ULTRASTRUCTURE, STEROIDOGENIC POTENTIAL, AND ENERGY METABOLISM OF THE SNELL ADRENOCORTICAL CARCINOMA 494
A Comparison with Normal Adrenocortical Tissue
G. L. KIMMEL, F. G. PÉRON, A. HAKSAR, E. BEDIGIAN,
W. F. ROBIDOUX, JR., and M. T. LIN
From the Worcester Foundation for Experimental Biology, Shrewsbury, Massachusetts 01545. Dr. Kimmel’s present address is the Research Triangle Institute, Research Triangle Park, North Carolina 27709.
ABSTRACT
Electron microscope studies were carried out with the adrenocortical carcinoma 494 and normal adrenal cortex tissue. The mitochondria of the tumor cells showed marked differences when compared with mitochondria from fasciculata cells of the normal adrenal cortex. These differences were primarily related to mitochondrial number and crista structure.
Corticosterone production in isolated tumor cells was extremely low and neither ACTH nor dibutyryl cyclic AMP had any stimulatory effect. Normal adrenal cells showed at least a tenfold increase under identical conditions. In the presence of corticosteroid precursors the amount of corticosterone produced by the tumor cells was much less than that produced by normal cells.
The results indicate a reduced capacity for 116-hydroxylation in the tumor mitochondria and a possible reduced capacity for biosynthetic steps before the 118-hydroxylation reaction. Glycolysis in isolated tumor cells was also lower than in normal cells.
Isolated tumor mitochondria oxidized succinate normally with a good degree of coupling with phosphorylation. However, unlike normal adrenal mitochondria, the tumor mitochondria showed little or no oxygen uptake with other Krebs cycle substrates. These data suggest that the tumor mitochondria may be lacking in the flavoprotein dehydrogenases responsible for the oxidation of NADH and NADPH, although other components of the respiratory chain may be intact.
INTRODUCTION
The adrenocortical carcinoma 494, originally dis- determine its steroid biosynthetic potential and covered in the rat by Snell and Stewart (1), has compare it with that in the normal adrenal. Ney et been studied by several laboratories (2-7) to al. (3) noted that the tumor was able to utilize
exogenous substrates for corticosteroid produc- tion. However, the rate of synthesis was greatly reduced (2, 3). This was especially true of the 118-hydroxylation of 11-deoxycorticosterone (DOC)1 to corticosterone (2).
The effect of adrenocorticotropin (ACTH) and adenosine-3’, 5’-monophosphate (CAMP) on the tumor has also been studied. Although ACTH was unable to stimulate steroidogenesis in this tissue (3), Schorr et al. (4, 6) demonstrated that the trophic hormone was capable of stimulating the tumor adenyl cyclase to produce fairly large amounts of CAMP. The only abnormality ap- peared to be in the lack of specificity for ACTH. Unlike the adenyl cyclase in the normal adrenal cortex, which is stimulated only by ACTH, the tumor adenyl cyclase was stimulated by a number of other trophic hormones (e.g., thyroid-stimulat- ing hormone, luteinizing, and follicle-stimulating hormone). From these results and the finding that large concentrations of exogenous CAMP failed to stimulate corticosteroidogenesis in vitro (3), it was concluded that reduced steroidogenesis in the tumor was due primarily to a defect(s) beyond the formation of CAMP (4).
Corticosteroidogenesis in the normal adrenal gland is a process dependent on several steps. Fractionation studies showed some of these to occur in the mitochondrial fraction and others to occur in the microsomal fraction (8, 9). Therefore, it is possible that the reduced rate of steroidogene- sis in the tumor may be a result of a defect(s) in one or more of the cell organelles. The present study was designed to correlate the ultrastructure of the rat adrenocortical carcinoma 494 with its corticosteroidogenic potential and to make an overall comparison with the normal rat adrenal cortex. The findings of an alteration in the tumor mitochondrial structure and the biosynthetic steps associated with it give evidence that the abnormal steroidogenesis of the tumor can be largely attrib- uted to abnormal mitochondrial function. In addi- tion, our results demonstrate a lowered capacity for specific mitochondrial oxidations in the tumor which suggest a possible lack of certain enzymes in the respiratory chain.
’ The generic names of substances for which trivial names have been given in the paper are: Pregnenolone, 38- hydroxy-5-pregnene-20-one; progesterone, 4-preg- nene-3, 20-dione; DOC, 11-deoxycorticosterone, 21- hydroxy-4-pregnene-3, 20-dione; corticosterone, 118, 21- dihydroxy-4-pregnene-3, 20-dione; 118-OH-progester- one, 118-hydroxy-4-pregnene-3, 20-dione.
MATERIALS AND METHODS Animals
The rat adrenocortical carcinoma 494 was trans- planted and maintained in a manner similar to that described by Ney et al. (3). Young male Sprague-Dawley rats (Charles River Breeding Laboratories, Inc., Wil- mington, Mass.), weighing 50-70 g, received viable tu- mor tissue subcutaneously through a small incision near the base of the tail. The tumor tissue for all studies was taken 4-6 wk after the initial transplant. The atrophic adrenals from the tumor-bearing animal were also re- moved at this time. Normal male rats of the same age were used as a source for normal adrenal tissue. Adre- nal tissue was also taken for electron microscope exami- nation from tumor-bearing and control rats which had received long-term ACTH treatment. Treatment con- sisted to three daily subcutaneous injections (morning, noon, and evening) of ACTH (Acthar gel, 40 USP U/ ml, Armour Pharmaceutical Co., Chicago, Il1.), for a 2-wk period.
Preparation of Tissue for Electron Microscopy
Tissues for ultrastructural studies were fixed in 5% glutaraldehyde (pH 7.2) for 2 h at room temperature. This was followed by postfixation in 2% OsO. (pH 7.2) for 1.5 h. Both fixatives were made in 0.1 M PO, buffer containing CaCl2. Samples were then dehydrated in a graded series of aqueous ethanol solutions and embedded in Epon 812 by the method of Luft (10). Ultrathin sections were cut on a Porter-Blum MT2-B ultrami- crotome (Ivan Sorvall, Inc., Newtown, Conn.), and stained with uranyl acetate and lead citrate. Sections were viewed with a Zeiss EM9S-2 electron microscope.
Cell Preparation and Incubations
Cells were isolated from the normal adrenal as previously described by this laboratory (11, 12). Isolated tumor cells were prepared both with and without the use of proteolytic enzymes. Tumor cells were isolated using trypsin and collagenase following a procedure identical to that for normal adrenal cell isolation. For tumor cells isolated without enzymes, viable tumor tissue was placed in a 50-ml erlenmeyer flask containing 20 ml Krebs- Ringer bicarbonate buffer plus glucose (KRBG) (11, 12). The flask was gassed with 95% O2-5% CO2, and the tissue stirred for 15-20 min at 37°℃ using a magnetic stirrer. This was repeated six to eight times, with fresh KRBG each time. The first two fractions were discarded due to a large amount of contamination with red blood cells. The remaining fractions were then centrifuged and concen- trated according to the procedure for normal adrenal cell isolation. All isolated cells demonstrated an ultrastruc- ture identical to that of their respective intact tissue. Similar results have been reported by Sharma et al. (13)
for normal adrenal cells and by Sharma and Hashimoto (5) for trypsinized adrenal tumor cells. There were no apparent structural differences between cells isolated with enzymes and those isolated in buffer alone.
All experiments were carried out using tumor cells isolated without enzymes, except in the case where cells isolated with and without enzymes were compared. Incubations of the isolated cells were always carried out in KRBG containing 5 mg/ml BSA. Lima bean trypsin inhibitor (1 mg/ml) was also added when cells were isolated with trypsin and collagenase. Incubations were carried out in 10-ml beakers in a final volume of 1.5 ml with shaking in a Dubnoff metabolic incubator (60 rpm) at 37℃ for 2 h under an atmosphere of 95% O2-5% CO2.
ACTH used in the incubation studies was USP corticotropin reference standard distributed by United States Pharmacopeial Convention, Inc. Dibutyryl cy- clic-3’, 5’-AMP (dbCAMP) and NADPH were obtained from Sigma Chemical Co., St. Louis, Mo .; trypsin and lima bean trypsin inhibitor from Worthington Biochemi- cal Corp., Freehold, N. J .; collagenase from Gallard- Schlesinger Chemical Mfg. Corp., Carle Place, N. Y .: and bovine serum albumin (BSA) fraction V powder, Pentex, from Miles Laboratories Inc. Kankakee, Ill. All other chemicals were of reagent grade purity obtained from various sources.
Steroid Measurement
Corticosterone was measured essentially by the method of Silber et al. (14). Incubation sample aliquots and corticosterone standards were dissolved in 13% EtOH and extracted with 4 ml of methylene dichloride. After removal of the ethanolic phase, 3 ml of 80% aqueous H2SO, were added to the methylene dichloride extract. After shaking, the fluorescence of the acid phase was measured on an Aminco-Bowman spectrophotofluo- rometer (American Instrument Co., Travenol Laborato- ries, Inc., Silver Spring, Md.), 50 min after addition of the acid. Appropriate blank values were subtracted for any interfering fluorescence contributed by pregnenolone or 118-OH progesterone.
Pooled incubation samples were also measured for DOC. After the addition of radiolabeled DOC (<0.001 ug, specific activity, 40 Ci/mmol, New England Nuclear, Boston, Mass.) as a recovery standard, the samples were extracted thoroughly with methylene dichloride. After evaporation of most of the organic solvent under a stream of N2, the concentrated extracts were run on paper chromatograms in the Bush B3 system (15). The A4-3-ketosteroids on the dried chromatograms were visualized under UV lights and tentative identification was made by comparison with known standards. A narrow strip from each chromatogram was immersed in blue tetrazolium reagent (16) for the detection of steroids containing «-ketol-reducing side chains; i.e., corticoster- one and DOC. If a blue spot corresponding to DOC was observed, the remaining area on the chromatogram was
eluted with ethanol and DOC measured quantitatively by the blue tetrazolium assay (16).
Oxygen Uptake Studies
Mitochondria from the tumor tissue and normal adrenal were isolated by a procedure described previ- ously (17). The one exception was that only four to eight passes of a loose-fitting pestle of the Potter-Elvehjem homogenizer were made for homogenization of the tissue. Although not usually done for preparing mito- chondria from normal adrenals (17), this was necessary to avoid excessive breakage of the apparently fragile tumor mitochondria. Electron microscope examination revealed isolated fractions rich in intact mitochondria in both normal and tumor preparations. However, there was a degree of breakdown of the internal matrix and breaking of the outer membrane in some of the tumor mitochondria. Oxygen utilization of the incubated mito- chondria was measured polarographically using a vibrat- ing microplatinum electrode in conjunction with an Oxygraph, model KM (Gilson Medical Electronics, Inc., Middleton, Wis.) as previously described (18, 19).
Pyruvate and Lactate Measurements
Incubations of normal and tumor cells were also carried out for pyruvate and lactate production. The effect of ACTH and dbCAMP were investigated, and arsenite was used as a blocking agent for pyruvate-lactate utilization. Pyruvate and lactate were measured by following the oxidation-reduction of NADH/NAD+ at 340 nm with lactate dehydrogenase as previously de- scribed (12, 20).
RESULTS
Ultrastructure
NORMAL ADRENAL TISSUE: The zona fas- ciculata of the normal adrenal gland showed ultrastructural features (Fig. 1) identical to those reported by others (21-24). The mitochondria were numerous and closely packed within the cell. They were circular or oval in shape and contained packed vesicular cristae. Lipid droplets were abun- dant and distributed irregularly in almost all cells. An agranular endoplasmic reticulum and free ribosomes were also apparent throughout the cytoplasm.
TUMOR TISSUE: The ultrastructure of the adrenocortical carcinoma tissue is shown in Fig. 2. The most striking intracellular change from that of the normal adrenal involves the mitochondria. The number of mitochondria per cell in the tumor was noticeably less than that of normal fasciculata cells (25% that of mitochondria from normal rat adre-
DB
M
SER
L
1
nal, based on milligrams of protein/original tissue wet weight)2. Many of the mitochondria were elongated and generally showed a much less uni- form shape. However, the most striking change was the predominance of sparse, lamella-shaped cristae. Many of the tumor cells contained round to oval vesicles, each possessing an electron-lucid matrix. In many cases these vesicles were so large and numerous that they occupied almost all of the cytoplasmic space. Occasionally, the membrane of one or more of these vesicles was seen to be continuous with the outer nuclear membrane. The endoplasmic reticulum of the tumor cells was dilated and often granular, and free ribosomes, often occurring in clusters, were quite apparent.
ADRENAL TISSUE FROM TUMOR-BEARING ANIMAL: The zona fasciculata of the atrophied adrenal gland from the tumor-bearing animal showed marked ultrastructural differences from both the normal adrenal and the adrenocortical tumor (Fig. 3). The cells appeared smaller, and the nuclear to cytoplasmic ratio appeared to increase. The nuclei often presented an irregular appear-
ance, in contrast to the generally oval nuclei of the tumor and normal adrenal cells.
The mitochondria of adrenal cells from tumor- bearing rats were less numerous and of less uniform shape than those found in normal fas- ciculata cells. Mitochondria from the atrophied adrenals had lost their characteristic vesicular cristae and developed lamellar cristae similar to those in tumor mitochondria. The mitochondrial matrix was more dense and few intramitochondrial inclusions were seen. Further, many mitochondria appeared to be associated with the formation of membranous whorls within the atrophied adrenal cells. The smooth endoplasmic reticulum of the atrophied adrenals was less developed than in normal fasciculata cells. Lipid droplets were pres- ent, and lysosome-like dense granules were numer- ous. The observations are not unlike those found in adrenocortical tissue from hypophysectomized an- imals (24).
LONG-TERM ACTH TREATMENT: The in vitro results of Ney et al. (3) on corticosterone production by the adrenal from the tumor-bearing
*
V
M
2
rat demonstrated that the decreased production of steroid by this adrenal returned to normal levels after pretreatment with ACTH. To study any ultrastructural changes which might occur after ACTH administration, we injected both normal and tumor-bearing rats with ACTH over a period of 2 wk. At the end of the treatment period the adrenals of both groups were examined.
The changes in ultrastructure of the normal adrenal observed between the control (Fig. 1) and ACTH-treated (Fig. 4) rats in this study are in accordance with previous findings (25-27). These include, most noticeably, an increase in the num- ber of mitochondria, lipid droplets, and smooth endoplasmic reticulum. In addition, the normal adrenal glands after ACTH treatment had in- creased to about five times their normal size.
The change in the adrenal from the tumor-bear- ing animal was striking after ACTH (Fig. 5). The cells appeared to regain their normal size and nuclear to cytoplasmic ratio. The nuclei presented a normal round to oval appearance. The mitochon- dria were of a normal shape and size, and they filled the cytoplasm. The mitochondrial cristae had
regained their vesicular character. The lipid drop- lets became noticeably larger, equal to those in the normal adrenal after ACTH treatment. The adre- nals had become greatly enlarged, the wet weight rising from approximately 8 mg/pair before ACTH to about 100 mg/pair after the 2-wk period of injections.
Steroidogenesis
We found that the adrenocortical tumor can be dissociated into isolated cells without the use of proteolytic enzymes. This is a property not found in most other tissues, and it permitted a compari- son of the steroidogenic capacity of cells isolated with and without proteolytic enzymes. When incu- bated with various steroid precursors, there was little or no difference in the steroidogenesis of the cells regardless of the method used for isolation, indicating that the use of trypsin and other proteo- lytic enzymes is unnecessary in the case of this adrenal tumor.
Table I shows the corticosterone production by isolated cells from the normal adrenal and the
DB
Wh
3
adrenal tumor. As noted in Materials and Methods, normal cells were isolated using trypsin- collagenase, the tumor cells in buffer alone. The normal adrenal cells synthesize only negligible amounts of corticosterone in control incubations. This is in agreement with our previously published data (20). Addition of ACTH or dbCAMP pro- duced at least a 10-fold increase in the levels of corticosterone. The addition of corticosterone pre- cursors, pregnenolone, DOC, and 118-OH-pro- gesterone, also led to a marked increase in corti- costerone production. Progesterone, however, seemed to be utilized less readily than other pre- cursors.
Corticosterone production by tumor cells was extremely low when compared to that by normal cells. Neither ACTH nor dbCAMP had any stimulating effect on the tumor cells. Among the corticosterone precursors, 116-OH-progesterone was utilized most efficiently followed by DOC and pregnenolone. Progesterone was again the least efficient precursor. It is clear from Table I that in the tumor cells, corticosterone levels, after a 2-h
incubation in the presence of DOC, progesterone, and pregnenolone, were only about 1.5-2.5% of those in normal cells, whereas with 118-OH-pro- gesterone, they were as much as 13%.
The results described above suggested that a major defect in the steroidogenic pathway in the tumor was at the 118-hydroxylation step which takes place in the mitochondria and is the last step in the conversion of DOC to corticosterone. If this were the case, DOC might be expected to accumu- late when the tumor cells were incubated with pregnenolone or progesterone. Table II shows the qualitative results after chromatographic analysis of the incubation media after incubation with these steroids and other substances. Normal cells did not produce detectable levels of corticosterone under control conditions, but did so under all other conditions tested. DOC, however, was only quali- tatively detectable in chromatograms of samples to which the steroid precursors, pregnenolone and progesterone, had been added. The tumor cells also showed detectable levels of DOC in chromato- graphed samples containing the steroid precursors,
1
4
but did not demonstrate any qualitatively detecta- ble corticosterone production under any of the other conditions tested. The eluted DOC regions from the chromatograms with detectable levels of DOC were quantitatively assayed by the blue-tet- razolium procedure. These values coupled with corticosterone levels (quantitative) provided a DOC to corticosterone ratio (Table II). The ratios for the tumor cells were 15-25 times higher than those for the normal cells, indicating a significant build up of DOC in the tumor cells. It must be mentioned that actual levels of DOC were lower in the tumor cell incubations than in those of the normal cell, indicating a decreased production of DOC or its increased metabolism by a pathway not leading to corticosterone. At present, however, we have been unable to detect any other 44 unconjugated 3-ketosteroid besides DOC which might be a metabolic product of the tumor cell incubations.
OXYGEN UPTAKE STUDIES: Isolated mito- chondria were studied for their ability to utilize oxygen and produce corticosterone. As shown in Fig. 6 A, the oxygraph trace obtained with adrenal
mitochondria from normal, nonstressed rats (nor- mal mitochondria) was as expected when succinate was the oxidizable substrate (18, 28). A P/O ratio of about 1.9 was obtained after the addition of 200 nmol of ADP, whereas 10.5 nmol of corticosterone were formed per minute per milligram mitochon- drial protein after all exogenous ADP had been utilized. Addition of 2, 4-dinitrophenol (2,4-DNP) led to a large increase in oxygen uptake, and as is usual, to an inhibition of corticosterone production brought about by the uncoupling effect of this substance (17). The picture presented by tumor mitochondria incubated with succinate was similar in terms of O2 consumption as well as to the degree of coupling of oxidation with phosphorylation, as indicated by a P/O ratio of about two (Fig. 6 B). On the other hand, it is clear that after the addition of DOC, little 118-hydroxylation of this steroid took place. What little corticosterone synthesis occurred with DOC was completely abolished by the addition of 2,4-DNP. Although not shown in the figures, addition of the respiratory chain inhibitors Antimycin A and KCN in similar exper- iments inhibited succinate-supported O2 uptake in
O
L
M
5
| Additions | Normal | Tumor |
|---|---|---|
| Control | 1.01 ± 0.17 (8)* | 0.07 ± 0.01 (8) |
| ACTH 800 _U/ml | 11.99 ± 1.20 (8) | 0.07 ± 0.01 (8) |
| ACTH 1,600 uU/ml | 13.56 ± 0.79 (8) | 0.08 ± 0.01 (8) |
| dbCamp 500 uM | 13.98 ± 0.45 (8) | 0.06 ± 0.01 (8) |
| Pregnenolone 80 µg/ml | 36.64 ± 1.88 (8) | 0.50 ± 0.05 (8) |
| Progesterone 80 µg/ml | 7.14 ±0.62 (8) | 0.17 ± 0.01 (8) |
| DOC 80 µg/ml | 37.17 ± 1.55 (8) | 0.61 ± 0.07 (8) |
| 118-OH-Progesterone 20 µg/ml | 37.08 ± 1.86 (7) | 4.79 ± 0.18 (8) |
* Mean ± standard error (N).
Either 1.5-2.3 × 105 (normal) or 1.0-1.5 x 106 (tumor) cells per beaker were incubated 2 h at 37°℃. Results are expressed as micrograms corticosterone/106 cells.
both the control and tumor mitochondria. Corti- costerone production was also completely inhib- ited by these two substances in the control and tumor mitochondria.
Other experiments carried out with the Krebs- cycle acids «-ketoglutarate, malate, and isocitrate showed expected O2 uptake rates and substrate- supported steroid 118-hydroxylation of DOC in normal mitochondria (18, 29). On the other hand,
little or no O2 uptake or corticosterone production was observed with these substrates when tumor mitochondria were incubated, regardless of whether Pi or ADP were added singly or together to the incubation medium. As will be discussed later, these observations could indicate a relative paucity in the tumor of some flavoprotein enzymes in both the respiratory chain and the cytochrome P450 chain involved with steroid hydroxylations.
| Normal | Tumor | |||||
|---|---|---|---|---|---|---|
| B* | DOC | DOC: B | B* | DOC | DOC: B | |
| Control | - | - | - | - | ||
| ACTH 1600 gU/ml | + | - | - | - | ||
| dbCamp 500 AM | + | - | - | - | ||
| Pregnenolone 80 µg/ml | + | + | 0.4 | - | + | 10.7 |
| Progesterone 80 µg/ml | + | + | 2.5 | - | + | 37.0 |
Steroids separated by paper chromatography and analyzed with blue tetrazolium.
* B, Corticosterone.
+, visible spot; - , no visible spot after treatment with the blue tetrazolium reagent.
A
P2
B
02 =100
Succinate
02=100
-P2
O2=9.7
Succinate
02=22.5
O2=6.1
+ADP
O2 =39.9
ADP
O2=43.8
”
H +DOC
O2 =52.1
O2 =22.5
4
B=10.5
+DOC
O2=42.4
O2 =29.5
B=0.5
2,4-DNP
2,4 -DNP
B=3.4
O2 = 41.9
B=0
O2-97.4-
O2=0
2.5
5.0
7.5
10.0
0 =0
2.5
5.0
7.5
10.0
12.5
Time (min)
Time (min)
PYRUVATE LACTATE ACCUMULATION: Table III compares the glycolytic response of tumor and normal adrenal cells to ACTH and dbCAMP. The doses of the two stimulating agents used are known to be maximally stimulating for normal adrenal cells (12, 20). Under these condi- tions the increases in pyruvate and lactate accumu-
lation due to both ACTH and dbCAMP were much smaller in tumor cells than in normal cells. This could not have been due to increased metabo- lism of pyruvate in the mitochondria by the enzyme pyruvate dehydrogenase, because in the presence of arsenite (which inhibits pyruvate dehy- drogenase) the increases in pyruvate and lactate in
| Additions | Normal | Tumor | ||
|---|---|---|---|---|
| Pyruvate | Lactate | Pyruvate | Lactate | |
| None | 52.1 ± 6.4 | 392.2 ± 18.9 | 37.7 ± 1.3 | 154.7 ± 10.7 |
| ACTH 2 mU/ml | 152.6 ± 18.4 | 536.1 ± 56.2 | 54.9 ± 1.7 | 172.0 ± 5.0 |
| (100.5) | (143.9) | (19.2) | (17.3) | |
| dbCamp 500 uM | 126.6 ± 12.0 | 534.8 ± 35.4 | 44.7 ± 1.3 | 179.3 ± 9.9 |
| (74.5) | (142.6) | (9.0) | (24.6) | |
| Arsenite 2.0 mM | 220.9 ± 13.6 (168.8) | 595.4 + 11.2 (202.2) | 88.2 ± 1.8 (52.5) | 234.2 ± 5.3 (79.5) |
| ACTH 2 mU/ml + | 229.6 ± 9.6 | 703.6 ± 17.2 | 83.1 ± 3.3 | 267.9 ± 11.2 |
| Arsenite 2 mM | (177.5) | (311.4) | (47.4) | (113.2) |
| dbCAMP 0.5 mM + | 248.2 ± 14.2 | 620.5 ± 7.6 | 86.7 ± 2.9 | 252.4 ± 17.6 |
| Arsenite 2 mM | (196.1) | (228.3) | (51.0) | (97.7) |
Each value expressed as nM/106 cells/2 h incubation time, is the mean ± SE (N = 4). Values in parenthesis denote the net increase over the incubated control. The incubations were carried out for 2 h at 37℃ under 95% O2-5% CO2, and pyruvate and lactate were measured on the total incubate; i.e., cells + medium.
the tumor cells were again much lower than those in the normal adrenal cells.
DISCUSSION
In the normal adrenal gland ACTH stimulates membrane adenyl cyclase which promotes the conversion of ATP to CAMP. This cyclic nucleo- tide, in turn, inceases the synthesis of corticoster- one by stimulating the conversion of cholesterol to pregnenolone in the following sequence: choles- terol _ pregnenolone 2, progesterone 3, DOC 4, corticosterone. Reactions nos. 1 and 4 are exclu- sively associated with the mitochondria in adrenal cells, whereas reactions nos. 2 and 3 occur in the cell cytosol and microsomes, respectively (8, 9). This sequence is severely limited in the adrenocor- tical carcinoma 494 (1-5), apparently at a step(s) beyond the formation of CAMP (4). It is also limited in the adrenal from the tumor-bearing animal (3).
Ney et al. (3) showed that the in vitro corticos- teroidogenesis of the adrenal from the tumor-bear- ing animal was stimulated by ACTH treatment. In the present study long-term ACTH treatment restored a normal ultrastructure to this adrenal. These results suggest that the limited steroidogenic capacity of this gland is directly related to reduced circulating levels of ACTH. Such a reduction could be in response to a negative feedback of high circulating levels of corticosterone as reported by Ney et al. (3) and also observed in our laboratory. Although the tumor on a per cell basis produces
very little corticosterone, apparently the great mass of the tumor tissue can produce relatively high circulating levels of corticosterone.
In the tumor the mitochondrial ultrastructure was markedly altered and the number of mito- chondria per cell in the tumorous tissue was seen to be considerably lower than that normally found in adrenal fasciculata cells. In addition, large elec- tron-lucid vesicles were quite often present within the cell. While the mitochondrial origin of these was not determined, the vesicles share some of the characteristics of the aminoglutethamide-induced abnormal mitochondria reported by Racela et al. (30).
In light of the altered crista structure and reduced corticosteroidogenesis of the tumor, it is possible that the vesicular conformation of rat adrenocortical mitochondrial cristae in some way enhances the ability of mitochondria to perform the associated steroidogenic conversion. If this is the case, then the lamellar cristae would not provide the optimum conditions for cholesterol side-chain cleavage and the 118-hydroxylation reactions. In the tumor, our results and those of others (2, 31) showed that this latter enzymatic conversion was severely limited. However, Sharma and Brush (31) reported that the cells of this adrenocortical tumor are capable of converting cholesterol to DOC, indicating that tumor mito- chondria possess the capacity to convert choles- terol to pregnenolone. If this is indeed the case, then it would certainly appear that the structural
alterations in the tumor mitochondria can be associated only with events leading to the 118- hydroxylation of DOC to corticosterone, and not with the conversion of cholesterol to pregnenolone. Further studies on this question would provide a better understanding of tumor and mitochondrial function.
The fact that in the tumor 21-hydroxylation of 118-OH-progesterone to corticosterone was al- tered to a much lesser extent than 118-hydroxyla- tion of DOC to corticosterone gives additional evidence that the major defect in the adrenocorti- cal carcinoma is mitochondrial in origin because the 21-hydroxylase enzyme is associated with the microsomal component of the fasciculata cells (8). The lack of corticosterone production from pro- gesterone via 118-OH-progesterone can be ex- plained by the fact that the conversion of proges- terone to 118-OH-progesterone is dependent on the mitochondrial 110-hydroxylase system (32).
The oxygraph studies also showed the tumor mitochondria to be grossly abnormal in terms of O2 utilization as a result of oxidation of Krebs- cycle substrates like isocitrate, malate, and @-keto- glutarate. The finding that succinate was effi- ciently oxidized and indeed led to efficient coupling of oxidation to phosphorylation, makes it highly probable that these mitochondria have a full complement of succinate dehydrogenase enzyme, coenzyme(s) Q, and cytochromes required for O2 utilization and ATP production. The finding that the other above-mentioned Krebs-cycle substrates were not utilized for 118-hydroxylation of DOC or O2 uptake suggests that the tumor mitochondria are lacking in flavoprotein enzymes required for NADH and NADPH oxidation (respiratory chain NADH dehydrogenase and the NADPH dehydro- genase of the steroid hydroxylation chain). This is supported by the fact that exogenous NADPH additions did not lead to the usual conversion of DOC into corticosterone as seen in normal mito- chondria (16). On the other hand, we cannot assume at the moment that the respective mito- chondrial dehydrogenases responsible for substrate oxidation (e.g., malate dehydrogenase) are present or were functional in the tumor mitochondria. Experiments are being carried out to investigate these questions.
Abnormal metabolic function of the tumor tissue was also apparent by a lower activity of the Embden-Meyerhof pathway. This was manifested by a lower production of lactate and pyruvate in the incubated tumor cells as compared with that
found in normal adrenal cells. Whereas ACTH, dbCAMP, and arsenite additions led to the usual large accumulation of lactate and pyruvate (20) in the incubated normal cells, stimulation of glycoly- sis by ACTH and dbCAMP or inhibition of pyruvate oxidation by arsenite was much less in the tumor cells. As mentioned above, these results suggest an abnormal metabolic function in the tumor. However, we cannot categorically rule out that the decrease in Embden-Meyerhof pathway activity (as reflected by lowered pyruvate and lactate production) is due to glycolysis being channeled through other pathways, rather than being totally blocked. One of these pathways would involve the conversion of glyceraldehyde- 3-phosphate to glycerol-3-phosphate via dihydrox- yacetone phosphate. Should glycerol-3-phosphate be one of the main products of glucose metabolism in the tumor tissue besides pyruvate (or lactate), it is conceivable that the glycerophosphate shuttle provides a mechanism for transporting cytoplas- mic reducing equivalents into the mitochondria. The product of this reaction, reduced FAD (the flavin coenzyme of mitochondrial glycerol-3-phos- phate dehydrogenase), could then be oxidized via CoQ and the respiratory chain to yield ATP necessary for growth. It would be desirable to know whether ATP production in the tumor cell arises mainly as a result of glycolysis per se or via a mechanism utilizing the glycerophosphate shuttle. In any case, unlike the virtual absence of the cytosol NAD-linked glycerol-3-phosphate dehy- drogenase in a wide variety of cancer tissues (33-35), we have found a normal degree of activity of NAD as well as a very high activity of the mitochondrial counterpart, FAD-linked enzyme, in tumor cells2. Further experiments are in pro- gress to test the physiological significance of the above findings.
In conclusion, it is clear that the reduced corticosteroidogenesis of the Snell adrenocortical carcinoma is, at least in part, due to reduced mitochondrial function. This specifically involves a lowered capacity for 118-hydroxylation and could possibly be linked to a lack of or reduction in the certain mitochondrial enzymes necessary for nor- mal oxidation. However, we do not wish to suggest that this is the only cellular function that could be altered in the tumor cell. A number of other steps
2 F. G. Peron, A. Haksar, M. T. Lin, D. Kupfer, G. Kimmel, and W. Robidoux, Jr. Manuscript in prepara- tion.
in steroidogenesis (e.g., CAMP-dependent protein kinase activation [36] or intracellular pooling of steroid precursors and products [37]), may be involved. The relative importance of these steps in steroidogenesis will be realized only with addi- tional investigation into the function of both the normal and tumorous adrenal.
We would like to express our sincere thanks to Dr. Robert Ney (University of North Carolina, Department of Medicine) for initially supplying the tumorous animals and for his comments during the preparation of the manuscript. We would also like to thank Drs. Kupfer and Luftig for their help in the preparation of the manuscript.
This work was supported by grants AM-04899 and 5T01 AM-05564-16 from the National Institutes of Arthritis, Metabolism and Digestive Diseases.
Received for publication 6 December 1973, and in revised form 11 March 1974.
REFERENCES
1. SNELL, K. C., and H. L. STEWART. 1959. J. Natl. Cancer Inst. 22:1119.
2. JOHNSON, D. F., K. C. SNELL, D. FRANCOIS, and E. HEFTMAN. 1961. Acta Endocrinol. 37:329.
3. NEY, R. L., N. J. HOCHELLA, D. G. GRAHAME- SMITH, R. N. DEXTER, and R. W. BUTCHER. 1969. J. Clin. Invest. 48:1733.
4. SCHORR, 1., and R. L. NEY. 1971. J. Clin. Invest. 50:1295.
5. SHARMA, R. K., and K. HASHIMOTO. 1972. Cancer Res. 32:666.
6. SCHORR, I., P. RATHNAM, B. B. SAXENA, and R. L. NEY. 1971. J. Biol. Chem. 246:5806.
7. SHARMA, R. K. 1973. Eur. J. Biochem. 32:506.
8. HAYANO, M., N. SABA, R. I. DORFMAN, and O. HECHTER. 1956. Recent Prog. Horm. Res. 12:79.
9. GRANT, J. K. 1968. J. Endocrinol. 41:111.
10. LUFT, J. H. 1961. J. Biophys. Biochem. Cytol. 9:409.
11. HAKSAR, A., S. BANIUKIEWICZ, and F. G. PÉRON. 1973. Biochem. Biophys. Res. Commun. 52:959.
12. HAKSAR, A., and F. G. PERON. 1972. J. Steroid Biochem. 3:847.
13. SHARMA, R. K., K. HASHIMOTO, and A. E. KITABCHI, 1972. Endocrinology. 91:994.
14. SILBER, R. H., R. D. BUSCH, and R. OSLAPAS. 1958. Clin. Chem. 4:278.
15. BUSH, I. E. 1952. Biochem. J. 50:370.
16. PÉRON, F. G. 1962. In Methods in Hormone Re- search. R. I. Dorfman, editor. Academic Press, Inc., New York. Chap. 6. 199.
17. PÉRON, F. G., and J. L. MCCARTHY. 1968. In Functions of the Adrenal Cortex. K. W. McKerns, editor. Appleton-Century-Crofts Inc., New York. Chap. 8. 261.
18. PÉRON, F. G., and B. V. CALDWELL. 1968. Biochim. Biophys. Acta. 164:396.
19. PÉRON, F. G., C. P. TSANG, and A. HAKSAR. 1972. Biochim. Biophys. Acta. 270:266.
20. TSANG, C. P., and F. G. PÉRON. 1970. Steroids. 15:251.
21. KAHRI, A. I. 1971. Anat. Rec. 171:53.
22. PORTER, K. R., and M. A. BONNEVILLE. 1968. Fine Structure of Cells and Tissues. Lea & Febiger, Philadelphia, Pa. 74-75.
23. MALAMED, S., G. SAYERS, and R. L. SWALLOW. 1970. Z. Zellforsch. Mikrosk. Anat. 107:447.
24. LUSE, S. 1967. In The Adrenal Cortex. A. B. Eisenstein, editor. Little, Brown and Co., Inc., Boston, Mass. 1-59.
25. IDELMAN, S. 1970. Int. Rev. Cytol. 27:181.
26. KAHRI, A. I. 1966. Acta Endocrinol. Suppl. 108. 52:1.
27. NISHIKAWA, M., I. MURONE, and T. SATO. 1963. Endocrinology. 72:197.
28. SAUER, L. A. 1970. Arch. Biochem. Biophys. 139:340.
29. PURVIS, J. L., R. G. BATTU, and F. G. PÉRON. 1963. In Functions of the Adrenal Cortex. K. W. McKerns, editor. Appleton-Century-Crofts Inc., New York. 1007-1055.
30. RACELA, A., D. AZARNOFF, and D. SVOBODA. 1969. Lab. Invest. 21:52.
31. SHARMA, R. K., and J. S. BRUSH. 1973. Arch. Biochem. Biophys. 156:560.
32. CALDWELL, B. V., F. G. PERON, and J. L. MCCARTHY. 1968. Biochemistry. 7:788.
33. BOXER, G. E., and C. E. SHONK. 1960. Cancer Res. 20:85.
34. RAO, G. A., and S. ABRAHAM. 1973. Lipids. 8:232.
35. BORST, P. 1962. Biochim. Biophys. Acta. 57:270.
36. GARREN, L. D., G. N. GILL, H. MASUI, and G. M. WALTON. 1971. Recent Prog. Horm. Res. 27:433.
37. WHITEHOUSE, B. J., and G. P. VINSON. 1972. Steroids Lipids Res. 3:110.
38. CHANCE, B., and G. R. WILLIAMS. 1955. Nature (Lond.). 175:1120.