THE ANATOMICAL RECORD 227:232-244 (1990)

Anatomy of the Pancreas and Langerhans Islets in Snakes and Lizards A. A. MOSCONA Cummings Life Science Center, The University of Chicago, Chicago, Illinois 60637

ABSTRACT The pancreas of snakes (18 species) was comparatively examined and classified into five major types, based on structure of the lobes and ducts, spatial relationships with the spleen and the gall bladder, and the disposition of islet cells. These types trend toward fusion of the pancreatic lobes and compaction of the pancreas-a progression that coincides with the phylogeny of the snakes. The more primitive pancreas of lizards (17 species) also was surveyed; that of Varanus is of special interest because its structure is intermediate between the extended, tri-lobate pancreas of lizards and the compact pancreas of snakes and may represent a transitional link in the evolution of this organ. Islet tissue is always confined to the dorsal lobe and is concentrated in its distal region adjacent to the spleen. In primitive snakes and in Varanus, a large islet mass is sequestered within a distinct juxtasplenic “islet body” distanced from the dorsal lobe and connected to it by a slender stalk. In some of the most advanced snake species, numerous islets of endocrine cells are found within the spleen. The occurrence and formation of these intrasplenic islets is described in detail. The anatomic “affinity” between spleen and the islet region of the pancreas is discussed. A hypothesis for the development of the pancreas from embryonal placodes on the mid-gut is presented; i t proposes that the exocrine and the endocrine components derive from different progenitor cells, and that the endocrine progenitors are located in the center of the dorsal placode. The hypothesis combines embryological and evolutionary views about the origin of the pancreas, and offers a rationale for differences in its structure and in the disposition of the islets.

INTRODUCTION

The remarkable morphological diversity of the pancreas in snakes and the occurrence of Langerhans islets within the spleen have attracted interest since before the turn of this century (Laguesse, 1893,1901;Gianelli and Giacomini, 1896; Diamare, 1899, 1905). Following the discovery of insulin, mammalian islets became the main focus of research, and interest in the pancreas of snakes diminished. Siwe’s (1926, 1937) comprehensive pancreas studies also included information on embryology and anatomy of this organ in snakes. Extending Laguesse’s work, Thomas (1942) reported massive concentrations of islets near the spleen and, in some snake species, also within the spleen. Moscona (1954) described the formation of intrasplenic islets in snakes and noted seasonal changes in their composition. More recent work on snake pancreas was concerned with histochemistry and electron microscopy of islet cells and with evolutionary considerations (Miller, 1962; Barrington, 1964; Hellerstrom and Asplund, 1966; Trandaburu and Calugareanu, 1969; Gabe, 1970; Miller and Lagios, 1970; Falkmer and Patent, 1972; Penhos and Ramey, 1973; Epple and Brinn, 1975; Bonner-Weir and Weir, 1979; Buchan, 1984; Rhoten, 1984). There have been relatively few recent studies on the microscopic anatomy of snake pancreas, but these contributed important insights into physiological and evolutionary Q 1990 WILEY-LISS. INC

aspects of this organ (Miller and Lagios, 1970; Epple and Brinn, 1975; Bonner-Weir and Weir, 1979). This paper describes the morphology of the pancreas and the disposition of the islet tissue in a number of snake species; i t classifies the pancreas of snakes into five major types, based on structure of the lobes and ducts, the distribution of islet tissue, and relationships with spleen and gall bladder. Also, the pancreas of lizards was examined, primarily a s a guide to the more complex anatomy of snake pancreas. The results provide a n overview of pancreas morphology in these groups, point to common patterns that underlie structural diversities, and furnish material for considerations of pancreas development, morphogenesis, and phylogeny. MATERIALS AND METHODS

This study is based on examination of 18 species of snakes (82 specimens) and 17 species of lizards (78 specimens). These species, except Python regius,are all native to the region of eastern Mediterranean and were collected in Israel. Most of the specimens examined were adults captured in the field. They were sacrificed within a few hours or days after capture by intramus-

Received July 5, 1989; accepted September 19, 1989.

PANCREAS AND ISLETS I N SNAKES AND LIZARDS

cular injection of a lethal dose of barbiturate, decapitated, dissected, and studied. In addition, several wellpreserved specimens from a collection in the Department of Zoology of The Hebrew University in Jerusalem were examined, through the courtesy of the late Professor Georg Haas, to whom I am grateful also for identifying the species used in this work. Python regius and Varanus griseus were available only as preserved specimens. The following species were investigated (number of specimens is given in parenthesis).

233

into a heart-shaped structure and attached to the duodenum posterior to the spleen and close to it. The spleen is spherical and lies behind a large, oblong gall bladder. In snakes the gall bladder is located outside the liver and the long hepatic duct splits into branches. The cystic duct emerges from the anterior pole of the gall bladder and i t also subdivides into several branches (Fig. 2 A-C). The branches of the two ducts form anastomoses, resulting in a biliary plexus (equivalent to the common bile duct) whose complexity varies among species and individuals. The ducts enter into Snakes the pancreas near the middle of the gland. A compact Typhlops vermicularis (4);Typhlops simoni (2);Lep- pancreas attached to the gut behind the spleen and the totyphlops macrorhynchus (1); Eryx jaculus (6);Python gall bladder represents the most specialized form of regius (1);Natrzx tesselatus ( 3 2 ) ;Coluber jugularis ( 3 ) ; this organ in snakes and is characteristic for ColuColuber najadum (4);Coluber nummifer ( 2 ) ;Oligodon bridae, Natricidae, Viperidae, and several species of (Holarchus) melanocephalum (1); Eirenis (Contia) de- Opistoglypha. It will be referred to as the Colubrid Eirenis (Contia) coronella (1);Malpolon type. cemlineata (4); monspessulanum (9); Tarbophis savignyi (4);Tarbophis guentheri (2);Psamophis schokari (2); Micrelaps mul- Typhlopid type leri ( 2 ) ;Vipera palaestinae ( 2 ) . In the Typhlops species examined (Figs. 2 D,E), the spleen is situated near the front end of the gall bladder Lizards and tends to be oblong. The pancreas is more elongated Stenodactylus stenodactylus (1);Tropiocolotes steud- and consists of two connected parts. The posterior part neri ( 2 ) ;Gymnodactylus kotschyi (1);Hemidactylus tur- (the ventral lobe; see below) is heart-shaped and is atcica (10); Ptyodactylus hasselquisti (4);Acanthodacty- tached to the duodenum immediately behind the gall lus schreiberi ( 5 ) ;Agama pallida (2); Agama savignyi bladder. The anterior part (the dorsal lobe) is connected (1);Agama stelio ( 8);Ablepharus kitaibeli ( 5 ) ;Mabuya to the posterior part by a n isthmus, or a short stalk (as vittatus ( 8 ) ;Eumeces schneideri (4);Chalcides ocellata in Typhlops simoni); it extends forward alongside the (10); Chalcides guentheri ( 2 ) ; Lacerta laevis (6); gall bladder toward the spleen and is closely juxtaposed to it. The cystic and hepatic ducts usually do not form Ophisops elegans (8); Varanus griseus (1). Following dissection, the pancreas and adjacent or- branches; they join near the center of the posterior part gans were drawn by the author with the help of camera of the pancreas and enter into the gland. Variations in lucida. The cystic and hepatic ducts were injected with this Typhlopid type of pancreas affect, mainly, the India ink to visualize their branches and plexuses. For length and thickness of the anterior part. histology, the excised pancreas was fixed in neutral formaldehyde or in Zenker’s fluid; sections were stained Boid type In the Boidae examined here, Eryx jaculus and Pywith hematoxylin-eosin or tri-chrome stain and were examined for distribution of the islet tissue and location thon regius, the spleen is located near the front end of of the ducts. This paper does not deal with the identities the gall bladder and the pancreas has two distinct parts of the various islet cell types, beyond what could be (Figs 2 F,G). The anterior part emerges from the posterior a s a slender process t h a t extends forward toinferred from conventional staining. wards the large spleen which is located near the front RESULTS of the gall bladder; this process expands into a bulky, Pancreas Types in Snakes: General Description bulbous body which is closely attached to the spleen. On the basis of this study and literature survey, the This anteriorly elongated part of the pancreas is revarious forms of snake pancreas can be grouped into ferred to a s the “splenic process” and its expansion a s several major types, according to anatomical features the juxtasplenic body. The stalk portion of the splenic and relation to the spleen and the gall bladder. The process can be very thin and seems to have been overtypes proposed here generally coincide with the phylo- looked in some of the previous descriptions of this type genetic classification of snakes (Underwood, 1967) ac- of pancreas. Thus, a distinct juxtasplenic body located cording to which, Colubridae and Natricidae represent near the front end of the gall bladder and connected to the most advanced families, Typhlopidae and Lepto- the posterior part of the pancreas by a long stalk-like typhlopidae are a more archaic group, and Boidae are process characterizes the Boid type pancreas. In Eryx jaculus (Fig. 2 F), the posterior part of the a n intermediate group. It is debatable whether the second or the third group is evolutionarily closest to the pancreas is compact and heart-shaped. The hepatic duct forms several long branches which join the cystic saurian ancestors of snakes. duct as they enter into the posterior part near its base. Colubrid type In Python regius (Fig. 2 G), the posterior part is unThe pancreas most commonly described a s “typical” usual; i t consists of several lobules that partly surfor snakes is small and compact, of the kind found in round the caudal pole of the gall bladder. One of these Coluber asianus, Natrix tesselatus, Holarchus melano- lobules is larger and elongates into a slender stalk that cephalum, and Viperapalaestinae (Figs. 1A, B, 2 A-C). extends toward the spleen and terminates in a juxtaWhereas in most higher vertebrates the pancreas is splenic body. The cystic duct does not exit from the elongated or diffuse, in these snakes it is compacted anterior pole of the gall bladder, as in other snakes, but

234

A.A. MOSCONA

Fig. 1 . Pancreas and adjacent organs in snakes. A: Coluber asianus fjugularisi ( x 1.5). B: Natrir tesselatus showing suprasplenic islets; B1, dorsal view; B2, side view of another specimen with prominent islets on the spleen surface. x 3. C: Malpolon monspessulanum, its

spleen invaded and “split” by islet tissue. x 1. D: Varanus griseus; ‘‘islet body” attached to the spleen and connected by stalk to the dorsal lobe. x 2.

PANCREAS AND ISLETS IN SNAKES AND LIZARDS

t

235

i

A Fig. 2. Pancreas and adjacent organs in snakes. A Coluber asianus; compact pancreas, spleen, gall bladder, and biliary ducts (scale, 0.5 cm). B: Holarchus melanocephalum (scale, 0.25 cm). C: Viperapalaestinae (scale, 0.5 cm). D: Typhlops uermicularis (scale, 0.15 cm). E:

Typhlops simoni (scale, 0.15 cm). F: Eryx jaculus (scale, 2 cm). G Python regius (scale, 2 cm). H: Leptotyphlops macrnrhynchus (scale, 0.2 cm). I: Malpolon monspessulanum (scale, 1 cm). (Fig. 2 F-I on following page.)

near its posterior end. Because of these differences, the Boid type of pancreas is subdivided into two sub-types, Ericid and Pythonid.

Leptotyphlopid type

A b breuiations cd dl dS dW gb h hd jx P SP spi vl

Cystic duct Dorsal lobe Duct of Santnrini Duct of Wirsung Gall bladder Liver Hepatic duct Juxtasplenic body Pancreas Spleen Intrasplenic islet tissue Ventral lobe

In Leptotyphlops macrorhynchus (Fig. 2 H), the spleen is situated in front of the gall bladder and is almost as large as the bladder. The posterior part of the pancreas is compact, heart-shaped, and is attached to the gut behind the gall bladder; it sends forward a slender splenic process that expands into a large juxtasplenic body. The hepatic and cystic ducts enter into the lower half of the posterior part. Thus, the Leptotyphlopid type pancreas is characterized by a splenic body attached to a large and elongated spleen located anterior to the gall bladder. Malpolon type

The pancreas of Malpolon monspessulanum (Opisthoglypha) is different from that of any other snake

A.A. MOSCONA

236

G

H

I Fig. 2 F-I.

I

PANCREAS AND ISLETS IN SNAKES AND LIZARDS

237

h F Fig. 3. Pancreas in lizards. A: Eumeces schneideri; dorsal lobe process terminates in juxtasplenic body (scale, 0.5 cm). B: Morphology in different species (1,Hemidactylus turcica; 2, Gymnodactylus kotschyi; 3, Stenodactylus sp.; 4, Tropzocolotes steudneri; 5, Pytodactylus hasselquisti; 6, Ablepharus kitaibeli; 7, Ophisops elegans; 8, Agama stelio). C: Diagram of ducts in lizard pancreas (dS, hatched; dW, white;

cd, black). D: Varanus griseus: a) ventral, b) dorsal aspects to show connection of juxtasplenic body with dorsal lobe (scale, 1 cm). E: Diagram of ducts in the pancreas of Varanus (dS, hatched; dW, white; cd, black). F Distribution of islet tissue (black) in Eumeces. G: Islet tissue (black) in Varanus (mass of islet tissue within the juxtasplenic body).

examined (Figs. lC, 21). As previously described (Moscona, 1954), it consists of two elongated, almost separate parts suspended in the loop of the duodenum and joined by their distal ends. In full-grown specimens (1.5 m) the total length of the pancreas was 10 cm, while in Colubrids of similar size the pancreas was about 2 cm long. The posterior part of the pancreas (ventral lobe) is attached to the rear portion of the duodenum. The anterior part (dorsal lobe) is attached below the pylorus,

immediately behind the spleen, and forms two short arms which partly surround the spleen. The hepatic and cystic ducts form a plexus and enter into the middle of the posterior part. This kind of pancreas appears not to have been described before in snakes. It is here designated as Malpolon type. It should be noted that the other Opisthoglypha examined, Tarbophis sauignyi, T. guentheri, and Psammophis sp.), the pancreas is of a Colubrid type.

A.A. MOSCONA

238 The Pancreas of Lizards

The extended pancreas of lizards is considered to be a more primitive form, a prototype from which the compact snake pancreas had evolved. Figure 3A depicts the pancreas of Eumeces schneideri (Scincidae). It consists of three elongated branches that extend from the “head” of the pancreas which is attached to the duodenum and contains the ampulla of Vater. Two of the branches are derived from the embryonic ventral pair of pancreatic primordia and represent the ventral lobe of the pancreas; the third branch is the dorsal lobe and it is derived from the dorsal primordium and (Brachet, 1896; Miller and Lagios, 1970). Whereas in many lizards the two ventral branches are, more or less, fused (Fig. 3B), in Eumeces they are distinct and provide a clearer image of the basic structure of the saurian pancreas in relation to embryonic primordia. In Eumeces, one of the ventral lobe branches extends anteriorly along the gut and the bile duct toward the gall bladder which, in lizards is located within the liver. The other ventral lobe branch is short and runs laterally across the gut. The third branch, the dorsal lobe, extends from the pancreas head into the mesentery and elongates towards the distant spleen. This dorsal lobe (sometimes referred to a s the splenic process) terminates in a juxtasplenic body similar to that of Boid-type pancreas. In the early embryo, the pancreatic primordia open separately into the gut (Brachet, 1896; Pictet and Rutter, 1972). In lizards these openings merge into a single outlet, the ampulla of Vater, into which empty the two pancreatic ducts. The duct of the ventral lobe is the duct of Wirsung, shown in the diagram in Figure 3C (it has two branches in Eumeces and other lizards with a dual ventral lobe). The duct of the dorsal lobe, the duct of Santorini, runs from the juxtasplenic body down the dorsal lobe into the ampulla. The bile duct does not traverse through the pancreas, as it in snakes, but enters directly into the ampulla (Fig. 3C). Variations among lizard species are mainly in the length and fusion of the pancreatic lobes, size of the juxtasplenic body and its proximity to the spleen, a s shown in Figure 3B. A particularly important variant of lizard pancreas is that of Varanus griseus (Varanidae) shown in Figures 1D and 3D. Here, the pancreatic lobes are fused and partly compacted; its structure is intermediate between the extended pancreas of lizards and the condensed pancreas of snakes. The pancreas is suspended in the mesentery and is attached to the duodenum by a short funnel-shaped conduit that leads into the ampulla of Vater. The part closest to the gut is drained by the duct of Wirsung and consists, largely, of the ventral lobe (its origin from two ventral primordia can not be discerned in adults); i t also includes a small portion of the dorsal lobe around the terminal sector of the duct of Santorini (Fig. 3E). The mid-part of the pancreas consists, predominantly, of the dorsal lobe; i t projects toward the spleen a thin, stalk-like process which expands into a large juxtasplenic body (found also in other Varanus species; Gabe, 1970). The dorsal lobe, including the splenic process are drained by the Santorini duct (Fig. 3E). The bile duct enters the ampulla separately. Thus, in contrast to the elongated pancreas lobes of more primitive lizards, in Varanus they are fused and

shortened. Further evolutionary progression in this direction might have lead to the even more compacted types of snake pancreas found in Boids and Leptotyphlopids. Such a progression is consistent with the postulated evolutionary origin of snakes from primitive varanid-type ancestors, suggested by other considerations (Walls, 1942; Romer, 1945). Distribution of Islets in Lizard Pancreas

In all the lizards examined, Langerhans islets were detected in the domain of the dorsal lobe. Their concentration was consistently greatest in the region closest to the sdeen. as shown for Eumeces schneideri in Figure 3F. The distribution of islets tissue in the pancreas of Varanuspriseus is shown in Figure 3G. There is a massive concktration of islet tissue in the large juxtasplenic body which constitutes, in effect, a n “islet body.” Some of the endocrine cells in this body are arranged in tubular configurations contiguous with ductal structures, probably the end branchings of the Santorini duct. A distinct boundary zone separates between the islet body and the spleen. The connecting stalk contains clusters of islets along the Santorini duct; smaller islets are present throught the dorsal lobe, their size and density decreasing in the direction of the ventral lobe. I

Pancreatic Ducts in Snakes

The pancreatic lobes in snakes are fused and compacted and their identity and relation to embryonic primordia is not readily apparent. However, by tracing the disposition of the pancreatic ducts, the domains of the lobes can be identified. The anterior part of the pancreas closest to the spleen is supplied by the duct of Santorini; accordingly, this domain represents the dorsal lobe. The posterior part is supplied by the duct of Wirsung and represents the domain of the ventral lobe which originates from two ventral embryonic primordia; not only do they fuse completely but, in most snakes they also become totally confluent with the dorsal lobe. Colubrid type

The disposition of the ducts in Colubrid-type pancreas is shown in Figure 4A. In most species with this type of pancreas both ducts open into a single outlet, the ampulla of Vater. The common bile duct traverses through the ventral lobe and opens into the ampulla. This ductal pattern was found in Coluber rauergieri, C. asianus, C. najadum, and C. numifer; Oligodon (Holarchus) melanocephalum; Eirenis (Contia) decemilineata; E . coronella; Natrix tesselatus; Tarbophis sauignyi. The Colubrid-type pancreas of the viper Vipera palaestinae has two separate outlets into the duodenum (also found in Vipera berus; Laguesse, 19011, shown in Figure 4B. The Wirsung duct of the ventral lobe enters together with the bile duct into a posteriorly located ampulla; the Santorini duct of the dorsal lobe opens into a separate anterior ampulla located near the pylorus. The two outlets correspond to the openings of the pancreatic rudiments in the embryo. In other species with Colubrid-type pancreas these openings merge into a single outlet, in vipers they remain separate. Their

PANCREAS AND ISLETS IN SNAKES AND LIZARDS

239

F

E Fig. 4. Diagrams of location of ducts and ampullae in pancreas of snakes (dS, hatched; dW, white; cd, black). A Coluber. B: Vipera. C: Typhlops. D: Eryx. E: Leptotyphlops. F Malpolon.

persistence provides confirmation for the identity of the ducts and the domains of the fused lobes. Typhlopid type

In the species examined here, Typhlops simoni and T . uermicularis, the pancreas has two separate outlets into the gut (Fig. 4C).The Santorini duct runs from the splenic end of the pancreas down through the elongated anterior part and opens a t its base into the duodenum; it identifies the domain of the dorsal lobe. The Wirsung duct branches in the posterior part and opens into the posterior ampulla; its domain represents, largely, the ventral lobe. The bile duct passes through the ventral lobe and enters into the posterior ampulla. Boid and leptotyphlopid types

These types are characterized by a distinct juxtasplenic body connected to the rest of the pancreas by a long, slender stalk. In Eryx jaculus (Boidae) and Leptotyphlops macrorhynchus (Leptotyphlopidae) there is a single ampulla; it is located near the base of the pan-

creas and into it open the two pancreatic ducts and the bile duct, a s shown in Figures 4D,E. The Santorini duct runs from the juxtasplenic body down the stalk and through the domain of dorsal lobe. The duct of Wirsung branches in the domain of the ventral lobe. The bile duct enters into the base of the ventral lobe. In Python regius, the Santorini duct runs from the juxtasplenic body down the stalk into a small anterior lobe and connects to a n opening into the duodenum. The rest of the pancreas, representing the ventral lobe, consists of a number of leaflet-like lobules with several openings into the intestine. The origin of these lobules was not investigated. It is possible that they start out in the embryo as multiple outpocketings from the gut and continue to develop a s individual subunits; alternatively, they may arise a s branches of the ventral lobe. Malpolon type

The two distinct and elongated lobes of the pancreas of Malpolon monspessulanus have separate outlets into

240

A.A. MOSCONA

from the islet-rich region of the dorsal lobe: a fingerlike process bulges out from this region, penetrates into the closely adjacent spleen and pushes through it; the growing mass of islet tissue forms branches some of which pierce the surface of the spleen. Previously reported intrasplenic islets were found in snakes which, according to the classification suggested here, have a Colubrid-type pancreas. Diamare (1899) found occasional islets in the spleen of large specimens of Vipera berus. Thomas (1942) mentioned intrasplenic islets in several Colubrid species. They were noted in adult Crotalus uiridzs (Hellerstrom and Asplund, 1966), Crotalus atrox (Buchan, 19841, and Bungarus fasciatus (Gabe, 1970). However, they do not appear to The Distribution of Islets in Snake Pancreas occur regularly and consistently. A comprehensive surIn the snake pancreas there is a massive concentra- vey of still other species with Colubrid-type pancreas tion of islet tissue in the region adjacent to the spleen would be desirable, with attention to size, gender, and (“principal islet”). In some species islet tissue was season. It should be noted here that I found intrafound also within the spleen itself (Laguesse, 1901; splenic islets only in one other pancreas type, that of Thomas, 1942; Moscona, 1954; Miller, 1962; Heller- Malpolon (see below); it is significant that only in Malstrom and Asplund, 1966; Buchan, 1984; Rhoten, polon and Colubrids the spleen is located behind the 1984); however, the occurrence of intrasplenic islets gall bladder immediately adjacent to the dorsal lobe. appears not to have been systematically investigated. In the present study, islet distribution and its relation Malpolon type In the pancreas of Malpolon monspessulanum (Fig. to the spleen were surveyed in the context of the pan1C) islets are confined to the dorsal lobe (Fig. 5F); their creas types described above. concentration is greatest in the region immediately Colubrid type next to the spleen and decreases in the posterior direcIn this compact pancreas, islets are confined to the tion. Since the two lobes are separate in this species, domain of the dorsal lobe (Fig. 5A); their largest con- the apparent absence of islets in the ventral lobe centration is always in the region closest to the spleen strongly favours the view that islet develop only in the (Laguesse, 1901; Trandaburu and Calugareanu, 1969). dorsal lobe. Smaller islets are distributed throughout the rest of The dorsal lobe in Malpolon forms two short arms the dorsal lobe, their size and frequency decreasing in which partly embrace the adjacent spleen. Already in the posterior direction (Miller, 1962). I found no islets young individuals one of the arms contains a massive in the domain of the ventral lobe; this agrees with ev- concentration of islet tissue (Fig. 5G). A finger-like proidence that islets develop predominantly, perhaps ex- cess rich in islet tissue penetrates from this arm into clusively, in the dorsal primordium of the pancreas the spleen and gives rise to massive intrasplenic islets (Frye, 1958; Dieterlen-Lievre, 1970; Pictet and Rutter, (Fig. 5H, I) a s described above for Natrix. In full-grown 1972). specimens, branches of the islet mass extend through Among all species with Colubrid-type pancreas sur- the spleen to its surface and form suprasplenic islets veyed in this study, I found intrasplenic islets only in (Figs. 55, 1C). The histology of these splenic islets is Natrix tesselatus, in 12 out of 32 specimens examined. similar to that described for Natrix; they are accompaThey were conspicuous in full-grown individuals, ab- nied by tubular and ductal elements, and are interconsent in the youngest specimens, and most massive in nected by fibrous tissue which links them to the dorsal egg-bearing females. Therefore, intrasplenic islets de- lobe. In large females caught dormant during the winvelop progressively as the snake grows and matures ter season, much of the intrasplenic islet mass was reand may be subject to seasonal and hormonal influ- placed by strands of ductal and connective tissue (Fig. ences (Moscona, 1954). 5K), a s previously described (Moscona, 1954). In full-grown Natrix, the intrasplenic islet tissue traversed the spleen and protruded from its surface a s Typhlopid type pinhead-sized buttons visible to the naked eye (Fig. In this type of pancreas, the spleen is located near 1B). Islet and spleen tissue were not directly confluent, the front end of the gall bladder, and the dorsal lobe but were separated by a boundary. The intrasplenic extends toward it. In Typhlops vermicularis (Fig. 5B) islets consisted, in part, of cells arranged in tubular there is a large islet mass within the dorsal lobe next to and cord-like configurations; their structure was simi- the spleen; however, I found no islets within the spleen. lar to that previously described for pancreatic islets in Small islets were distributed in the rest of the dorsal Natrix natrix (Trandaburu and Calugareanu, 1969). lobe and near its base, close to the anterior ampulla. Histological staining detected A and B cell types, A cells predominating (detailed examination of cell types Boid and leptotyphlopid types was not pursued in this study). Strands of ductal tissue Characteristic of these types is the presence of a and fibrocytic cells interlinked the islets within the large mass of islet tissue in the juxtasplenic body which spleen and connected them to the dorsal lobe. is situated near the front pole of the gall bladder and is Examination of a n age series of Natrix specimens connected to the rest of the dorsal lobe by a slender revealed that intrasplenic islets arose a s a projection stalk. This “islet body” is embedded in a matrix of tu-

the gut (Fig. 4F). The Santorini duct of the anterior, dorsal lobe opens into the pyloric part of the duodenum. The Wirsung duct connects to a posterior ampulla at the base of the ventral lobe that opens into the lower part of the duodenum. The bile duct enters into the ventral lobe and joins the posterior ampulla. As described above, also in vipers and Typhlops, there are two ampullae, but they are relatively close, while in Malpolon they are separated by a long stretch of duodenum. Apparently, in Malpolon embryo the duodenum between dorsal and ventral pancreatic primordia elongates rapidly, distancing the separately developing lobes from one another.

PANCREAS AND ISLETS I N SNAKES AND LIZARDS

B

V

id\

Di

24 I

\

E

Fig. 5. Distribution of islet tissue (black) in pancreas of snakes. A: Coluber. B: Typhlops. C: Leptotyphlops. D: Eryx (1, mature; 2, juvenile). E: Python; F Mulpolon. G-J: Formation of intrasplenic islet mass in Mulpolon. K: Its regression in a wintering female (description in text).

bular-acinar tissue and ductal structures, some of which lead into terminals of the duct of Santorini. In full-grown Eryx (Fig. 5 C ) the islet body partly surrounds the spleen; in Leptotyphlops i t is almost cornpletely enclosed by the very large spleen (Fig. 5D).

However, the two tissues are separated by a distinct boundary, and I found no islets within the spleen itself. Islets were present in the stalk along the Santorini duct, and a t its base within the domain of the dorsal lobe, and were especially large and numerous in Eryx.

A.A. MOSCONA

242

Figure 5C also depicts the pancreatic region and islet localization in a young Eryx, before the splenic process had fully extended. The juxtasplenic body in Python regius (Fig. 5E) also contains a large mass of islet tissue, but islets were not detected within the spleen. Numerous islets were found in the lobule to which the stalk is connected, but not in the other lobules of the ventral lobe domain. DISCUSSION

The large “islet-body” located in the juxtasplenic portion of Boid and Leptotyphlopid pancreas constitutes a distal extension of the dorsal lobe; i t is homologous, on one hand, with the islet mass (“principal islet”) located within the compact dorsal lobe of Colubrid-type pancreas and, on the other, with the juxtasplenic body of Varanus and the splenic process of lizard pancreas. Hence, the following anatomic-evolutionary progression can be envisaged from the extended lizard pancreas through Varanus, Leptotyphlopids, Boids, and Typhlopids to the compact Colubrid type: the gall bladder moves out from the liver toward the gut; the spleen translocates toward the gall bladder and, eventually, lies behind i t near the pancreas; the islet-rich splenic process contracts and becomes incorporated within the compacted dorsal lobe in a region close to the spleen; this region retains a tendency to project a short process into the adjacent spleen, as in Natrix and Malpolon. The presence of a n elongated dorsal lobe in primitive snakes is consistent with their postulated descent from lizard-like ancestors. In this respect, the pancreas of Varanus is noteworthy as a candidate evolutionary link from the extended pancreas of lizards to the compacted pancreas of snakes. There is, of course, a paradoxical incongruity between the extreme length of the snake body and the compacted form of snake pancreas. This evolutionary puzzle is further compounded by Malpolon, a snake in which both pancreatic lobes are unusually elongated. There is, a t present, no physiological rationale for the directional extension of the dorsal lobe toward a distant spleen, or for the close juxtaposition of the isletrich region and the spleen in snakes, lizards, and birds (Moscona and Zajicek, 1954). This anatomical “affinity” raises speculations about some, as yet unknown, functional relationship between islet tissue and spleen. Although this possibility may not be apparent in higher vertebrates because of different arrangement of these structures, it deserves interest in light of the situation in reptiles. Concerning intrasplenic islets, they occur only in snakes in which the spleen is situated behind the gall bladder very close to the pancreas, i.e., those with a compact Colubrid-type pancreas and in Malpolon. Their formation in Natrix and Malpolon was described in Results. During juvenile and post-juvenile growth, a finger-shaped short projection grows out from the isletrich region of the dorsal lobe, penetrates into the adjacent spleen, and gives rise to intrasplenic islets. I suggest that this projection is homologous with the dorsal lobe process of primitive snakes and lizards which extends toward a distant spleen. In Natrix and Malpolon the spleen abuts on the dorsal lobe and the projection pushes into it. Accordingly, intrasplenic islets represent a n atavistic phenomenon, a tendency of the dorsal

lobe to recapitulate its ancestral spleenward elongation. Their irregular occurrence and their absence in species with a distant spleen are consistent with this suggestion. The massive concentration of islet tissue in the distal part of the dorsal lobe, in snakes and lizards, raises questions about the embryonic origin and distribution of islets within the pancreas. The following comments address these questions. First, there is experimental evidence that islets arise or develop only in the lobe derived from the embryonic dorsal pancreatic primordium (Pictet and Rutter, 1972). The present observations are consistent with this view. Second, islet cells are thought to be evolutionarily descendent from gastrointestinal hormone-producing cells (Steiner et al., 1969). In some primitive fish, such cells are dispersed in the gut epithelium within a field occupied also by zymogen-secreting cells (Barrington, 1964; Epple and Brinn, 1975; Bonner-Weir and Weir, 1979). It has been hypothesized that, in the course of tetrapod evolution, ancestral progenitors of islet cells became aggregated into glandular clusters and associated with a n intestinal diverticulum, the antecedent of exocrine pancreas (Gorbman et al., 1983).Extrapolating this to ontogeny, I suggest that, also in embryos of higher vertebrates, these two components originate from disparate precursors, and that differences in their morphogenetic relationships result in architectural variations of the isletpancreas system (Epple and Brinn, 1975). Hence, the polymorphism of snake pancreas might be traceable to differences in composition and growth patterns of the pancreatic primordia. The pancreas of reptiles and higher vertebrates arises from three placodes located in the embryonic mid-gut, one dorsal and two ventral (in some cases, there is one ventral placode). The placodes evaginate into diverticular outpocketings, the primordia of the dorsal and ventral lobes, which subsequently merge to a n extent depending on the species (Pictet and Rutter, 1972). According to Siwe (1962, 1937), the progenitors of islet cells are located in the center of the dorsal placode as a row of cell clusters; the rest of this placode, and the other two placodes consist of exocrine tissue precursors. Inferring from earlier descriptions (Siwe’s, 1926; Pictet and Rutter, 1972), the dorsal placode evaginates center first (Fig. 6A,B) and the central cell clusters are lifted up on the apices of the outpocketings (Fig. 6C,D). As these elongate, most of the apical cells are carried toward the distal region of the dorsal lobe and give rise to endocrine islets. The stems of the outpocketings continue to branch and give rise to exocrine acini, tubules and ducts. The main lumen of the dorsal diverticulum becomes the duct of Santorini. This developmental model accounts for the abundance of islets in the distal part of the dorsal lobe and can explain its structural variations in snakes. It suggests that, in Boids and Leptotyphlopids (as well as in lizards) the center of the dorsal placode grows out and elongates into a n extended process (Fig. 6E) with an “islet body” at its end (Fig. 6F). In Colubrid-type pancreas the center of the placode does not elongate as much, and the dorsal lobe is compact (Fig. 6G). The above model implies that the dorsal placode is constituted as a compound “morphogenetic field”: developmental competence for islet cells is maximal in

PANCREAS AND ISLETS I N SNAKES AND LIZARDS

243

E

C

Fig. 6. Proposed developmental model for the dorsal lobe and islet tissue. For explanations, see Discussion. Black shading indicates location of islet-forming cells in the dorsal placode and dorsal lobe. A-C: Evagination of the dorsal placode; islet cells in apices of evaginations. D: Islet cells are carried distally on central evaginations.

E,F:In pancreas of Boids, Leptotyphlopids, and Vurunus, a rapidly elongating process distances most islet cells from the main body of the dorsal lobe (E) and sequesters them in a juxtasplenic “islet body” (F). G In Colubrid-type pancreas, islet tissue remains incorporated within the dorsal lobe.

the center and decreases centripetally; competence for exocrine and ductal cells resides throughout the field. Accordingly, morphogenesis of the dorsal pancreas reflects the characteristics of its placodal field, i.e., its overall size, proportions, and organization of the two kinds of progenitor cells, and relative growth rates of its parts. The anatomical location of the spleen may also be a factor, as mentioned earlier. Differences in these parameters could account for the various forms of the islet-pancreas system in snakes and in lizards. The concepts advanced here might apply also to the pancreas of other vertebrates.

fessor Heinz Medelsohn of Tel Aviv University. The author thanks Lyle L. Fox for printing the figures, and Linda Degenstein for dedicated assistance. Costs of manuscript preparation were partly paid by a stipend from the Louis Block Fund of the University of Chicago.

ACKNOWLEDGMENTS

This study was made possible thanks to the courtesy and guidance provided by the late Professor Georg Haas of the Hebrew University in Jerusalem, and Pro-

LITERATURE CITED Barrington, E.J. 1964 Hormones and Evolution. English University Press, London. Bonner-Weir, S., and G.C. Weir 1979 The organization of the endocrine pancreas: a hypothetical unifying view of the phylogenetic differences. Gen. Comp. Endocrinol. 38:28-37. Brachet, A. 1986 Recherches sur le development du pancreas et du foie (Selaciens, reptiles, mammiferes). Journ. de I’Anat. et de la Physiol., 32:620-696. Buchan, A.M. 1984 An immunocytochemical study of endocrine pancreas of snakes. Cell Tiss. Res., 2351657-661.

244

A.A. MOSCONA

Diamare, V. 1899 Studii comparativi s u l k isole di Langerhans del pancreas. Memoria I Intern. Mschr. Anat. Physiol., 16:155-209. Diamare, V. 1905 Studii comparativi sulk isole di Langerhans del pancreas. Mem. I1 Intern. Mschr. Anat. Physiol., 22,129-187. Dieterlen-Lievre, F. 1970 Tissus exocrine et endocrine du pancreas chez l’embryon de poulet: origine et interactions tissulaires dans la differenciation. Devel. Biol., 22t138-156. Epple, A,, and J.E. Brinn 1975 Islet histophysiology: evolutionary correlations. Gen. Comp. Endocr., 27:320-349. Falkmer, S., and G.J. Patent 1972 Comparative and embryological aspects of the pancreatic islets. In: Handbook of Physiology: The Endocrine Pancreas. D.F. Steiner and N. Freinkel, eds. Williams and Wilkins, Baltimore, Vol. 1, pp. 1-24. Frye, B.E. 1958 Development of the pancreas in Ambystoma opacum. Amer. J. Anat., 102t117-139. Gabe, M. 1970 Pancreas Endocrine. In: Traite de Zoologie. P. Grasse, ed. Masson et Cie., Paris. pp. 1333-1399. Gianelli, L. et E. Giacomini 1896 Recerche istologiche sul tube digerante dei Reitti 3. Intestino medio e terminale, fegato, pancreas. C. Sci. R. Acad. Fisiocrit. Siena, 8:3-11. Gorbman, A., W. Dickhoff, S. Vigna, N. Clark, and C. Ralph 1983 The endocrine pancreas. In: Comparative Endocrinology. John Wiley and Sons, New York, pp. 349-347. Hellerstrom, C., and K. Asplund 1966 The two types of A-cells in the pancreatic islets of snakes. Z. Zellforsch., 70r68-80. Laguesse, M.E. 1893 Sur la formation des ilots de Langerhans dans le pancreas. Comp. Rend. Soc. Biol., 45:819-920. Laguesse, M.E. 1901 Sur la structure de pancreas chez quelques ophidiens et particulierment sue les ilots endocrines. Arch. Anat. Microsc. Morphol. Exp., 4t157-218. Miller, M.R. 1962 Observations on the comparative histology of the reptilian pancreatic islets. Gen. Comp. Endocrinol. 2:407-414. Miller, M.R., and M. Lagios 1970 The pancreas. In: Biology of the

Reptilia. C. Gans. ed. Academic Press, New York, Vol3, pp. 319346. Moscona, A.A. 1954 Seasonal Changes in the Islets of Langerhans in snakes. Bull. Res. Council Israel, 4:253-255. Moscona, A.A., and A. Zajicek 1954 Note on the endocrine tissue in the pancreas of the chick. Bull. Res. Council Israel., 4:313-314.i Penhos, J., and E. Ramey 1973 Studies on the endocrine pancreas of amphibians and reptiles. Amer. Zool., 13:667-698. Picket, R., and W.J. Rutter 1972 Development of the embryonic pancreas. In: Handbook of Physiology: The Endocrine Pancreas. D.F. Steiner and N. Freinkel, eds. Williams and Wilkins, Baltimore, Vol 1, pp. 25-66. Rhoten, W.B. 1984 Immunocytochemical localization of four hormones in the pancreas of the garter snake, Thamnophis sirtalis. Anat. Rec., 208t233-242. Romer, A S . 1945 Vertebrate Paleontology. University of Chicago Press, Chicago. Siwe, S.A. 1926 Pankreasstudien. Gegenbaurs Morphologisches Jahrbuch, 57:84-307. Siwe, S.A. 1937 V. Die Grossen Driisen des Darmkanals, A. Die Leber. In: Handbuch der vergleichenden Anatomie der Wirbeltiere. Vol 3, pp. 725-774. Steiner, D.F., J.L. Clark, C. Nolan, A.H. Rubenstein, E. Margoliash, B. Aten, and P.E. Oyer 1969 Proinsulin and the biosynthesis of insulin. Recent Prog. Horm. Res., 25:207-282. Thomas, T.B. 1942 The pancreas of snakes. Anat. Rec., 82t327-345. Trandaburu, T., and L. Calugareanu 1969 Light and electron microscopic investigation of the endocrine pancreas of the grass-snake [Natrix n. natrix(L.11. Z. Zellforsch, 97:212-225. Underwood, G. 1967 A contribution to the classification of snakes. British Museum of Natural history, London. Walls, G.L. 1942 The vertebrate eye and its adaptive radiation. Cranbrook Institute of Science, Bloomfield Hills, Michigan.