13C NMR spectroscopy

1
13C NMR spectroscopy

• In 1957 Lauterbur demonstrated the unique
  chemical specificity of 13C NMR (Lauterbur, J
  Chem Phys 26, 217-218, 1957)
• 12C is NMR silent, 13C is spin ½ and NMR visible
• Only 1 of carbon is 13C and its sensitivity is
  low
• Few opportunities for in vivo studies at natural
  abundance
• Labelled substrates required for studies of
  cerebral metabolism

2
13C MRS and cerebral metabolism

• 1986 first application to brain tissue slices
  (Morris et al, Biochem Soc Trans 14, 1720-1721,
  1986)
• 1986 first in vivo application in rabbit brain
  (Behar et al, Magn Reson Med 3, 911-920, 1986)
• 1991 first application to human brain (Beckman
  et al, Biochemistry 30, 6362-6366, 1991)
• 1994 measurement of incorporation of label from
  glucose into glutamate, glutamine and aspartate
  (Gruetter et al, J Neurochem 63, 1377-1385, 1994)

3
Cerebral glucose metabolism
Glucose
Glutamate C4
GLYCOLYSIS
Glutamine C4
1
Gln C2
lactate
Glu C2
Glu C3
Gln C3
3
pyruvate
citrate
Glutamate
TCA
4
a-keto- glutarate
time of infusion
oxalo- acetate
4
succinate
Glutamine
4
TCA cycle redistribution of label
Glucose
GLYCOLYSIS
1
lactate
3
pyruvate
citrate
Glutamate
TCA
4
3
a-keto- glutarate
2
oxalo- acetate
4
3
2
succinate
Glutamine
5
Glutamate isotopomers
(Gruetter et al, J Neurochem 63, 1377-1385, 1994)
6
Determination of human TCA cycle rates
7
1-13C Glucose infusion
Blood sampling
Hot box
Magnet bore
Water lines
Infusion line
Hot box water pump
8
Infusion protocol
10
Infusion Profile (arbitrary scale)
5
99 1-13Cglucose
glucose
99 Bolus
NMR
NMR (baseline)
Shim
MRI
10 20 30 40
50 60 70
80 90 100 110
Time (min)
9
1H / 13C coil design
2 L NaCl
Plastic former
3 cm diam. glass sphere 1 M 1-13C glucose 1 M
2-13C acetate
Proton coil 14 cm diam.
Carbon surface coil 7 cm diam.
(Adriany and Gruetter, JMR 125, 178-184, 1997)
10
1H / 13C coils 1H coils
11
Acquisition sequence
phase cycling x -x y -y
13C RF
WALTZ-16 decoupling (120 ms)
1H RF
TR 3 s
NOE generation (538 ms WALTZ-16)
512 data points
Acquisition
Surface gradient
12
13C spectra of human brain during infusion of
1-13C glucose

• In vivo decoupled 13C NMR spectra at 3 Tesla
• Glucose C1 and glutamate C4 data used to
  determine FTCA

13
Metabolic modelling

• Developed by
• Mason et al (1992, 1995)
• Yu et al (1999)
• Enables determination of rates for
• Glucose utilisation
• TCA cycle
• ?-Ketoglutarate/Glutamate exchange
• Glutamine synthesis

14
Single compartment model
15
Metabolic modelling

• Several assumptions can be used to simplify model
• Differential equations used to describe
  mathematically flow of label through metabolite
  pools

Mason et al, J Cereb Blood Flow and Metab 12,
434-447, 1992 Mason et al, J Cereb Blood Flow
and Metab 15, 12-15, 1995 Yu et al, Biophys J
69, 2090-2102, 1995 Yu et al, Am J Physiol 41,
C2037-C2048, 1997
16
Glutamate and glutamine fractional enrichments
? GLC/2, ? GLUC4, X GLUC2, GLNC4, ?
GLNC2 Model fits for glu and gln shown as solid
and dashed lines respectively (Chhina et al, J
Neurosci Res 66, 737-746, 2001)
17
Human TCA cycle rates

• FTCA measured by 13C MRS
• 0.5 mmolmin-1g-1 (n1, Rothman et al
  1992)
• 0.73 0.19 mmolmin-1g-1 (n5, Mason et al 1995)
• 0.77 0.07 mmolmin-1g-1 (n6, Shen et al 1999)
  
• 0.75 0.08 mmolmin-1g-1 (n3, Chhina et al
  2001)
• 0.7 0.2 mmolmin-1g-1 (n5, Halliday 2003)

• CMRglc measured by PET
• 0.42 mmolmin-1g-1 (Fox et al 1998)
• 0.3 mmolmin-1g-1 (Rosenberg et al 1990)
• FTCA approx twice CMRglc

18
Effect of activation
19
Link between brain activity and BOLD response
fMRI BOLD intensity
Relationship between fMRI signal and rate of
energy processes
Blood Magnetic Susceptibility Effects
CBV
CBF
CMRO2
Relationship between energetic processes and
activity of neuronal processes
CMRglc
Neuronal Activity
20
Measurement of human TCA cycle rates during
visual activation

• Investigation using 13C NMR at 3 Tesla
• direct detection of 13C nucleus in GLUC4, GLNC4,
  GLUC2 and GLNC2 over whole of visual cortex
• control period (25 mins) followed by active
  period (30-40 mins)
• red LED goggles flashing at 8 Hz for activation
  period
• formate as external standard

Chhina et al, J Neurosci Res 66, 737-746, 2001
21
Functional MRI and 13C MRS during visual
activation
Visual Cortex
Centre of 13C Coil
Chhina et al, J Neurosci Res 66, 737-746, 2001
22
Glutamate and glutamine fractional enrichments
durng activation
? GLC/2, ? GLUC4, X GLUC2, GLNC4, ?
GLNC2 Model fits for glu and gln shown as solid
and dashed lines, respectively (Chhina et al, J
Neurosci Res 66, 737-746, 2001)
23
FTCA and Fcyc rates
24
Oxidative uncoupling during activation?

• PET measurements of CMRglc and CMRO2 show that
  glucose metabolism is oxidative under basal
  conditions
• On visual activation, PET shows an increase in
  CMRglc of 51 but only 5 for CMRO2
• Suggests that glucose metabolism and oxygen
  consumption are uncoupled during activation
• Fox et al calculate that 91 of DCMRglc is non
  oxidative

Fox et al, Proc Natl Acad Sci USA 83, 1140-1144,
1986 Fox et al, Science 241, 462-464, 1988
25
Oxidative metabolism during cerebral activation?

• 13C MRS studies show increase in oxidative
  glucose metabolism
• No evidence of label in LACC3
• Small amount of newly synthesised glycogen

Chhina et al, J Neurosci Res 66, 737-746, 2001
26
Oxidative metabolism during cerebral activation?

• 13C NMR spectroscopic studies suggest that
  glucose metabolism is essentially oxidative, even
  during strong activation of the visual cortex
• Situation may be different in older subjects or
  diseased states
• Large increase in CMRO2 on forepaw stimulation in
  anaesthetised rats provides further support for
  oxidative glucose metabolism (Hyder et al, Proc
  Natl Acad Sci USA 93, 7612-7617, 1996)

27
Neurotransmitter cycling and energy metabolism

• Early work on brain slices confirmed two
  compartments
• Glucose supplies both pools, acetate supplies
  only glial pool (Badar-Goffer et al, Biochem J
  266, 133-139, 1990)
• Glial pool 15-20 of total cerebral FTCA (Bluml
  et al, NMR in Biomed 15, 1-5, 2002)
• Can study neuronal glial interaction, and their
  relationships with neuronal energetics (McLean et
  al, Proc 12th Ann Meeting Soc Magn Reson Med 511,
  1993)
• Neurotransmitter cycling and energy metabolism
  could be tightly coupled (Magistretti et al,
  Science 283, 496-497, 1999)

28
Neurotransmitter cycling and energy metabolism
NEURONS
GLIA
GLUCOSE
pyruvate / lactate
pyruvate / lactate
aKG
OAA
aKG
OAA
glutamate
glutamate
aspartate
aspartate
glutamine
GABA
glutamine
29
Neurotransmitter cycling rates

• Human measurements under basal conditions
• 0.47 mmolmin-1g-1 (n4, Mason et al 1995)
• 0.4 0.1 mmolmin-1g-1 (C2, n4, Gruetter et al
  1998)
• 0.18 0.03 mmolmin-1g-1 (C3, n4, Gruetter et
  al 1998)
• 0.32 0.05 mmolmin-1g-1 (n4, Shen et al 1999)
• 0.25 0.08 mmolmin-1g-1 (n3, Chhina et al
  2001)
• 0.32 0.07 mmolmin-1g-1 (n8, Lebon et al 2002)

30
Neurotransmitter cycling and energy metabolism

• Amount of energy required in various signalling
  processes in grey matter of rat cortex (Attwell
  et al, J Cereb Blood Flow and Metab 21,
  1133-1145, 2001)
• 47 Action potential
• 34 Postsynaptic glutamate effects
• 13 Maintenance of membrane potential
• 3 Glutamate recycling
• Implies some proportionality in relationship
  between FTCA and Fcyc, but not as close as
  originally proposed

31
Glycogen
32
13C Parameters

• 13C - 1H pulse sequence
• WALTZ-8 decoupling (766Hz bandwidth)
• TR 360ms data points 512
• 3,000 acquisitions in 18 minutes
• SAR was within the recommended NRPB safety
  guidelines

33
Volunteers
34
Protocol
MR Spectroscopy
0 1 2 4 5
6 8 (hours)
meal
meal
190.5g carbohydrate, 41.0g fat, 28.8g protein,
1253 kcal
35
Changes in muscle glycogen
Glycogen (mmol/l)
Carey et al 2002
36
Protocol
MR Spectroscopy
Calf Muscle
Liver
0 2 4 5
6 8 24
Time (hours)
Breakfast 3g Labelled Lipid
Meal
141.8g carbohydrate, 24.6g fat, 15.4g protein,
1232 kcal
101.5g carbohydrate, 21.1g fat, 30.5g protein,
733kcal
37
Liver Spectra

• Unsaturated Carbons Saturated
  Carbons

38
13C Lipid Turnover in Liver
39
Postprandial changes in labelled triglycerides in
liver and skeletal muscle
Postprandial incremental change in liver
triglyceride (left) and skeletal muscle
triglyceride (right) in control (dashed line) and
diabetic subjects (solid line). Data are shown as
mean ?SE.
40
Summary

• Rapid incorporation of 13C enriched fatty acids
  into liver (maximum at 6 hours plt0.01). Label
  quickly displaced by subsequent unlabelled meals
• Consistent with presence of high level of 13C
  label in VLDL triglyceride at 8 hours
• Substantial 13C level of VLDL throughout the 24
  hours due to recycling of fatty acids
• Modest uptake of labelled lipid in skeletal muscle