Available via license: CC BY 4.0
Content may be subject to copyright.
Page 1/26
Hyperpolarized 13C metabolic imaging detects long-
lasting metabolic alterations following mild
repetitive traumatic brain injury
Myriam Chaumeil ( myriam.chaumeil@ucsf.edu )
University of California, San Francisco https://orcid.org/0000-0002-2110-4613
Caroline Guglielmetti
University of California, San Francisco https://orcid.org/0000-0003-0305-1631
Kai Qiao
University of California, San Francisco
Brice Tiret
University of California, San Francisco
Mustafa Ozen
Bay Area Institute of Science, Altos Labs https://orcid.org/0000-0002-7708-6549
Karen Krukowski
Bay Area Institute of Science, Altos Labs
Amber Nolan
University of Washington
Maria Serena Paladini
Bay Area Institute of Science, Altos Labs
Carlos Lopez
Bay Area Institute of Science, Altos Labs
Susanna Rosi
Bay Area Institute of Science, Altos Labs
Article
Keywords:
Posted Date: August 14th, 2023
DOI: https://doi.org/10.21203/rs.3.rs-3166656/v1
License: This work is licensed under a Creative Commons Attribution 4.0 International License.
Read Full License
Page 2/26
Abstract
Career athletes, active military, and head trauma victims are at increased risk for mild repetitive traumatic
brain injury (rTBI), a condition that contributes to the development of epilepsy and neurodegenerative
diseases. Standard clinical imaging fails to identify rTBI-induced lesions, and novel non-invasive
methods are needed. Here, we evaluated if hyperpolarized 13C magnetic resonance spectroscopic
imaging (HP 13C MRSI) could detect long-lasting changes in brain metabolism 3.5 months post-injury in a
rTBI mouse model. Our results show that this metabolic imaging approach can detect changes in cortical
metabolism at that timepoint, whereas multimodal MR imaging did not detect any structural or contrast
alterations. Using Machine Learning, we further show that HP 13C MRSI parameters can help classify rTBI
vs. Sham and predict long-term rTBI-induced behavioral outcomes. Altogether, our study demonstrates
the potential of metabolic imaging to improve detection, classication and outcome prediction of
previously undetected rTBI.
INTRODUCTION
Individuals subject to frequent concussions such as career athletes (e.g. football players, boxers),
accidental head trauma victims, domestic abuse victims, or active military are some of the population at
risk for mild repetitive traumatic brain injury (rTBI). Indeed, rTBI is being steadily recognized as a risk
factor for the development of epilepsy and neurodegenerative diseases, particularly chronic traumatic
encephalopathy (CTE)1, 2, 3. However, to date, non-invasive diagnostic biomarkers of rTBI are lacking. In
particular, clinical computed tomography (CT) and magnetic resonance imaging (MRI), the standard
imaging methods for trauma patients, are unable to detect rTBI-induced pathology4, 5, 6. This lack of
imaging techniques hampers proper diagnosis and appropriate clinical care, as a result new approaches
are critically needed.
To further our understanding of rTBI, several models were developed over the past years7, 8. The closed-
head impact model of engineered rotational acceleration (CHIMERA) device was designed to deliver
multiple subconcussive mild TBI in a controlled and reproducible manner9. CHIMERA-induced rTBI has
been shown to lead to reproducible pathological and behavioral changes up to several months following
impacts10, mirroring the long-term effects of rTBI seen in the clinic (review by McNamara
et al.
11). Only a
few studies have investigated the use of magnetic resonance imaging (MRI) to detect CHIMERA-induced
rTBI. Diffusion MRI, which is sensitive to water diffusion in tissue and changes in tissue microstructure,
detected subtle differences in rTBI animals at 7 days following injury and in the optic tract, brachium of
the superior colliculus, corpus callosum and hippocampus regions12, 13; however, long-lasting changes
were not studied. T2-weighted MRI, which is sensitive to changes in tissue microstructure, edema, and
myelination, did not detect signs of brain injury at 7 days and 40 days post-injury10, 12. All these studies
are in line with clinical ndings, and further highlight the need for more sensitive approaches to detect
and monitor long-term brain changes after rTBI.
Page 3/26
Metabolic impairment following TBI has been well documented in patients and animal models in the
hours following TBI using 13C-labelled substrates infusion and metabolomics approaches (reviews by
Jalloh, Dermers-Marcil, and Carpenter14, 15, 16). Notably, cerebral microdialysis studies have identied that
the lactate / pyruvate ratio (Lac/Pyr) parameter is associated with poor outcome17, 18. However, as
cerebral microdialysis is an invasive method, its use at chronic timepoints following TBI, or in closed
head injury and concussion is not feasible. Hyperpolarized 13C magnetic resonance spectroscopic
imaging (HP 13C MRSI) is a unique technology that allows to measure metabolic uxes
in vivo
, and to
compute lactate / pyruvate ratio values as well. HP 13C MRSI applications have been extensively
described in the oncology eld19, 20, and its use is emerging to probe brain metabolism in health and
diseases21. HP 13C MRSI of [1-13C]pyruvate enables to monitor the conversion of this key metabolite into
its product(s) [1-13C]lactate and/or [13C]bicarbonate in the brain, which provides unprecedented metabolic
information22, 23. Prior studies of moderate TBI have shown changes in the HP 13C lactate / pyruvate
ratio (HP 13C Lac/Pyr), and HP 13C bicarbonate / lactate ratio at early timepoints (4 hours up to 7 days)
following injury in preclinical models and in patients24, 25, 26, demonstrating the potential value of this
technique. Furthermore, recent studies have combined injection of HP [1-13C]pyruvate with HP [13C]urea, a
metabolically inactive probe, to simultaneously evaluate the pyruvate to lactate ux and tissue perfusion,
respectively27, 28, 29, 30, 31. However, it remains to be determined whether HP 13C MRSI of HP [1-
13C]pyruvate might be useful to the detection of rTBI at chronic timepoints, and inform on potential
mechanism of metabolic injury. Furthermore, co-injection of HP [1-13C]pyruvate and HP [13C]urea has
never been tested in any TBI model or patient.
Here, we questioned if advanced imaging approaches that have never been applied to the study of rTBI
could detect long-lasting alterations following rTBI in the CHIMERA model. In addition to the above
described HP 13C MRSI of HP [1-13C]pyruvate and HP [13C]urea, we also evaluated Susceptibility-weighted
imaging (SWI) and T1 mapping. SWI is an MRI method particularly sensitive to iron that can inform on
venous deoxygenated blood and iron deposition in tissue, and which has proven very valuable to identify
microbleeds in TBI, but has not yet been investigated in the CHIMERA model32, 33. Recent reports have
highlighted the potential of T1 mapping to detect oxidative stress in the rodent brain34, and thus this
technique holds great potential to detect the production of reactive oxygen species that may occur
following diffuse axonal injury and inammatory processes observed in the CHIMERA model11.
As shown in Fig.1, we induced rTBI in two-month old male mice using the CHIMERA apparatus, tested
risk-taking behavior 3 months post rTBI, and performed four MRI-based scans. We investigated the
potential of 13C MRSI of HP [1-13C]pyruvate and [13C]urea to detect metabolic and tissue perfusion
impairment, T2-weighted MRI to assess structural changes (as clinical standard of MR imaging), T1-
mapping to evaluate tissue microstructure alterations and oxidative stress, and SWI MRI to detect
changes in tissue microstructure, microbleed and tissue oxygenation following rTBI. Last, brain tissue
was collected to evaluate changes in enzymes activity and transporter protein expression. Given the
Page 4/26
multidimensional nature of the data, we used a Machine Learning (ML) approach to identify how
measured parameters could best predict changes in risk-taking behavior and HP 13C MRSI.
RESULTS
HP 13C MRSI detects long-lasting metabolic alterations
following rTBI
We investigated whether HP 13C MRSI could be used as a non-invasive tool to detect rTBI-induced long-
lasting changes, and specically to question whether metabolic alterations are present at chronic time
points following injury.
Following intravenous co-injection of HP [1-13C]pyruvate and HP [13C]urea, we observed signals from both
substrates, as well as from the metabolic product HP [1-13C]lactate in the cortex (Fig.2.a) and subcortex
(Fig.2.b) of Sham and rTBI mice at 3.5 months post-injury. Upon quantication, we observed signicant
metabolic differences between the cortex of rTBI compared to Sham mice: HP [1-13C]lactate levels were
1.09 fold lower in rTBI (Fig.3.a, p = 0.0073), HP [1-13C]pyruvate levels were 1.05 fold higher (Fig.3.b, p =
0.0073), and HP 13C Lac/Pyr was 1.15 fold lower (Fig.3.c, p = 0.0071). No signicant differences in HP
[13C]urea levels were observed in the cortical area (Fig.3.c), although a trend towards lower HP [13C]urea
levels in rTBI was observed (p = 0.0794). In subcortical areas, we did not detect any differences in HP [1-
13C]lactate, HP [1-13C]pyruvate, HP 13C Lac/Pyr and HP [13C]urea between rTBI and Sham mice (Fig.3.e-
h). In agreement with the metabolic quantications, HP 13C heatmaps clearly show lower HP [1-
13C]lactate, higher HP [1-13C]pyruvate and lower HP 13C Lac/Pyr in the cortex of rTBI mice (Fig.3.i).
Altogether, our results indicate that HP 13C MRSI can detect region-specic long-lasting metabolic
changes following mild rTBI.
Multimodal MRI does not detect long-lasting effect of injury in rTBI
We evaluated whether a comprehensive multimodal MRI approach could detect changes between rTBI
and Sham mice at 3.5 months post-injury, when metabolic alterations where detected by HP 13C MRSI.
We rst used T2-weighted MRI, the clinical standard of MRI approach, that is sensitive to inammation
and/or changes in myelin content. We found that T2 signal intensities were not different between rTBI
and Sham mice in any of the regions studied (prefrontal cortex, cortex, hippocampus and thalamus
(subcortex)) (Fig.4.a-b). In addition, we did not detect any differences in brain region volumes between
groups (Supplementary Fig.1). Next, we used a T1 mapping sequence that was shown to be sensitive to
changes in microstructure or alterations related to oxidative stress. Similar to the T2 intensities, we did
not observed any differences in the T1 values in any of the region studied between rTBI and Sham mice
(Fig.4.c-d). Lastly, a SWI sequence was used, as it is capable of detecting microbleeds as well as
Page 5/26
potential changes in oxygenation following injury. Once again, no differences in SWI values were
observed between rTBI and Sham mice in any region (Fig.4.e-f). We did not detect any microbleed lesions
in any of the studied animals.
Altogether, our results indicate that a comprehensive multimodal MRI approach combining T2-weighted
MRI, T1 mapping and SWI was not able to detect any signs of injury in rTBI mice at 3.5 months post-
injury, unlike HP 13C MRSI.
Disrupted enzymatic activity, but not transporter expression,
is observed at chronic time points after rTBI
To further investigate potential underlying mechanisms responsible for the observed changes in HP 13C
MRSI readouts, we evaluated the activity of enzymes responsible for pyruvate conversion into its
downstream metabolites and the expression of transporters that control the entry of pyruvate into cells
and the eux of metabolites outside of the cells.
In the brain, lactate dehydrogenase (LDH) converts pyruvate into lactate, and pyruvate dehydrogenase
(PDH) controls pyruvate entry into the tricarboxylic cycle and its conversion into acetyl-coA. We observed
that PDH was 1.6 fold lower in the prefrontal cortex and 1.7 fold lower in the cortex of rTBI compared to
Sham mice (Table1, p = 0.0044 and p = 0.0375, respectively). No differences in PDH were observed in
subcortical areas that include the hippocampus and thalamus. The activity of LDH was not signicantly
different between rTBI and Sham mice for cortical and subcortical areas.
Table 1
PDH and LDH enzyme activity.
PDH activity LDH activity
mean ± SD
P-value
mean ± SD
P-value
Prefrontal cortex rTBI 0.0007 ± 0.0001 0.0044 (**) 0.0070 ± 0.0007
0.3345
Sham 0.0011 ± 0.0003 0.0078 ± 0.0020
Cortex rTBI 0.0012 ± 0.0005 0.0375 (*) 0.0022 ± 0.0004
0.5837
Sham 0.002 ± 0.0008 0.0023 ± 0.0003
Hippocampus rTBI 0.0074 ± 0.003
0.4171
0.0020 ± 0.0002
0.1602
Sham 0.0091 ± 0.005 0.0023 ± 0.0005
Thalamus rTBI 0.00013 ± 0.00003
0.6509
0.0023 ± 0.0004
0.1282
Sham 0.00014 ± 0.00006 0.0021 ± 0.0002
Unpaired t-test.
*p ≤ 0.05, **p ≤ 0.01.
Page 6/26
As levels of HP metabolites and HP 13C Lac/Pyr also depend on HP [1-13C]pyruvate intake by cells and on
the eux of HP [1-13C]lactate, we evaluated the expression of monocarboxylate transporters (MCTs).
MCT1 is primarily responsible for pyruvate entry into the cell, and MCT4 is principaly involved in the eux
of lactate outside of the cell. Protein quantication of MCT1 and MCT4 performed for the prefrontal
cortex, cortex, hippocampus and thalamus did not show any differences between rTBI and Sham mice
(Table2), indicating that the differences observed with HP 13C MRSI are likely not due to MCT1 and
MCT4 expression.
Table 2
MCT1 and MCT4 protein expression.
MCT1 MCT4
mean ± SD
P-value
mean ± SD
P-value
Prefrontal cortex rTBI 1.3 ± 0.31
0.0685
1.59 ± 0.72
0.1109
Sham 1 ± 0.32 1 ± 0.54
Cortex rTBI 1 ± 0.14
0.6399
0.72 ± 0.35
0.1502
Sham 1 ± 0.07 1 ± 0.33
Hippocampus rTBI 0.91 ± 0.31
0.5090
1 ± 0.77
0.9147
Sham 1 ± 0.17 1 ± 0.58
Thalamus rTBI 0.89 ± 0.34
0.4891
0.75 ± 0.59
0.3361
Sham 1 ± 0.24 1 ± 0.27
Unpaired t-test.
Altogether, our results indicate that PDH activity is decreased in the cortical areas following rTBI, while
LDH activity, MCT1 and MCT4 expression are not signicantly different compared to Sham at 3.5 months
post-injury.
Machine learning identies rTBI/Sham classiers, and
predictors of behavior and HP 13C readouts
The machine learning (ML) analysis included all the data described above, as well as behavioral data we
previously reported in Krukowski
et al
.35, which showed that mild rTBI leads to increased risk-taking
behavior in male mice at 100 days post-injury.
Given n = 20 (10 for rTBI and 10 for Sham) mice along with 44 measured variables (see Table3 for list of
variables and abbreviations), we used ML to perform two types of analyses. First, we wanted to identify
the best classifying variables allowing for separation of the two groups (rTBI vs Sham). Second, we
aimed to nd the best predictors of changes in risk-taking behavior, as it recapitulates a key behavioral
Page 7/26
component observed in rTBI patients, and of cortical HP 13C Lac/Pyr, due to its potential to serve as a
novel biomarker for long-lasting consequences of rTBI.
Page 8/26
Table 3
List of variables used for ML analyses.
Variable abbreviation Variable description
EPM frequency extreme Number of entries into the extreme zones on the EPM
EPM duration extreme Time (sec) in the extreme zones on the EPM
EPM frequency openandcenter Number of entries in the Open + Center on EPM
EPM duration openandcenter Time (sec) in the open + Center on the EPM
EPM totaldistance centerpoint Total distance traveled on EPM, measured by centerpoint, units =
cm
EPM averagevelocity
centerpoint Average animal velocity on the EPM, units = cm/sec
HP 13C Urea Ctx Hyperpolarized [13C]urea level in cortical area
HP 13C Lac/Pyr Ctx Hyperpolarized lactate/pyruvate ratio in cortical area
HP 13C Urea Subctx Hyperpolarized [13C]urea level in subcortical area
HP 13C Lac/Pyr Subctx Hyperpolarized lactate/pyruvate ratio in subcortical area
PDH Pfc PDH activity in prefrontal cortex
PDH Ctx PDH activity in cortex
PDH Hp PDH activity in hippocampus
PDH Thal PDH activity in thalamus
nT2 Ctx Normalized T2 intensity value in cortex
nT2 Hp Normalized T2 intensity value in hippocampus
nT2 Pfc Normalized T2 intensity value in prefrontal cortex
nT2 Subctx Normalized T2 intensity value in subcortical areas
Volume Ctx Volume of cortex calculted from T2w MRI
Volume Hp Volume of hippocampus calculated from T2w MRI
Volume Pfc Volume of prefrontal cortex calculated from T2w MRI
Volume Subctx Volume of subcortical areas calculated from T2w MRI
Volume Ventricle Volume of ventricles calculated from T2w MRI
Volume Brain Volume of whole brain calculated from T2w MRI
Page 9/26
Variable abbreviation Variable description
T1 Ctx T1 value in cortex
T1 Hp T1 value in hippcampus
T1 Pfc T1 value in prefrontal cortex
T1 Subctx T1 value in subcortex
MCT1 Thal MCT1 expression in thalamus
MCT1 Hp MCT1 expression in hippocampus
MCT1 Ctx MCT1 expression in cortex
MCT1 Pfc MCT1 expression in prefrontal cortex
MCT4 Thal MCT4 expression in thalamus
MCT4 Hp MCT4 expression in hippocampus
MCT4 Ctx MCT4 expression in cortex
MCT4 Pfc MCT4 expression in prefrontal cortex
LDH Thal LDH activity in thalamus
LDH Hp LDH activity in hippocampus
LDH Ctx LDH activity in cortex
LDH Pfc LDH activity in prefrontal cortex
SWI Ctx SWI value in cortex
SWI Hp SWI value in hippcampus
SWI Pfc SWI value in prefrontal cortex
SWI Subctx SWI value in subcortex
Various classication and feature extraction methods were implemented to identify the best classifying
variables between Sham and rTBI mice. Consequently, we identied ve triplets of variables that could
accurately distinguish between either group and ranked them based on their feature importance scores
computed using various feature extraction algorithms (see methods). The top two triplets with high
feature scores are presented in Fig.5.a and the others are shown in Supplementary Fig.2. The top two
triplets are: 1) PDH Pfc, LDH Thal, and LDH Hp, and 2) PDH Ctx, LDH Thal, and HP 13C Lac/Pyr Ctx.
These ndings suggest that although one single feature is not sucient to identify the difference
between rTBI and Sham, the combined PDH and LDH activity, as well as HP 13C imaging readouts can
help distinguish differences between the two conditions. The three lower-tier triplets consisted of the
subsequent variables: 3) PDH Pfc, EPM duration openandcenter, and LDH Thal, 4) PDH Pfc, EPM duration
Page 10/26
openandcenter, and HP 13C Lac/Pyr Ctx, and 5) PDH Pfc, MCT1 Pfc, and HP 13C Lac/Pyr Ctx. These three
last triplets further idenfy the risk-taking behavior and MCT1 expression in the prefrontal cortex as
important variables to classify Sham and rTBI.
Next, we identied the best predictors of the changes in risk-taking behavior and HP 13C Lac/Pyr Ctx
presented in Fig.5.b and 5.c, respectively. Interestingly, four variables, namely EPM frequency
openandcenter, MCT1 Pfc, nT2 Pfc, and HP 13C Lac/Pyr Ctx, were sucient to predict risk-taking behavior
(Fig.5.b bottom panel) with similar accuracy than when all variables were used for prediction (Fig.5.b
top panel). This result shows that a systems approach comprising behavioral, transporter expression,
structural and HP 13C imaging measures yields a tangible prediction for changes in risk-taking behavior.
Similarly, four variables, namely EPM duration openandcenter, HP 13C Lac/Pyr Subctx, PDH Hp, and SWI
Hp, were enough to predict the HP 13C Lac/Pyr Ctx (Fig.5.c bottom panel) with similar accuracy than
when all the variables were used for prediction (Fig.5.c top panel). This result demonstrates that using a
systemic combination of behavioral, enzyme activity, and advanced MRI measures best predicts the HP
13C Lac/Pyr Ctx.
DISCUSSION
In this study, we demonstrated that HP 13C MRSI of [1-13C]pyruvate can detect metabolic changes 3.5
months following rTBI when structural T2 MRI, T1 mapping, SWI MRI and HP 13C MRSI of [13C]urea did
not. Specically, we measured a signicantly lower HP 13C Lac/Pyr in the mouse cortex 3.5 months post-
injury, which was associated with lower PDH activity. Using a ML approach, we further validated that the
HP 13C Lac/Pyr is among the best classiers of rTBI and Sham groups, and is a predictor of the risk-
taking behavior observed in this rTBI model. Our ndings demonstrate the ability of HP 13C MRSI of [1-
13C]pyruvate to detect rTBI-induced damages and highlight promising potential to improve diagnosis and
monitoring of rTBI patients at chronic time points when other imaging techniques are insucient.
Cerebral metabolism has been probed using HP [1-13C]pyruvate both in preclinical and clinical settings in
healthy and diseased brain (see Le Page & al. for recent literature review21). Prior studies performed in
moderate TBI preclinical models of contusion injury have reported increased HP 13C Lac/Pyr at early
timepoints following injury (4 hours up to 7 days following injury), but no signicant changes at chronic
timepoints (28 days post-injury)24, 25. In contrast, in this study we observed a decreased HP 13C Lac/Pyr
at chronic timepoints, suggesting different underlying pathological changes between contusion injury
and rTBI. Studies performed using positron emission tomography (PET) imaging with the glucose
analogue 18F-uorodeoxyglucose (18F-FDG) have detected long-term brain hypometabolism following
TBI4, 5, 36, which is in line with the lower HP 13C Lac/Pyr measured here at chronic timepoints. To the best
of our knowledge, 18F-FDG PET imaging has never been applied to CHIMERA rTBI model. Furthermore, the
use of this method for TBI is limited by ionizing radiations, and the high background of 18F-FDG PET
signal in the brain tissue, hampering the detection of small changes in glucose uptake.
Page 11/26
Decreased HP 13C bicarbonate / lactate ratio was found in moderate TBI models25, and decreased HP 13C
bicarbonate levels were found in TBI patients26, highlighting possible changes in mitochondrial function
and aerobic versus anaerobic respiration following trauma. In agreement with these ndings, a decreased
of PDH activity after injury has been previously described24, 37, including in our current ndings in rTBI. In
this study, we were not able to detect 13C bicarbonate, likely due to the low signal to noise and the fast T1
relaxation rate of this metabolite at ultra-high eld (14.1 Tesla). HP 13C Lac/Pyr can be inuenced by the
activity of LDH and/or PDH, the availability of their cofactors38, 39, 40, as well as by MCTs expression,
where increased MCT1 expression leads to increased HP 13C Lac/Pyr in selected cell lines41. In this study,
we did not nd any signicant differences in MCT1 and MCT4 between rTBI and Sham mice, suggesting
that they do not play a prominent role in the changes observed in the HP 13C Lac/Pyr at this late time
post injury.
It has previously been shown that the blood-brain-barrier (BBB) limits the entry of HP probes, which could
in turn inuence the measured HP 13C Lac/Pyr42, 43. To evaluate potential changes in perfusion and
delivery, we co-injected [1-13C]pyruvate with HP [13C]urea, a metabolically inactive probe. Although
decreased cerebral blood ow alteration has been reported following rTBI44, 45, we did not observe any
differences in HP [13C]urea levels between Sham and rTBI, indicating that perfusion was likely unaltered
at 3.5 months post-injury. To the best of our knowledge, this is the rst study reporting the use of HP
[13C]urea in a rTBI model. Additional studies at earlier timepoints following injury, and in other TBI
models, are needed to evaluate the full potential of HP [13C]urea in detecting BBB alterations and/or
vasculature changes following brain injury.
Conventional and advanced anatomical MRI did not detect any differences between Sham and rTBI at
3.5 months post-injury. In agreement with these ndings, Haber
et al
. and our group previously reported
no differences using T2 MRI at 7 days and 40 days post-injury, respectively10, 12. These results suggest
that conventional T2 MRI may not be able to detect rTBI patholology-induced using the CHIMERA device,
either at early or late timepoints following injury. Using T1 mapping we investigated whether we could
detect changes in tissue microstructure and reactive oxygen species production, but found no differences
between Sham and rTBI34. Oxidative stress and reactive oxygen species have been shown to play an
important role in TBI pathogenesis and in mediating axonal degeneration46, 47, 48. However it remains
unclear if these events may predominantly occur at early timepoints following injury and would have
resolved by the time we performed our imaging study (3.5 months injury), or whether T1 mapping was not
able to detect these events in this rTBI model. HP [1-13C]dehydroascorbic acid (DHA) and HP [1-13C]N-
acetyl cysteine (NAC) have been shown to be sensitive probes to investigate redox changes
in vivo
49, 50,
and therefore represent attractive probes to further interrogate the involvement of oxidative stress using
HP 13C MRSI. We also included SWI MRI exams as this method has been shown to improve the detection
of microbleeds and hemorrhagic diffusive axonal injury after TBI, which was associated with neurologic
decits and long-term outcome in human TBI51, 52. However, we did not detect any differences in SWI MRI
Page 12/26
between Sham and rTBI. As for T1 mapping, it remains to be determined whether no microbleeds or
oxygenation changes occurs in this model at early timepoints, or whether potential changes have
resolved by the time we conducted our imaging exams.
We took a systems approach to measure behavior outcomes using ML analyses. We found that sets of
variables were able to classify rTBI and Sham mice. These variables are linked to cognitive abilities (risk-
taking behavior), metabolism and molecule transport (PDH and LDH activity, HP 13C Lac/Pyr, and MCT1),
highlighting the importance of long-term metabolic impairment in rTBI and suggesting their potential as
injury biomarkers. Interestingly, ML was able to identify that enzymatic changes in the thalamus and
hippocampus regions, which were not statistically signicant using conventional unpaired t-test in
isolation, became important variables to classify rTBI and Sham when considered together. This is in
agreement with changes in hippocampal function that have been previously reported in rTBI up to 6
months post-injury53. Future HP 13C imaging studies will aim to also include the thalamus and
hippocampus to determine whether
in vivo
metabolic changes can be detected in these regions. ML was
also used to determine which variables are best predictors of the risk-taking behavior. We found that the
changes in risk-taking behavior were best predicted by variables from the cortical areas, including HP 13C
Lac/Pyr, MCT1 expression, structural MRI, and behavioral parameters. Similarly, cerebral microdialysis
studies have shown that the Lac/Pyr is an important variable associated to clinical outcome. High
Lac/Pyr within the rst days after injury was associated to poor clinical outcome 6 months later18, while
here we show that a lower HP 13C Lac/Pyr is associated to higher risk-taking behavior 3.5 months after
injury. This discrepancy might be explained by different timing of the Lac/Pyr measurement (days versus
months after injury), and the injury severity (severe TBI versus mild rTBI). Nonetheless, our study provides
further evidence that the Lac/Pyr is a useful marker to predict behavioral outcome, which can now be
measured in a non-invasive manner using HP 13C MRSI, thus opening new avenues to evaluate metabolic
alterations months after trauma. Last, ML was used to determine the best predictors of the HP 13C
Lac/Pyr in the cortex, and identied subcortical variables, including PDH activity, SWI MRI and metabolic
MRSI, as well as risk-taking behavior. Altogether, these results demonstrate the importance of multimodal
approaches to detect rTBI pathology and associated long-lasting changes.
Thanks to the recent efforts of the community to enable easy data sharing though data repository54,
future studies with higher sample size will become possible, which in turn can lead to improvement of our
understanding of biological and functional pathway involved in rTBI, and help identify novel biomarkers.
In summary, our ndings demonstrate the potential of HP [1-13C]pyruvate to detect long-lasting metabolic
alterations in a mouse model of rTBI. In addition, ML identied HP 13C MRSI as a key parameter to predict
long-term rTBI-induced behavioral outcomes. Over the past few years, the use of HP 13C MRSI in clinical
trials worldwide has been rapidly expanding, and the injection of HP [1-13C]pyruvate has proven feasible
and safe, with no reported side effects55. In this study, we were able to measure changes in HP 13C MRS
parameters from two regions, the cortex which is closest to the impact, and the subcortex, which is more
remote. We were not able to differentiate between smaller brain regions (e.g. prefrontal cortex,
Page 13/26
hippocampus, thalamus) due to the large voxel size used in this study relative to the size of the mouse
brain. Current sequences available on clinical scanners can achieve up to 1 cm3 spatial resolution and
cover the entire human brain, thus providing metabolic information from brain areas close and remote to
the site of injury. With the growing availability of the HP 13C MRS technology, our ndings provide a
strong rationale to translate its use in patients suffering from rTBI, with the aim to improve the detection
of rTBI-induced damages, help in understanding metabolic pathways involved in rTBI pathogenesis, and
eventually aid the development of treatment strategies.
METHODS
Animals and rTBI model induction
All animal research was approved by the Institutional Animal Care and Use Committee of the University of
California, San Francisco. Mice were given one week of acclimation and housed with a reversed 12-h
light/12-h dark cycle and provided food and water ad libitum. At 8 weeks of age, mice were randomly
assigned to the rTBI or sham control group. Animals were anesthetized using isourane (2–3%) in
oxygen 1 L/min during the procedure. rTBI animals were subjected to multiple, mild, closed-head injuries
using the CHIMERA device as previously reported10, 35. Briey, rTBI animals were placed supinely into an
angled holding platform without any shaving of the head or incision into the skin so that the head was
level with the piston target hole while aligning the eyes, ears, and nose such that the impact was centered
on the dorsal convexity of the skull, targeting a 5-mm area surrounding bregma. A nose cone delivering
isourane was removed just prior to the impact. Impact was initiated using RealTerm software, which
was connected to a system including air tank, pressure regulator, digital pressure gauge, two-way
solenoid valve, and piston. The impact was administered with a velocity range of 3.9–4.5 m/sec,
resulting in an impact energy of 0.5 J from the 5 mm, 50 g piston10, 35. Animals were moved to an
incubator immediately after the impact and monitored until fully recovered. rTBI animals received an
injury once per day for 5 days with a 24 h interval in between impacts. Five repeated hits were chosen to
specically focus on the effects of repeated exposure to TBI, as athletes, veterans and sometimes trauma
victims are exposed to constant and repeated blows, even without experiencing concussive symptoms.
Sham mice were exposed to the same isourane anesthesia paradigm without sustaining an impact.
Skull fractures, seizures, apnea, or mortality were not observed in any animals, and no animals were
excluded from the study due to injury parameters.
Risk-taking behavioral test
For all behavioral assays, the experimenters were blinded to surgery. Before behavioral analysis, animals
were inspected for gross motor impairments. Animals were inspected for whisker loss, limb immobility
(including grip strength), and eye occlusions. If animals displayed any of these impairments, they were
eliminated from the study. Behavioral tests were recorded and scored using a video tracking and analysis
setup (Ethovision XT 8.5, Noldus Information Technology). If tracking was unsuccessful, videos were
Page 14/26
scored by two individuals blinded to surgery. Risk-taking behavioral phenotype was evaluated using the
Elevated Plus Maze (EPM) at ~ 100 days (3 months) post-injury (counted from the day of the rst injury)
as described previously10, 35. The EPM consists of two exposed, open arms (35 cm) opposite each other
and two enclosed arms (30.5 cm) also across from each other. The four arms are attached to a center
platform (4.5 cm square), and the entire maze is elevated 40 cm off the oor. Bright white lights
illuminated both ends of the open arm. Mice were placed individually onto the center of the maze and
allowed to explore the maze for 5 min, and their activity was recorded. The maze was cleaned with 70%
ethanol between animals. Risk-taking behavior was measured by changes in time spent in the open arms
+ center of the EPM. We also measured the number of entries into the extremes zones on the EPM, the
time spent in the extreme zons on the EPM, the number of entries in the open arms + center on the EPM,
the total distance traveled on EPM measured by centerpoint, and the average animal velocity on the EPM.
These behavioral data were previously reported in Krukowski
et al
.35.
Magnetic resonance imaging
Mice were anesthetized using isourane (1.5-2% in O2) and a 27-gauge catheter was placed in the tail
vein to allow for intravenous (i.v.) injection. Next, animals were placed in a dedicated cylindrical cradle
allowing for reproducible positioning of the mouse head; which was subsequently inserted inside a dual
tune 1H-13C volume coil (ØI = 40 mm) or a single tuned 1H proton coil (ØI = 40 mm) in a 14.1 T vertical MR
system (Agilent Technologies). Respiration rate was continuously monitored through the PC-sam
software interface (SA Instrument, NY, USA).
First, T2-weighted images from the entire brain were acquired for adequate positioning of the grid used
for hyperpolarized 13C acquisitions using the following parameters: repetition time (TR) = 1200 ms, echo
time (TE) = 20 ms, slice thickness 1.8 mm, 2 averages, matrix 256 × 256, eld of view (FOV) 30 × 30 mm².
For HP 13C MRSI acquisitions, 24 µL of [1-13C]pyruvate and 55 µL [13C]urea preparation were co-polarized
using a Hypersense DNP polarizer (Oxford Instruments) for one hour. After dissolution, the HP [1-
13C]pyruvate and [1-13C]urea preparation was rapidly dissolved in isotonic buffer (pH ~ 7) to a nal
concentration of 80 mM and 78 mM, respectively. A nal volume of 300 µL of the HP [1-13C]pyruvate and
[13C]urea solution was then injected i.v. through the tail vein catheter. 2D dynamic chemical shift imaging
13C data were acquired 16 seconds post i.v. injection of the HP [1-13C]pyruvate and [13C]urea solution
using the following parameters: TR = 67 ms, TE = 0.58 ms, spectral width 5000 Hz, 256 points, ip angle
10°; matrix = 16 ×16, eld of view (FOV) = 32 × 32 mm²; slice thickness 4 mm.
Next, for T2-weighted MRI, T1 MRI and SWI acquistions, the dual tune 1H-13C volume coil (ØI = 40 mm)
was removed and replaced by a 1H volume only coil (ØI = 40 mm). T2-weighted MRI was acquired using a
2D fast spin-echo, with effective echo time (TEeff)/TR = 11.80/2006 ms, FOV = 25 × 25 mm2, in 256 × 256
array and 0.5 mm slice thickness. T1-mapping data were acquired using fast spin-echo with inversion
recovery: TEeff/ TR = 7.44/10000 ms, 8 inversion times (TI): 100, 170, 310, 530, 940, 1640, 2900, 5000 ms,
Page 15/26
FOV = 30 × 30 mm2, in 128 × 128 array and 1 mm slice thickness. SWI acquisitions were acquired using
TEeff/TR = 4.64/111.29 ms, FOV = 20 × 20 mm2, in a 256 × 256 array and 0.4 mm slice thickness.
Magnetic resonance imaging data analysis
Brain regions were manually delineated on T1 maps, T2-weighted and SWI magnitude images for each
mouse based on the Allen Adult Mouse Brain atlas (Allen Institute) using the Aedes region of interest
package for MATLAB (Mathworks). For each region, brain volumes were calculated with the T2-weighted
data, as well as the mean T2-weighted values which were normalized to the mean of the cerebrospinal
uid signal from the ventricles as signal value standard. The mean T1 relaxation times was calculated
from T1 maps generated in VNMRj by pixel-wise tting according to (Eq. 1).
Where y is the measured signal from fast spin echo with multiple inversion recovery, and the three t
parameters: relaxation time T1, equilibrium longitudinal magnetization M0, and pre-inversion recovery
longitudinal relaxation M(0). The TI list is used as input for time t.
SWI data were processed as previously described32 with phase images unwrapped by PRELUDE (FSL),
High pass Gaussian ltered with pixel size 32 x 32, and positive phase map scaling used (Eq. 2).
Normalized positive phase map φpos(t), where φ(t) is the ltered, unwrapped phase at location t, and φmax
is the maximum phase of the slice of interest. The positive phase map is a spatial map varying between
zero and one, with higher phase approaching zero and thus increasing contrast on the nal merged SWI
data. The phase map is multiplied with the magnitude image four times to create nal SWI data56. The
mean SWI intensity was calculated for each mouse for each brain region and normalized to the mean of
the Sham for each region, which corresponds to 1.
HP 13C MRS imaging data was analyzed using the in-house SIVIC software
(http://sourceforge.net/apps/trac/sivic/) and custom-built programs written in MATLAB (MATLAB
R2011b, The MathWorks Inc.). The area under the curve (AUC) of HP [1-13C]pyruvate, AUC of HP [1-
13C]lactate and AUC of HP [13C]urea were measured for each voxel and normalized to the noise level. To
account for variations in polarization levels and delivery, HP [1-13C]pyruvate signal and HP [1-13C]lactate
signal were calculated by dividing each value by the sum of HP [1-13C]pyruvate and HP [1-13C]lactate for
Page 16/26
each voxel. HP [13C]urea signal was obtained by normalizing HP [13C]urea signal from each brain voxel to
the sum of [13C]urea signal from tissue surrounding the brain. HP 13C Lac/Pyr was calculated as the ratio
of the AUC. Next, average from voxels containing cortex or subcortex was calculated, and the obtained
mean values were used to evaluate statistical signicance between sham and rTBI groups. Color
heatmaps of HP [1-13C]pyruvate, HP [1-13C]lactate, HP [13C]urea and HP 13C Lac/Pyr were generated
using a linear-based interpolation of the 13C 2D CSI data to the resolution of the anatomical images using
custom-built programs written in MATLAB and SIVIC.
Ex vivo analyses of brain samples
Following the imaging session, mice were transcardially perfused with ice-cold phosphate buffered
saline, and brains were rapidly dissected. Next, prefrontal cortex, cortex, thalamus and hippocampus were
isolated and snap-frozen in liquid nitrogen. Samples were stored at − 80°C until further processing for
activity assays and western blots.
LDH and PDH activities were evaluated using spectrophotometric activity assay kits according to
manufacturer’s guidelines (ab102526 and ab109902; Abcam, respectively), and normalized to protein
concentration determined by the Bradford protein assay method.
For western blot analyses, frozen brain samples were then homogenized with RIPA buffer (Pierce, 89900)
and protease inhibitor Halt TM Protease Inhibitor Cocktail (Thermo Scientic, 1862209) using a
TissueLyser II (Qiagen). Lysate were incubated on ice for 15 minutes and then centrifuged at 14,000 rpm
for 10 min at 4°C. Protein concentration in supernatants was determined using the Bradford protein
assay method. Equal amount of proteins was loaded on Mini-Protean® TGX TM Precast gels 12%
(BioRad, 456–1043) for 34 minutes at 200 V and 400 mA in Tris/Glycine/SDS Buffer, (BioRad, 1610732).
Proteins were transferred onto 0.2 µm PVDF membranes (BioRad, 1704157) using Trans-Blot Turbo
(BioRad) for 7 minutes using the following settings: 1.3 A, 25 V. Next, the membranes were blocked for 1
hour using Tris-buffered saline supplemented with 0.1% Tween 20% and 5% milk (Research Product
International, M17200-500). Membranes were incubated with the primary antibodies: rabbit anti-
MCT1/SLC16A1 pAb (Novus, NBP1-59656, lot C, 1:400), rabbit anti-MCT4 (Bethyl, A304-439A-M, 1:1000),
rabbit mAb anti-beta-tubulin (9F3) (Cell Signaling, 2128S, 1:2000) diluted in Tris-buffered saline
supplemented with 0.1% Tween 20% and 5% milk overnight. An anti-rabbit IgG HRP-linked secondary
antibody (Cell Signaling, 7074S, 1:3000) was used to detect immune-reactive bands using enhanced
chemiluminescence (ECL Western Blotting Substrate, Pierce, 32209) according to the manufacturer
instructions. Quantication of protein bands was done by measuring band intensities using ImageJ
software. MCT1 and MCT4 levels were normalized to beta-tubulin expression and expressed as the levels
relative to the expression of Sham mice, which corresponds to 1.
Machine Learning analyses
The ML pipeline is summarized in Supplementary Fig.3. Preprocessing of the raw experimental data was
performed before ML analyses to (i) scale the measurements for fair comparison and (ii) predict the
Page 17/26
missing measurements. The raw measurements were rescaled using Scikit-learn python library’s standard
scaler57 (see Supplementary Fig.4 for an example of original versus scaled data distributions). For a few
mice, we were not able to measure some of the variables due to tissue isolation (n = 2 rTBI and n = 2
Sham missing for PDH activity, n = 3–4 rTBI and n = 3 Sham missing for LDH activity, MCT1 and MCT4
expression) and low signal to noise of the HP [13C] imaging data (n = 1 rTBI mouse excluded). To impute
the missing measurements we separated the data into rTBI and Sham groups, and within each group the
data was split as training and testing. Mice with all measured variables were used to train a Random
Forrest Regressor (RFR). The trained regressor was then used to predict missing measurements
associated with the testing data.58 To identify the variables that are the best classiers of Sham and rTBI
groups, and to eliminate sensitivity of the results to different methods, we used multiple Scikit-learn
classiers and feature selection algorithims that are neural network, logistic regression, recursive feature
elimination (RFE), SelectKBest, feature importance with ExtraTreesClassier, and LASSO. Using each
method, all the variables were scored based on their importance in the classication of Sham versus rTBI
and a pool of variables with high importance scores was created. From this pool, we created groups of
best classiers with the minimum number of variables, that is three in this case, and ranked those triplets
of variables based on their individual importance scores across different methods. Similarly, to identify
the best predictors of the risk-taking behavior and HP 13C Lac/Pyr in the cortex, we trained multiple
regression algorithms of Scikit-learn library that are LASSO, RFR, Ridge Regression, Support Vector
Regressor, using the measurements of variable to be predicted as the y-values and all the remaining
variables as the x-values (predictors). Then, the variables with high absolute coecients were identied
as the most important features that contribute to the prediction of the regressor. After pooling and
grouping the variables with high feature scores across different algorithms, we observed the best
predictor group with the minimum number of variables.
Statistical analysis
Results are expressed as mean ± standard deviation (SD). Statistical analyses of MRI, behavioral, and
ex
vivo
parameters was performed using unpaired t-test (GraphPad Prism (v 9.1.2), (*p ≤ 0.05, **p ≤ 0.01,
***p ≤ 0.001, ****p ≤ 0.0001).
Declarations
Acknowledgements
This work was supported by research grants: NIH R01NS102156 (MC), NIH NIA grants R01AG056770
(SR) and NIH R21NS096718 (SR, MC), Dana Foundation: The David Mahoney Neuroimaging program
(MC) and the NIH Hyperpolarized MRI Technology Resource Center P41EB013598. K.K. was supported by
an NRSA post-doctoral fellowship from the NIA F32AG054126. AN was supported by the NINDS K08
NS114170.
Author contributions
Page 18/26
M.M.C., and S.R. designed research; C.G., K.Q., B.T., K.K., A.N., and M.S.P performed research; C.G., K.Q.,
B.T., K.K., A.N., M.O., and C.F.L. analyzed data; C.G. and M.M.C. wrote the paper.
Competing interests
The authors declare that they have no competing interests.
References
1. Brett BL, Gardner RC, Godbout J, Dams-O'Connor K, Keene CD. Traumatic Brain Injury and Risk of
Neurodegenerative Disorder.
Biol Psychiatry
91, 498-507 (2022).
2. Ding K, Gupta PK, Diaz-Arrastia R. Epilepsy after Traumatic Brain Injury. In:
Translational Research in
Traumatic Brain Injury
(eds Laskowitz D, Grant G) (2016).
3. Hay J, Johnson VE, Smith DH, Stewart W. Chronic Traumatic Encephalopathy: The
Neuropathological Legacy of Traumatic Brain Injury.
Annu Rev Pathol
11, 21-45 (2016).
4. Ng TS
, et al.
Neuroimaging in repetitive brain trauma.
Alzheimers Res Ther
6, 10 (2014).
5. Asken BM, Rabinovici GD. Identifying degenerative effects of repetitive head trauma with
neuroimaging: a clinically-oriented review.
Acta Neuropathol Commun
9, 96 (2021).
. Toth A. Magnetic Resonance Imaging Application in the Area of Mild and Acute Traumatic Brain
Injury: Implications for Diagnostic Markers? In:
Brain Neurotrauma: Molecular, Neuropsychological,
and Rehabilitation Aspects
(ed Kobeissy FH) (2015).
7. McAteer KM, Turner RJ, Corrigan F. Animal models of chronic traumatic encephalopathy.
Concussion
2, CNC32 (2017).
. Hoogenboom WS, Branch CA, Lipton ML. Animal models of closed-skull, repetitive mild traumatic
brain injury.
Pharmacol Ther
198, 109-122 (2019).
9. Namjoshi DR
, et al.
Merging pathology with biomechanics using CHIMERA (Closed-Head Impact
Model of Engineered Rotational Acceleration): a novel, surgery-free model of traumatic brain injury.
Mol Neurodegener
9, 55 (2014).
10. Nolan A
, et al.
Repeated Mild Head Injury Leads to Wide-Ranging Decits in Higher-Order Cognitive
Functions Associated with the Prefrontal Cortex.
J Neurotrauma
35, 2425-2434 (2018).
11. McNamara EH, Grillakis AA, Tucker LB, McCabe JT. The closed-head impact model of engineered
rotational acceleration (CHIMERA) as an application for traumatic brain injury pre-clinical research: A
status report.
Exp Neurol
333, 113409 (2020).
12. Haber M
, et al.
Dening an Analytic Framework to Evaluate Quantitative MRI Markers of Traumatic
Axonal Injury: Preliminary Results in a Mouse Closed Head Injury Model.
eNeuro
4, (2017).
13. Komlosh ME
, et al.
Using double pulsed-eld gradient MRI to study tissue microstructure in traumatic
brain injury (TBI).
Microporous Mesoporous Mater
269, 156-159 (2018).
14. Jalloh I, Carpenter KL, Helmy A, Carpenter TA, Menon DK, Hutchinson PJ. Glucose metabolism
following human traumatic brain injury: methods of assessment and pathophysiological ndings.
Page 19/26
Metab Brain Dis
30, 615-632 (2015).
15. Demers-Marcil S, Coles JP. Cerebral metabolic derangements following traumatic brain injury.
Curr
Opin Anaesthesiol
35, 562-569 (2022).
1. Carpenter KL
, et al.
(13)C-labelled microdialysis studies of cerebral metabolism in TBI patients.
Eur J
Pharm Sci
57, 87-97 (2014).
17. Guilfoyle MR
, et al.
Characterising the dynamics of cerebral metabolic dysfunction following
traumatic brain injury: A microdialysis study in 619 patients.
PLoS One
16, e0260291 (2021).
1. Timofeev I
, et al.
Cerebral extracellular chemistry and outcome following traumatic brain injury: a
microdialysis study of 223 patients.
Brain
134, 484-494 (2011).
19. Nelson SJ
, et al.
Metabolic imaging of patients with prostate cancer using hyperpolarized [1-(1)
(3)C]pyruvate.
Sci Transl Med
5, 198ra108 (2013).
20. Wang ZJ
, et al.
Hyperpolarized (13)C MRI: State of the Art and Future Directions.
Radiology
291, 273-
284 (2019).
21. Le Page LM, Guglielmetti C, Taglang C, Chaumeil MM. Imaging Brain Metabolism Using
Hyperpolarized (13)C Magnetic Resonance Spectroscopy.
Trends Neurosci
43, 343-354 (2020).
22. Li Y, Vigneron DB, Xu D. Current human brain applications and challenges of dynamic hyperpolarized
carbon-13 labeled pyruvate MR metabolic imaging.
Eur J Nucl Med Mol Imaging
48, 4225-4235
(2021).
23. Hurd RE, Yen YF, Tropp J, Pfefferbaum A, Spielman DM, Mayer D. Cerebral dynamics and metabolism
of hyperpolarized [1-(13)C]pyruvate using time-resolved MR spectroscopic imaging.
J Cereb Blood
Flow Metab
30, 1734-1741 (2010).
24. Guglielmetti C
, et al.
In vivo metabolic imaging of Traumatic Brain Injury.
Sci Rep
7, 17525 (2017).
25. DeVience SJ
, et al.
Metabolic imaging of energy metabolism in traumatic brain injury using
hyperpolarized [1-(13)C]pyruvate.
Sci Rep
7, 1907 (2017).
2. Hackett EP
, et al.
Imaging Acute Metabolic Changes in Patients with Mild Traumatic Brain Injury
Using Hyperpolarized [1-(13)C]Pyruvate.
iScience
23, 101885 (2020).
27. Lau AZ, Miller JJ, Robson MD, Tyler DJ. Simultaneous assessment of cardiac metabolism and
perfusion using copolarized [1-(13) C]pyruvate and (13) C-urea.
Magn Reson Med
77, 151-158
(2017).
2. Qin H
, et al.
Simultaneous Metabolic and Perfusion Imaging Using Hyperpolarized (13)C MRI Can
Evaluate Early and Dose-Dependent Response to Radiation Therapy in a Prostate Cancer Mouse
Model.
Int J Radiat Oncol Biol Phys
107, 887-896 (2020).
29. Qin H
, et al.
Clinical translation of hyperpolarized (13) C pyruvate and urea MRI for simultaneous
metabolic and perfusion imaging.
Magn Reson Med
87, 138-149 (2022).
30. von Morze C
, et al.
Investigating tumor perfusion and metabolism using multiple hyperpolarized
(13)C compounds: HP001, pyruvate and urea.
Magn Reson Imaging
30, 305-311 (2012).
Page 20/26
31. Bok R
, et al.
The Role of Lactate Metabolism in Prostate Cancer Progression and Metastases
Revealed by Dual-Agent Hyperpolarized (13)C MRSI.
Cancers (Basel)
11, (2019).
32. Haacke EM, Xu Y, Cheng YC, Reichenbach JR. Susceptibility weighted imaging (SWI).
Magn Reson
Med
52, 612-618 (2004).
33. Haller S, Vernooij MW, Kuijer JPA, Larsson EM, Jager HR, Barkhof F. Cerebral Microbleeds: Imaging
and Clinical Signicance.
Radiology
287, 11-28 (2018).
34. Tain RW, Scotti AM, Li W, Zhou XJ, Cai K. Imaging short-lived reactive oxygen species (ROS) with
endogenous contrast MRI.
J Magn Reson Imaging
47, 222-229 (2018).
35. Krukowski K
, et al.
Integrated Stress Response Inhibitor Reverses Sex-Dependent Behavioral and Cell-
Specic Decits after Mild Repetitive Head Trauma.
J Neurotrauma
37, 1370-1380 (2020).
3. Koerte IK
, et al.
A review of neuroimaging ndings in repetitive brain trauma.
Brain Pathol
25, 318-
349 (2015).
37. Sharma P, Benford B, Li ZZ, Ling GS. Role of pyruvate dehydrogenase complex in traumatic brain
injury and Measurement of pyruvate dehydrogenase enzyme by dipstick test.
J Emerg Trauma Shock
2, 67-72 (2009).
3. Brindle KM. Imaging metabolism with hyperpolarized (13)C-labeled cell substrates.
J Am Chem Soc
137, 6418-6427 (2015).
39. Day SE
, et al.
Detecting tumor response to treatment using hyperpolarized 13C magnetic resonance
imaging and spectroscopy.
Nat Med
13, 1382-1387 (2007).
40. Witney TH
, et al.
A comparison between radiolabeled uorodeoxyglucose uptake and hyperpolarized
(13)C-labeled pyruvate utilization as methods for detecting tumor response to treatment.
Neoplasia
11, 574-582, 571 p following 582 (2009).
41. Rao Y
, et al.
Hyperpolarized [1-(13)C]pyruvate-to-[1-(13)C]lactate conversion is rate-limited by
monocarboxylate transporter-1 in the plasma membrane.
Proc Natl Acad Sci U S A
117, 22378-22389
(2020).
42. Miller JJ
, et al.
(13)C Pyruvate Transport Across the Blood-Brain Barrier in Preclinical Hyperpolarised
MRI.
Sci Rep
8, 15082 (2018).
43. Hackett EP
, et al.
Probing Cerebral Metabolism with Hyperpolarized (13)C Imaging after Opening the
Blood-Brain Barrier with Focused Ultrasound.
ACS Chem Neurosci
12, 2820-2828 (2021).
44. Clark AL, Weigand AJ, Bangen KJ, Merritt VC, Bondi MW, Delano-Wood L. Repetitive mTBI is
associated with age-related reductions in cerebral blood ow but not cortical thickness.
J Cereb
Blood Flow Metab
41, 431-444 (2021).
45. Ojo JO
, et al.
Chronic Repetitive Mild Traumatic Brain Injury Results in Reduced Cerebral Blood Flow,
Axonal Injury, Gliosis, and Increased T-Tau and Tau Oligomers.
J Neuropathol Exp Neurol
75, 636-655
(2016).
4. Frati A
, et al.
Diffuse Axonal Injury and Oxidative Stress: A Comprehensive Review.
Int J Mol Sci
18,
(2017).
Page 21/26
47. Delage C, Taib T, Mamma C, Lerouet D, Besson VC. Traumatic Brain Injury: An Age-Dependent View of
Post-Traumatic Neuroinammation and Its Treatment.
Pharmaceutics
13, (2021).
4. Ismail H
, et al.
Traumatic Brain Injury: Oxidative Stress and Novel Anti-Oxidants Such as Mitoquinone
and Edaravone.
Antioxidants (Basel)
9, (2020).
49. Keshari KR, Kurhanewicz J, Bok R, Larson PE, Vigneron DB, Wilson DM. Hyperpolarized 13C
dehydroascorbate as an endogenous redox sensor for in vivo metabolic imaging.
Proc Natl Acad Sci
U S A
108, 18606-18611 (2011).
50. Yamamoto K
, et al.
Real-Time insight into in vivo redox status utilizing hyperpolarized [1-(13)C] N-
acetyl cysteine.
Sci Rep
11, 12155 (2021).
51. Park JH, Park SW, Kang SH, Nam TK, Min BK, Hwang SN. Detection of traumatic cerebral
microbleeds by susceptibility-weighted image of MRI.
J Korean Neurosurg Soc
46, 365-369 (2009).
52. Tong KA
, et al.
Diffuse axonal injury in children: clinical correlation with hemorrhagic lesions.
Ann
Neurol
56, 36-50 (2004).
53. Chen H, Desai A, Kim HY. Repetitive Closed-Head Impact Model of Engineered Rotational Acceleration
Induces Long-Term Cognitive Impairments with Persistent Astrogliosis and Microgliosis in Mice.
J
Neurotrauma
34, 2291-2302 (2017).
54. Chou A
, et al.
Empowering Data Sharing and Analytics through the Open Data Commons for
Traumatic Brain Injury Research.
Neurotrauma Rep
3, 139-157 (2022).
55. Vaeggemose M, R FS, Laustsen C. Comprehensive Literature Review of Hyperpolarized Carbon-13
MRI: The Road to Clinical Application.
Metabolites
11, (2021).
5. Haacke EM, Mittal S, Wu Z, Neelavalli J, Cheng YC. Susceptibility-weighted imaging: technical
aspects and clinical applications, part 1.
AJNR Am J Neuroradiol
30, 19-30 (2009).
57. Pedregosa. F
, et al.
Scikit-learn: Machine Learning in Python.
Journal of Machine Learning Research
12, 2825-2830 (2011).
5. Emmanuel T, Maupong T, Mpoeleng D, Semong T, Mphago B, Tabona O. A survey on missing data in
machine learning.
J Big Data
8, 140 (2021).
Figures
Figure 1
Page 22/26
Study outline.
Experimental timeline of the study. Two-month old male mice received a rTBI using the CHIMERA device
or underwent a Sham procedure (no impact). Risk-taking behavior was evaluated at 3 months post-injury
using the Elevated Plus Maze. MR imaging was performed 3.5 months after Sham or rTBI, and included
HP 13C MRSI, T2-weighted MRI, T1 mapping MRI, and SWI MRI. Tissue was collected 4 months after
Sham or rTBI procedures to evaluate PDH and LDH activities, and expression of MCT1 and MCT4. ML
analyses methods were used to identify the best classiers between rTBI and Sham, and the best
predictors of the risk-taking behavior and HP 13C Lac/Pyr in the cortex.
Figure 2
HP 13C spectra following co-injection of HP [1-13C]pyruvate and [13C]urea.
Page 23/26
Representative T2-weighted MR image overlaid with the grid used for HP 13C MRSI acquisitions.
Representative 13C spectra showing HP [1-13C]pyruvate, HP [13C]urea and HP [1-13C]lactate in the (a)
cortex (red voxel) and (b) subcortex (red voxel) for a Sham and a rTBI mouse.
Figure 3
HP 13C MRSI detects long-lasting metabolic alterations following rTBI.
Quantitative analyses of (a) HP [1-13C]lactate levels, (b) HP [1-13C]pyruvate levels, (c) HP 13C Lac/Pyr, and
(d) HP [13C]urea for the cortex (highlighted red voxels), revealed lower HP [1-13C]lactate levels (p =
0.0073), higher HP [1-13C]pyruvate levels (p = 0.0073), and lower HP 13C Lac/Pyr (p = 0.0071) in rTBI
Page 24/26
compared to Sham mice. In contrast, quantitative analyses of (e) HP [1-13C]lactate levels, (f) HP [1-
13C]pyruvate levels, (g) HP 13C Lac/Pyr, and (h) HP [13C]urea for the subcortex (highlighted red voxels),
did not detect differences between rTBI and Sham mice. (i) Representative HP 13C heatmaps for a Sham
and a rTBI mouse, highlighting lower HP [1-13C]lactate, higher HP [1-13C]pyruvate and lower HP 13C
Lac/Pyr in cortical areas in rTBI mice.
N
= 9 rTBI and 10 Sham mice. Unpaired
t
-test
(**p ≤ 0.01);
data are
expressed as means ± SD.
Figure 4
Multimodal MRI does not detect long-lasting effect of injury in rTBI.
(a) Representative T2-weighted MRI data and corresponding manual brain masking. (b) Quantitative
analyses of T2-weighted signal intensity revealed no signicant differences for brain subregions between
Sham and rTBI. (c) Representative T1 map and corresponding manual brain masking. (d) Quantitative
analyses of T1 maps revealed no signicant differences for brain subregions between Sham and rTBI. (e)
Representative SWI data and corresponding manual brain masking. (f) Quantitative analyses of SWI
intensity revealed no signicant differences for brain subregions between Sham and rTBI. Brain masking
color code: yellow: cortex, green: light blue: prefrontal cortex; hippocampus; dark blue: thalamus.
N
= 10
rTBI and 10 Sham mice. Unpaired
t
-test; data are expressed as means ± SD.
Page 25/26
Figure 5
ML analyses identify best rTBI and Sham classiers and best predictors of changes in risk-taking
behavior and HP 13C Lac/Pyr Ctx.
(a) Top two triplets that can classify rTBI (red) and Sham (black) mice. Here, circles represent the mice for
which all three variables are measured whereas triangles represent mice for which at least one of the
Page 26/26
three variables were missing and predicted by ML algorithms (see Methods). (b-c) Prediction performance
of the best predictors of risk-taking behavior (b, bottom panel), and HP 13C Lac/Pyr Ctx (c, bottom panel)
compared to the prediction performance of the case in which all variables are used (b-c top panel).
N
= 10
rTBI and 10 Sham mice.
Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.
SupplementaryInformation.pdf
