranes and is similar to that seen in ZG membranes (Fig. 4D). When cryosections of the pancreatic tissue were exposed to the antizymogen antibody

and the distribution of bound immunoglobulin then revealed by the protein A-coated gold particles, the large vacuoles appeared specifically labeled (Fig. 4E). To ob- tain a quantitative estimate of the immunolabeling of the vacuoles in comparison to other cytoplasmic structures, a morphometric analysis was carried out. As can be seen in Table 1, we found that the density (number per unit area) of gold particles in the large vacuoles was approximately as high as in condensing vacuoles and Golgi cisternae and about 40% of that in ZG. In contrast, a much lower labeling was observed in the surrounding cytoplasmic ground substance, in the organelles known to contain no zymogens (nuclei and mitochondria), and also in ER cisternae (where zymogens are highly diluted (20, 25, 29). Moreover, within the vacuoles, the distribution of the label was uneven, inasmuch dense regions appeared labeled more heavily than clear regions. Such a gold particle distribution is specific for the antizymogen serum because with preimmune or unspecific immune serums the labeling of all compartments was minimal and uniform (Table 1). These observations indicate that the dense content of vacuoles contains (and might be composed of) concentrated zymogens.

Further information on the nature of large vacuoles was obtained by acid phosphatase cytochemistry. In fact, the vacuoles were found to be positive for this enzyme activity (Fig. 4, F and G) as are lysosomes, trans Golgi cisternae, and some condensing vacuoles (see Ref. 23). The observed similarity in zymogen and acid phosphatase labeling between vacuoles and condensing vacuoles raises the question of the possible identity of these two organelles. This possibility, however, seems quite unlikely. In fact, besides obvious differences in size, shape, and appearance, the two types of organelles differ in localization. Thus, condensing vacuoles are restricted to the Golgi area (even after CDE feeding), whereas vacuoles are spread into the cytoplasm. This wide distribution seems to correlate well with that of lysosomes, which in acinar cells are known to be present also between granules and ER cisternae (24). Taken together, our observations are therefore consistent with the view that large vacuoles are formed by crinophagy (31), i.e., by fusion of ZG with lysosomes, with consequent discharge of zymogens into the lysosomal compartment.

Ten to twenty-four hours later (i.e., 34-48 h after withdrawal of the CDE diet and 10-24 h after resumption of regular laboratory chow), the appearance of the acinar cells has been further modified (Fig. 5) w The number of large, acid phosphatase-positive vacuoles is increased, and these vacuoles now contain other structures in addition to the filamentous material labeled by antizymogen serum. At this stage, the vacuoles also contain membranes and a variety of organelIes, including small vesicles adjacent to the limiting membrane, larger vesicles, membrane fragments and whorls, stacks of recognizable ER, mitochondria, and ZG surrounded by their limiting membrane (Fig. 5, A and B). In addition, the limiting membrane of the acid phosphatase-positive vacuoles is heavily stained by OsOd-ferricyanide fixation, which is a characteristic of lysosomal membranes (cf. Fig. 5A with Fig. 4A). Freeze-fracture studies of these cells reveal that the general structure of their membranes is well pre-

by 10.220.33.1 on April 5, 2017


ownloaded from



FIG. 3. Fine structure of pancreatic acinar cells of control mice (A- C) and mice fed with CDE diet for 24 h (D). In control cells, images of exocytosis (*) are very commonly seen in both thin sections (A) and freeze fracture (C), although with the former technique continuity of discharged granules with acinar lumen (AL) often lies outside section plane. As a consequence of this ongoing exocytotic activity, acinar lumen is filled with dense filamentous material (discharged zymogens) (A and B), and numerous large coated pits and vesicles (which are responsible for recycling membrane from luminal surface to cytoplasm) appear, especially at surface infoldings that probably correspond to

previously discharged zymogen granules (arrows in A and B). In acinar cells of CDE-fed mice (D), general ultrastructure is well preserved, but number of zymogen granules (ZG) is greatly increased. These organelles are seen all over in cytoplasm, sometimes very close to plasmalemma (double arrow), but never in process of being discharged by exocytosis. Acinar lumen contains only sparse ftiamentous material. Large coated vesicles are no longer seen. Fixation of thin-sectioned specimens was by 1% 0~0~. GC, Golgi complex; CV, condensing vacuole. A: magnification, ~42,0OQ B: ~50,ooO; C: ~45,ooO; D: ~16,000.

by 10.220.33.1 on April 5, 2017


ownloaded from



FIG. 4. Fine structure of pancreatic acinar cells of mice fed for 24 h with CDE diet and then starved for an additional 24 h. At this stage, picture is characterized by appearance of membrane-bound vacuoles (V), many of which containing little or no segregated membranes, but dense, filamentous material often arranged into discrete masses, remi- niscent of partially disarranged zymogen granule cores (arrows in A; views at higher magnification in B and C). Freeze-fracture study (D) reveals irregularly faceted shape and low density in intramembrane particles of vacuolar limiting membrane and confnms paucity of their

segregated membranes (arroz&eads). Cytochemical analyses revealed that vacuoles contain zymogens [because they are specifically labeled by indirect immunolabeling, using antizymogen antibodies and protein A-coated colloidal gold particles (E)] and that they are positive for acid phosphatase activity whether they contain little (5’) or abundant (G) segregated membranes. Other labeling is as in Fig. 3. Sample of A was fmed with Os04-ferricyanide; those of B and C with 1% 0~04 alone. A: magnification, x18,ooO; B and C: ~26,ooO; D and E: x32,ooO; F and G: x26,000.

by 10.220.33.1 on April 5, 2017


ownloaded from



served. Both the luminal and basolateral regions of the plasmalemma appear normal. IMP are randomly distributed, and the IMP density is the same as that noted for control diet-fed animals. The organization of tight junctions and gap junctions also appears normal (not shown in figures). Exocytosis images are not seen at either the luminal or the lateral plasmalemma. The cytoplasmic “regular” organelles also appear normal. Vacuoles are now frequently cross fractured, and both the limiting and the segregated membranes can be seen. As noted at

earlier times after administration of the CDE diet, the limiting membrane has only few IMP and these are located on the P face, whereas the segregated membranes are almost always IMP free (Fig. 5C).

Pancreas samples obtained from animals killed 58 and 72 h after withdrawal of the CDE diet (approximately 1 day before the expected onset of AHPN) have a more heterogeneous appearance (Fig. 6, A-C). Some cells contain ZG (frequently smaller than the ZG of control samples) as well as vacuoles containing membrane fragments.

FIG. 5. Pancreatic acinar cells in mice fed CDE diet for 24 h, starved zymogen granules (*)I (A and B). In samples fixed with OsO1-ferricy- for 24 h, and finally reestablished on their normal laboratory diet for anide, limiting membrane of the vacuoles stains very dark at this stage another 24 h. Cytoplasm appears tilled with zymogen granules and with (A). In freeze fracture (C), this membrane is characterized by a low vacuoles (V) containing segregated membranes (often small vesicles, complement of intramembrane particles, wherease segregated mem- located preferentially close to limiting membrane) and recognizable branes are completely smooth. Other labeling is as in Figs. 3 and 4. A membrane-bound organelles [endoplasmic reticulum cisternae and and B: magnification, ~24,ooO; C: ~22,000.

by 10.220.33.1 on April 5, 2017


ownloaded from



FIG. 6. Pancreatic acinar cells in mice fed CDE diet for 24 h, starved for 24 h, and finally reestablished on laboratory diet for 48 h. Marked

greatly decreased in number and also in size. Usually, they are located

cellular heterogeneity of pancreas at this stage of experiment is shown. all around acinar lumina. Observation of large globular infoldings of

Many cells exhibit vacuoles containing membrane remnants and debris latter (*) as well as coated pits and vesicles (double arrow) suggests

(arrows in A and C). Some have a well-developed system of rough- that in these cells some exocytotic activity might have been resumed

surfaced cisternae (RER) (A), whereas others (C) contain only rough- (B). Intermingled with apparently well-preserved cells are areas occu- pied by necrotic cells and clusters of cell debris (s in A). Fixation was

surfaced vesicles and tubules and free polysomes, and accumulate lipid droplets (L). With respect to earlier stages, zymogen granules have

by 1% 0~0~. N, nucleus (C). Other labels are as in previous figures. A: ~14,000, B: x28,ooo; c: ~14,000.

by 10.220.33.1 on April 5, 2017


ownloaded from


ETHIONINE EFFECTS UN MOUSE PANCREAS G305

In these cells, images of exocytosis as well as large coated pits and vesicles at the apical pole can sometimes be seen (Fig. 6B). A number of small cells containing lipid droplets and disordered arrays of ER but few or no ZG can also be observed (Fig. 6C). Finally, some grossly altered, apparently necrotic cells can be seen, and clusters of free organelles can be observed intermingled with well-preserved cells (Fig. 6A).

DISCUSSION

Zymogen transport and discharge. The digestive enzyme content of the pancreas is increased 24 h after young female mice are fed a CDE diet and several days before the appearance of AHPN. Previous studies (8) have indicated that the rise in zymogen content and number of ZG results from a reduction in zymogen discharge in the face of continued zymogen synthesis. Zym- ogen discharge is known to occur by exocytosis (i.e., by fusion-fission of the ZG and luminal plasma membranes). The transfer of newly synthesized zymogens from the ER to Golgi and from Golgi to ZG also involves fusion-fission of membranes, but these events are not blocked after ingestion of the CDE diet. Thus, these observations indicate that the CDE diet specifically interferes with the fusion-fission event associated with exocytosis. Exo- cytosis of zymogens, in contrast to the earlier steps in the synthesis-transport-discharge sequence, is most sensitive to stimulatory and inhibitory influences (20, 25, 29). Moreover, there is evidence that, in CDE-treated mice, the pancreas becomes relatively insensitive to secreta- gogues (CCK-PZ octapeptide (8) and bethanechol (J. Meldolesi, unpublished observatons). Thus, it appears that the CDE diet might affect primarily the process of stimulated exocytosis. The mechanism of this effect has not been elucidated. It is unlikely that this blockade results from reduced cellular energy stores, inasmuch as other energy-requiring processes (synthesis and intracellular transport) are not markedly altered. Protein car- boxy1 methylation may have a regulatory role in exocytosis (6, 32). Indeed, the CDE diet has been found to reduce the level of pancreatic protein carboxyl methylase activity (9). Alternatively, it was suggested that the block of exocytosis might be due to alterations of the phospho- lipid metabolism in membranes, possibly the zymogen granule and plasma membranes (17).

Morphological studies. No gross alterations in acinar cell ultrastructure were observed 24 h after exposure to the CDE diet. The only changes that were noted were an increase in the number of ZG, disappearance of exocytotic images, and absence of both large coated pits and vesicles. This combination of changes would not be un- expected. The increase in ZG number was previously reported by Lombardi et al. (17) and is the consequence of blocked exocytosis in the face of continued zymogen synthesis and intracellular transport. Failure to see exocytotic images is merely the morphological counterpart of blocked exocytosis. Finally, because ZG-plasma membrane fusion is blocked, recycling to the cytoplasm of excess membrane inserted at the cell surface (12) is not required, and large coated vesicles do not form.

Accumulation of ZG within the acinar cell cytoplasm apparently cannot proceed indefinitely. Crinophagy, viz.,

discharge of the granule contents into the lysosomal compartment by fusion of the corresponding limiting membranes, is the process by which secretory cells elim- inate their excess stored secretion products. This phenomenon was originally identified by Smith and Far- quhar (31) in pituitary mammotroph cells of lactating rats suddenly switched from a high to a low rate of prolactin release by removal of their litter. Later, it was recognized in other systems, including the B cells of the endocrine pancreas and the parathyroid gland (5, lo), and might be among the physiological mechanisms involved in the regulation of hormone release (5, 11). The large, membrane-bound vacuoles, poor in or entirely devoid of segregated membranes, which appear in the cytoplasm of mouse acinar cells around 24 h after the withdrawal of CDE diet, appear to have originated from crinophagy because I) some of them contain dense masses of filamentous material of approximately the same size as ZG and that look very similar to ZG stripped of their limiting membrane by in vitro detergent treat- ment (4); 2) all of them are positive for the presence of zymogens (revealed by immunocytochemistry) and acid phosphatase; and 3) they can be distinguished from condensing vacuoles (the only other large-sized organelles positive for acid phosphatase and zymogen immunolabeling) by their different appearance, organization of the segregated content, and wide distribution in the cytoplasm.

Crinophagy has not been observed in pancreatic acinar cells previously. Such a failure is probably due to the relatively low frequency of vacuoles containing immedi- ately recognizable membraneless ZG (which might be rapidly disassembled after engulfment) and to the fragil- ity of the vacuoles themselves, which appear well preserved and entirely membrane bound only in perfectly fixed specimens. In their previous studies on the pancreas of CDE-fed mice, Lombardi et al. (17) interpreted large vacuoles as focuses of ER degeneration, probably because the limiting membrane was not well preserved in their samples. Immuno- and acid phosphatase cytochemistry were not used in those studies.

At later times after withdrawal of the CDE diet (34-48 h), the large cytoplasmic vacuoles appear to have accu- mulated membranes and even recognizable organelles within their content. This could occur by two separate, yet apparently concomitant, processes, i.e., vesiculation and pinching in of the limiting membrane of crinophagic vacuoles and autophagocytosis. Autophagic and crinophagic vacuoles could then fuse, leading to large structures containing both zymogens (in soluble form or inte- grated into the filamentous material) as well as organelles and debris.

Acinar cells were also studied by freeze fracture 24 and 48 h after CDE diet withdrawal. Two of the observations made using this technique are worth noting. The first concerns membrane structure in crinophagic and autophagic vacuoles. In their limiting membranes, IMP are present in low density and are more concentrated on the P- than on the E-fracture face. Segregated membranes of autophagic vacuoles, in contrast, are nearly always IMP free. These observations, which are consistent with others recently made by R&z and Meldolesi (28), might have some bearing on the problem, as yet unresolved, of

by 10.220.33.1 on April 5, 2017


ownloaded from


G306

the subcellular origin of vacuole-limiting membranes. The second finding concerns the plasmalemma. At these, as well as at later time points, the structure of this membrane, including its junctional elements, was well preserved, and fusions of ZG with the lateral plasmalemma (i.e., lateral exocytoses) were never observed. This is important because the possibility of lateral exocytosis has been hypothesized recently to explain the onset of another form of AHPN, which develops in the rat after prolonged, supramaximal stimulation by a CCK-PZ analogue (I, 16).

The morphological observations described above indicate that a massive involvement of lysosomes occurs at stages that precede the onset of AHPN. Previous studies by Rao et al. (26, 27) demonstrated intraparenchymal activation of zymogens during CDE diet feeding and provided evidence that the lysosomal enzyme cathepsin B1 could be responsible for triggering this process, Among the possibilities these authors considered to account for intraparenchymal zymogen activation was “inclusion of zymogen granules in autophagic vacuoles” (27). However, those studies were performed using animals continuously fed the CDE diet throughout the experiment, and therefore the possibility that these effects are due to the direct action of ethionine could not be excluded. In contrast, our studies indicate that lysosomes can be involved long after the CDE diet is withdrawn. The intriguing possibility that seems to emerge from this observation is that lysosomal involvement and possibly also AHPN are con- sequences of the only phenomenon detected during exposure to the CDE diet, i.e., the block of zymogen discharge. Crinophagy and autophagocytosis are physiological processes (at least to a certain extent), and therefore it is not clear why their activation should have deleterious effects. One possibility previously suggested by Rao et al. (26) is that the capacity of pancreatic acinar cells to control the activated zymogens is limited and can be overwhelmed if the level is massive. Based on our present results, it can now be suggested that crinophagic and autophagic vacuoles are the intracellular sites where activation of zymogens might start and eventually be- come uncontrolled. In this case, some vacuoles might

REFERENCES

1. ADLER, G., W. BIEGER, AND H. KERN. Amino acid transport in the rat exocrine pancreas. III. Effect of maximal and supramaximal hormonal stimulation in vivo. CeZZ Tissue Res. 194: 447-462, 1978.

2. BERNFELD, P. Amylases a and p. In: Methods in EnzymoZogy, edited by S. P. Colowick and N. 0. Kaplan. New York: Academic, 1955, vol. I, p. 149457,

3. BURTON, K. A study of the conditions and mechanisms of the diphenylamine reaction for the calorimetric estimation of deoxyri- bonucleic acid. Biochem. J. 62: 315-324, 1956.

4, ERMAR, T. H., AND S. S. ROTHMAN. Internal organization of zymogen granules. Formation of reticular structures in vitro. J. Ultra- struct. Res. 64: 98413, 1978.

5. FARQUHAR, M. G. Secretion and crinophagy in prolactin cells. In: Comparative EndocrinoZugy of Prolactin, edited by H. D. Dell- mann, J. A. Johnson, and D, M. Klachko. New York: Plenum, 1977, p. 37-94.

6. GAGNON, C., AND S. HEISLER. Protein carboxyl methylation: role in exocytosis and chemotaxis. Life Sci, 25: 9934000, 1979.

7. GEUZE, J. J., J. W. SLOT, AND K. T. TOKUYASU. Immunocytochem- ical localization of amylase and chymotrypsinogen in the exocrine pancreatic cell with special attention to the Golgi complex. J, Cell BioZ. 82: 697-707, 1979.

KOIKE, STEER, AND MELDOLESI

rupture or release their content of activated digestive enzymes to the interstitium, thus initiating autodigestion of the gland and adjacent tissues. Although this is ad- mittedly still hypothetical, the stepwise development of the CDE-induced AHPN might be supported by the following arguments: I) pancreas toxicity of ethionine is prevented if protein synthesis (and therefore also ZG accumulation) is blocked (18), and 2) morphological alterations, up to the complete necrosis of cells and acini, were observed in all CDE-fed mice. However, only 40% of the mice that were not killed developed AHPN. The damage in the others may have been insufficient to overwhelm the protective mechanisms of the gland.

It should also be mentioned that large vacuoles, apparently containing zymogens, form in the cytoplasm in another model of experimental AHPN: that appearing in overstimulated animals (16). Morphological evidence has been presented suggesting that these vacuoles can release their contents into the interstitial space. These vacuoles were interpreted by Kern and associates (I, 16) as result- ing from the uncontrolled intracytoplasmic mutual fusion of ZG. Because, however, these vacuoles have not been studied by acid phosphatase cytochemistry, the possibility that they originate by crinophagy cannot be excluded.

Clearly, additional studies on the effects of the CDE diet in the mouse pancreas, particularly on the mechanisms by which this diet blocks exocytosis, wilI be needed to complete our understanding of this form of AHPN. Other important problems that deserve specific investi- gation are the possible difference in sensitivity between “basal” and %imulated” zymogen discharge, the pro- posed lysosomal activation of zymogens, and the relations between the latter process and the eventual development and spreading of cellular necrosis.

We thank Drs. J, Slot and H. Gueze, Department of Histology and Cell Biology, University of Utrecht, for helpful suggestions and advice in the immunocytochemical technique.

Present address of H. Koike: Dept. of Oral Pathology, Tokyo Dental College, Tokyo, Japan.

Received 9 June 1981; accepted in final form 17 September 1981.

8. GILLILAND, L., AND M. L. STEER. Effects of ethionine on digestive enzyme synthesis and discharge by mouse pancreas. Am. J. Physiol. 239 ( Gastointest. Liver PhysioZ. 2): G418-G426, 1980.

9. GILLILAND, L., N, TURNER, AND M, L. STEER. The effects of ethionine administration and choline deficiency on protein carbox- ylmethylase activity in mouse pancreas, Biochim. Biophys, Acta 672: 280-287, 1981.

10. HABENER, J. F., B. KEMPER, AND J. T. POTTS. Calcium-dependent intracellular degradation of parathyroid hormone: a possible mechanism for the regulation of hormone storage. Endocrinology 97: 431-441, 1975.

Il. HALBAN, P. A., AND C. B. WOLLHEIM. Intracellular degradation of insulin stores by rat pancreas islets in vitro, An alternative pathway for homeostasis of pancreatic insulin content. J. BioZ. Chem. 255: 6003-6006,198O.

12, HERZOG, V., AND H. REGGIO. Pathways of endocytosis from the luminal plasma membrane in rat exocrine pancreas. Eur, J. Cell BioZ. 21: X41-150, 1980.

13. JAMIESON, J, D., AND G. E. PALADE. Intracellular transport of secretory proteins in pancreatic acinar cells. I. Role of the periph- eral elements of the Golgi complex. J. CeZZ BioL 34: 577-596, 1967.

14. JAMIESON. J. D.. AND G. E. PALADE. Intracellular transDort of

by 10.220.33.1 on April 5, 2017


ownloaded from



secretory proteins in pancreatic acinar cells, II. Transport to condensing vacuoles and zymogen granules, J, Cell BioZ. 34: 597-615, 1967.

15. KOIKE, K., Y. TANAKA, G. RBz, AND J. MELDOLESI. Membrane interactions in pancreatic acinar cells: exocytosis, recycling and autophagocytosis. In: Biology of Normal and Cancerous Exocrine Pancreatic Cells, edited by A. Ribet, L. Pradayrol, and C. Susini. Amsterdam: Elsevier/North-Holland, 1981, p. 215-228.

16. LAMPEL, M., AND H. KERN. Acute interstitial pancreatitis in the rat induced by excessive doses of a pancreatic secretagogue. Vir- chows Arch. PathoZ. Anat. HistoZ. 373: 97-117, 1977.

17. LOMBARDI, B., L. W. ESTES, AND D. S. LONGNECKER. Acute hemorrhagic pancreatitis (massive necrosis) with fat necrosis induced in mice by DL-ethionine fed with a choline deficient diet. Am. J. Pathol: 79: 465-476, 1975,

18. LOMBARDI, B., AND N, K, RAO. Acute hemorrhagic pancreatic necrosis in mice. Influence of the age and sex of the animals and of dietary ethionine, choline, methionine and adenine sulfate. Am. J. Pathol. 81: 87-100, 1975,

19. LOWRY, 0. H., N. J. ROSEBROUGH, A. L. FARR, AND R. J. RANDALL. Protein measurement with the Folin phenol reagent. J. BioZ. Chem. 193: 265-275, 1951.

20. MELDOLESI, J., N. BORGESE, P. DE CAMILLI, AND B. CECCARELLI. Cytoplasmic membranes and the secretory process. In: Membrane Fusion, edited by G. Poste and G. N. Nicolson. New York: Elsevier/ North-Holland, 1978, p. 509-627.

21. MELDOLESI, J., J. D. JAMIESON, AND G. E. PALADE. Composition of cellular membranes in the pancreas of the guinea pig. I. Isolation of membrane fractions. J. Cell BioZ. 49: 109-129, 1971.

22. NOVIKOFF, A. B. Lysosomes in the physiology and pathology of cells. Contributions of staining methods. Ciba Foundation Sym- posium on Lysosomes. London: Churchill, 1963, p. 36-52.

23. NOVIKOFF, A. B., M. MORI, N. QUINTANA, AND A, YAM. Study on the secretory process in the mammalian exocrine pancreas. I. The condensing vacuoles. J. CeZZ BioZ. 75: 148-165, 1978,

24. OLIVER, C. Cytochemical localization of acid phosphatase and trimetaphosphatase activities in exocrine acinar cells. J, Histochem. Cytochem. 28: 78-84,198O.

25, PALADE, G. E. Intracellular aspects of the process of protein secretion. Science 189: 347-358, 1975,

26. RAO, K. N., J. TUMA, AND B. LOMBARDI. Acute hemorrhagic pancreatic necrosis in mice. Intraparenchymal activation of zymogens and other enzyme changes in pancreas and serum. Gastroen- teroZogy 70: 720-726, 1976.

27. RAO, K. N., M. F. ZURETTI, F. M. BACCINO, AND B. LOMBARDI, Acute hemorrhagic pancreatic necrosis in mice. The activity of lysosomal enzymes in the pancreas and the liver, Am. J. PathoZ. 98: 45-59, 1980,

28, RI&, G., AND J. MELDOLESI. Freeze-fracture of drug-induced autophagocytosis in the mouse exocrine pancreas. Lab, Invest, 43: 269-277,198O.

29. SCHEELE, G. A. Biosynthesis, segregation, and secretion of export- able proteins by the exocrine pancreas. Am. J. PhysioZ. 238 (Gas- trointest. Liver Physiol, 1): G467-G477, 1980.

30. SCHEELE, G. A., G. E. PALADE, AND A, M, TARTAKOFF. Cell fractionation studies on the guinea pig pancreas. Redistribution of exocrine proteins during tissue homogenization. J. Cell BioZ. 78: HO-130,1978.

31. SMITH, R. E., AND M. G. FARQUHAR~ Lysosome function in the regulation of the secretory process in cells of the anterior pituitary gland. J, Cell BioZ. 31: 319-347, 1966.

32. STRITTMATTER, W. J., C. GAGNON, AND J. AXELROD. P-Adrenergic stimulation of protein carboxyl methylation and amylase secretion. J. Pharmacol. Exp. Ther, 207: 419-424, 1978.

33. TANAKA, Y,, P. DE CAMILLI, AND J. MELDOLESI. Membrane interactions between secretion granules and plasmalemma in three exocrine glands. J. CeZZ BioZ, 84: 438-453, 1980.

34, TOKUYASU, K. T. Immunochemistry on ultrathin frozen sections. Histochem. J. 12: 381-404, 1980,

by 10.220.33.1 on April 5, 2017


ownloaded from


Documents

Pancreatic effects of ethionine: blockade of exocytosis ......nophagy) can be observed, and, 24 h later, autophagocytosis is noted. Finally, shortly before the onset of pancreatic