Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
(CANCER RESEARCH 43, 3497-3503, August 1983] Ketone Body, Glucose, Lactic Acid, and Amino Acid Utilization by Tumors in Vivo in Fasted Rats1 Leonard A. Sauer2 and Robert T. Dauchy Mary Imogens Basse«Hospital, Cooperstown, New York 13326 ABSTRACT Arteriovenous differences for acetoacetate, ß-hydroxybutyrate, glucose, lactic acid, and glutamine and other amino acids were measured across Morris hepatomas 5123C, 7777, and 7288CTCF and Walker sarcocarcinoma 256 in vivo in rats fasted for 2 days. The acetoacetate and ß-hydroxybutyrate concentra tions in arterial whole blood of fasted tumor-bearing rats were 0.52 ±0.06 and 1.82 ±0.19 HIM (S.E., n = 38), respectively. Both ketone bodies were utilized by the tumors, and the rates of utilization were directly related to the rates of supply. The mean utilization rates for acetoacetate and 0-hydroxybutyrate were 13.9 ±2.9 (range, 0 to 64; n = 30) and 24.7 ±4.4 (range, 0 to 145; n = 38) nmol/min/g tumor wet weight, respectively. Eight of the tumors produced acetoacetate, presumably from utilized ß-hydroxybutyrate. An average of 52% of the acetoace tate and 30% of the 0-hydroxybutyrate carried in the arterial blood was removed during one pass through the tumors. The concentrations of glucose and glutamine in the arterial whole blood of fasted tumor-bearing rats (n = 38) were 6.55 ±0.3 and 0.76 ±0.02 HIM, respectively; both of these substrates were utilized at rates that were directly proportional to the rates of supply. The mean rates of glucose and glutamine utilization for all tumors in fasted rats were 101 ±11 (range, 3 to 313) and 8.2 ±1.1 (range, 0 to 25.1) nmol/min/g tumor wet weight, respectively. Thirty-six % of the glucose and 25% of the gluta mine supplied to the tumors was utilized. Comparison (by linear regression and analysis of covariance) of the rates of supply and utilization of glucose and glutamine in tumors growing in fasted versus fed rats indicated that these substrates are utilized more efficiently by tumors growing in fasted animals. Lactic acid was either produced or utilized, depending on the arterial whole-blood concentration. Production or utilization occurred, respectively, when the arterial lactate concentration was less or greater than 1 to 3 HIM. The arterial whole-blood amino acids (except gluta mine) were utilized at rates that ranged from 1 to 4 nmol/min/g tumor wet weight. The results indicate that energy production for tumor growth in fasted rats is supported, in part, by an increased availability of ketone bodies, by an increased efficiency of utilization of glucose and glutamine, and, under certain circum stances, by utilization of lactic acid. INTRODUCTION Tumors have the remarkable ability to sustain a rapid rate of growth during starvation or semistarvation of the host (2, 16). The tumor growth occurs as the host tissues, especially fat and skeletal muscle, decrease in weight (4, 5), and it is obvious that 1Supported by Grant CA 27809 from the USPHS, NIH, and the Stephen C. Clark Research Fund of the Mary Imogene Bassett Hospital. 2 To whom requests for reprints should be addressed. Received January 28, 1983; accepted April 21,1983. the nutrient needs of the tumor are adequately supplied and that they are extracted from the diminishing resources of the host. The host-derived metabolic fuels utilized by the tumor in the fasted animal are not known. Arterial blood glucose, the putative major fuel for fast-growing tumors, is decreased in concentration during fasting, and, as we (26) and others (9) have shown, the rate at which tumors utilize glucose in vivo is directly dependent on the rate of supply. Consequently, arterial glucose is unlikely to be the only fuel utilized by a tumor during fasting. Fenselau ef al. (6) made the important observation that succinylCoA:acetoacetyl-CoA transferase, the initial enzyme for utiliza tion of acetoacetate by peripheral tissues, is increased in activity with increased hepatoma growth rate. ß-Hydroxybutyrate dehydrogenase is present but relative to liver is reduced in activity in Morris hepatomas (20). These in vitro results suggested that ketone bodies may be important fuels for tumor growth in vivo in the fasted or semifasted animal. In this study, we measured the arteriovenous differences for the ketone bodies, lactic acid, glucose, and amino acids across 4 rat tumors in fasted rats. We found that both acetoacetate and /3-hydroxybutyrate were uti lized by rat tumors in vivo and that the rates of utilization were directly proportional to the rates of supply. Lactate was utilized at arterial lactate levels greater than 1 to 3 mw. Evidence is also presented which suggests that arterial glucose and glutamine are utilized more efficiently by tumors growing in fasted rats relative to tumors growing in fed rats. These properties of the tumor enhance access to available nutrients and may help ex plain the rapid tumor growth during fasting. MATERIALS AND METHODS Animals, Tumors, and Reagents. Adult male and female Buffalo and HarÃ-anSprague-Dawley rats were obtained from colonies established here. The rats were fed a standard laboratory chow (Charles River Rat, Mouse, Hamster Formula; Agway, Inc., Syracuse, N. Y.) and water ad libitum and were subjected to alternate 12-hr periods of dark and light. Tumor implantation, growth of "tissue-isolated" tumors, preparation of the animal for tumor harvest, and collection of arterial and venous blood samples were as described previously (26). The Morris hepatomas were originally obtained from the late Dr. Harold P. Morris, Morris Hepatoma Program, Howard University Cancer Center, Washington, D. C. The Walker sarcocarcinoma 256 was obtained from the E. G. and G. Mason Research Institute, Worcester, Mass. These tumors have been carried in this laboratory for about 4 years. The mean tumor blood flow rate was 107 ±4 /al/min (n = 38). Food was removed from fasted animals at 8 a.m. after the overnight feeding period, and the tumors were harvested 48 hr later. Blood Sample Preparation and Assay. Perchloric and trichloroacetic acid extracts of arterial and venous whole blood were prepared as described previously (26). Acetoacetate (18) and fj-hydroxybutyrate (33) were assayed fluorometrically by enzymatic methods using an Eppendorf photometer with fluorometric attachment. Acetoacetate was measured immediately after neutralization of the perchloric acid extract. Glucose, AUGUST 1983 Downloaded from cancerres.aacrjournals.org on April 29, 2017. © 1983 American Association for Cancer Research. 3497 L. A. Sauer and R. T. Dauchy lactic acid, glutamine, and amino acids were assayed as described previously (26). Expression and Evaluation of Results. For interpretation of the arteriovenous difference measurements, we assume that the tumor is in a steady state with regard to nutrient supply and metabolism and that this steady state continues during the period of sample collection, usually about 5 min. Utilization or production of a nutrient by the tumor occurs when the arteriovenous difference is positive or negative, respectively. Utilization (or use) of a nutrient includes inter alia oxidation to CO2 and/ or conversion via biosynthesis to form protoplasm and secretions. Only the overall process of utilization was measured in these experiments. Utilization and production rates are given in nmol substrate utilized or produced per min per g tumor, wet weight and were calculated from the arteriovenous differences and the tumor blood flow rate (¿Ã-l/min/g). Supply rates are expressed as nmol/min/g tumor, wet weight and were calculated from the arterial whole-blood concentration and the tumor blood flow rate. Errors in assay of amino acids and other substrates and in measurement of tumor blood flow rates and the effects of these errors on calculations of utilization and production rates are the same as those described earlier (26). Data are presented as mean ±S.E. Groups of results were compared by linear regression and analysis of covariance (27). The same caveats are in effect here with regard to interpretation of data by linear regression as were in effect in the previous report (26). RESULTS AND DISCUSSION Ketone Body Utilization. Ketone bodies were utilized by 37 of the 38 tumors examined. The rates of utilization by Morris hepatomas 5123C, 7777, and 7288CTCF and Walker sarcocarcinoma 256 were directly dependent on the rate of supply (Chart 1, A and B). The mean rate of acetoacetate and /3-hydroxybutyrate utilization for all tumors was 13.9 ±2.9 (n = 30) and 24.7 ±4.4 (n = 38) nmol/min/g tumor, wet weight, respectively. Eight of the tumors produced acetoacetate at slow rates (not shown), presumably from /3-hydroxybutyrate utilized (see below). The fastest rate of acetoacetate utilization (64 nmol/min/g tumor, wet weight) was observed in a hepatoma 7288CTCF, and the fastest rate of /3-hydroxybutyrate utilization (145 nmol/min/g tumor, wet weight) occurred in a hepatoma 5123C (not shown in Chart 10). The rate of /3-hydroxybutyrate supply to the latter WALKER 256 7288CTCF 7777 IOO SUPPLY, O IOO 200 nmol/min/g Chart 1. Regression of the rate of acetoacetate and /J-hydroxybutyrate utiliza tion on the rate of supply in rat tumors in fasted rats in vivo. Each point represents the determination for a single tumor. , least-squares fit to the data points. A, acetoacetate [y = 0.53x - 0.92, n = 30, r = 0.961 (p < 0.05)]; B, #-hydroxybutyrate [y = 0.30X - 0.4, n = 38, r = 0.869 (p < 0.05)]. 3498 tumor was 403 nmol/min/g tumor, wet weight, a value that agrees reasonably well with the predicted supply rate (482 nmol/ min/g tumor, wet weight) based on the linear regression analysis. It is interesting that, ¡nthe Morris hepatomas, the activity of succinyl-CoA:acetoacetyl-CoA transferase is increased (6) and that the activity of /3-hydroxybutyrate dehydrogenase is de creased (20) during tumor progression. Of the hepatomas stud ied here, the CoA transferase activity has been reported to be most active in hepatoma 7288CTCF (6). Total enzyme activity of the homogenate was 5.1 //mol/min/g tumor, wet weight (measured in the direction of acetoacetate formation). Assuming an equivalent reaction rate in the direction of acetoacetyl-CoA formation, the fastest acetoacetate utilization rate observed in a hepatoma 7288CTCF was less than 5% of the enzyme activity measured in vitro. The mean /3-hydroxybutyrate dehydrogenase activity of hepatoma 5123C has been reported to be 2.5 ¿¿mol/ min/g tumor, wet weight (20). The maximum rate of /3-hydroxy butyrate utilization by hepatoma 5123C in this study was 145 nmol/min/g tumor, wet weight, a rate that is less than 10% of the maximum in vitro enzyme activity. It is worth noting that the in vitro enzyme activity determinations were made at saturating levels of substrates and other cofactors and were performed with isolated hepatoma mitochondria solubilized by either deter gent treatment (6) or sonication (20). The in vitro enzyme activity measurements indicate that the specific enzyme content of the hepatoma mitochondria is more than adequate to explain the observed ketone body utilization rates and that the potential exists for even greater utilization rates than those measured here. The apparent straight-line relationship between ketone body utilization and supply illustrated in Chart 1 indicates that the rate of transport of the substrates into the tumor cells and the availability of required cofactors in the mitochondria are also adequate. The enzyme phenotype for mitochondrial energy pro duction that develops during tumor progression in hepatomas results from both the synthesis of new enzymes, e.g., CoA transferase, and the persistence of constitutive enzymes, e.g., 0-hydroxybutyrate dehydrogenase. Though adequate for the needs of the tumor, the final enzyme activities in the tumor mitochondria may be low relative to the activities found in mito chondria of the parent organ or in mitochondria of other special ized organs as with, for example, the /3-hydroxybutyrate dehy drogenase in liver (20, 32), kidney, and brain (32), and CoA transferase in kidney (32) and heart muscle (6, 32). The NAD(P)dependent malic enzyme (25) and glutaminase (17), 2 other mitochondrial enzymes newly synthesized by Morris hepatomas during tumor progression, are additional examples. Both of these enzymes are much higher in activity in mitochondria from small intestinal mucosa (22, 24), a tissue in which glutamine is known to be a major respiratory fuel (34). Arterial whole-blood concentrations of acetoacetate and ßhydroxybutyrate in several 2-day-fasted control Buffalo rats and tumor-bearing Buffalo and Sprague-Dawley rats are listed in Table 1. Total arterial whole-blood ketone body concentrations for control and tumor-bearing animals were 3.75 ± 0.85 RIM (range, 2.0 to 8.7; n = 7) and 2.33 ±0.22 ITIM(range, 0.4 to 7.1 ; n = 38), respectively, suggesting that the presence of the tumor decreased the ketogenic response of the host to starvation. The fat stores of tumor-bearing rats are known to decrease during tumor growth (4, 5), and the lower arterial whole-blood ketone body concentrations in tumor-bearing animals probably result, in part, from a decreased availability of substrate for ketogenesis. CANCER RESEARCH VOL. 43 Downloaded from cancerres.aacrjournals.org on April 29, 2017. © 1983 American Association for Cancer Research. Tumor Metabolism in Vivo Table 1 Arterial whole-blood concentrations and in vivo utilization or production rates for ketone bodies, glucose, láclate, ano glutamina by tumors in rats tasted tor 2 days AcetoacetateTumorA. tion(nmol/min/gtumor,wet tion(nmol/min/gtumor.wet tion(nmol/min/gtumor,wet tion(nmol/min/gtumor.wet tion(nmol/min/gtumor,wet tion(mu)3.03±0.661.220.721.940.611.882.422.010.232.230.791.060.992.332.313.22Utiliza tration(ttlM)4.83±0.364.889.95.627.336.965.316.506.376.288.867.328.087.635.617.38Utiliza tration(rtiM)0.72±0.050.500.550.880.710.630.800.880.670.540.770. tration(mu)1.24±0.360.951.11.082.581.6 ment69709778112113929310481987479105107eArterialconcentra tion (rriM)0.72±0.13"1.010.570.930.240.710.660.330.201.310.060.140.220.200.110.36Utiliza wt)644566222445282e31e21e6e0-HydroxybutyrateArterialconcentra wt)30536538702949222015242829GlucoseArterialconcen wt)16247176921261241147465187313187959396GlutamineArterialconcen wt)151136717981161017776LáclateArterialconcen wt)1891411175126810556 Control"B. Lactate-producing tumorsHepatoma7288CTCFHepatoma 123CWalker5 carcinosar-coma 256C. Lactate-utilizing tumorsHepatoma7288CTCFHepatoma 123CWalker5 carcinosar-coma256Experi "Buffato rats (n = 7). 6 Mean ±S.E. c These represent production rates. d These represent utilization rates. 8 Fasted for 96 hr. The acetoacetate:/î-hydroxybutyrate ratios in the arterial blood of the 38 animals examined ranged from 0 to 1.42. In some animals, significant differences were noted between the acetoacetate:0-hydroxybutyrate ratio in the carotid arterial blood and the ratio in the tumor venous blood. Assuming that these ratios are measures of the oxidation-reduction state in the animal carcass and in the tumor, respectively, we may conclude that the tumor is able to maintain an oxidation-reduction state differ ent from that of the rest of the animal. An interesting relationship was noticed between the arterial and venous acetoacetate:/?hydroxybutyrate ratio differences across the tumor and whether a tumor was a lactate producer or a lactate utilizer. For the purposes of this discussion, we divided the tumors into 3 groups. Group 1 tumors were termed strong lactate producers and showed production rates of greater than 20 nmol lactate pro duced per min per g tumor, wet weight. Group 2 tumors either produced or utilized less than 20 nmol lactate per min per g tumor, and Group 3 tumors were strong lactate utilizers (greater than 20 nmol lactate utilized per min per g tumor). The acetoacetate:0-hydroxybutyrate ratios in the carotid arterial whole blood for animals carrying Group 1, 2, and 3 tumors were 0.56 ±0.08 (n = 15), 0.36 ±0.09 (n = 8), and 0.12 ±0.02 (n = 11), respectively. The ratios in the tumor venous whole blood for Group 1, 2, and 3 tumors were 0.34 ±0.04, 0.39 ±0.14, and 0.23 ±0.05, respectively. The arterial whole-blood lactic acid concentrations for the animals carrying Group 1,2, and 3 tumors were 1.27 ±0.11,1.64 ±0.31, and 3.45 ±0.42 mw, respectively. As expected, the arterial lactic acid concentration was decreased the more oxidized the oxidation-reduction state of the animal carcass. In animals carrying Group 1 tumors, the tumor was more reduced (p < 0.05) than the animal carcass, arterial wholeblood lactic acid was low, and the tumors released lactate. Table 1B shows acetoacetate and /3-hydroxybutyrate utilization and lactate production rates and other data for 6 representative tumors (Experiments 69, 93, 97, 104, 112, and 113) from this group. Both ketone bodies were utilized by the hepatomas and by the Walker sarcocarcinoma. In this group, an average of 56% of the arterial acetoacetate and of 30% of the 0-hydroxybutyrate was removed during one pass through the tumor. Group 3 tumors were more oxidized (p < 0.05) than the animal carcass, the arterial whole-blood lactic acid concentration was high, and the tumors utilized arterial lactic acid. Data for 6 representative experiments from this tumor group are shown in Table 1C. Arterial whole-blood acetoacetate concentrations were lower in these animals, and, as expected, 8 of these tumors released acetoacetate (see Experiments 74, 81, 105, and 107) into the venous blood. Utilization of /3-hydroxybutyrate remained high, however, and 30% of the 0-hydroxybutyrate supplied to the tumors was utilized. Group 2 tumors were intermediate between the tumors in Groups 1 and 3, and the oxidation-reduction states of the tumor and the host were not different (p > 0.05). Three representative experiments from this tumor group are shown in Table 1B. Tumors in Groups 1, 2, and 3 did not differ significantly in oxidation-reduction state, suggesting that the tumor attempts to maintain a relatively fixed oxidation-reduction state despite wide fluctuations in the oxidation-reduction state of the host. This could provide a stable environment ensuring continued reductive biosynthesis and growth. We assume that the aceto acetate and 0-hydroxybutyrate reached equilibrium during pas sage through the tumor and that the ratio is an accurate measure of the tumor oxidation-reduction state. The high enzyme activi ties (see above) of the tumors relative to the ketone body supply rates to the tumors suggest that this is a reasonable assumption. It is not clear why more than one-half of the control (5 of 7) AUGUST 1983 Downloaded from cancerres.aacrjournals.org on April 29, 2017. © 1983 American Association for Cancer Research. 3499 L A Sauer and R. T. Dauchy and about half of the tumor-bearing fasted rats had acetoacetate:ß-hydroxybutyrate ratios of less than 0.25. Considering that certain animals may be more susceptible to respiratory depres sion and tissue hypoxia during pentobarbital anesthesia, we routinely supplied all anesthetized animals with oxygen through a loose-fitting nose cone (26). In other experiments, we found arterial whole-blood lactate concentrations of 2 mw and greater and low acetoacetate:/i-hydroxybutyrate ratios in anesthetized rats breathing 100% oxygen in a closed system with respiration assisted via a rodent respirator (data not shown). It does appear, therefore, that tissue hypoxia is not responsible for the elevated lactate and ß-hydroxybutyrate levels. The acetoacetate:/S-hydroxybutyrate ratio is decreased in rats by starvation. The animal carcass, particularly skeletal muscle, becomes more reduced as a response to fasting (8) and may explain the low acetoacetate:/3hydroxybutyrate ratios and high lactic acid concentrations ob served. A more reduced oxidation-reduction state in muscle during fasting might help conserve muscle protein (8). In addition to the oxidation-reduction state data discussed above, other metabolic data indicate that the Group 3 tumor is more oxidized than the rest of the body. For example, lactate, /3-hydroxybutyrate, glucose, and glutamine were all utilized (and presumably oxidized) at substantial rates in these tumors (Table 1C). It seems very unlikely, therefore, that the tumors listed in Table 1C were anoxic. Gullino ef al. (10) have shown that a decrease in oxygen availability decreases the rate of glucose utilization by tumors in vivo. Lactic Acid Production and Utilization. In a previous study (26) performed with fed tumor-bearing rats, we found that tumors in vivo could either produce or utilize lactic acid. Which of these processes occurred depended on the arterial lactic acid concen tration. Production of lactate was most often observed at lactic acid concentrations below 2 mM, and lactate utilization generally occurred at lactate levels above 2 mM. In the range of 1 to 3 mM arterial lactate, tumors were often found that neither removed nor added (despite the fact that glucose utilization had occurred) significant amounts of lactate to the blood. Chart 2 shows that an identical relationship was observed in the tumors growing in the 38 fasted rats studied here. The data for both the fasted and fed animals (see Ref. 26) were combined in Chart 2 to illustrate the identical nature of the results obtained and to emphasize the frequency with which in vivo lactate utilization occurred. A total of 71 tumors were sampled; 24 were found to utilize arterial whole-blood lactate. Since lactic acid is a metabolic end product, it can reenter metabolism only through oxidation via lactic dehydrogenase; therefore, the capacity for NADH plus pyruvate oxidation in these tumors must be high. For example, assuming that 1 and 2 mol of pyruvic acid were formed, respectively, from each mol of arterial lactate and glucose utilized, the hepatoma 7288CTCF listed in Table 1C (Experiment 98) may have oxidized more than 650 nmol of pyruvate per min per g of tumor. These oxidations occurred simultaneously with the utilization of 10 and 20 nmol/min/g of tumor of glutamine and ß-hydroxybutyrate, respectively. Unlike normal liver, the pyruvate dehydrogenase activity of fast-growing Morris hepatomas is not decreased by fasting (11). Several of the tumors that produced lactate at low rates (Group 2 tumors) were high-glucose utilizers (see Table 1B, Experiments 70, 78, and 92). The great potential for lactate production was not expressed, presumably because the oxidation-reduction states of tumor and host were similar and because arterial 3500 200z053 ••0 'a*;-*• 100fiCo" «L^w^?5f> i•» oy „ Õ 2^X5I23C *A icotz 0o r« DA 0A" ûA 0Ooo£ = •, 7777 = •, 7288CTCF = A , WALKER 256 * T , v . 200TA O 2 ARTERIAL 4 6 LACTIC 8 IO ACID, mM Chart 2. Lactic acid production and utilization by rat tumors in vivo plotted against the arterial lactic acid concentration. Each point represents the determina tion for a single tumor. Fed animals had free access to food, and fasted animals were fasted for 48 hr. lactate-derived and glucose-derived NADH and pyruvate gener ated intracellularly were being utilized at the same time, and only a small amount of lactate was released into the tumor vein. Windmueller and Spaeth (34) have shown in rat small intestinal mucosa that arterial lactate may be utilized at the same time that glucose-derived lactate is released into the venous blood. When the arterial lactate concentration was greater than 1.2 to 1.6 mw, net lactate utilization occurred simultaneously with glucose utili zation (34). In tumor, as in small intestinal mucosa, it would appear that the arterial lactic acid concentration plays an impor tant role in determining if intracellular lactate (as NADH and pyruvate) is oxidized or is formed and released. More than 50 years ago, Warburg (29) described the produc tion of large amounts of lactic acid from glucose by tumor slices incubated aerobically. Since then, numerous other in vitro ex periments confirmed this finding, and a high rate of aerobic glycolysis became an acknowledged biochemical hallmark of tumor tissues (15). More recently, however, high rates of aerobic glycolysis were described in normal tissues (3, 13, 19, 21, 23), and then, following the development of the Morris hepatomas, it was discovered that slow-growing tumors were poor lactate producers (1). Although these findings did alter the view that aerobic glycolysis was associated only with malignant cells, the idea that a high rate of lactate production is a property of a fastgrowing tumor cell has persisted and is supported by in vitro experiments (31). We believe, however, that this idea can no longer be considered as an accurate representation of glucose and lactate metabolism by tumors in vivo. In the total of 71 tumors examined in vivo (Chart 2), about 50% either utilized arterial lactate or released only a small amount of lactic acid into the venous blood. These tumors included Morris hepatoma 7288CTCF, a fast-growing hepatoma, and Walker sarcocarcinoma 256, an undifferentiated tumor. High arterial lactate con centrations inhibited release of lactate even though glucose had CANCER RESEARCH VOL. 43 Downloaded from cancerres.aacrjournals.org on April 29, 2017. © 1983 American Association for Cancer Research. Tumor Metabolism been utilized (Table 1C). On the other hand, low arterial lactate concentrations promoted lactate release (Table 1B). Since in vitro experiments of glucose oxidation and lactate production are most often performed in the absence of added lactate, high rates of lactate formation are expected (Chart 2). There are no discrepancies between results obtained in vitro and the in vivo results shown in Chart 2 if it is recognized that, in the presence of high arterial lactate, tumor tissues may remove lactate from the arterial blood. The uptake, metabolism, and oxidation of lactate utilized by tumors in vitro have been described by Spen cer and Lehninger (28), Katz et al. (14), and Weinhouse (30), respectively. Glucose Utilization. The mean arterial whole-blood glucose concentrations in fasted control and tumor-bearing rats were 4.83 ±0.36 mw (n = 8) (Table 1) and 6.55 ±0.3 mw (n = 38), respectively. The same relative differences were observed in control and tumor-bearing fed rats (26), suggesting that the presence of the tumor increases the arterial whole-blood glucose concentration of the fasted host. All 38 of the tumors growing in the fasted rats utilized glucose; the rate of glucose utilization for each tumor plotted against the rate of glucose supply to the tumor via the arterial blood is shown in Chart 3. Chart 3 also shows regressions of the rates of glucose utilization on the rates of glucose supply for tumors growing in the fasted rats ( ) and for tumors growing in fed animals ( , data taken from Ref. 26). The same apparent direct dependence of the glucose utilization rate on the rate of supply observed in vivo ¡ntumors growing in fed rats (9, 26) was observed for tumors growing in fasted rats. The results for the 2 animal groups were not identical, however, as shown by the shift to the left of the regression line for tumors grown in the fasted rats. Comparison of the 2 regres sion lines by analysis of covariance (27) indicated that the differ ence between the regression lines (and, hence, between the sample populations) is significant (p < 0.05). At any rate of glucose supply, tumors growing in fasted rats increased their glucose utilization rates by about 50 nmol/min/g tumor over those of tumors growing in fed rats. This apparent increase in efficiency of glucose utilization by tumors in fasted animals may in Vivo compensate for the decreased arterial glucose concentration observed in fasted (6.55 mw, n = 38) versus fed (7.4 ±0.4 mw, n = 26; Ref. 26) tumor-bearing rats. Amino Acid Utilization. Glutamine, the most abundant amino acid in rat arterial blood, was the most rapidly utilized. The rate of glutamine utilization by tumors growing in fasted rats appeared to be directly dependent on the rate of glutamine supply to the tumor. An identical relationship was noted previously for tumors growing in fed rats (26). Chart 4 shows the rates of glutamine utilization by 38 tumors growing in fasted rats plotted against the rates of glutamine supply to the tumors. The regression line for these data points is shown by the solid line. The dashed line is the regression line for the data obtained for tumors growing in fed animals (data not shown; see Ref. 26). As was noted for the relationship between glucose utilization and supply by tumors growing in fed and fasted animals (Chart 3), the regression line for glutamine utilization on glutamine supply for fasted rats was shifted to the left of that for fed rats. These results suggest that tumors growing in fasted rats are able to increase the efficiency of the processes by which arterial glutamine and glucose are utilized. Since the rate-limiting steps for glucose and glutamine utilizations in vivo are not known, no decisions can be made about the mechanisms of these apparent activations of substrate uptake. Arterial whole-blood amino acid levels in fasted control and tumor-bearing Buffalo rats are listed in Table 2. The concentra tions of threonine, serine, glycine, and alanine were increased, and the concentration of lysine was decreased in the blood of the hepatoma 7288CTCF-bearing animals. Serine was the only amino acid the concentration of which was increased in the blood of hepatoma 5123C-bearing rats. Amino acid utilization and production by tumors growing in fasted rats (not shown) were similar to those observed for tumors growing in fed rats (26). Each of the amino acids measured was utilized by hepatoma 5123C (n = 9); the utilization rates ranged from 0 to 3 nmol/min/ g tumor, except for the utilization rates of lysine and glutamine, which were 5.5 and 8.2, respectively. Similar rates of amino acid utilization were noted in the faster-growing tumors, hepatoma 7288CTCF (n = 9) and Walker sarcocarcinoma 256 (n = 14). V = WALKER 256 «7288CTCF A o = 7777 = 5I23C o 400 GLUCOSE SUPPLY, 800 0 nmol/min/g GLUTAMINE SUPPLY, nmol/min/g Chart 3. Regression of the rate of glucose utilization on the rate of glucose supply in rat tumors in fasted rats in vivo. Each point represents the determination for a single tumor. , least-squares fit to the data points shown for the tumors growing in fasted animals [y = 0.38x - 4.7, n = 38, r = 0.864 (p < 0.05)]; , least-squares fit for data from tumors growing in fed rats taken from Ref. 26 (y = 0.36x - 36.9). Comparison by analysis of covariance indicated that the 2 regression lines were different (p < 0.05). AUGUST 40 80 Chart 4. Regression of the rate of glutamine utilization on the rate of glutamine supply in rat tumors in fasted rats in vivo. Each point represents the determination for a single tumor. , least-squares fit to the data points shown for the tumors growing in fasted rats [y = 0.23x + 0.30. n = 38, r = 0.631 (p < 0.05)]; , least-squares fit for data from tumors growing in fed rats taken from Ref. 26 (y = 0.24x - 4.4). Comparison by analysis of covariance indicated that the 2 regression lines were different (p < 0.05). 1983 Downloaded from cancerres.aacrjournals.org on April 29, 2017. © 1983 American Association for Cancer Research. 3501 L A. Sauer and R. T. Dauchy Table 2 Amino acid levels in arterial whole blood of normal and tumor-bearing fasted Buffalo rats Whole-blood samples were removed from the carotid arteries of anesthetized animals, and amino acid analyses were performed as described in 'Materials and Methods/ 2-dayhepatomafasted rats(i/mol/liter)49 bearing rats " acidAspartic Amino Oimol/liter)37 hepatoma-bearing rats Ornol/liter)45 of substrates (glucose and glutamine). An increased efficiency of glucose and glutamine utilization during fasting is important, because it suggests that the signals generated in the host during fasting which cause hypometabolic changes in some host tissues will promote a hypermetabolic state in the tumor. ACKNOWLEDGMENTS acidThreonineSerineGlutamic579± 3110± ± 6133±18C231 ± Specialthanks are due Dr. Walter Nagelfor numeroushelpfuldiscussions during 5153± 13198 this research and for careful reading of the manuscript. Dr. Lorann Stallones also 7e234± 16C283 ± ±10239 provided valuable help and advice in the statistical analyses. acidGlutamineGlycineAlanineIsoleucineLeucineTyrosinePhenylalanineLysineAmmoniaNormal 16724± 13775± 31807 ± 50241 ± 37255± ±64351 23e224 ± 15148 ± ±29200 27°76 ± 1068 ± REFERENCES ±2064± 799± 795 7119± ± 1255 ± 1060± ± 1559 1. Aisenberg, A. C., and Morris, H. P. Energy pathways of hepatoma No. 5123. Science (Wash. D. C.), 797:1314-1315,1961. 649± 450 862± 3516± 4513± 7442± 2. Buzby, G. P., Mullen, J. L, Stein, T. P., Miller, E. E., Hobbs, C. L. and Rosato, ±13°239 E. F. Host-tumor interaction and nutrient supply. Cancer (Phila.),45: 2940±28154 ±22210 ±315123C ±257288CTCF ±27 2948, 1980. Significant difference at p < 0.05 level as compared to normal 2-day-fasted rats. Again, glutamine and lysine were the most rapidly utilized. Low production rates (less than 3 nmol/min/g tumor) for glycine and alanine were noted in 5 of the 9 hepatomas 7288CTCF and in 5 of the 14 Walker sarcocarcinomas 256. Production of these amino acids by a few of these tumors also occurred in fed rats (26) as well. The source of these amino acids is not known. In this paper, we have shown that tumors growing in fasted rats are able to utilize the ketone bodies suppled via the arterial blood. To our knowledge, this is the first demonstration of the utilization of these substrates by tumors in vivo. Fields ef al. (7) were unable to demonstrate ketone body oxidation in hepatoma 7777 in vitro, a tumor that we found utilized ketone bodies in vivo (Chart 1). Although we assumed that the utilized acetoacetate and /tf-hydroxybutyrate were oxidized via established mito chondria! reactions (6, 20), we have no evidence as yet that the oxidations occurred. The same statement may be made con cerning the in vivo lactic acid, glucose, and glutamine utilizations. Experiments in which radioactive ketone bodies (and other sub strates) are infused into the tumor arterial blood followed by isolation, identification, and measurement of C02 and other products in the tumor venous blood are required to determine the extent of substrate oxidation in vivo. Despite our present lack of information on the extent of the oxidation of these substrates, it seems likely that the utilization of the ketone bodies and lactic and amino acids and the apparent enhanced utiliza tions of glucose and glutamine described herein are responsible for the rapid rate of tumor growth observed in the fasted host (2, 16). In a subsequent report,3 we will show that the growth rate of s.c. tumor implants and of tumors implanted on the epigastric vascular pedicle (26) is increased by fasting. Refeeding decreases the rate of tumor growth to the prefasted growth rate. Holley (12) has proposed that the growth rate of tumors is directly related to the intracellular concentrations of the nutrients needed for energy production and for the synthesis of cell constituents. An accelerated tumor growth rate during acute fasting, therefore, could result from both increased substrate availability (ketone bodies and lactate) and from an increased efficiency of utilization 3L. A. Sauer and R. T. Dauchy, manuscript in preparation. 3502 3. Cohn, D. V., and Forscher, B. K. Aerobic metabolism of glucose by bone. J. Biol. Chem., 237: 615-618, 1962. 4. Costa. G. Cachexia, the metabolic component of neoplastic diseases. Cancer Res., 37: 2327-2335,1977. 5. Ekman, L., Kariberg, I., Edstrom, S., Undmark, L., Schersten, T., and Lundholm, K. Metabolic alterations in liver, skeletal muscle, and fat tissue in response to different tumor burdens in growing sarcoma-bearingrats. J. Surg. Res., 33: 23-31, 1982. 6. Fenselau, A., Wallis, K., and Morris, H. P. Subcellular localization of acetoacetate coenzymeA transferase in rat hepatomas. CancerRes., 36:4429-4433, 1976. 7. Fields. A. L. A., Wolman. S. L., Cheema-Dhadli,S., Morris, H. P., and Halperin, M. L. Regulation of energy metabolism in Morris hepatomas 7777 and 7800. Cancer Res., 41: 2762-2766,1981. 8. Goodman, M. N., Larsen, P. R., Kaplan, M. M., Aoki, T. T., Young, V. R., and Ruderman, N. B. Starvation in the rat. Am. J. Physiol.,239: E277-E286,1980. 9. Cullino, P. M., Grantham, F. H.. and Courtney, A. H. Glucose consumption by transplanted tumors in vivo. Cancer Res., 27: 31-40,1967. 10. Cullino, P. M., Grantham, F. H., Courtney, A. H., and Losonczy, I. Relationship between oxygen and glucose consumption by transplanted tumors in vivo. Cancer Res., 27: 1041-1052, 1967. 11. Halperin, M. L., Taylor, W. M., Cheema-Dhaldi,S., Morris, H. P., and Fritz, I. B. Effects of fasting on the control of fatty acid synthesis in hepatoma 7777 and host liver. Eur. J. Biochem., 50: 517-522,1975. 12. Holley,R. W. A unifying hypothesis concerningthe nature of malignantgrowth. Proc. Nati. Acad. Sei. U. S. A., 69: 2840-2841,1972. 13. Hume, D. A., Radik, J. L., Ferber, E., and Weidemann,M. J. Aerobic glycolysis and lymphocyte transformation. Biochem. J., 774: 703-709, 1978. 14. Katz, J., Brand, K., Golden, S., and Rubenstein, D. Lactate and pyruvate metabolism and reducing equivalent transfer in Ehrlich ascites tumor. Cancer Res., 34. 872-877, 1974. 15. Lehninger,A. L. Biochemistry, pp. 849-850. New York: Worth Publishers,Inc., 1975. 16. LePage, G. A., Potter, V. R., Busch, H., Heidelberger, C., and Huribert, R. B. Growth of carcinoma implants in fed and fasted rats. Cancer Res., 12: 153157, 1952. 17. Linder-Horowitz, M., Knox, W. E., and Morris, H. P. Glutaminaseand growth rates of rat hepatomas. Cancer Res., 29:1195-1199,1969. 18. Mellanby, J., and Williamson, D. H. Acetoacetate. In: H. U. Bergmeyer (ed.), Methods of Enzymatic Analysis, Vol. 4, pp. 1840-1843. New York: Academic Press, Inc., 1974. 19. Neuman, W. F., Neuman, M. W., and Brommage, R. Aerobic glycolysis in bone: lactate production and gradients in calvarÃ-a. Am. J. Physiol.. 234: C41C50, 1978. 20. Ohe, K.. Morris. H. P., and Weinhouse, S. ff-Hydroxybutyrate dehydrogenase activity in liver and liver tumors. Cancer Res., 27:1360-1371,1967. 21. Paul. R. J.. Bauer, M.. and Pease. W. Vascular smooth muscle: aerobic glycolysis linked to sodium and potassium transport processes. Science (Wash. D. C.), 206: 1414-1416,1979. 22. Pinkus, L. M., and Wmdmueller.H. G. Phosphate-dependent glutaminase of small intestine: localization and role in intestinal glutamine metabolism. Arch. Biochem. Biophys., 182: 506-517, 1977. 23. Rhoades, R. A., Shaw, M. E., Eskew, M. L., and Wali, S. Lactate metabolism in perfused rat lung. Am. J. Physiol., 235: E619-E623, 1978. 24. Sauer, L. A., Dauchy, R. T., and Nagel, W. O. Identification of an NAD(P)dependent malic enzyme in small-intestinal-mucosalmitochondria. BiochemJ., 784: 185-188, 19.79. 25. Sauer, L. A., Dauchy, R. T., Nagel, W. O., and Morris, H. P. Mitochondrial malic enzymes. J. Biol. Chem., 255. 3844-3848,1980. 26. Sauer, L. A., Stayman, J. W., Ill, and Dauchy, R. T. Amino acid, glucose, and CANCER RESEARCH VOL. 43 Downloaded from cancerres.aacrjournals.org on April 29, 2017. © 1983 American Association for Cancer Research. Tumor Metabolism lactic add utilization in vivo in rat tumors. Cancer Res., 42: 4090-4097,1982. 27. Snedecor, G. W., and Cochran, W. G. Statistical Methods. Ames, Iowa: The Iowa State University Press, 1976. 28. Spencer, T. L, and Lehninger, A. L. L-Lactate transport in Ehrlich ascitestumour cells. Biochem. J., 754: 405-414,1976. 29. Warburg, O. The Metabolism of Tumors. London: Arnold Constable, 1930. 30. Weinhouse, S. Oxidative metabolism of neoplastic tissues. Adv. Cancer Res., 3:269-325, 1955. 31. Weinhouse, S. Glycolysis, respiration, and anomalous gene expression in experimental hepatomas: G. H. A. Clowes MemorialLecture. Cancer Res., 32: in Vivo 2007-2016,1972. 32. Williamson, D. H., Bates, M. W., Page, M. A., and Krebs, H. A. Activities of enzymes involved in acetoacetate utilization in adult mammalian tissues. Biochem. J., J2Õ.-41-47,1971. 33. Williamson, D. H., and Mellanby, J. D-(-)-3-Hydroxybutyrate. In: H. U. Bergmeyer (ed.), Methods of Enzymatic Analysis, Vol. 4, pp. 1836-1839. New York: Academic Press, Inc., 1974. 34. Windmueller, H. G., and Spaeth, A. E. Identification of ketone bodies and glutamine as the major respiratory fuels in vivo for postabsorptive rat small intestine. J. Biol. Chem., 253: 69-76, 1978. AUGUST 1983 Downloaded from cancerres.aacrjournals.org on April 29, 2017. © 1983 American Association for Cancer Research. 3503 Ketone Body, Glucose, Lactic Acid, and Amino Acid Utilization by Tumors in Vivo in Fasted Rats Leonard A. Sauer and Robert T. Dauchy Cancer Res 1983;43:3497-3503. Updated version E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/43/8/3497 Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected]. Downloaded from cancerres.aacrjournals.org on April 29, 2017. © 1983 American Association for Cancer Research.