The Antiketogenic Effect of Alanine in Normal Man: Evidence for an

Alanine-Ketone Body Cycle

R.Nosadini, K. G. M. M. Alberti, D. G. Johnston, S. Del Prato, C. Marescotti, and E. Duner

The effect of alanine on ketone body levels, independent of hormonal changes, in normal man has been investigated. Five
normal subjects were given somatostatin infusions (200 /.rg/hour) for 3 hr. After 1 hr alanine or isotonic saline was infused
for 2 hr. With saline blood &hydroxybutyrate and acetoacetate levels rose steadily to a peak of 0.230 + 0.063 and 0.112 i
0.023 mmole/l respectively. With alanine fi-hydroxybutyrate and acetoacetate levels plateaued at 0.099 & 0.020 and
0.055 + 0.006 mmole/l respectively. Alanine levels reached nearly 1 mmole/l but a significant effect on ketone body levels
was apparent at physiologic levels (~0.6 mmole/l). Plasma fatty acid and glycerol levels did not change significantly.
Insulin C-peptide and glucagon levels were suppressed to a similar extent in both experiments. These results support the
view that alanine suppresses ketogenesis in man by a direct hepatic effect independent of insulin and glucagon. It is
suggested that this forms part of a negative feedback substrate cycle between alanine and ketone bodies.

T HE INVERSE RELATION between circulating venous blood samples were taken from the third cannula at 15, 0.

ketone body and alanine levels has been reported 15, 30.60 and 75 min, and every I5 mm thereafter until I80 min.
An aliquot, 1.5 ml, of blood was placed immediately in 5 ml %
in both rat and man.‘,’ Considerable interest has been
(v/v) perchloric acid for subsequent assay of glucose, lactate,
shown in the possible effect of ketone bodies in inhibit- pyruvate, alanine, glycerol, 3-hydroxybutyrate and acetoacetate:“.‘4
ing proteolysis3 and branched amino acid breakdown 2.25 ml was added to 0.25 ml aprotin (10,000 units/ml) containing
in muscle4 and inhibiting alanine release from extrahe- 25 pmole disodium-EDTA and centrifuged immediately for gluca-
patic tissues.5 Until recently, less attention has been gon assay,15 and a further 2 ml blood was put in heparinized tubes,
centrifuged and the serum stored at -20“ for insulin” and C-
paid to the converse situation: the effect of alanine on
peptide assay.” Results are expressed as mean r SEM. Results
hepatic ketogenesis. In the rat Ozand et al. have from the two infusions have been compared using the Student’s t
reported an antiketogenic effect of alanine6,’ which we test.
have shown to be a direct inhibition of ketogenesis, in
all probability, secondary to increased availability of RESULTS
oxaloacetate.’ In man investigations have been more Infusion of somatostatin alone caused no change in
difficult, because of the simultaneous effect of alanine alanine concentration but, after a 30 minute time lag a
in promoting glucagong and insulin secretion.” The steady rise in iijl-hydroxybutyrate and acetoacetate
present study was undertaken to establish whether concentration occurred reaching a plateau of approxi-
alanine could affect ketone body levels independent of mately 0.2 and 0.1 mM respectively after 90 min (Fig.
changes in glucagon and insulin in man. Alanine was I). Plasma NEFA levels showed a small initial rise but
therefore infused against a background of somatosta- remained stable thereafter. Blood glycerol levels
tin which simultaneously suppresses glucagon” and showed a similar pattern. A slight decrease in blood
insulin” secretion. An unequivocal hypoketonaemic glucose levels was observed at 15 and 30 min during
effect of alanine is described. somatostatin infusion. Thereafter glucose concentra-
tions rose to a peak at 120 min with a subsequent
MATERIALS AND METHODS slight decrease.
Subjects Plasma insulin and C-peptide levels showed imme-
diate suppression after onset of somatostatin infusion.
Six healthy male subjects, age 29 ? 4 (SEM) yr, 107 + 4% ideal
body weight, were studied, each on two occasions. None were The ratio between C-peptide and insulin plasma
suffering from any disease, liver function tests were normal, and
none were taking any medication. All gave fully informed consent.
From the Istituto di Medicina Clinica, Divisione di Gerontologia
e Malattie del Ricambio. Policlinico Universitb di Padova. Via
Protocol
Giusliniani , 2 35100 Padova. Italy and the Department of Clinical
Studies were commenced at 8 a.m. after an overnight fast. Biochemistry and Metabolic Medicine. Rqval Victoria 1nfirmar.v.
Subjects underwent the two studies in random order. A Venflon Newcastle Upon Tyne NEI 4LP. United Kingdom.
cannula (I7 gauge) was placed in each antecubital vein and a Supported in part by the British Diabetic Association, the British
further indwelling cannula in a forearm vein. In one arm somatosta- Council and Novo Industries. Copenhagen. Denmark.
tin in isotonic saline (Wyeth Lab. Inc., Philadelphia, Pa., U.S.A.) Receivedfor publication June 12, 1980.
was infused at a rate of 200 rg/hr (9.8 ml/hr) through one Address reprint requests to R. Nosadini. Istituto di Medicina
cannula from Ok 180 min after two basal blood samples were drawn, Clinica. Divisione di Gerontologia e Malattie del Ricambio, Poli-
and saline (0.154 M) or alanine from 60-180 min through the clinic0 Universitir di Padova, Via Giustiniani. 2 35100 Padova,
second cannula. A loading dose of L-(+)-alanine (10% w/v, in Italy.
water) of 0.03 mmole/kg body weight was given in 2 min. followed ‘: 1981 by Grune & Stratton, Inc.
by a continuous infusion of 0.01 mmole/kg/min. Free flowing 0026-n495/81/3006~006$01.00/0

Mefabolrsm, Vol. 30. No. 6 (June). 1981 563

564
NOSADINI ET AL.

I L C*I ALANINE 0.01 mmat. kci’ mm-’

Fig. 1. Plasma NEFA, blood

alanine. glycerol. &hydroxybu-
tyrate. acetoacetate concen-
trations and B-hydroxybuty-
rete/acetoacetate fj3OHBI
AcAc) ratio during somatosta-
d tin plus alanine IO---O) or
f somatostetin plus saline (O-
015
-0) infusions. Alanine (0.01
f f 0.10 mmol/kg body weight/mini or
saline were infused from 60 to
4 1 005
180 minutes while somatosta-
m om I tin (200 Ag/hr) was infused
from 0 to 160 minutes follow-
ing an overnight fast.

l*p < 0.01

lp < 0.05 I
comparing ketone body values
in the presence or absence of
alanine using the Student
paired t test. Results are
TIME tmin.3 expressed as mean + SEM.
THE ANTIKETOGENIC EFFECT OF ALANINE 565

Table 1. Mean NM) Blood Metabolite and Plasma Hormone Levels During Sometostetin Plus Alanine or Saline Infusions. Somatostatin
(SRIF) was Infused at 200 pg/hr From O-180 min. Saline (0.154) Ml or Li+) Alanine I+ALA) Were Infused From SO-180 min.

Blood Blood Blood Lactate

Time Glucose Lactate Pyruvate
Min mM mM mM Pyruvate Ratio

SRIF SRIF+ALA SRIF SRIF+ALA SRIF SRIF+ALA SRIF SRIF $-ALA

~15 5.35 -+ 0.17 5.28 t 0.19 0.83 f 0.09 0.83 + 0.11 0.08 r 0.02 0.07 + 0.00 1.9 f 1.2 11.9 + 0.6
0 5.26 + 0.27 5.27 + 0.17 0.69 + 0.17 0.79 * 0.13 0.07 + 0.01 0.06 + 0.00 1 1.9 * 1.4 11.9 + 0.8
15 4 94 r 0.23 5.01 + 0.18 0.71 + 0.08 0.81 t 0.07 0.06 + 0.00 0.07 t 0.01 1.7 i- 1.7 12.3 + 1.1
30 5.06 + 0.29 5.10 t 0.51 0.75 f 0.08 0.87 + 0.09 0.07 * 0.00 0.07 -t 0.01 2.2 k 1.8 13.0 t 0.9
60 5.32 r 0.37 5.39 + 0.61 0.69 + 0.05 0.72 + 0.08 0.04 f 0.00 0.05 k 0.00 12.3 1 1.5 12.8 k 0.9
90 5 67 + 0.54 6.07 ?- 0.38 0.71 + 0.07 0.69 t 0.10 0.06 + 0.01 0.05 + 0.01 13.6 _+ 3.3 13.4 +- 0.8
120 6.89 - 0.59 7.01 t 0.45 0.84 -t 0.08 0.91 k 0.12 0.06 f 0.01 0.06 + 0.00 13.9 + 3.2 14.4 t 0.2
150 6.05 + 0.27 6.99’ c 0.25 0.75 + 0.06 0.93 f 0.11 0.06 + 0.00 0.07 t 0.01 13.7 r 2.2 12.9 +- 0.9
180 5.89 i 0.69 6.89 f 0.57 0.89 k 0.10 0.99 + 0.25 0.06 + 0.01 0.07 t 0.01 14.9 t 2.0 13.9 k 0.9

lf’ < 0.05 usmg the student paired t test. For other details see the test.

-__
Plasma Plasma C-Peptide Plasma
Time Insulin C-Peptide - Glucagon
Min nglml nglml Insulin Ratio pg/ml

SRIF SRIF+ALA SRIF SRIF+ALA SRIF SRIF+ALA SRIF SRIF +ALA

-15 0.30 i- 0.08 0.31 + 0.09 1.53 -t 0.36 1.59 + 0.62 5.22 k 1.03 5.13 f 0.99 115 zk 32 109 i 39
0 0.29 i 0.08 0.31 + 0.08 1.54 k 0.33 1.61 + 0.59 5.14 k 1.11 5.19 k 1.13 110 2 27 111 +41
15 -

30 0.19 + 0.07 0.18 + 0.07 0.48 k 0.23 0.54 + 0.21 4.65 + 0.72 4.80 + 0.75 75 k 21 64 + 17
60 0.11 k 0.05 0.10 + 0.05 0.47 * 0.22 0.55 + 0.21 4.92 * 0.71 4.91 k 0.69 56 f- 18 55 + 21
90 0.13 + 0.08 0.12 i 0.04 0.41 t 0.21 0.39 t 0.20 4.99 k 1.02 4.12 + 0.99 54* 12 59 t 19
120 0.10 + 0.07 0.11 + 0.05 0.40 k 0.18 0.37 t 0.12 3.73 k 0.79 3.90 t 0.87 49i 9 44 t 17
150 0.12 + 0.05 0.12 + 0.07 0.41 * 0.17 0.39 + 0.17 3.39 k 0.66 3.08 t 0.59 50 f 11 47i 7
180 0.11 + 0.04 0.12 k 0.07 0.40 _c 0.16 0.40 + 0.13 3.28 + 0.69 3.33 t 0.67 47 _t 12 49i 8

‘P .: 0.05 using the student paired t test. For other details see the test

concentrations tended to decrease slowly throughout gon levels showed no significant differences in the two
all the study period. Plasma glucagon levels were experiments.
suppressed by approximately 50%.
When alanine infusion was commenced 60 min
after somatostatin, alanine levels reached 0.6 mmole/ DISCUSSION
1 30 min later, and then rose further reaching a Using somatostatin to maintain low insulin and
plateau of 1 mmole/ I (Fig. 1). fl-hydroxybutyrate and glucagon levels the results clearly show a marked
acetoacetate levels which had risen sharply before hypoketonaemic effect of alanine with suppression
alanine to 0.096 * 0.015 and 0.055 + 0.014 mmole/ 1 already significant with physiological elevations of
respectively, remained stable at that concentration for alanine concentration. At higher levels of alanine both
a further hour and then fell towards the end of the P-hydroxybutyrate and acetoacetate were suppressed
alanine infusion. Values were significantly lower than by 60%-80’S There was no difference in plasma
in the somatostatin control experiment throughout the NEFA and blood glycerol levels suggesting that
period of alanine infusion. In contrast, plasma NEFA change in substrate supply is not the cause of the
and blood glycerol levels were similar with and hypoketonaemia.
without alanine. Blood glucose showed a slight There is clear evidence that ketone bodies inhibit
increase during alanine infusion compared with peripheral amino acid4 and protein metabo1ism,3 and
somatostatin alone, while lactate was not significantly Hall et al.” have shown a reciprocal relationship
different as well as pyruvate. Lactate-pyruvate and between alanine turnover and blood ketone body
P-hydroxybutyrate-acetoacetate ratios during somato- levels. Less attention has been paid to possible effects
statin and saline infusion were comparable to those of alanine on ketone body metabolism. Genuth”,”
observed during somatostatin and alanine infusion. suggested an inhibitory effect of alanine on ketone
Insulin, C-peptide, C-peptide/insulin ratio and gluca- body metabolism in man but changes in portal vein
566 NOSADINI ET AL.

insulin and glucagon concentration could have The observation that insulin and C-peptide were
occurred. Ozand et al.6-7 have shown that alanine equally suppressed suggests that the hypoketonaemic
infusion in the rat causes a fall in blood ketone body effect of alanine is a hormone independent reciprocal
levels which they attributed to a change in the distri- negative feedback alanine-ketone body cycle. We have
bution of reducing equivalents in the liver. Also using previously shown that induction of hyperglycaemia in
the rat we have shown a fall in ketone body production diabetic man with relative hypoinsulinaemia causes an
rate with alanine infusion, but could not confirm a increase in alanine levels, ” due probably to a direct
change in redox status’ neither in blood nor inside the effect on peripheral glucose metabolism.22 In this situ-
liver. Our results which could be mimicked by aspar- ation ketone body levels did not rise despite the rela-
tate infusion suggested that increased oxoloacetate tive insulin deficiency. Thus, when alanine and glucose
may be related to the decreased ketogenesis. are plentiful, as after feeding, ketogenesis is
In the present experiments although a change in suppressed independent of hormone action, and fatty
ketone body utilization could not be ruled out the acids are diverted to triglyceride. Conversely in fasting
results are most suggestive of a direct effect on hepatic when glucose and alanine levels are low ketone body
ketone body production. This conclusion is supported levels rise and further inhibit alanine production thus
by previous studies in rats in which alanine caused a preserving extrahepatic protein. In moderately hyper-
72 i 9% decrease in the rate of production of ketone glycaemic diabetes, excessive hyperketonaemia is
bodies whereas their metabolic clearance was prevented, when insulin is not totally deficient, by
unchanged indicating that the primary effect of alan- glucose induced alanine production. In normal man
ine was inhibition of hepatic ketogenesis.’ However, effects of hormones, such as glucagon, insulin and
the hypoketonaemic effect, which we have attributed cortisol will predominate but further modulation of
to alanine, might be mediated by intraportal increase ketogenesis and alanine production is produced by this
of insulin levels, not readily apparent at peripheral substrate cycle.
levels. To rule out this possibility we have measured
simultaneously insulin and C-peptide concentrations, ACKNOWLEDGMENT
since C-peptide, which is not degraded by the liver, is The authors are grateful to Dr. J. M. Burrin, Dr. McCulloch, and
mirroring more closely P-cell secretion. A. Cornell for advice and assistance.

REFERENCES
1. Blackshear PB, Alberti KGMM: Sequential amino acid hypothalamic inhibitor of the endocrine pancreas. Science 184:482-
measurements during experimental diabetic ketoacidosis. Am J 484
Physiol 228:205-21 I, 1975 12. Alberti KGMM, Christensen NJ, Christensen SF, et al:
2. Felig P, Marliss E, Ohman JL, et al: Plasma amino acid levels Inhibition of insulin secretion by somatostatin. Lancet 2: 1299-l 301,
in diabetic ketoacidosis. Diabetes 19:727-728, 1970 1973
3. Palaiologos G, Felig P: Effect of ketone bodies on amino acid 13. Lloyd B, Burrin J, Smythe P, et al: Enzymic Ruorometric
metabolism in isolated rat diaphragm. Biochem J I54:709-716, continuous-flow assays for blood glucose. lactate, pyruvate, alanine,
1976 glycerol and 3-hydroxybutyrate. Clin Chem 24:1724-29. 1978
4. Landaas S: Inhibition of branched chain amino acid degrada- 14. Price CP, Lloyd B. Alberti KGMM: A kinetic spectropho-
tion by ketone bodies. &and J Clin Invest 37:4l l-418, 1977 tometric assay for rapid determination of acetoacetate in blood. Clin
5. Sherwin RS, Hendler RG, Felig P: Effect of ketone infusions Chem 23:1893-97,1979
on amino acid and nitrogen metabolism in man. J Clin Invest 15. Faloona GR, Unger R: Glucagon, in Jaffe BM, Behrmann
22: I 382-90,1975 HR (eds): Methods of hormone radio-immunoassay. New York,
6. Ozand PT. Reed WD. Girard J, et al: Hypoketonaemic effect London, Academic Press, 1974, pp 317-330
of L-alanine. Specific decrease in blood concentrations of 3-hy- 16. Herbert V. Kam Seng L, Gottlieb CW, et al: Coated charcoal
droxybutyrate in the rat. Biochem J 164:557-564, 1977 immunoassay of insulin. J Clin Endocrinol 25:1375-84, 1965
7. Ozand PT. Reed WD, Hawkins RL, et al: Effect of L-alanine
17. Kuzuya T, Matsuda T, Yoshida S: Human C-peptide immu-
infusion on glucogenogenesis and ketogenesis in the rat and in vivo.
noreactivity (CPR) in blood and urine. Evaluation of a radioimmu-
B&hem J 170:583-591, 1978
noassay method and its clinical applications. Diabetologia I2:5 I I --
8. Nosadini R, Datta H, Hodson A, et al: Antiketogenic action of
518.1976
alanine in the rat: possible mechanism. Biochem J 190:323-332,
18. Hall SEH, Braaten JT, McKendry JBR, et al: Normal
1980
alanine-glucose relationships and their changes in diabetic patients
9. Muller WA, Faloona GR, Unger RH: The effect of alanine on
before and after insulin treatment. Diabetes 28:737-745, 1979
glucagon secretion. J Clin Invest 50:2215-28, 1971
IO. Wise JK, Hendler R, Felig P: Evaluation of alpha-cell 19. Genuth SM: Effect of oral administration of alanine in
function by infusion of alanine in normal, diabetic, and obese fasting obese subjects. Metabolism 22:927-934, 1973
subjects. New Engl J Med 288:487-490. 1973 20. Genuth SM. Castro J: Effect of oral alanine on blood
1 I. Koerber DJ, Ruth W, Chideckel E, et al: Somatostatin: beta-hydroxybutyrate and plasma glucose. insulin, free fatty acids,
THE ANTIKETOGENIC EFFECT OF ALANINE 567

and growth hormone in normal and diabetic subjects. Metabolism 22. Liljenquist JE, Shulman Cl, Williams PE, et al: Hypergly-
231375-376, 1974 caemia per se is an important regulator of alanine and lactate
21. Nosadini R, Noy GA, Alberti KGMM, et al: The metabolic metabolism in man: in Waldhaus W, Alberti KGMM (eds): 10th
response to hyperglycaemic clamping in insulin-dependent diabetes. Congress of the International Diabetes Federation. Amsterdam,
Diabetologia 19:13~19, 1961 Excerpta Medica, Int Congr Series 48 1, I4 I, 1980