Simultaneous fMRI and metabolic MRS of hyperpolarized [1‐13C]pyruvate during nicotine stimulus in rat

Functional MRI (fMRI) and MRS (fMRS) can be used to noninvasively map cerebral activation and metabolism. Recently, hyperpolarized 13C spectroscopy and metabolic imaging have provided an alternative approach to assess metabolism. In this study, we combined 1H fMRI and hyperpolarized [1‐13C]pyruvate MRS to compare cerebral blood oxygenation level‐dependent (BOLD) response and real‐time cerebral metabolism, as assessed with lactate and bicarbonate labelling, during nicotine stimulation. Simultaneous 1H fMRI (multislice gradient echo echo‐planar imaging) and 13C spectroscopic (single slice pulse‐acquire) data were collected in urethane‐anaesthetized female Sprague–Dawley rats (n = 12) at 9.4 T. Animals received an intravenous (i.v.) injection of either nicotine (stimulus; 88 μg/kg, n = 7, or 300 μg/kg, n = 5) or 0.9% saline (matching volume), followed by hyperpolarized [1‐13C]pyruvate injection 60 s later. Three hours later, a second injection was administered: the animals that had previously received saline were injected with nicotine and vice versa, both followed by another hyperpolarized [1‐13C]pyruvate i.v. injection 60 s later. The low‐dose (88 μg/kg) nicotine injection led to a 12% ± 4% (n = 7, t‐test, p ~ 0.0006 (t‐value −5.8, degrees of freedom 6), Wilcoxon p ~ 0.0078 (test statistic 0)) increase in BOLD signal. At the same time, an increase in 13C‐bicarbonate signal was seen in four out of six animals. Bicarbonate‐to‐total carbon ratios were 0.010 ± 0.004 and 0.018 ± 0.010 (n = 6, t‐test, p ~ 0.03 (t‐value −2.3, degrees of freedom 5), Wilcoxon p ~ 0.08 (test statistic 3)) for saline and nicotine experiments, respectively. No increase in the lactate signal was seen; lactate‐to‐total carbon was 0.16 ± 0.02 after both injections. The high (300 μg/kg) nicotine dose (n = 5) caused highly variable BOLD and metabolic responses, possibly due to the apparent respiratory distress. Simultaneous detection of 1H fMRI and hyperpolarized 13C‐MRS is feasible. A comparison of metabolic response between control and stimulated states showed differences in bicarbonate signal, implying that the hyperpolarization technique could offer complimentary information on brain activation.


INTRODUCTION
Functional MRI (fMRI) is one of the most well established pre-clinical and clinical research methods to noninvasively map cerebral activation and connectivity in the resting state or during a task or stimulus. 1,2It is based on the local blood oxygenation level-dependent (BOLD) signal changes due to haemodynamic response, thuss giving an indirect readout of the brain function. 1Potentially, a more direct readout of brain metabolism can be achieved using functional MRS (fMRS), which allows direct detection of several cellular metabolites, such as lactate, glutamate and GABA (gamma-aminobutyric acid), in steady or dynamic state. 3While technically demanding, it can offer noninvasive metabolic details behind the BOLD effect.A combination of fMRI and fMRS can give an even more detailed view of brain activation. 46][7] The hyperpolarization via dissolution dynamic nuclear polarization (dDNP) method can dramatically increase the nuclear spin polarization, and thus signal intensity, of 13 C-labelled molecules, making their rapid detection feasible. 6Upon [1- 13 C]pyruvate injection, labelling of target metabolites, most importantly lactate and bicarbonate, is driven by intracellular enzymatic reactions, suggesting that it could be used to measure cellular events directly. 8The technique has become a powerful nonradioactive tool that allows, in near real time, the monitoring of biological processes relevant to human diseases, in a pathway-specific manner, 8 and can also allow real-time metabolic imaging with whole brain coverage during a short lifetime of hyperpolarized signal. 9dDNP MRS has been widely used in basic and clinical research to study cancer and cardiac diseases. 8,10,11With the advent of human-scale studies, there has been an increasing interest in studying the brain. 5Consistent lactate and bicarbonate signals can be observed in the brain after injecting [1- 13 C]pyruvate, suggesting that alterations due to disease, for example, can be detected. 12However, the acute brain metabolic response to a stimulus, combining functional and hyperpolarized metabolic imaging simultaneously, remains unexplored.
Nicotine can be used as a cerebral stimulant in fMRI, since it causes a strong BOLD effect which is also dose dependent. 13,14Several studies have demonstrated that the physiological nicotinic impact in the brain is mediated through binding to, and activation of, nicotine acetylcholine receptors (nAChRs). 14,15The nAChRs are a family of excitatory cationic ion channels, divided into 17 identified pentameric subunits, broadly expressed in the central nervous system (CNS) and periphery in animals and humans. 15,16The experimental use of nicotine triggers the release of several neurotransmitters and affects the electrophysiological properties of postsynaptic neurons, inducing neuronal membrane depolarization.This depolarization leads to brain activation, especially in brain regions with high nAChR expression levels (for example, the prefrontal cortex). 17,18rthermore, neural activity has a verified connection to haemodynamic changes, via locally increased glucose utilization on nicotine binding sites. 17,19In addition to BOLD, nicotine also increases local field potential (LFP), cerebral blood flow (CBF), cerebral blood volume (CBV) and mean arterial blood pressure (MABP); all these effects have the most potent response within 1-2 min. 13,20,21 this study, we investigated the cerebral metabolic response in rats, under urethane anaesthesia, to a pharmacological brain stimulation induced by the systemic administration of nicotine, or saline as a control, using a simultaneous detection of 1 H fMRI and 13 C-MRS of hyperpolarized [1-13 C]pyruvate.The aim was to compare the BOLD response and metabolic changes upon stimulus, to assess the suitability of dDNP MRS technique for functional studies.Nicotine was selected due to its robust effect on fMRI.

Animal preparation
Twelve adult female Sprague-Dawley rats (Envigo Laboratories B. V.; w = 263-346 g) were used.The animals were maintained on a 12/12 h light-dark cycle at 22 ± 2 C with 50%-60% humidity.Food and water were available ad libitum, including before anaesthesia.The experiments were all done approximately at the same time of the light cycle phase, so that the animals would be in the same state of alertness.All animal procedures were approved by the Animal Experiment Board in Finland and conducted in accordance with European Commission Directive 2010/63/ EU guidelines (license number R-ESAVI-2020-006286).
On the day of the experiment, the animals were transferred to the experimental room in their home box and allowed to acclimatize for at least 20 min, before the procedure was started.The rats were initially anaesthetized with isoflurane (Baxter, Lessines, Belgium), 5% for induction and 2% for maintenance, in a carrier N 2 /O 2 gas (70/30) mixture via a snout mask.Thereafter, urethane (Sigma-Aldrich, Espoo, Finland) anaesthesia was induced by 1250 mg/kg intraperitoneal injection in three portions, 5 min between each injection, under gradually decreasing isoflurane anaesthesia.The isoflurane level was first lowered to 1%, followed 0.5% and finally to 0% upon the third portion of urethane administration.The chosen protocol for anaesthesia was based on the previous results from pharmacological fMRI studies, 2,20 which demonstrated that the combination of urethane anaesthesia with BOLD contrast acquisition can offer a robust practice for fMRI studies using nicotine as a stimulus.After switching the anaesthetic and complete induction of the urethane anaesthesia (monitored by tail and paw withdrawal verification by pinching, and lack of blink reflex), the animals were submitted to a surgery for femoral vein and artery catheterization.Under aseptic conditions, and keeping the animal warm using a heating pad, a catheter (BD Intramedic PE-50, Franklin Lakes, NJ, USA) was inserted into the femoral vein for drug and pyruvate administration.A second catheter (BD Intramedic PE-10, Franklin Lakes, NJ, USA), was inserted into the femoral artery for blood sampling used in blood gas analysis (i-STAT machine Model 300, Abbott Point of Care Inc., Princeton, NJ, USA).The animals were placed inside the magnet using a holder equipped with ear and bite bars.The animals were kept at $37 C using circulating warm water.The temperature and breathing rate of the animals were monitored during the experiments.Blood samples were collected $20 min prior to injection of saline/nicotine and pyruvate (see Section 0) to assess the overall physiological status of the animals.

Nicotine preparation and administration
Nicotine hydrogen tartrate salt (Sigma-Aldrich, Espoo, Finland), a nonselective nAChR agonist, was dissolved at 0.5 mg/mL into saline 0.9% (Natriumklorid Braun 9 mg/mL, B. Braun Medical, Helsinki, Finland), and pH was adjusted with sodium hydroxide (NaOH) to $7.5. 14During the experiments, the animals received either a nicotine dose of 88 μg/kg free base (0.25 mg/kg salt form, n = 7) or a nicotine dose of 300 μg/kg free base (0.85 mg/kg salt form, n = 5) in a volume based on rat body weight, or a corresponding dose of saline as a control (0.5 mL/kg of body weight).20]22 The schedule of the injections is shown in Figure 1.First each animal received either a nicotine dose or saline injection 60 s before hyperpolarized [1-13 C]pyruvate injection.Three hours later, the experiment was repeated, with animals that previously received saline receiving nicotine and vice versa, 60 s before [1-13 C]pyruvate injection (n = 12; in one animal with low nicotine dose 13 C data collection failed).The injection timing was based on the estimations of the strongest nicotine response 13,20,21 .Additionally, a 1 mL bolus of saline was given between test injections, to maintain hydration levels.In one animal, a failure of pyruvate dissolution required a second low nicotine dose and pyruvate experiment to be performed about 3 h after a previous nicotine injection.
The 3 h interval between the injections in the same animal was based on the nicotine pharmacokinetic information, in order to have extra time beyond the half-life of the nicotine in the brain. 23I G U R E 1 A schematic illustration of the experimental protocol.Copenhagen, Denmark).The sample was dissolved using 13 mL of 200 mM TRIS (tris(hydroxymethyl)aminomethane) buffer (pH 7.3) containing 0.1 g/L EDTA (ethylenediaminetetraacetic acid), and neutralized with 434 μL of 2 M NaOH to yield a $160 mM pyruvate solution with a pH of $7.5 and 20%-30% polarization levels, calculated from the ratio between the hyperpolarized signal and the corresponding thermal equilibrium signal observed in separate in vitro experiments.The hyperpolarized [1-13 C]pyruvate dose (0.8 mmol/kg) was injected i.v.into the animals at a rate of 10 mL/min. 7 1 HfMRI and MRS MR data were collected at 9.4 T (DirectDrive2 console; Agilent, Palo Alto, CA, USA) using a 1 H/ 13 C transmit/receive surface coil (23 mm; Neos Biotec, Pamplona, Spain).The pulse sequence was developed based on the concept suggested by Gordon et al. 24 and consisted of an fMRI (GE-EPI (gradient echo echo-planar imaging), nominal flip angle 60 , echo time, T E , 20 ms, 10 slices of 1.5 mm, FOV (field of view) 32 Â 16 mm 2 , matrix 80 Â 40) block followed by one slice-selective 13 C MRS spectrum (T E = 670 μs, slice 14 mm covering the same region as fMRI slices, nominal flip angle 25 at level of hippocampus) with one fMRI/MRS cycle taking 2 s.Fifteen minutes of data were collected, with the nicotine/saline injection performed at 5 min and the pyruvate injection at 6 min.

Data analysis
fMRI data were checked for motion and slices showing visible artefacts were rejected from further analysis.Regions of interest covering cortical, subcortical and muscle regions were analysed for fMRI.The signal time courses were baseline-corrected by fitting a line to the first 5 min of the data and subtracting it from the data.Since the aim of the paper was to compare fMRI response with fMRS collected from a slice-selective spectrum, no extensive pixel-level fMRI analysis was performed.
For 13 C, the first 90 s following pyruvate injection were summed and phase-and baseline-corrected peak integrals were calculated.
Metabolite-to-total carbon and bicarbonate-to-lactate ratios were calculated.It is worth noting that the labelling of cerebral metabolites is normally relatively modest, so the majority of carbon signal is due to the pyruvate peak.The data were analysed using both Student's t-test assuming equal group variances and, because normality cannot be guaranteed in small sample numbers, Wilcoxon's signed-rank paired test.The results were reported as mean ± standard deviation (SD) and p < 0.05 were taken as statistically significant.

RESULTS
The data produced using interleaved 1 H and 13 C sequences were of a quality similar to those collected using either fMRI or 13 C MRS alone.The blood gases did not vary significantly during the experiments and were within normal physiological range (Table 1).Upon injection of nicotine at 88 μg/kg, a transient increase in breathing rate was observed.Concurrently, a marked increase in cortical BOLD signal was observed (12% ± 4%, n = 7, t-test, p $ 0.0006 (t-value À5.8, degrees of freedom 6), Wilcoxon p $ 0.0078 (test statistic 0)) (Figure 2).In contrast, there were no increases observed for either subcortical or intramuscular signals.A control injection of saline (0.9%; 1 mL/kg) did not change the BOLD signal (Figure 2).The order of injections (saline/nicotine or nicotine/saline) did not cause a difference in the observed BOLD responses (p > 0.5 when comparing response between animals receiving nicotine first, n = 4, and animals receiving saline first, n = 3).However, two animals showed more T A B L E 1 The pH and arterial blood gases measured before the saline or nicotine injections.sudden signal variation in BOLD images after nicotine injection.The SD of BOLD signal variation in the time period from 1 to 2 min after nicotine injection, that is, the time period when 13 C is collected, was around 1% in four animals and more than 10% in the other two animals, suggesting a different physiological response to nicotine injection.Increased BOLD signal was also retained for longer in these two animals, whereas in the other four animals BOLD signal was returning towards the baseline during the 10 min observation window.
All the expected resonances were observed after injection of [1- 13 C]pyruvate (Figure 3A).There was no difference in overall carbon signal levels between spectra collected after saline and nicotine (88 μg/kg) injections.Likewise, the linewidths were similar between experiments.A comparison of spectra following either saline or nicotine injection showed that, while there was no change in lactate signal, the bicarbonate signal was elevated in four out of six animals (Figure 3B-D).The animals with no change in bicarbonate matched those with altered BOLD response.
A 300 μg/kg nicotine dose (n = 5) led to a severe respiratory depression in all the animals, resulting in the death of two rats soon after the nicotine injection.Both negative (n = 2) and positive (n = 1) BOLD responses were observed in the surviving animals.Two animals with negative BOLD showed minimal bicarbonate signal, elevated lactate and increased peak linewidths, whereas the results for animals with positive BOLD resembled those for the low-dose animals (Figure 4).
The data that support the findings of this study are available from the corresponding author upon reasonable request.

DISCUSSION
In the present study, we performed simultaneous fMRI and hyperpolarized 13 C MRS following nicotine stimulus.This approach was selected to allow comparison of the two functional responses, while also confirming successful nicotine stimulus during MRS.According to the previous literature, nicotine produces a robust BOLD effect in brain, which we also observed.We also observed increased bicarbonate labelling in four out of six animals receiving 88 μg/kg nicotine, but no overall increase in lactate labelling, resulting in altered bicarbonate-to-lactate ratios.In contrast, a higher 300 μg/kg dose of nicotine led to severe respiratory depression and variable BOLD and metabolic response.
The labelling of lactate and bicarbonate occurs intracellularly by lactate dehydrogenase and pyruvate dehydrogenase complex + carbonic anhydrase enzymes, respectively.Therefore, it is assumed that these signals better reflect intracellular metabolism, although, especially in the case of lactate, circulating metabolites can also influence the result. 25It is important also to note that metabolite labelling observed in dDNP 13 C experiment does not directly represent underlying concentration but is strongly limited by the pyruvate influx, membrane transport via monocarboxylate transporters (MCTs), 26 and enzymatic activities, and the observed signals are also modulated by T 1 relaxation.
The observed increase in bicarbonate labelling could, therefore, imply increased mitochondrial activity, possibly in neurons 27 1 min after nicotine injection.[21] On the other hand, two animals showed no increased bicarbonate labelling.Both animals showed more artefacts also in BOLD signal after the nicotine injection; thus, it is possible that their response to the nicotine dose was different to the other animals-for example, due to anaesthesia, or biological variance. 16,20,21,28,29The brain nicotine response is complex and can depend on multiple factors and network dynamics, leading to a range of effects on the brain activation and function. 30,31This can be explained by the large number of nAChRs subtypes that are widely expressed throughout the CNS. 15,31The neuronal nAChRs are a superfamily of ligand-gated channel receptors that exist as a peculiar pentameric structure composed of single subunits α (α2-α10) or a combination of α and β (β2-β4) subunits, with different subunit combinations giving variation in their pharmacological and functional properties. 16,31One of the animals with no bicarbonate response also received two nicotine injections due to a failed 13 C experiment, which may further contribute to the different bicarbonate labelling, although BOLD responses were similar after the two nicotine injections.Finally, the observed bicarbonate signal is due to the release of labelled CO 2 , which is rapidly equilibrated with bicarbonate by carbonic anhydrase enzyme based on pH.Therefore, it is possible that alterations in pH could contribute to the signal change, but an Metabolic response after injection of 300 μg/kg nicotine.Higher dose caused severe respiration depression leading to both negative (n = 2) and positive (n = 1) BOLD responses.Increased lactate, loss of bicarbonate and increase of peak linewidth were observed during negative BOLD.
increase in bicarbonate signal would also require an increase in pH.While dDNP techniques can be used to assess pH via bicarbonate/CO 2 ratio, 17,32 the 13 CO 2 signal is too small to be reliably detected in the present experiments.
Interestingly, we did not observe major change in apparent lactate labelling, even though increased lactate levels have been reported during 1 H MRS experiments following stimuli. 33Transient increases (typically around 10%) in lactate have been observed in humans using different stimuli. 3,34,35A recent study combining fMRI and 1 H MRS showed that changes in local cerebral lactate levels occur, for example, during whisker stimulation of awake rats, and the change in lactate is closely associated with BOLD response. 4However, it should be noted that there are highly varying reports on the lactate levels following stimulus. 3The reliable quantification of lactate signal in 1 H MRS can be challenging due to factors such as peak overlap, signal-to-noise ratio, especially in rodents, and variations in acquisition and analysis methods. 36,37Similarly to 1 H, traditional 13 C MRS has given conflicting results. 33,38Because of the limited signal-to-noise ratio, the time scales of MRS are usually also much longer than those of the typical fMRI studies or hyperpolarized experiment used here.Furthermore, in pre-clinical studies the choice of anaesthetics can affect the results. 39In the current study, we chose to use urethane because it provides a long-lasting, stable physiological state and also produces a robust fMRI response. 2,20,40However, we cannot rule out that urethane does not interfere with lactate production upon stimulus, thus explaining the lack of lactate response.To our knowledge, there are no reported studies using a combination of 1 H MRS and acute pharmacological brain stimulation by nicotine injection.A limitation to this study is that our experimental setup did not allow satisfactory spectroscopic quality to be reached for 1 H MRS experiments, so we could not directly investigate cerebral lactate levels during the nicotine experiment.
The physiological statuses of animals were similar immediately before the control and nicotine injections, suggesting that possible alterations in blood lactate labelling are not likely to explain the lactate response.Likewise, possible CBV and CBF effects should influence both lactate and bicarbonate labelling similarly.It would be expected that increased CBV would lead to a higher pyruvate signal in comparison to lactate and bicarbonate and therefore to a reduction in corresponding ratios.Our results with higher nicotine dose suggest that breathing difficulties can lead to markedly increased lactate and reduced bicarbonate signals and that the experiment can detect changes in lactate.On the other hand, cerebral lactate is not a single pool but a dynamic system.Lactate can cross the blood-brain barrier and is also shuttled between astrocytes and neurons based on different cellular distributions of MCTs, with this interaction playing an important role in lactate concentration gradient in the brain and brain function. 4,27This dynamic nature of lactate metabolism impacts the interpretation of MRS signal, as they may reflect acute changes, rather than representing steady-state levels. 41Lactate can be present in at least four different compartments (blood pool, astrocyte, neuron and extracellular space) within the brain. 42While all four pools are potentially visible in 1 H-based lactate measurement, the extracellular space, for example, is not expected to significantly participate in lactate labelling during the dDNP experiment.Assuming a simple presentation of the astroglialneuronal lactate shuttle, 43 the majority of lactate production occurs in astrocytes via glycolysis, and lactate is then transported to neurons to be used for oxidative metabolism in mitochondria.However, as only a fraction of the lactate pool is labelled, an increase in extracellular lactate pool during 1 H MRS, over a time scale of several minutes, may not result in a similar acute change in hyperpolarized lactate labelling.Interestingly, a recent study suggested that dDNP experiments may be directly sensitive to the effects of lactate shuttling, but this is not expected to contribute to the current study due to the low flip angles used. 44As discussed above, unlike the case of the bicarbonate signal, where a CO 2 is released directly because of enzymatic activity and rapidly equilibrated with bicarbonate, the observed lactate signal mainly reflects labelling of the preexisting lactate pool, both signals being heavily modulated by pyruvate uptake. 26Recent reports on mice suggest that hyperpolarized glucose might show greater response to alterations in cellular energy metabolism, due to it entering the cells via more effective glucose transporters and reporting on actual de novo synthesis of lactate. 45,46However, experimental use of hyperpolarized glucose is complicated by short T 1 and has so far been successfully demonstrated only in mice.
The timing of the pyruvate injection may play a role in the observed response.We carefully considered the timing of the nicotine and pyruvate injections.The strongest nicotine responses were estimated to take place 1-2 min after nicotine injection, based on formerly reported time estimations.Paasonen et al. have reported the highest CBF, CBV and BOLD responses after 1 min and observed changes in LFP in the same time period, 20 Gozzi et al. reported the highest MABP and rCBV responses after 1.5 min 21 and King et al. reported the highest BOLD response at $2 min. 13However, local CBV peaks have been reported even after 10 min as individual effects. 47In neural level electrode measurements, the highest activity was observed at 1 min for Type II neurons and at 5 min for Type I neurons. 18A study with radio-labelled glucose showed an elevated glucose uptake at time points of 2 min or later. 19Since the signal of hyperpolarized pyruvate relaxes in about 1 min, we chose to inject it 1 min after the nicotine, to cover for the time when the strongest nicotine response takes place.Nevertheless, we cannot exclude the possibility that a different time point would have allowed the detection of more prominent changes in metabolite labelling.
Nicotine is known to have an anaesthesia-and nicotine-dose-dependent BOLD response. 13,14We chose urethane because it allows longterm stable anaesthesia with only a minor effect on physiology. 48Furthermore, it has been previously shown that a robust nicotine-induced BOLD response is observed under urethane anaesthesia, 20 and the metabolic response of urethane-anaesthetized rats resembled awake rats more than that with isoflurane. 7In the current study, we chose the lowest possible nicotine dose that has shown significant CBF, CBV and BOLD effect; a robust BOLD response was observed using the dose 88 μg/kg. 20Higher doses may have had unwanted side effects on the respiratory system, as we also saw in our experiments with higher dose.Since the half-life of nicotine, in the rat, is 45 min, 23 the gap between experiments was set to 3 h, to avoid residual nicotine effects during the second experiment, and the order of injections was also varied.Similar BOLD responses were obtained regardless of the injection order.29,34 It is worth noting that previous nicotine studies 14,[20][21][22] have all used male Wistar and Sprague-Dawley rats, whereas female Sprague-Dawley rats were used here.This might introduce some differences in the overall response and contribute to the severe respiratory depression observed at the 300 μg/kg dose previously used in several studies. 14,21,22For example, sex hormones can influence pharmacology and behaviour after acute nicotine administration, inducing functional changes in nAChR that could affect nicotine sensitivity, and probably also cerebral metabolic profile. 29Using female rats might also cause some variance in the metabolic data, since the oestrous cycle might slightly change the energy metabolism 49 and cerebral haemodynamics. 50The oestrous cycle has no proven influence on the anaesthetic effect of urethane. 51The similar amplitudes of BOLD responses, although with slightly slower signal recovery, observed here and in the study by Paasonen et al., 20 suggest that the overall effects are similar.
The observed metabolite resonances are relatively small, and the observed effect is modest.In the current study, we chose to analyse baseline-corrected peak integrals because the BOLD effect could lead to small alterations in peak linewidths, as was reported in the highnicotine-dose group (Figure 4).In low-dose experiments, the peak linewidths were similar after saline and nicotine injections.We also used a fixed flip angle scheme and data summation, because this gives a robust readout while still reflecting the underlying labelling process without the need for a full kinetic model.Finally, only the MRS approach was used to avoid the further complexity of an interleaved pulse sequence and to maximize bicarbonate signal-to-noise ratio.However, this may lead to a confounding contribution from signals outside the brain.Previous studies have shown that it is possible to image bicarbonate signal, although great care will be needed when optimizing the experiment. 44Increased pyruvate polarization levels will also help, as polarization levels more than double those achieved here have already been reported. 52

CONCLUSIONS
Taken together, our results suggest that a combination of fMRI and hyperpolarized MRS may offer complementary information on alterations in cerebral activity upon stimulus.Any alterations in bicarbonate signals may be of particular interest, since they should be more directly linked to mitochondrial activity and difficult to measure using other techniques.While rodents and humans have different energy demands and metabolic rates, the results suggest that simultaneous 1H fMRI and hyperpolarized 13 C-MRS of [1- 13 C]pyruvate could open new avenues for future clinical research on normal and impaired cognitive performance in humans.In addition, it could be a promising tool to investigate spectral patterns for neurodegenerative diseases and neuropsychiatric disorders, through a noninvasive, and real-time study of the brain tissue metabolism upon a stimulus.

[ 1 -
13 C]pyruvate hyperpolarization via dDNP [1-13 C]pyruvic acid (80 mg, 14 M, Sigma-Aldrich, Espoo, Finland) with 25 mM radical AH11501 (EPA, GE Healthcare, Chicago, IL, USA) was hyperpolarized via dDNP at 1.35 K, 6.7 T and 188 GHz for 1.5 h in an experimental SpinAligner hyperpolarizer (Technical University of Denmark, B, A representative signal change (BOLD %) in one animal 1 min after injection of saline (0.9%; 0.5 mL/kg, n = 7) or nicotine (88 μg/kg, n = 7): BOLD response (A) and the corresponding average response curves (blue, saline; red, nicotine) (B).The time of pyruvate injection at 6 min is also shown.Shaded areas represent one SD from the average.F I G U R E 3 A, Average summed (60 s)13 C spectra following hyperpolarized [1-13 C]pyruvate injection given 1 min after injection of saline (0.9%; 0.5 mL/kg, n = 6) or nicotine (88 μg/kg, n = 6).B-D, Metabolite responses normalized to total carbon (TC) for bicarbonate (B) and lactate (C) as well as bicarbonate-to-lactate ratio (D).No changes were seen in lactate, but increased HCO 3 /TC was observed in four out of six animals.Shaded areas in A represent one SD from the average.