J Neurosurg 49:169-178, 1978
Surface topography of normal and neoplastic human anterior pituitary cells maintained in vitro ROBERT D. HARRIS, CDR, MC, USN, EDWARDL. SELJESKOG, M.D., PH.D., KENNETH J. MURRAY, M.D., PH.D., SHELLEY N. CHOU, M.D., PH.D., WILLIAM P. CUNNINGHAM, PH.D., AND STEVEN D. DOUGLAS, M.D. Departments of Neurosurgery and Genetics and Cell Biology, and Section of Immunology, Department of Medicine, University of Minnesota Medical School, Minneapolis, Minnesota Pituitary tissues were obtained from 25 patients who underwent surgery for excision of pituitary macroadenomas, selective excision of microadenomas, or removal of a normal gland for palliation of metastatic cancer. Cells thus obtained were maintained in vitro for varying intervals, fixed, and examined by light (phase contrast), microscopy, transmission electron microscopy (TEM), and scanning electron microscopy (SEM). Previous SEM reports indicate that surface topography of in vitro neoplastic cells displays features that may correlate with neoplastic behavior. Cultured normal and pituitary tumor cells did not display these surface differences, with one exception, a Prolactin-secreting microadenoma. Characteristic patterns for the cell populations were identified. Certain cell types appeared in all the cultures: 1) large and small granule-containing cells; 2) fiat and irregular agranular cells; 3) spindle-shaped cells; and 4) spherical, irregularly surfaced cells. In one case of an endocrine-inactive juvenile pituitary chromophobe adenoma, unique cells were observed. Surface topography did not appear to be of predictive value in determining the neoplastic character of pituitary tumors. KEY WORDS 9 neoplastic characterization 9 pituitary cell culture scanning electron microscopy 9 phase contrast microscopy 9 transmission electron microscopy 9 pituitary adeuoma 9 pituitary microadenoma
p
ITUITARY adenomas constitute approximately 15% to 25% of intracranial tumors. The ability to predict neoplastic behavior of these lesions on a morphological basis remains elusive despite modern advances in electron microscopy and immunohistochemistry. Scanning electron microscopy (SEM) has made it possible to visualize intact cell surfaces at a resolution approaching 10 to 20 nm. Using SEM, several investigators have described in vitro differences in surface topographical anatomy J. Neurosurg. / Volume 49 / August, 1978
between normal and malignant cells. These reports suggest that "increased surface activity" of malignant cells is an indicator of malignant potential. 6 8,1o To determine whether neoplastic and normal pituitary cells in vitro differ in their surface topographical anatomy, we cultured pituitary tissue obtained from 25 consecutive patients at hypophysectomy or adenoma excision. We have attempted to relate our observations to those described for other tumor cells in order to assess whether SEM 169
R. D. Harris, et al. topography is of predictive value for the neoplastic potential of a pituitary tumor. Simultaneous phase contrast and transmission electron microscopic (TEM) studies were recorded to follow cell growth and aid in the identification of cell types seen in scanning photographs. Pituitary adenoma cells, in general, did not demonstrate the increased surface topographical activity reported for some tumor systems. However, there were different patterns of cell types observed between the normal and adenoma cultures. In addition, an unusual degree of surface activity occurred in one microadenoma and very unusual and distinctive cells were seen in a culture obtained from a chromophobe adenoma removed from a 14-year-old boy. This report describes the SEM morphology of representative cells derived from normal and abnormal pituitary tissue. Materials and Methods
Patients
This investigation included 25 patients operated on consecutively at the University of Minnesota Hospitals during'a period of several months. Culture Techniques
Tumor specimens and normal pituitary tissue obtained directly at surgery were diced into 1 cu mm fragments and several fragments placed in culture discs. The culture medium described by Kohler, et al.,4 was used to maintain the cultures, and daily phasecontrast microscopy was used to follow cell growth. Representative samples were fixed at 2 weeks and at variable intervals thereafter. Morphological Studies
Standard electron microscopic fixation techniques were employed using 1.5% glutaraldehyde containing 6% sucrose in 0.1% sodium cacodylate (pH 7.4). Flasks were scored under phase-contrast microscopy and discs conforming to the scanning microscope stubs were excised. The remaining portion of the flask was submitted for staining with Masson's trichrome stain in the case of light microscopy, or Epon embedding for TEM. For SEM, discs were dehydrated in graded alcohol and the alcohol replaced with freon in a graded fashion. Freon critical-point drying ]70
was then accomplished and the specimens were coated with gold and palladium. In all, 150 stubs were prepared and examined with an ETEC Autoscanner,* and 310 photographs were taken. For TEM, fixation was as described. This portion of the flask was dehydrated in graded alcohol. The monolayers were postfixed in 1% OsO4 in 0.1 M sodium cacodylate with enbloc staining using 1% uranyl acetate. A portion of the original surgical specimen was similarly fixed and prepared for correlation with the in vitro specimen. These dehydrated specimens were then embedded in Epon 812. Beem capsules, partially filled with Epon, were inverted over the scored areas and, after hardening, thin sections were prepared. These were examined in the Siemens 102 transmission electron microscope.t For standard light microscopic studies, fixation was on plastic discs and the cells were stained en bloc with a routine Masson trichrome stain. This permitted identification of intracellular granules. Results
Categorization of Tissue
The results of this study fell into five categories as determined by the patients' endocrine evaluation, operative findings, and light microscopy of the surgical specimen. These included: 1) normal pituitary tissue; 2) non-secreting chromophobe (macro-) adenomas in adults; 3) prolactin-secreting chromophobe microadenomas; 4) an adrenocorticotropic hormone (ACTH)-secreting basophilic adenoma; and 5) a non-secreting juvenile pituitary chromophobe adenoma. All specimens were observed and photographed by phase-contrast light microscopy and by SEM. Selected specimens were examined by TEM and by standard light microscopy after staining. Correlation of the TEM and SEM mlcrographs for various specimens demonstrated that two sizes of granules were present in granular cells and that these granules were lost after the cultures had been maintained for several weeks. Additionally, the granules correlated in size to those seen in *ETEC Autoscanner made by ETEC Corp., Hayward, California. tSiemens 102 transmission electron microscope made by Siemens Corp., Berlin, West Germany. J. Neurosurg. / Volume 49 / August, 1978
N o r m a l a n d n e o p l a s t i c p i t u i t a r y cells TEM micrographs of the original surgical specimens. Standard light microscopy of the stained monolayers could not be precisely correlated with the SEM studies, since the SEM discs were coated with a gold/palladium film, thereby obscuring their intracellular light-staining characteristics. The Masson-stained monolayers did, however, reveal the presence of stainable granular and agranular cells in all specimens examined. Similarly these intracytoplasmic granules were noted to disappear as the culture aged. The intent of this investigation was to study the surface characteristics of normal and abnormal pituitary cells in culture, and this was possible despite the difficulties in precise correlation of various techniques. Terminology for these scanning characteristics has been defined previously1~ and is as follows:
attach to the culture substrate or to other cells. In addition, they may extend over the surface of an underlying third cell. Short broken sections of microfibrils can become detached to lie free. We have attempted to determine whether SEM would reveal any unique characteristics of pituitary tumors that could be of possible neoplastic predictive value.
Normal Pituitary
Scanning Electron Microscopic Studies. The explant from which cells were noted to bud exhibited prominent ruffling (Fig. 1 A), and at 2 weeks three types of cells were evident. First there were large (20 to 25/~), irregularly-shaped, smooth-surfaced, relatively agranular cells (Fig. 1 B). The remainder of the cell population was comprised Blebs: A nonspecific term for the spherical equally of granule-containing cells (Fig. 1 C) and cylindrical projections from the cell and spherical, irregularly-surfaced cells. surface. These cell types were similar to those of the Filopodia: Hair or thread-like projections chromophobe adenoma discussed below. from the cell surface containing micro- Within the granular cells we observed both spikes and retraction fibrils. large (> 1/z) and small (< 0.5/~) granules. In Lamellar Cytoplasm: The leading edge of addition, occasional spindle-shaped cells were fibroblast-like cells which is flat and thin. seen. As noted previously, the granular cells Irregularities and waves of this edge give decreased in number as the culture aged, and rise to ruffles. by 4 to 6 weeks the culture monolayer had a Lamellipodia: Transistory projections from homogeneous appearance. the cell that are thin, sheet-like, and Light Microscopic Studies. Phase-contrast mobile. These occur at, or near, the end of microscopy of the cultures revealed both a fibroblast. These projections are syn- granular and agranular cells. A smooth onymous with ruffles or microridges and agranular cell is shown adjacent to a cell with contain a subplasmalemmal micro- perinuclear granules in Fig. 3 H. filament network, but not the oriented bundles seen in a microfilament sheath. Microspikes: Rigid, spike-like projections Chromophobe Adenoma (Non-Secreting) from the edges of cells from the flat subA previous report has indicated that when strates that contain microtubules and examined by phase-contrast microscopy microfilaments. These projections undergo chromophobe adenomas contain three cell a probing movement and may be seeking types? These were: 1) large (25 to 30 ~z) cells an attachment site for the cell. In contrast with a granular appearance; 2) small (15 to 20 to a microvillus, they have a wide base and u) cells with a granular appearance; and 3) inare shorter. termediate sized cells with variable granule Microvilli: Finger-like projections from any content. The SEM observations in this report portion of the cell surface that are greater also described two groups of cells: large than 0.1 u in diameter. These projections irregular cells with peripheral granules and have a uniform cross-sectional diameter small cells with a smooth surface. and a subplasmalemmal microfilament Scanning Electron Microscopic Studies. array. Our studies confirmed these findings and as Retraction Microfibrils: Long, thin strands of noted previously, granular cells with large material that are drawn out from the and small granules were evident (Fig. 1 F), as retracting edge of a cell. These strands may well as especially prominent small rounded J. Neurosurg. / Volume 49 / August, 1978
]7]
R. D. Harris, et al.
FIG. 1. Upper:Normal pituitary cells in vitro at 2 weeks. A: Surface detail of normal pituitary explant demonstrating prominent ruffling, X 4000. B: Smooth-surfaced pituitary cell, X 1000. Note its infrequent granules and irregular shape. C: Granular type of normal pituitary cell demonstrating large (> 1 u) and small (< 0.5 u) granules, • 4000. This type of cell is similar to the granular type of cell noted in the chromophobe adenoma (F). Center."Chromophobe adenoma in vitro at 2 weeks. D: Small, rounded adenoma cell with a smooth but irregular surface ("mulberry" cell), x 3000. E: Large pitted adenoma cell, • 1000. F: Granular type of adenoma cell, X 4000. This cell is similar to the granular cell of the normal pituitary (C). Lower." Chromophobe adenoma in vitro at 2 weeks. G: Irregular, somewhat smooth-surfaced adenoma cell. The cell is flattened and demonstrates filopodia, • 1000. H: Intermediate size adenoma cell with an irregular smooth surface and filopodia, X 1000. I: Small, round adenoma cell with surface disruption and granules corresponding in size to surface alterations, x 8500.
172
J. Neurosurg. / Volume 49 / August, 1978
N o r m a l a n d n e o p l a s t i c p i t u i t a r y cells smooth-surfaced "mulberry-shaped" cells microvilli near the site of cellular budding (Fig. 1 D). In addition to these cells we have (Fig. 2 A and B). This unusual observation observed by SEM three additional groups of was made only once in a culture 3 weeks old. cells. These three "new" types of cells include Its significance at this time is not known, but one group displaying a pitted surface (Fig. 1 may represent increased metabolic activity E). This type of cell has been previously associated with cellular division. described in an endocrine active tumor. 9 AdLight Microscopic Studies. Phase-contrast ditionally, there were two populations of cells microscopy demonstrated rounded cells condisplaying prominent basilar microspikes. taining granules, as well as spindle-shaped One group had a smooth surface and cells. irregular shape (Fig. 1 G); the other was of an intermediate variety, rounded, and Basophilic Adenoma (A CTH-Secreting) demonstrating an irregular but generally This specimen was obtained from a patient smooth-surface topography (Fig. 1 H). One with Nelson's syndrome. A previous report of cell photographed had an obvious surfacecells associated with Nelson's syndrome was membrane rupture and small spherical from an ACTH-secreting chromophobe granules were present extracellularly. The adenoma? The predominant cell in this granules correspond in size to the pit left in report (as studied by SEM) was large, fiat, the surface of the small cell (Fig. 1 I). As irregular, granular, and pitted, resembling the the culture aged, steadily increasing numbers type that we observed in the non-secreting of smooth-surfaced, flat, irregular cells chromophobe adenoma (Fig. 1 E). appeared. At 4 to 6 weeks these were the Scanning Electron Microscopic Studies. predominant cells. The predominant cells were small, spherical, Light Microscopic Studies. Our phase- smooth, and irregularly surfaced, containing contrast findings agreed with prior obser- few granules (Fig. 2 F). Additional cells vations. 9 Standard light microscopy of a resembled the spindle-shaped cells and culture at 2 weeks demonstrated the expected smooth-surfaced flat cells described above for intracytoplasmic granules (Fig. 3 I). This corchromophobe adenomas (Figs. 1 E and 2 D). related well with the phase-contrast, SEM, Light Microscopic Studies. Phase-contrast and TEM observations. microscopic observations were indisTransmission Electron Microscopic tinguishable from standard chromophobe Studies. Figure 3 D and E represents non- adenoma. There was insufficient culture secreting chromophobe adenomas after 2 material to stain for standard light weeks in culture; the granules are still present. microscopy, but hematoxylin and eosin After a series of four passages over 4 months, (H & E) staining of the surgical specimen these chromophobe adenoma cultures displayed vacuolization (Fig. 3 F) and loss of demonstrated the expected basophilic adenoma. granules (Fig. 3 G), where the cells now resemble fibroblasts. Juvenile Pituitary Chromophobe Adenoma (Non-Secreting) Chromophobe Microadenoma (ProlactinSecreting) This chromophobe tumor was removed
Scanning Electron Microscopic Studies. At 2 weeks, the microadenoma cultures demonstrated three distinct cell types. First, there were small round cells with an irregularly smooth surface containing areas of granules (Fig. 2 C). Second, there were large spindle-shaped cells containing surface granules and polar basilar microspikes (Fig. 2 D). Finally, there were groups of cells similar to the first group, but much larger in size without granules and displaying polar filopodia (Fig. 2 E). One of the microadenoma explants revealed a profusion of J. Neurosurg. / Volume 49 / August, 1978
from a 14-year-old boy. The cells displayed a markedly different morphology from other cells in this study. Scanning Electron Microscopic Studies. After 2 weeks of culture there were especially prominent large microexplants with polar basilar microspikes (Fig. 2 G). Large stellate cells were also abundant, with a relatively smooth surface and eccentrically placed nucleus with many filopodia (Fig. 2 H). There were also flat cells, regularly shaped with central nuclei, resembling a "fried egg" (Fig. 3 B). These cells were either agranular (Fig. 3 ]73
R. D. Harris, et al.
FIc. 2. Upper." Microadenoma explant and cell types in vitro at 3 weeks. A: Profusion of microvilli on explant surface near the budding cell, X 900. This may be a reflection of increased metabolic activity of the cell. B: High power of microvilli noted in A, X 5000. C: Small, rounded "mulberry-shaped" cell with surface granules, X 1500. This cell is similar to that observed in the non-secreting adenoma (Fig. 1 D). Center."Microadenoma and basophilic adenoma in vitro at 2 weeks. D: Large granular, spindleshaped cell with filopodia. Microadenoma culture, • 250. E: Large, rounded, smooth-surfaced microadenoma cell with filopodia, X 650. F: Small, round, irregularly-surfaced cell from a patient with Nelson's syndrome: a basophilic adenoma, X 2500. An SEM of this tumor revealed that this cell is similar in appearance to those small, round "mulberry" cells observed in the microadenoma (C) and the chromophobe non-secreting adenoma (Fig. 1 D). Phase contrast observations in this case were identical to those of the chromophobe adenoma. Lower." Juvenile pituitary adenoma in vitro at 2 weeks. G: Microexplant of a "juvenile'pituitary adenoma, X 85. H: Large (50 #) stellate cell, with prominent filopodia, X 150. I: Surface detail of a granular type flat cell showing large and small granules, X 4500. This cell is similar to that observed in the normal pituitary gland (Fig. 1 C), and the chromophobe adenoma (Fig. l F).
174
J. Neurosurg. / Volume 49 / August, 1978
Normal and neoplastic pituitary cells
FIG. 3. Upper." Juvenile pituitary adenoma SEM and phase contrast in vitro. A: Small, rounded "mulberry" cell at 2 weeks, X 3500. This cell is similar to those noted in the adult chromophobe adenoma (Fig. 1 D), as well as the microadenoma (Fig. 2 C), and the basophilic adenoma (Fig. 2 F). B: Flat, regularly-shaped agranular cell with a central nucleus ("fried egg"), X 1000. Granular cells of similar shape were also noted (Fig. 2 I). C: Phase contrast photograph of microexplant attaching to the culture flask (3 days), X 360. Center: Transmission electron microscopy of adult and juvenile pituitary adenomas. D: Surgical specimen of a juvenile chromophobe adenoma demonstrating large and small granules, X 10,000. E: A juvenile pituitary chromophobe adenoma in vitro at 2 weeks showing the presence of granules, x 25,000. F: An adult chromophobe adenoma at 4 months after four serial passages revealing vacuolization, x 10,000. Lower." G: Transmission electron microscopy of monolayer cells from the same culture as Fig. 3 F. Granules have disappeared and the cells display characteristics of fibroblasts, X 12,500. H: Phase contrast of normal pituitary cells in vitro at 2 weeks. An agranular cell is shown adjacent to a cell with perinuclear granules, X 1000. I: Light micrograph of a chromophobe adenoma 2 weeks old. Intracytoplasmic granules are prominent in the cells in the upper right and on the mid-left side. Masson's trichrome stain, X 500. J. Neurosurg. / Volume 49 / August, 1978
175
R. D. Harris, et al. B) or contained granules (Fig. 2 I). Finally, we observed small cells with an irregularly smooth surface and spherical configuration (Fig. 3 A). Light Microscopic Studies. By phasecontrast studies microexplants were visible as early as 3 days (Fig. 3 C). The distinct character of the various cells seen by SEM was not appreciated by phase-contrast studies. Transmission Electron Microscopic Studies. Observations by TEM were identical
to those in the general group~of non-secreting chromophobe adenomas.
Discussion
Clinically, two broad categories of pituitary tumors have been recognized. First are those pituitary adenomas that are slow-growing and, regardless of endocrine activity, are noninvasive and lacking in metastatic capabilities. These lesions are often cured by complete resection. This group may encompass microadenomas, although their natural history is by no means certain at this time. The second category of pituitary tumor is a more aggressive lesion; it may display invasiveness and frequently recurs despite wide excision. Occasionally, pituitary carcinoma may occur from which there are rare reports of metastasis? Using traditional histological techniques it has largely been impossible to differentiate between the two clinical groups of tumors. This morphological inability to predict neoplastic behavior poses a continuing problem in the postoperative management of the patient. This study was performed to determine whether SEM characterization of normal and tumor cell surfaces would be of value in predicting the neoplastic behavior of these tumors. Abnormal surface properties of malignant cells were described in 1954 by Coman and Anderson) They postulated that absence of normal adhesive sites on neoplastic cells accounted, in part, for their independent invasive activity. The advent of the SEM permitted remarkable visualization of the intact surfaces of cells in a three-dimensional image. Scanning studies of cell-surface anatomy must utilize in vitro methods to allow visualization of maximal surface area. Solidtissue-fragment scanning does not permit the 176
study of all cell surfaces, since cells are adherent to each other. Furthermore, cells in the interior of the fragment will not be seen; only fragment-surface cells can be identified. Unfortunately, in vitro methods introduce an artificial environment for these cells; however, by maintaining constant culture conditions (such as, media, pH, and temperature) certain differences in other tumor systems have been detected between normal and neoplastic cells. Porter and Fonte 7 described in vitro cellsurface topography for several cultured transplantable mouse and rat tumor cell lines. Surface morphology was influenced by pH variation and media composition; however, the extent was not determined. Using identical culture conditions, they demonstrated that tumor cell surfaces displayed a greater surface activity than normal cells. This was evidenced by increased numbers of microvilli, surface blebs, ruffles, and other lamellipodia. In tumor cells, these changes persisted throughout the growth cycle. Vesely and Boyde1~ also examined surfaces of cultured normal and tumor cells in rats; they suggested that differences in the cell-surface activity may account for the ability to invade, metastasize, and generate abnormal tissues, noted with persistent growth in vitro. They were careful to point out that any study must distinguish between "tumorgenicity" (that is, neoplastic transformation) and cellular adaptation for survival in vitro. Thus, any report describing morphological abnormalities in vitro must always be interpreted carefully, as some changes may simply reflect the artificial environment. Human cells have also been studied by SEM and distinctive morphological features are noted. 8'8 Murray, et al., 6 described the surface architecture of human malignant gliomas and cerebellar astrocytomas, and confirmed the presence of "increased surface activity" observed for other malignant cells. Cell-surface alterations have generally been considered to be related to pinocytosis (blebs) and increased surface absorptive area (microvilli and ruffles). That neoplastic and dividing cells are more metabolically active may account for the increased number of blebs, ruffles, and microvilli. Wetzel, in a cautionary note, stressed that while surface morphology of a cell may reflect its type, J. Neurosurg. / Volume 49 / August, 1978
Normal and neoplastic pituitary cells neoplastic state, and the stage in the cell cycle, these characteristics cannot be determined by SEM alone. 1~ The culture environment, especially the type of substrate and media, the concentration of serum, and the presence or absence of various metabolized enzymes and drugs have all been reported to alter cell-surface activity. T M Our study demonstrates new morphological patterns observed in cultures of normal and neoplastic pituitary cells. We wished to describe the surface topographical features of these cells under identical conditions .and determine whether the topographical anatomy varied between the two groups as reported in other in vitro studies of neoplasmsY 8,10 The results demonstrate that pituitary cells exhibit a varied appearance in culture. Summarizing the findings, certain consistent patterns were discerned in the various classes of the neoplasms and representative micrographs are presented that illustrate these findings. Cell types appearing in all cultures included: 1) cells containing large and small granules; 2) fiat, irregular, agranular cells; 3) spindle-shaped cells; and 4) spherical, irregularly-surfaced cells. The spindle-shaped cells and the fiat, irregularagranular cells were prominent at 2 weeks, and by 6 weeks a monolayer was usually present consisting of the latter fiat, irregular, agranular cells containing a homogeneous cytoplasm. With continued in vitro maintenance, the homogeneity of the cell population increased in all cultures, and distinguishing characteristics were lost by 6 weeks. This most likely represents an overgrowth of mesenchymal cells in the culture. In the juvenile pituitary adenoma, striking SEM differences were noted in comparison to other cell cultures. This culture exhibited large microexplants, large stellate cells, and both granular and agranular cells resembling "fried eggs." The only expression of increased surface activity was exhibited in one prolactin-secreting chromophobe microadenoma explant. This may only represent increased surface activity associated with cellular budding. Light microscopy confirmed the presence of intracytoplasmic granules in cultures, while transmission electron microscopy revealed two sizes of granules in these cells, both in the surgical specimens and in the cultures. This correlated well with our SEM observation. J. Neurosurg. / Volume 49 / August, 1978
Conclusion
This report describes in vitro morphological features of normal and neoplastic human pituitary cells. We were unable to demonstrate prominent alterations of cellsurface activity that have been previously noted between malignant and benign cells. The sole exception was found in one microadenoma. Scanning electron microscopy does reveal several distinctive cell-type patterns, which are more conspicuous than surface-activity differences. Thus far it does not offer a direct method for neoplastic characterization of pituitary tumors. Future studies, which include media modifications, new culture techniques, TEM, and light microscopic correlation and immunohistochemical surface labeling markers, may further the characterization of normal and neoplastic pituitary cells. References
1. Baechler CA, Chen S, Bergner R, et al: Importance of the environment to surface features of cells in situ and in culture, in Scanning Electron Microscopy/1975 Part I. Chicago: liT Research Institute, 1975, pp 393-402 2. Coman DR, Anderson TF: A structural difference between the surfaces of normal and of carcinomatous epidermal cells. Cancer Res 15:541-543, 1955 3. Hallowes RC, Mistry D: Use of the scanning electron microscope to characterize cell types grown in cultures of human epithelial tissues, in Scanning Electron Microscopy/1976 Part V. Chicago: IIT Research Institute, 1976, pp 121-126 4. Kohler PO, Bridson WE, Rayford PL, et al: Hormone production by human pituitary adenomas in culture. Metabolism 18: 782-788, 1969 5. Landor AM: Ultrastructure of human sella tumors. Correlations of clinical findings and morphology. Aeta Neurochir (Suppl 22):1-167, 1975 6. Murray KJ, Chou SN, Douglas SD: Topographical anatomy of human cerebral and cerebellar astrocytomas in vitro. Surg Neurol 6:337-340, 1976 7. Porter KR, Fonte VG: Observations on the topography of normal and cancer cells, in Scanning Electron Microscopy/1973 Part IIl. Chicago: IIT Research Institute, 1973, pp 683-688 8. Spring-Mills E, Elias J J: Cell surface changes associated with human breast cancer, in Scanning Electron Microscopy/1976 Part V. 177
lq. D. Harris, et al.
9. 10.
11.
12.
Chicago: liT Research Institute, 1976, pp 1-10 Teraoka A: Morphological and functional properties of pituitary adenomas in culture. Nenrol Med Chir (Tokyo) 12:52-63, 1972 Vesely P, Boyde A: The significance of SEM evaluation of the cell surface for tumor cell biology, in Scanning Electron Microscopy/1973 Part III. Chicago: IIT Research Institute, 1973, pp 689-696 Westbrook E, Wetzel B: Impact of culture conditions on the surface morphology of cells in vitro, in Scanning Electron Microscopy/1975 Part I. Chicago: IIT Research Institute, 1975, pp 351-360 Wetzel B: Cell kinesies, an interpretative review of the significance of cell surface form,
in Scanning Electron Microscopy/1976 Part V. Chicago: IIT Research Institute, 1976, pp 135-144
178
Dr. Harris is a U. S. Navy-sponsored postdoctoral fellow in Neurosurgery at the University of Minnesota. The opinions expressed herein are those of the authors and do not necessarily reflect the opinions of the Department of Defense, the Bureau of Medicine and Surgery of the Department of the Navy, or those of the Sea Service at large. This research was supported by grants from the Minnesota Medical Foundation, MMF-HL-7377, from the National Institutes of Health, AI 12478, and from the Kroc Foundation. Address reprint requests to: Robert D. Harris, M.D., Department of Neurosurgery, Box 96, Mayo, University of Minnesota Hospitals, 420 Delaware Street, S. E., Minneapolis, Minnesota 55455.
J. Neurosurg. / Volume 49 / August, 1978
Surface topography of normal and neoplastic human anterior pituitary cells maintained in vitro ROBERT D. HARRIS, CDR, MC, USN, EDWARDL. SELJESKOG, M.D., PH.D., KENNETH J. MURRAY, M.D., PH.D., SHELLEY N. CHOU, M.D., PH.D., WILLIAM P. CUNNINGHAM, PH.D., AND STEVEN D. DOUGLAS, M.D. Departments of Neurosurgery and Genetics and Cell Biology, and Section of Immunology, Department of Medicine, University of Minnesota Medical School, Minneapolis, Minnesota Pituitary tissues were obtained from 25 patients who underwent surgery for excision of pituitary macroadenomas, selective excision of microadenomas, or removal of a normal gland for palliation of metastatic cancer. Cells thus obtained were maintained in vitro for varying intervals, fixed, and examined by light (phase contrast), microscopy, transmission electron microscopy (TEM), and scanning electron microscopy (SEM). Previous SEM reports indicate that surface topography of in vitro neoplastic cells displays features that may correlate with neoplastic behavior. Cultured normal and pituitary tumor cells did not display these surface differences, with one exception, a Prolactin-secreting microadenoma. Characteristic patterns for the cell populations were identified. Certain cell types appeared in all the cultures: 1) large and small granule-containing cells; 2) fiat and irregular agranular cells; 3) spindle-shaped cells; and 4) spherical, irregularly surfaced cells. In one case of an endocrine-inactive juvenile pituitary chromophobe adenoma, unique cells were observed. Surface topography did not appear to be of predictive value in determining the neoplastic character of pituitary tumors. KEY WORDS 9 neoplastic characterization 9 pituitary cell culture scanning electron microscopy 9 phase contrast microscopy 9 transmission electron microscopy 9 pituitary adeuoma 9 pituitary microadenoma
p
ITUITARY adenomas constitute approximately 15% to 25% of intracranial tumors. The ability to predict neoplastic behavior of these lesions on a morphological basis remains elusive despite modern advances in electron microscopy and immunohistochemistry. Scanning electron microscopy (SEM) has made it possible to visualize intact cell surfaces at a resolution approaching 10 to 20 nm. Using SEM, several investigators have described in vitro differences in surface topographical anatomy J. Neurosurg. / Volume 49 / August, 1978
between normal and malignant cells. These reports suggest that "increased surface activity" of malignant cells is an indicator of malignant potential. 6 8,1o To determine whether neoplastic and normal pituitary cells in vitro differ in their surface topographical anatomy, we cultured pituitary tissue obtained from 25 consecutive patients at hypophysectomy or adenoma excision. We have attempted to relate our observations to those described for other tumor cells in order to assess whether SEM 169
R. D. Harris, et al. topography is of predictive value for the neoplastic potential of a pituitary tumor. Simultaneous phase contrast and transmission electron microscopic (TEM) studies were recorded to follow cell growth and aid in the identification of cell types seen in scanning photographs. Pituitary adenoma cells, in general, did not demonstrate the increased surface topographical activity reported for some tumor systems. However, there were different patterns of cell types observed between the normal and adenoma cultures. In addition, an unusual degree of surface activity occurred in one microadenoma and very unusual and distinctive cells were seen in a culture obtained from a chromophobe adenoma removed from a 14-year-old boy. This report describes the SEM morphology of representative cells derived from normal and abnormal pituitary tissue. Materials and Methods
Patients
This investigation included 25 patients operated on consecutively at the University of Minnesota Hospitals during'a period of several months. Culture Techniques
Tumor specimens and normal pituitary tissue obtained directly at surgery were diced into 1 cu mm fragments and several fragments placed in culture discs. The culture medium described by Kohler, et al.,4 was used to maintain the cultures, and daily phasecontrast microscopy was used to follow cell growth. Representative samples were fixed at 2 weeks and at variable intervals thereafter. Morphological Studies
Standard electron microscopic fixation techniques were employed using 1.5% glutaraldehyde containing 6% sucrose in 0.1% sodium cacodylate (pH 7.4). Flasks were scored under phase-contrast microscopy and discs conforming to the scanning microscope stubs were excised. The remaining portion of the flask was submitted for staining with Masson's trichrome stain in the case of light microscopy, or Epon embedding for TEM. For SEM, discs were dehydrated in graded alcohol and the alcohol replaced with freon in a graded fashion. Freon critical-point drying ]70
was then accomplished and the specimens were coated with gold and palladium. In all, 150 stubs were prepared and examined with an ETEC Autoscanner,* and 310 photographs were taken. For TEM, fixation was as described. This portion of the flask was dehydrated in graded alcohol. The monolayers were postfixed in 1% OsO4 in 0.1 M sodium cacodylate with enbloc staining using 1% uranyl acetate. A portion of the original surgical specimen was similarly fixed and prepared for correlation with the in vitro specimen. These dehydrated specimens were then embedded in Epon 812. Beem capsules, partially filled with Epon, were inverted over the scored areas and, after hardening, thin sections were prepared. These were examined in the Siemens 102 transmission electron microscope.t For standard light microscopic studies, fixation was on plastic discs and the cells were stained en bloc with a routine Masson trichrome stain. This permitted identification of intracellular granules. Results
Categorization of Tissue
The results of this study fell into five categories as determined by the patients' endocrine evaluation, operative findings, and light microscopy of the surgical specimen. These included: 1) normal pituitary tissue; 2) non-secreting chromophobe (macro-) adenomas in adults; 3) prolactin-secreting chromophobe microadenomas; 4) an adrenocorticotropic hormone (ACTH)-secreting basophilic adenoma; and 5) a non-secreting juvenile pituitary chromophobe adenoma. All specimens were observed and photographed by phase-contrast light microscopy and by SEM. Selected specimens were examined by TEM and by standard light microscopy after staining. Correlation of the TEM and SEM mlcrographs for various specimens demonstrated that two sizes of granules were present in granular cells and that these granules were lost after the cultures had been maintained for several weeks. Additionally, the granules correlated in size to those seen in *ETEC Autoscanner made by ETEC Corp., Hayward, California. tSiemens 102 transmission electron microscope made by Siemens Corp., Berlin, West Germany. J. Neurosurg. / Volume 49 / August, 1978
N o r m a l a n d n e o p l a s t i c p i t u i t a r y cells TEM micrographs of the original surgical specimens. Standard light microscopy of the stained monolayers could not be precisely correlated with the SEM studies, since the SEM discs were coated with a gold/palladium film, thereby obscuring their intracellular light-staining characteristics. The Masson-stained monolayers did, however, reveal the presence of stainable granular and agranular cells in all specimens examined. Similarly these intracytoplasmic granules were noted to disappear as the culture aged. The intent of this investigation was to study the surface characteristics of normal and abnormal pituitary cells in culture, and this was possible despite the difficulties in precise correlation of various techniques. Terminology for these scanning characteristics has been defined previously1~ and is as follows:
attach to the culture substrate or to other cells. In addition, they may extend over the surface of an underlying third cell. Short broken sections of microfibrils can become detached to lie free. We have attempted to determine whether SEM would reveal any unique characteristics of pituitary tumors that could be of possible neoplastic predictive value.
Normal Pituitary
Scanning Electron Microscopic Studies. The explant from which cells were noted to bud exhibited prominent ruffling (Fig. 1 A), and at 2 weeks three types of cells were evident. First there were large (20 to 25/~), irregularly-shaped, smooth-surfaced, relatively agranular cells (Fig. 1 B). The remainder of the cell population was comprised Blebs: A nonspecific term for the spherical equally of granule-containing cells (Fig. 1 C) and cylindrical projections from the cell and spherical, irregularly-surfaced cells. surface. These cell types were similar to those of the Filopodia: Hair or thread-like projections chromophobe adenoma discussed below. from the cell surface containing micro- Within the granular cells we observed both spikes and retraction fibrils. large (> 1/z) and small (< 0.5/~) granules. In Lamellar Cytoplasm: The leading edge of addition, occasional spindle-shaped cells were fibroblast-like cells which is flat and thin. seen. As noted previously, the granular cells Irregularities and waves of this edge give decreased in number as the culture aged, and rise to ruffles. by 4 to 6 weeks the culture monolayer had a Lamellipodia: Transistory projections from homogeneous appearance. the cell that are thin, sheet-like, and Light Microscopic Studies. Phase-contrast mobile. These occur at, or near, the end of microscopy of the cultures revealed both a fibroblast. These projections are syn- granular and agranular cells. A smooth onymous with ruffles or microridges and agranular cell is shown adjacent to a cell with contain a subplasmalemmal micro- perinuclear granules in Fig. 3 H. filament network, but not the oriented bundles seen in a microfilament sheath. Microspikes: Rigid, spike-like projections Chromophobe Adenoma (Non-Secreting) from the edges of cells from the flat subA previous report has indicated that when strates that contain microtubules and examined by phase-contrast microscopy microfilaments. These projections undergo chromophobe adenomas contain three cell a probing movement and may be seeking types? These were: 1) large (25 to 30 ~z) cells an attachment site for the cell. In contrast with a granular appearance; 2) small (15 to 20 to a microvillus, they have a wide base and u) cells with a granular appearance; and 3) inare shorter. termediate sized cells with variable granule Microvilli: Finger-like projections from any content. The SEM observations in this report portion of the cell surface that are greater also described two groups of cells: large than 0.1 u in diameter. These projections irregular cells with peripheral granules and have a uniform cross-sectional diameter small cells with a smooth surface. and a subplasmalemmal microfilament Scanning Electron Microscopic Studies. array. Our studies confirmed these findings and as Retraction Microfibrils: Long, thin strands of noted previously, granular cells with large material that are drawn out from the and small granules were evident (Fig. 1 F), as retracting edge of a cell. These strands may well as especially prominent small rounded J. Neurosurg. / Volume 49 / August, 1978
]7]
R. D. Harris, et al.
FIG. 1. Upper:Normal pituitary cells in vitro at 2 weeks. A: Surface detail of normal pituitary explant demonstrating prominent ruffling, X 4000. B: Smooth-surfaced pituitary cell, X 1000. Note its infrequent granules and irregular shape. C: Granular type of normal pituitary cell demonstrating large (> 1 u) and small (< 0.5 u) granules, • 4000. This type of cell is similar to the granular type of cell noted in the chromophobe adenoma (F). Center."Chromophobe adenoma in vitro at 2 weeks. D: Small, rounded adenoma cell with a smooth but irregular surface ("mulberry" cell), x 3000. E: Large pitted adenoma cell, • 1000. F: Granular type of adenoma cell, X 4000. This cell is similar to the granular cell of the normal pituitary (C). Lower." Chromophobe adenoma in vitro at 2 weeks. G: Irregular, somewhat smooth-surfaced adenoma cell. The cell is flattened and demonstrates filopodia, • 1000. H: Intermediate size adenoma cell with an irregular smooth surface and filopodia, X 1000. I: Small, round adenoma cell with surface disruption and granules corresponding in size to surface alterations, x 8500.
172
J. Neurosurg. / Volume 49 / August, 1978
N o r m a l a n d n e o p l a s t i c p i t u i t a r y cells smooth-surfaced "mulberry-shaped" cells microvilli near the site of cellular budding (Fig. 1 D). In addition to these cells we have (Fig. 2 A and B). This unusual observation observed by SEM three additional groups of was made only once in a culture 3 weeks old. cells. These three "new" types of cells include Its significance at this time is not known, but one group displaying a pitted surface (Fig. 1 may represent increased metabolic activity E). This type of cell has been previously associated with cellular division. described in an endocrine active tumor. 9 AdLight Microscopic Studies. Phase-contrast ditionally, there were two populations of cells microscopy demonstrated rounded cells condisplaying prominent basilar microspikes. taining granules, as well as spindle-shaped One group had a smooth surface and cells. irregular shape (Fig. 1 G); the other was of an intermediate variety, rounded, and Basophilic Adenoma (A CTH-Secreting) demonstrating an irregular but generally This specimen was obtained from a patient smooth-surface topography (Fig. 1 H). One with Nelson's syndrome. A previous report of cell photographed had an obvious surfacecells associated with Nelson's syndrome was membrane rupture and small spherical from an ACTH-secreting chromophobe granules were present extracellularly. The adenoma? The predominant cell in this granules correspond in size to the pit left in report (as studied by SEM) was large, fiat, the surface of the small cell (Fig. 1 I). As irregular, granular, and pitted, resembling the the culture aged, steadily increasing numbers type that we observed in the non-secreting of smooth-surfaced, flat, irregular cells chromophobe adenoma (Fig. 1 E). appeared. At 4 to 6 weeks these were the Scanning Electron Microscopic Studies. predominant cells. The predominant cells were small, spherical, Light Microscopic Studies. Our phase- smooth, and irregularly surfaced, containing contrast findings agreed with prior obser- few granules (Fig. 2 F). Additional cells vations. 9 Standard light microscopy of a resembled the spindle-shaped cells and culture at 2 weeks demonstrated the expected smooth-surfaced flat cells described above for intracytoplasmic granules (Fig. 3 I). This corchromophobe adenomas (Figs. 1 E and 2 D). related well with the phase-contrast, SEM, Light Microscopic Studies. Phase-contrast and TEM observations. microscopic observations were indisTransmission Electron Microscopic tinguishable from standard chromophobe Studies. Figure 3 D and E represents non- adenoma. There was insufficient culture secreting chromophobe adenomas after 2 material to stain for standard light weeks in culture; the granules are still present. microscopy, but hematoxylin and eosin After a series of four passages over 4 months, (H & E) staining of the surgical specimen these chromophobe adenoma cultures displayed vacuolization (Fig. 3 F) and loss of demonstrated the expected basophilic adenoma. granules (Fig. 3 G), where the cells now resemble fibroblasts. Juvenile Pituitary Chromophobe Adenoma (Non-Secreting) Chromophobe Microadenoma (ProlactinSecreting) This chromophobe tumor was removed
Scanning Electron Microscopic Studies. At 2 weeks, the microadenoma cultures demonstrated three distinct cell types. First, there were small round cells with an irregularly smooth surface containing areas of granules (Fig. 2 C). Second, there were large spindle-shaped cells containing surface granules and polar basilar microspikes (Fig. 2 D). Finally, there were groups of cells similar to the first group, but much larger in size without granules and displaying polar filopodia (Fig. 2 E). One of the microadenoma explants revealed a profusion of J. Neurosurg. / Volume 49 / August, 1978
from a 14-year-old boy. The cells displayed a markedly different morphology from other cells in this study. Scanning Electron Microscopic Studies. After 2 weeks of culture there were especially prominent large microexplants with polar basilar microspikes (Fig. 2 G). Large stellate cells were also abundant, with a relatively smooth surface and eccentrically placed nucleus with many filopodia (Fig. 2 H). There were also flat cells, regularly shaped with central nuclei, resembling a "fried egg" (Fig. 3 B). These cells were either agranular (Fig. 3 ]73
R. D. Harris, et al.
FIc. 2. Upper." Microadenoma explant and cell types in vitro at 3 weeks. A: Profusion of microvilli on explant surface near the budding cell, X 900. This may be a reflection of increased metabolic activity of the cell. B: High power of microvilli noted in A, X 5000. C: Small, rounded "mulberry-shaped" cell with surface granules, X 1500. This cell is similar to that observed in the non-secreting adenoma (Fig. 1 D). Center."Microadenoma and basophilic adenoma in vitro at 2 weeks. D: Large granular, spindleshaped cell with filopodia. Microadenoma culture, • 250. E: Large, rounded, smooth-surfaced microadenoma cell with filopodia, X 650. F: Small, round, irregularly-surfaced cell from a patient with Nelson's syndrome: a basophilic adenoma, X 2500. An SEM of this tumor revealed that this cell is similar in appearance to those small, round "mulberry" cells observed in the microadenoma (C) and the chromophobe non-secreting adenoma (Fig. 1 D). Phase contrast observations in this case were identical to those of the chromophobe adenoma. Lower." Juvenile pituitary adenoma in vitro at 2 weeks. G: Microexplant of a "juvenile'pituitary adenoma, X 85. H: Large (50 #) stellate cell, with prominent filopodia, X 150. I: Surface detail of a granular type flat cell showing large and small granules, X 4500. This cell is similar to that observed in the normal pituitary gland (Fig. 1 C), and the chromophobe adenoma (Fig. l F).
174
J. Neurosurg. / Volume 49 / August, 1978
Normal and neoplastic pituitary cells
FIG. 3. Upper." Juvenile pituitary adenoma SEM and phase contrast in vitro. A: Small, rounded "mulberry" cell at 2 weeks, X 3500. This cell is similar to those noted in the adult chromophobe adenoma (Fig. 1 D), as well as the microadenoma (Fig. 2 C), and the basophilic adenoma (Fig. 2 F). B: Flat, regularly-shaped agranular cell with a central nucleus ("fried egg"), X 1000. Granular cells of similar shape were also noted (Fig. 2 I). C: Phase contrast photograph of microexplant attaching to the culture flask (3 days), X 360. Center: Transmission electron microscopy of adult and juvenile pituitary adenomas. D: Surgical specimen of a juvenile chromophobe adenoma demonstrating large and small granules, X 10,000. E: A juvenile pituitary chromophobe adenoma in vitro at 2 weeks showing the presence of granules, x 25,000. F: An adult chromophobe adenoma at 4 months after four serial passages revealing vacuolization, x 10,000. Lower." G: Transmission electron microscopy of monolayer cells from the same culture as Fig. 3 F. Granules have disappeared and the cells display characteristics of fibroblasts, X 12,500. H: Phase contrast of normal pituitary cells in vitro at 2 weeks. An agranular cell is shown adjacent to a cell with perinuclear granules, X 1000. I: Light micrograph of a chromophobe adenoma 2 weeks old. Intracytoplasmic granules are prominent in the cells in the upper right and on the mid-left side. Masson's trichrome stain, X 500. J. Neurosurg. / Volume 49 / August, 1978
175
R. D. Harris, et al. B) or contained granules (Fig. 2 I). Finally, we observed small cells with an irregularly smooth surface and spherical configuration (Fig. 3 A). Light Microscopic Studies. By phasecontrast studies microexplants were visible as early as 3 days (Fig. 3 C). The distinct character of the various cells seen by SEM was not appreciated by phase-contrast studies. Transmission Electron Microscopic Studies. Observations by TEM were identical
to those in the general group~of non-secreting chromophobe adenomas.
Discussion
Clinically, two broad categories of pituitary tumors have been recognized. First are those pituitary adenomas that are slow-growing and, regardless of endocrine activity, are noninvasive and lacking in metastatic capabilities. These lesions are often cured by complete resection. This group may encompass microadenomas, although their natural history is by no means certain at this time. The second category of pituitary tumor is a more aggressive lesion; it may display invasiveness and frequently recurs despite wide excision. Occasionally, pituitary carcinoma may occur from which there are rare reports of metastasis? Using traditional histological techniques it has largely been impossible to differentiate between the two clinical groups of tumors. This morphological inability to predict neoplastic behavior poses a continuing problem in the postoperative management of the patient. This study was performed to determine whether SEM characterization of normal and tumor cell surfaces would be of value in predicting the neoplastic behavior of these tumors. Abnormal surface properties of malignant cells were described in 1954 by Coman and Anderson) They postulated that absence of normal adhesive sites on neoplastic cells accounted, in part, for their independent invasive activity. The advent of the SEM permitted remarkable visualization of the intact surfaces of cells in a three-dimensional image. Scanning studies of cell-surface anatomy must utilize in vitro methods to allow visualization of maximal surface area. Solidtissue-fragment scanning does not permit the 176
study of all cell surfaces, since cells are adherent to each other. Furthermore, cells in the interior of the fragment will not be seen; only fragment-surface cells can be identified. Unfortunately, in vitro methods introduce an artificial environment for these cells; however, by maintaining constant culture conditions (such as, media, pH, and temperature) certain differences in other tumor systems have been detected between normal and neoplastic cells. Porter and Fonte 7 described in vitro cellsurface topography for several cultured transplantable mouse and rat tumor cell lines. Surface morphology was influenced by pH variation and media composition; however, the extent was not determined. Using identical culture conditions, they demonstrated that tumor cell surfaces displayed a greater surface activity than normal cells. This was evidenced by increased numbers of microvilli, surface blebs, ruffles, and other lamellipodia. In tumor cells, these changes persisted throughout the growth cycle. Vesely and Boyde1~ also examined surfaces of cultured normal and tumor cells in rats; they suggested that differences in the cell-surface activity may account for the ability to invade, metastasize, and generate abnormal tissues, noted with persistent growth in vitro. They were careful to point out that any study must distinguish between "tumorgenicity" (that is, neoplastic transformation) and cellular adaptation for survival in vitro. Thus, any report describing morphological abnormalities in vitro must always be interpreted carefully, as some changes may simply reflect the artificial environment. Human cells have also been studied by SEM and distinctive morphological features are noted. 8'8 Murray, et al., 6 described the surface architecture of human malignant gliomas and cerebellar astrocytomas, and confirmed the presence of "increased surface activity" observed for other malignant cells. Cell-surface alterations have generally been considered to be related to pinocytosis (blebs) and increased surface absorptive area (microvilli and ruffles). That neoplastic and dividing cells are more metabolically active may account for the increased number of blebs, ruffles, and microvilli. Wetzel, in a cautionary note, stressed that while surface morphology of a cell may reflect its type, J. Neurosurg. / Volume 49 / August, 1978
Normal and neoplastic pituitary cells neoplastic state, and the stage in the cell cycle, these characteristics cannot be determined by SEM alone. 1~ The culture environment, especially the type of substrate and media, the concentration of serum, and the presence or absence of various metabolized enzymes and drugs have all been reported to alter cell-surface activity. T M Our study demonstrates new morphological patterns observed in cultures of normal and neoplastic pituitary cells. We wished to describe the surface topographical features of these cells under identical conditions .and determine whether the topographical anatomy varied between the two groups as reported in other in vitro studies of neoplasmsY 8,10 The results demonstrate that pituitary cells exhibit a varied appearance in culture. Summarizing the findings, certain consistent patterns were discerned in the various classes of the neoplasms and representative micrographs are presented that illustrate these findings. Cell types appearing in all cultures included: 1) cells containing large and small granules; 2) fiat, irregular, agranular cells; 3) spindle-shaped cells; and 4) spherical, irregularly-surfaced cells. The spindle-shaped cells and the fiat, irregularagranular cells were prominent at 2 weeks, and by 6 weeks a monolayer was usually present consisting of the latter fiat, irregular, agranular cells containing a homogeneous cytoplasm. With continued in vitro maintenance, the homogeneity of the cell population increased in all cultures, and distinguishing characteristics were lost by 6 weeks. This most likely represents an overgrowth of mesenchymal cells in the culture. In the juvenile pituitary adenoma, striking SEM differences were noted in comparison to other cell cultures. This culture exhibited large microexplants, large stellate cells, and both granular and agranular cells resembling "fried eggs." The only expression of increased surface activity was exhibited in one prolactin-secreting chromophobe microadenoma explant. This may only represent increased surface activity associated with cellular budding. Light microscopy confirmed the presence of intracytoplasmic granules in cultures, while transmission electron microscopy revealed two sizes of granules in these cells, both in the surgical specimens and in the cultures. This correlated well with our SEM observation. J. Neurosurg. / Volume 49 / August, 1978
Conclusion
This report describes in vitro morphological features of normal and neoplastic human pituitary cells. We were unable to demonstrate prominent alterations of cellsurface activity that have been previously noted between malignant and benign cells. The sole exception was found in one microadenoma. Scanning electron microscopy does reveal several distinctive cell-type patterns, which are more conspicuous than surface-activity differences. Thus far it does not offer a direct method for neoplastic characterization of pituitary tumors. Future studies, which include media modifications, new culture techniques, TEM, and light microscopic correlation and immunohistochemical surface labeling markers, may further the characterization of normal and neoplastic pituitary cells. References
1. Baechler CA, Chen S, Bergner R, et al: Importance of the environment to surface features of cells in situ and in culture, in Scanning Electron Microscopy/1975 Part I. Chicago: liT Research Institute, 1975, pp 393-402 2. Coman DR, Anderson TF: A structural difference between the surfaces of normal and of carcinomatous epidermal cells. Cancer Res 15:541-543, 1955 3. Hallowes RC, Mistry D: Use of the scanning electron microscope to characterize cell types grown in cultures of human epithelial tissues, in Scanning Electron Microscopy/1976 Part V. Chicago: IIT Research Institute, 1976, pp 121-126 4. Kohler PO, Bridson WE, Rayford PL, et al: Hormone production by human pituitary adenomas in culture. Metabolism 18: 782-788, 1969 5. Landor AM: Ultrastructure of human sella tumors. Correlations of clinical findings and morphology. Aeta Neurochir (Suppl 22):1-167, 1975 6. Murray KJ, Chou SN, Douglas SD: Topographical anatomy of human cerebral and cerebellar astrocytomas in vitro. Surg Neurol 6:337-340, 1976 7. Porter KR, Fonte VG: Observations on the topography of normal and cancer cells, in Scanning Electron Microscopy/1973 Part IIl. Chicago: IIT Research Institute, 1973, pp 683-688 8. Spring-Mills E, Elias J J: Cell surface changes associated with human breast cancer, in Scanning Electron Microscopy/1976 Part V. 177
lq. D. Harris, et al.
9. 10.
11.
12.
Chicago: liT Research Institute, 1976, pp 1-10 Teraoka A: Morphological and functional properties of pituitary adenomas in culture. Nenrol Med Chir (Tokyo) 12:52-63, 1972 Vesely P, Boyde A: The significance of SEM evaluation of the cell surface for tumor cell biology, in Scanning Electron Microscopy/1973 Part III. Chicago: IIT Research Institute, 1973, pp 689-696 Westbrook E, Wetzel B: Impact of culture conditions on the surface morphology of cells in vitro, in Scanning Electron Microscopy/1975 Part I. Chicago: IIT Research Institute, 1975, pp 351-360 Wetzel B: Cell kinesies, an interpretative review of the significance of cell surface form,
in Scanning Electron Microscopy/1976 Part V. Chicago: IIT Research Institute, 1976, pp 135-144
178
Dr. Harris is a U. S. Navy-sponsored postdoctoral fellow in Neurosurgery at the University of Minnesota. The opinions expressed herein are those of the authors and do not necessarily reflect the opinions of the Department of Defense, the Bureau of Medicine and Surgery of the Department of the Navy, or those of the Sea Service at large. This research was supported by grants from the Minnesota Medical Foundation, MMF-HL-7377, from the National Institutes of Health, AI 12478, and from the Kroc Foundation. Address reprint requests to: Robert D. Harris, M.D., Department of Neurosurgery, Box 96, Mayo, University of Minnesota Hospitals, 420 Delaware Street, S. E., Minneapolis, Minnesota 55455.
J. Neurosurg. / Volume 49 / August, 1978
