AIBP Regulates Metabolism of Ketone and Lipids but Not Mitochondrial Respiration

Round 1

Reviewer 1 Report

This is an interesting manuscript that evaluates the metabolic and functional consequences of APOA1 binding protein (AIBP) deletion in the cardiovascular system.  The authors present several novel findings that are of interest to investigators in this field.  Several suggestions include:

-        - An expanded description of the knockout mice would be helpful.  Were these global or cardiomyocyte-specific knockout animals?  If these have been published previously, a citation in the Methods section would be beneficial.

-        -The investigators alternate between AIBP and NAXE in the manuscript.  It would be helpful to use one nomenclature.

-        -For some of the experiments, the investigators assessed cardiac tissue.  Was this whole hearts or specific portions of the heart?

-        -The investigators indicate these were combined data from male and female animals.  Were the two genders represented equally in the samples?

-        -In some of the experiments (Figure 1, for instance), the investigators alternate between calling the samples cardiac tissue and cardiomyocytes.  If the samples are not purified cardiomyocytes, stick to cardiac tissue.

Author Response

This is an interesting manuscript that evaluates the metabolic and functional consequences of APOA1 binding protein (AIBP) deletion in the cardiovascular system.  The authors present several novel findings that are of interest to investigators in this field. 

Response: We thank this reviewer for the complimentary comments!

Several suggestions include:

1. An expanded description of the knockout mice would be helpful.  Were these global or cardiomyocyte-specific knockout animals?  If these have been published previously, a citation in the Methods section would be beneficial.

Response: As this reviewer suggested, we added description of the of global AIBP knockout mice as following: “In systemic NAXE knockout mice, in contrast to humans [13], other than increases retinal angiogenesis in development and facilitates adult angiogenesis and blood flow recovery following hindlimb ischemia [8], no apparent neurological defects were observed.  This sentence is in page 2, lines 45 – 47 of the revised manuscript. So far, no cardiomyocyte-specific knockout animal is available. As suggested, we have put this reference (Mao R et al., Circ Res 2017) in the Method section (page 4, line 160 of the revised manuscript).

2. The investigators alternate between AIBP and NAXE in the manuscript.  It would be helpful to use one nomenclature.

Response: Thank you for this suggestion. Other than the title and initial introduction, we have changed the name to NAXE in the manuscript. While calling it NAXE conforms to the current nomenclature, the name AIBP seems to be more biologically relevant in this manuscript. Thus, we chose to keep AIBP in the title.

3. For some of the experiments, the investigators assessed cardiac tissue.  Was this whole hearts or specific portions of the heart?

Response: For all the metabolite measurements, we used the whole cardiac tissue. But to perform the Oroboros analysis (Fig. 9D), we dissected the cardiac muscle fibers from the left ventricle and used them for this analysis. Please also see Response 5.

4. The investigators indicate these were combined data from male and female animals.  Were the two genders represented equally in the samples?

Response: We used 4 female and 2 male AIBP knockout mice. To generalize our conclusions on AIBP, we carried out the analyses by combining the datasets of both genders.

5. In some of the experiments (Figure 1, for instance), the investigators alternate between calling the samples cardiac tissue and cardiomyocytes.  If the samples are not purified cardiomyocytes, stick to cardiac tissue.

Response: This is a good point. As elaborated above in Response 3, we used the whole cardiac tissue for the metabolite measurements and used the cardiac muscle fibers for mitochondrial respiration analysis. In the revised manuscript, we changed “cardiomyocytes” to “cardiac muscle fiber” (p17, line 349 of the revised manuscript), cardiac myofiber (p19, line 430)” or “cardiac tissue” (p17, line 359 and p18, lines 403&404) when it is applicable.

Reviewer 2 Report

Well written manuscript and definately well presented.

I would ask for some clarifications regarding data analysis in the 3 following points.

1. An examination of the supplementary material shows that the estimated p values were based on two sided independent samples t tests assuming equal variances. Still, a descriptive approach shows that the variances in the two groups, especially in somes indices, including the ones that statistically significant differences were detected, are rather different.  Based on which criterion did the authors choose to carry out an analysis assuming equal variances in, unexceptionally, all differences examined?

2. Since the sample size would not allow for reliable normal normal distribution testing, were non parametric tests considered? And if so why did the authors chose to carry on based on t tests instead?

3. The methodology indicates that multiple comparisons were carried out based the same samples. Was there an adjustment for type I error considered, to account for this? 

Author Response

Reviewer 2.

Well written manuscript and definitely well presented.

Response: We thank this reviewer for the complimentary comments!

I would ask for some clarifications regarding data analysis in the 3 following points.

1. An examination of the supplementary material shows that the estimated p values were based on two-sided independent samples t tests assuming equal variances. Still, a descriptive approach shows that the variances in the two groups, especially in some indices, including the ones that statistically significant differences were detected, are rather different. Based on which criterion did the authors choose to carry out an analysis assuming equal variances in, unexceptionally, all differences examined?

Response: These is a good suggestion!  We have reanalyzed the data below as suggested using non-parametric T-tests (shown as new Table S1 in the revised manuscript). The criteria of significance are p < 0.05 and q < 0.25.

As shown below, the statistical significance did not show much difference when analyzed using parametric T-test or non-parametric Mann-Whitney test. Therefore, our conclusions remain unchanged.

2. Since the sample size would not allow for reliable normal distribution testing, were non-parametric tests considered? And if so why did the authors chose to carry on based on t tests instead?

Response: We performed the nonparametric test as this reviewer suggested, and the statistical differences are similar to that measured using parametric T-test. Please see the Response 6.

3. The methodology indicates that multiple comparisons were carried out based the same samples. Was there an adjustment for type I error considered, to account for this? 

Response: This is a good suggestion! We reanalyzed the data with Mann-Whitney multiple comparisons based on the metabolites present in the same group of animals, followed by Benjamini-Hochberg correction. Only changes of 3HB and cholesterol levels are statistically significant (p<0.05 and q<0.25). However, the other metabolite changes, such as glucose, fructose, galactose, tagatose and hexose occur in the same trend (Fig. 1E & Table S1). In particular, the rise in 3HB content, which is expected to increase the easier accessible substrate for the mitochondria, is consistent with the overall reduction of cellular hexose levels.