PREreview de Novel in vivo measurement of muscle total carnitine concentration reveals potential mechanism linking mitochondrial dysfunction and lipid accumulation

1. Overall assessment

In this manuscript, the authors present a new method for measuring total and free carnitine in skeletal muscle using proton magnetic resonance spectroscopy (¹H-MRS). This method accounts for muscle fiber orientation and possible interference from other metabolites in the same spectral region. The authors compare their measurements with mass spectrometry results from muscle biopsies and then apply the method to patients with mitochondrial disease. This work focuses on 8 patients, 4 with A to G mtDNA mutation, 1 with A to T mtDNA mutation and 3 with single large-scale mtDNA deletion, associated with chronic progressive external ophthalmoplegia.

The results show lower muscle carnitine concentrations and higher muscle lipid content in patients with mitochondrial disease, along with reduced mitochondrial oxidative capacity. These findings support a relationship between muscle carnitine levels, mitochondrial function, and lipid accumulation. However, more evidence is needed to support the hypothesis that mitochondrial dysfunction leads to carnitine depletion through increased acylcarnitine export.

In my view, the main contribution of this paper is the development of a non-invasive method that could allow muscle carnitine concentrations to be measured repeatedly without taking biopsies. The authors may want to take another look at a few limitations in the validation and the clinical study, since addressing them would strengthen the paper. The association between carnitine depletion and lipid accumulation may be worth framing as suggestive rather than as evidence of a mechanism given the data presented.

2. Major strengths

1. The method addresses an important measurement limitation. Plasma carnitine does not necessarily reflect the amount present in skeletal muscle. A method that measures carnitine directly in muscle could provide more useful information about tissue carnitine availability.

2. They account for muscle fiber orientation. This is an important part of the method because the MRS signal can vary depending on the orientation of muscle fibers relative to the magnetic field. They show that this variation can occur even within the same muscle group, so correcting for it is necessary when comparing measurements between individuals. The trimethylamine (TMA) resonance used to estimate carnitine can contain contributions from other metabolites, like carnosine. Considering the interference from carnosine improves the specificity of the measurement and makes the concentration estimates more reliable.

3. The method is compared with muscle biopsy measurements. They report good agreement between MRS and mass spectrometry measurements, with an intraclass correlation coefficient of 0.905. This supports the potential of the method, even though only four patients were involved in the validation.

3. Major comments

1. The validation is based on a very small number of samples. They reported good agreement between the MRS method and mass spectrometry measurements, but the comparison was performed in only four patients. Although the reported intraclass correlation coefficient is encouraging, I am not sure if a sample size of 4 is sufficient to establish how well the method performs across a wider range of carnitine concentrations. Clarifying this limitation in the discussion would help with understanding potential systematic bias and how large the differences are between individual measurements.

2. Their data in figure 3 show increased intramyocellular lipid (IMCL) and intramuscular fat in patients with mitochondrial disease. These measurements provide useful information about muscle lipid content, but they do not establish why the lipids accumulate or if the accumulated lipids are responsible for metabolic dysfunction. I suggest they make a clearer distinction between IMCL and the fat measured by MRI, because these measurements reflect different tissue compartments. In addition, increased lipid content is not necessarily evidence of impaired fatty acid oxidation. Providing additional supporting data would increase confidence in this conclusion. Here they could do direct measurements of fatty acid oxidation, such as oxidation rates using labeled fatty acids or measurements of substrate oxidation in muscle. They could also do mitochondrial respiratory measurements, including oxygen consumption supported by fatty acid substrates, to assess whether mitochondrial capacity for fatty acid oxidation is reduced. Alternatively, acknowledging the limitations of the approach will prevent the findings from being interpreted more broadly than the data support.

4. Minor comments

1. Reproducibility: The authors report coefficients of variation of 7.8% for the soleus and 12.2% for the tibialis anterior. These results are useful, but the reproducibility assessment was performed during a fasting intervention. I wonder how that affects the interpretation of the results. The authors could clarify which sources of variability were tested and how reproducibility would be assessed under the conditions in which the method is intended to be used. To test the reproducibility across sessions, they could repeat the measurement on different days, to assess whether the method produces consistent results under realistic conditions.

2. Assumptions used in calculating the concentration: The method relies on assumptions about muscle water content and other factors used to calculate absolute concentrations. This is shown in the method section, where the assumptions used to convert the MRS signal into absolute muscle carnitine concentrations. The authors could add a brief discussion on how variation in these assumptions could affect the final estimates, especially when comparing healthy individuals with patients who have muscle disease.

3. Interpretation of total and free carnitine: This paper would benefit from a clearer explanation of what the MRS method measures directly and how total and free carnitine estimates are obtained. This would make it easier to understand the analytical basis of the measurements.

I reviewed this preprint as part of the Transparent and Accountable Peer Review training program. This report was written with the support of my research mentor

Competing interests

The author declares that they have no competing interests.

Use of Artificial Intelligence (AI)

The author declares that they did not use generative AI to come up with new ideas for their review.