72&MU: the occurrence of different lines of differentiation is not restricted to combinations of carcinoma and car- cinoid. Carcinosarcoma or sarcomatoid changes ...
APUD
Cells
in Teratomas
FRED T. BOSMAN, MD, and JAN-WILLEM K. LOUWERENS
From the Department of Pathology, University Medical Center,
Leiden, The Netherlands
The origin of the endocrine cells in the respiratory tract and the gastrointestinal tract is still a matter of debate. In the original concept of the amine precursor uptake and decarboxylation (APUD) system, all APUD cells were considered to be derived from the neural crest. More recently it has been proposed that the APUD cell types of the gastrointestinal and respiratory tracts originate from neuroendocrine-programmed ectoblast. Still other investigators have reported observations that favor a direct endodermal origin of these cell types. Based on the assumption that in teratomas different tissue types which in normal embryogenesis are derived from the neuroectoderm might be expected to occur together, we investigated a series of cystic ovarian tera-
tomas and testicular teratocarcinomas for the presence of brain tissue and of different types of APUD cells. In the ovarian teratomas, intestinal and respiratory APUD cell types were found almost exclusively without coexistence of brain tissue, whereas melanocytes, which are of neuroectodermal origin, occurred mostly together with brain tissue. In the testicular teratocarcinomas, intestinal types of APUD cells occurred without brain tissue. Peptide hormone production was found in appropriate tissues. It can therefore be concluded that in teratomas appropriate intestinal and respiratory APUD cells differentiate in and presumably descend directly from intestinal and respiratory epithelium. (Am J Pathol 1981, 104:174-180)
THE APUD (amine precursor uptake and decarboxylation) concept was proposed by Pearsel in order to link together a whole spectrum of cell types in different organs which shared amine processing characteristics, as expressed by the acronym. Initially it was suggested that all these cells were of neuroecto-
embryonic "epiblast" and share common neuroendocrine programming of which the final expression is determined by microenvironmental circumstances. APUD cells have been reported to occur in teratomas. 1720 Although the development of different tissue types in teratomas may not (completely) parallel normal embryogenesis, it seems reasonable to assume that if APUD cells in teratomas were derived from neuroectodermal-type tissues, such APUD cells would occur together wth other tissues showing neuroectodermal differentiation. Conversely, APUD cells occurring consistently without these elements would more likely not be of neuroectodermal origin. We therefore studied a series of ovarian teratomas and testicular teratocarcinomas with (immuno-) histochemical methods in order to establish 1) which types of APUD cells occurred together with brain tissue and 2) whether the existing peptide hormone producing APUD cell types occurred in appropriate histologic structures.
dermal origin. Neuroectodermal origin is well established for the neurohypophysis. Neuroectodermal origin of the hypothalamus has been proven,2 but the evidence for the neuroectodermal origin of the adenohypophysis has been contradicted by others.3 The avian allograft technique has shown that the APUD cells of the thyroid, carotid body, adrenal medulla, and skin derive from the neural crest.5-8 The origin of the APUD cells of the respiratory and digestive tracts, however, has remained a subject of discussion. Pearse and Polak9 and Pearse et al'0 reported indirect evidence that APUD cells of the respiratory and gastrointestinal tracts and some pancreatic islet cell types could have derived from precursor cells that had migrated from the neural crest. A large body of evidence, however, has been published suggesting that gastroenteropancreatic APUD cells do not arise from the neural crest."-" In a more recent and modified version of the APUD concept'6 Pearse proposed that all APUD cells derived (at different periods) from the
Supported by a grant from the Queen Wilhelmina Cancer Foundation (PA 77.52). Accepted for publication April 28, 1981. Address reprint request to Fred T. Bosman, MD, PhD, Department of Pathology, Wassenaarseweg 62, Leiden, The Netherlands.
0002-9440/81/0807-0174$00.85 © American Association of Pathologists
174
APUD CELLS IN TERATOMAS
Vol. 104aNo. 2
Materials and Methods From the files of the Department of Pathology 170 ovarian teratomas (both cystic "dermoids" and solid teratomas) that had occurred between 1970 and 1978 were retrieved. Sections stained with hematoxylin and eosin (H & E) were reviewed. All cases were stained with silver stains for melanocytes, and all teratomas containing other tissue elements in addition to epidermal structures were also investigated for other APUD cells. In addition, 6 teratocarcinomas of the testis, which contained abundant differentiated epithelial tissues, were studied. Serial 4-,u sections were cut from available paraffin blocks. Sections were stained for argyrophil cells according to the methods of Sevier and Munger2" and Grimelius.22 In cases with tissue types containing APUD cells other than melanocytes, additional sections were stained for the presence of polypeptidehormone-producing cells, using an indirect peroxidase-labeled antibody method as described previously.23 Specimens containing thyroid tissue were stained for calcitonin, whereas specimens containing gastrointestinal tissues were stained for gastrin, insulin, glucagon, somatostatin, and pancreatic polypeptide. The use of the anticalcitonin, antiglucagon, antiinsulin, and antigastrin antibodies has been described previously.23'24 The antisecretin antibody (a gift from Dr. J. S. Levine, University of Colorado Medical School) was prepared against porcine secretin. The antisomatostatin antibody was obtained from U.C.B. (Brussels, Belgium). The anti-bovine-pancreaticpolypeptide antibody was a gift from Dr. R. Chance (Eli Lilly, Indianapolis, Indiana). The peroxidaselabeled antibody was prepared by conjugation of goat anti-rabbit IgG (Miles Yeda, Israel) to horseradish peroxidase (Sigma, Type VI) with the periodate method.25 In some cases, double staining for two different antigens in the same section was performed as described by Nakane.26 Immunohistochemical control procedures involved staining of known positive and negative tissues and omission of the first antibody or its replacement by nonimmune serum or immune serum preabsorbed with the appropriate antigen. Substitution of the first antibody by nonimmune serum or serum preabsorbed with the appropriate antigen invariably abolished positive immunohistochemical staining.
Results The tissue types found in 170 ovarian teratomas are listed in Table 1. Other tissue types in addition to epidermis were found in 68 cases.
175
Table 1 -Tissue Types Present in 170 Cystic Ovarian Teratomas 170 Epidermis Skin appendages Brain
Respiratory epithelium Bone, cartilage, teeth Thyroid Intestinal epithelium
Salivary gland Thymus Carcinoid Mucinous cystadenoma
68 23 22 13 10 8 2 2 2 2
Table 2 shows the relation between the presence of different kinds of APUD cells and the presence of brain tissue in the ovarian teratomas. In 15 of 20 cases (75 7V) containing silver-reactive melanocytes in the epidermis, brain tissue was also present. Of the remaining 150 cases without melanocytes, brain tissue was found in only 8 (6%o). Melanocytes always showed an uneven patchy distribution (Figure 1). In 10 cases varying amounts of thyroid tissues were found. In none of these cases were argyrophil cells found. Also, immunostaining for calcitonin failed to demonstrate the presence of parafollicular (C) cells. In one case a carcinoid existed adjacent to the thyroid tissue. No brain tissue was found in any of these teratomas. In 22 cases ciliated columnar epithelium of the respiratory type was found, arranged partly in combination with acinar glands and cartilage as in larger bronchi. In 8 of these a small number of argyrophil
cells occurred (Figure 2), with coexisting brain tissue in only one case. In 8 cases intestinal type epithelium with goblet cells and brush border cells was found. In several cases small intestinal mucosal structure with crypts and/or villi was mimicked. In other cases ductlike structures or cystic spaces lined with a single layer of tall columnar epithelium occurred. In all these epithelia argyrophil cells were present (Figure 3), but brain tissue was invariably absent. In addition, a range of polypeptide-hormone-producing cells occurred in most of these neoplasms (Table 3; Figure 4). Whenever organoid differentiation was recognizable, the different APUD cell types were usually found at locations at which they occur in normal tissue (eg, argyrophil cells in crypts, secretin cells on villous surfaces, PP-cells in ductlike structures with tall columnar epithelial cells). In most cases more than one endocrine celltype occurred. Two cystic teratomas occurred together with mucinous cystadenomas. In both cases, silver-reactive and polypeptide-hormone-containing cells were found in
176
BOSMAN AND LOUWERENS
AJP * August 1981
Table 2-APUD Cells and Brain Tissue in Cystic Ovarian Teratomas With APUD cells Without APUD cells No. Brain tissue Brain tissue Brain tissue Brain tissue Tissue type cases present absent present absent Epidermis Thyroid Respiratory epithelium Intestinal epithelium *
170 10 22 8
15 0 1 0
5 0 7 8
8 0 0 0
142 10 14 0
APUD cell type Melanocyte C cell Argyrophil G, EG, S, D, PP*
G = gastrin cell; EG = entroglucagon cell; S = secretin cell; D = somatostatin cell; PP = pancreatic polypeptide cell.
the cystadenomas, but no brain tissue was found in the teratomas. The range of tissue types in testicular teratocarcinomas was much more limited. In all 6 cases intestinal-type epithelium was found, mostly arranged in tubular structures without cryptlike or villuslike formations. In four of these, argyrophil cells were found in combination with the presence of immunoreactive polypeptide hormones (Table 4, Figure 5). Neither silver-reactive cells nor immunoreactive cells were ever seen in areas of embryonal carcinoma. In none of the teratocarcinomas was mature brain tissue found.
Discussion In this study of teratomas and teratocarcinomas two different categories of APUD cells were en-
countered, namely, those of proven neuroectodermal origin (melanocytes) and those whose origin has been disputed (respiratory and gastrointestinal APUD cells). Cystic ovarian teratomas that contained melanocytes usually also contained brain tissue. The fact that brain tissue was not always found can be due to sampling error and to insensitivity of the detection method: most teratomas were not serially sectioned,
ri
'I-2'/u
$twi'~~~ 1. st
Al b
.,i.A"
Af X
>
E7 f r *4. O'
Figure
lA-Epidermis with sebagland in an ovarian teratoma. (H&E, x 100) B-Melanocytes in the same epidermis (Grimelius stain, x 100; inset, x 400) C-Cerebellar tissue in same ovarian teratoma. (H&E, x 100) ceous
, .7
s
.....
s ,iirj.l; *
.~q'0 %gt 4i'-> ;
,-..
APUD CELLS IN TERATOMAS
Vol. 104 * No. 2
177
Figure 2A-Ciliated respiratory epithelium in an ovarian teratoma. (H&E, x 100) B-Argyrophil cells in the same epithelium. (Grimelius stain, x 100; inset, x 400)
and no histochemical methods for the detection of brain tissue were used. Conversely, the small number of cases with brain tissue but without melanocytes could also be due to sampling error, even more so as in most cases melanocytes occurred in a rather patchy distribution. The finding that melanocytes and brain tissue most frequently occur together is in keeping with the neuroectodermal origin of melanocytes. In normal embryogenesis melanoblasts migrate from their origin in the neural crest and invade the epidermis. Whether in teratomas melanocytes also migrate into epidermal tissues or, alternatively, differentiate in situ cannot be determined from the present study.
In teratomas with thyroid tissue neither C cells nor brain tissue were found. This result is again in keeping with the neuroectodermal origin of C cells. C cells have been reported to occur in thyroid tissue in an ovarian teratoma in the presence of brain tissue.20 Similarly, different polypeptide-hormone-producing cell types in a pituitarylike structure were reported to occur in an ovarian teratoma containing brain tissue.18 All of these observations support our assumption that, although differentiation along a single line can be found in teratomas, different tissue types, which in normal embryogenesis are derived from the neuroectoderm, frequently occur together.
1..
Figure 3-Argyrophil cells in intestinal epithelium. (Sevier and Munger stain, counterstained with hematoxyin, x 100; inset, x 630)
*M. A,.
-
2
F.
-
178
BOSMAN AND LOUWERENS
AJP * August 1981
Table 3-Polypeptide APUD Hormones in Intestinal Epithelium GLI G SS
SECR PP*
Ovarian teratomas 1 2 3 4 5 6 7 8
+
+
+
+
+
+ +
+ +
+
+
+
+
+ + +
+
Testicular teratomas 1 2 3 4 5 6
+
+ +
+
+
+ +
+
+
+
+
* GLI = glucagonlike immunoreactivity; G = gastrin; SS = somatostatin; SECR = secretin; PP = pancreatic polypeptide.
In all ovarian teratomas with intestinal epithelium APUD cells were found. The cell types and their localization was mostly appropriate to the tissue type in which they occurred. In none of these cases was brain tissue found. Similarly, in four of the testicular teratocarcinomas APUD cells occurred in intestinal epithelium without coexisting brain tissue. APUD cells occurred somewhat less frequently in respiratory epithelium. However, argyrophil cells are very difficult to find in the adult respiratory tract,26 and therefore their
1A-Argyrophil cell (x440) Immunoreactive somatostatin-contamning cell. (x400) C-Immunoreactive gastrin containing cell. (x 400) D-lmmunoreactive secretin-containing cell. (x 630)
relative scarcity in teratomas is not surprising. The finding that in teratomas APUD cells of the intestinal and respiratory tracts frequently occur without brain tissue, in marked contrast to the frequent association of melanocytes and brain tissue described above, argues against a developmental relation between intestinal and respiratory APUD cells and neuroectodermal tissue in these neoplasms. Several observations indicate that intestinal and respiratory tract APUD cells might be of endodermal origin. Cheng and Leblondt2'27 demonstrated in mouse intestinal mucosa that all epithelial cell types, including enteroendocrine cells, probably develop from a common stem cell, the crypt-base columnar cell. Similar results were reported in colonic mucosa.'328 Pictet et all' studied the in vitro development of the pancreas and found that neither acinar nor insular tissue ever developed alone, suggesting that these also have a common origin. The occurrence of endocrine and exocrine characteristics within one cell29 provides further evidence that acinar and insular cells probably have a common origin. Finally, numerous tumors have been reported with both carcinomatous and carcinoidal differentiation,30 again suggesting that these different cell types can differentiate from the same progenitor cell. Observations concerning the development of cell and tissue types in neoplasms have to be cautiously interpreted, however, and should not without reservation be extrapolated to patterns of differentiation in normal embryogenesis. In malignant neoplasms
B
APUD CELLS IN TERATOMAS
Vol. 104 *No. 2
179
4.~~~~~~~~~~~~~~~~~~.
-~~~~ Figure 5-APUD cell types in testicu""'p.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~M :~Vt lar teratocarcinoma. A-H&E, x 100 ~ .A~ ~ ~ ~ ~ ~ ~ ~~, CB-Argyrophil cells. (xlOO0) Immunoreactive gastrin-containing 'P ~ ~ ~ ~ ~ ~~~~"h D-Immunoreactive I. (x 250) k cell. ...
0
...
~~~~~~~~~~~~~~~~~~~~el somatostatin-containing
VIT:~~~~~
(x 630)
72&MU:
the occurrence of different lines of differentiation is not restricted to combinations of carcinoma and carcinoid. Carcinosarcoma or sarcomatoid changes in carcinoma do not infrequently occur, and merely indicate the capricious character of differentiation in malignant neoplasms. However, in mature teratomas a different situation exists, because normal organoid development is found and APUD cells occur in appropriate tissues and at appropriate localizations. Against this background we feel justified to conclude that in teratomas intestinal APUD cells differentiate in situ in intestinal epithelium and respiratory APUD cells in ciliated columnar epithelium. These findings suggest that also in normal embryogenesis these APUD cells may not be of neuroectodermal origin but differentiate in situ from endodermal progenitor cells.
5.
6.
7.
8.
9. 10.
References 11.
1. Pearse AGE: The cytochemistry and ultrastructure of polypeptide hormone-producing cells of the APUD series and the embryologic, physiologic, and pathologic implications of the concept. J Histochem Cytochem 1969, 17:303-313 2. Takor Takor T, Pearse AGE: Neuroectodermal origin of avian hypothalamo-hypophyseal complex: The role of the ventral neural ridge. J Embryol Exp Morphol 1975, 34:311-325 3. Hammond WS: Early hypophysial development in the chick embryo. Am J Anat 1974, 141:303-315 4. Polak JM, Pearse AGE, Le Lievre C, Fontaine J, Le Douarin NM: Immunocytochemical confirmation of
12.
13.
14.
the neural crest origin of Avian calcitonin producing cells. Histochemistry 1974, 40:209-214 Le Douarin N, Le Lievre G: L'Origine des cellules a calcitonine du corps ultimobranchial de l'embryon d'oiseau. C R Assn Anat 1971, 56:558-568 Le Douarin N, Le Lievre C, Fontaine J: Recherches experimentales sur l'origine embryologique du corps carotidien chez les oiseaux. C R Acad Sci (D) (Paris) 1972, 275:583-586 Le Douarin N, Teillet MA: Sur quelques aspects de la migration des cellules neurales chez l'embryon de Poulet etudiee par la methode des greffes heterospecifiques de tube nerveux. C R Soc Biol (Paris) 1970, 164:390-397 Teillet MA, Le Douarin N: La migration des cellules pigmentaires etudiee par la methode des greffes heterospecifiques de tube nerveux chez l'embryon d'oiseau. C R Acad Sci (D) (Paris) 1970, 270:3095-3098 Pearse AGE, Polak JM: Neural crest origin of the endocrine polypeptide (APUD) cells of the gastrointestinal tract and pancreas. Gut 1971, 12:783-788 Pearse AGE, Polak HM, Heath CM: Development, differentiation, and derivation of the endocrine polypeptide cells of the mouse pancreas: Immunofluorescence, cytochemical, and ultrastructural studies. Diabetologia 1973, 9:120-129 Pictet RL, Rall LB, Phelps P, Rutter WJ: The neural crest and the origin of the insulin-producing and other gastrointestinal hormone-producing cells. Science 1976, 191:191-192 Cheng H, Leblond CP: Origin, differentiation, and renewal of the four main epithelial cell types in the mouse small intestine: III. Enteroendocrine cells. Am J Anat 1974, 141:503-520 Cheng WWL, Leblond CP: Renewal of the epithelium in the descending colon of the mouse: I. Presence of three cell populations: Vacuolated-columnar, mucous, and argentaffin. Am J Anat 1971, 131:73-100 Goldenberg DM, Fisher ER: Histogenetic relationship between carcinoids and mucin-secreting carcinomas of
180
15. 16.
17. 18. 19.
20.
21.
22.
23.
24. 25.
AJP * August 1981
BOSMAN AND LOUWERENS
colon as revealed by heterotransplantation. Br J Cancer 1970, 24:610-614 Sidhu GS: The endodermal origin of digestive and respiratory tract APUD-cells. Am J Pathol 1979, 96:5-20 Pearse AGE: The diffuse neuroendocrine system and the APUD concept: Related "endocrine" peptides in brain, intestine, pituitary, placenta, and anuran cutaneous glands. Med Biol 1977, 55:115-125 Fox H, Kazzaz B, Langley FA: Argyrophil and argentaffin cells in the female genital tract and in ovarian mucinous cysts. J Pathol Bacteriol 1964, 88:479-488 McKeel DW, Askin FB: Ectopic hypophyseal hormonal cells in benign cystic teratoma of the ovary. Arch Pathol Lab Med 1978, 102:122-128 Greko MA, LiVolsi VA, Pertschuk LP, Bigelow B: Strumal carcinoid of the ovary: An analysis of its components. Cancer 1979, 43:1380-1388 Dayal Y, Tashjian AH, Wolfe HJ: Immunocytochemical localization of calcitonin-producing cells in a strumal carcinoid with amyloid stroma. Cancer 1979, 43:1331-1338 Grimelius L: A silver nitrate stain for a2-cells in human pancreatic islets. Acta Soc Med Ups 1968, 73:243-270 Sevier A, Munger BC: A silver method for paraffin sections of neural tissue. J Neuropathol Exp Neurol 1965, 24:130-135 Nieuwenhuijzen Kruseman AC, Knijnenburg G, Brutel de la Riviere G, Bosman FT: Morphology and immunohistochemically-defined endocrine function of pancreatic islet cell tumors. Histopathology 1978, 2:389399 Nieuwenhuiizen Kruseman AC, Knijnenburg G: Immunohistochemical diagnosis of parafollicular carcinomas (Corres). Lancet 1977, 2:825 Wilson MB, Nakane PK: Recent developments in the periodate method of conjugating horseradish peroxidase (HRPO) to antibodies, Immunofluorescence and
26. 27.
28. 29.
30. 31.
32.
Related Staining Techniques. Edited by W Knapp, K Holubar, G Wick. Amsterdam, Elsevier, 1978, pp 215-224 Hage E: Endocrine cells in the bronchial mucosa of human fetuses. Acta Pathol Microbiol Scand [A] 1972, 80:225-234 Cheng H, Leblond CP: Origin, differentiation, and renewal of the four main epithelial cell types in the mouse small intestine: V. Unitarian theory of the origin of the four epithelial cell types. Am J Anat 1974, 141:537-562 Chang WWL, Leblond CP: Renewal of the epithelium in the descending colon of the mouse: I. Presence of three cell populations: Vacuolated-columnar, mucous, and argentaffin. Am J Anat 1971, 131:73-100 Brown RE, Schweisthal MR, Still WJS: Ultrastructural observations on the embryogenesis of islet cells in the rat pancreas in vitro. Proc Soc Exp Biol Med 1971, 136:441-445 Haggani MT, Williams G: Mucin producing carcinoid tumors of the vermiform appendix. J Clin Pathol 1977, 30:473-480 Hernandez FJ, Reid JD: Mixed carcinoid and mucussecreting intestinal tumors. Arch Pathol Lab Med 1969, 88:389-496 Kubo T, Watanabe H: Neoplastic argentaffin cells in gastric and intestinal carcinomas. Cancer 1971, 27:447-454
Acknowledgments The expert technical assistance of Georgette Cramer and of John van Bergen Henegouw is gratefully acknowledged. Drs. A. Schaberg, A. C. Nieuwenhuijzen Kruseman, and S. Dawsey critically read the manuscript. Klaas van der Ham prepared the photographs, and the manuscript was typed by Andrea Chaudron.
