Changes in PGC-1α-Dependent Mitochondrial Biogenesis Are Associated with Inflexible Hepatic Energy Metabolism in the Offspring Born to Dexamethasone-Treated Mothers

Round 1

Reviewer 1 Report

SUMMARY OF MAJOR FINDINGS:

In utero exposure to dexamethasone causes low birth weight and negatively affects metabolism in adulthood, specifically promoting glucose intolerance through upregulating enzymes involved in gluconeogenesis like PEPCK. However, the underlying molecular mechanisms contributing to gluconeogenic pathways is not well characterized. Campos et al sought to uncover these molecular mechanisms, particularly the involvement of PGC-1alpha, in upregulating hepatic gluconeogenic pathways in prepubertal offspring exposed to dexamethasone in utero. Examining multiple targets involved in these pathways, the authors demonstrate that increase in OXPHOS, gluconeogenesis, and triglyceride synthesis correlate with increased PGC1a protein levels.

 

MAJOR STRENGTHS:

The paper examines multiple pathways that PGC1a can modulate to influence host metabolism and promote lipogenesis, mitochondrial biogenesis, and gluconeogenesis. Specifically, the authors first validated their model by examining the effects of in utero dexamethasone exposure to increasing the transcript expression, protein levels, and enzymatic activities of key enzymes involved in gluconeogenesis (PEPCK, G6Pase). The authors also conducted functional metabolic tests to demonstrate glucose intolerance and enhanced gluconeogenesis in postnatal day (PND) 21 rats, which is a major strength in the paper. Another strength of the paper was the inclusion of epigenetic modifications that contribute to these metabolic effects, specifically examining the expression of miRNAs and methylation status of promoter regions of interest. Lastly, the paper is well written, flows well and the data is presented in a logical manner. It was an enjoyable read.

 

CRITICISMS:

• Further description regarding the relevance of the model is needed in the manuscript. Specific details regarding: the amount of dexamethasone in circulation in mothers, quantification of water intake, and importance of dose and window of exposure is critical. Does the dose used correlate to levels seen in the clinic? What is the percentage of women using glucocorticoids/dexamethasone in pregnancy? Is the litter size affected between both groups? Why is the exposure period only between embryonic day 14-19 and not include lactation? In the discussion, can the authors provide insight as to whether there is an opportunity to intervene postnatally to prevent this enhanced hepatic lipogenesis and gluconeogenesis in dexamethasone-exposed offspring?

 

• Throughout the paper, bar graphs quantifying data do not have an x-axis, including figures 1-4, 6-9.

 

• Although whole body insulin tolerance is not affected, can the authors comments on whether fasting insulin levels differ between groups? In addition, is there any evidence to suggest there is hepatic insulin resistance (assessed via Western blot)? This data will help contribute to our understanding that in utero exposure to dexamethasone mainly enhances gluconeogenesis and complement the functional metabolic tests performed (specifically the pyruvate tolerance test).

 

• Can the authors provide a rationale as to why male pups were only used in this study (not female)?

 

• In the discussion the authors provide a clear explanation regarding the importance of most mitochondrial measures included in the paper, except for citrate synthase. Could the authors please include a statement to discuss the importance of enhanced citrate synthase in DEX-exposed offspring?

 

• Some minor grammatic issues:
  2. Line 430: replace comma with period for “ATP/ADP ratio was 2.9 times higher…”
  3. Line 497: Replace “already” with “previously”
  4. Line 505: Remove “little” with “unexplored”
  5. Line 513: Define “GR” acronym – “glucocorticoid receptor”?
  6. Line 559-560: Sentence slightly awkward – rephrase

 

Author Response

Dear reviewer, thank you for your valuable comments. We have made changes to the text in orer to address your concerns. Please, find below the detailed answers to your concerns.

 

• Further description regarding the relevance of the model is needed in the manuscript.

Answer: We agree with the concern. We have added new sentences to the introduction in order to make clearer the relevance of the model (please, see lines 48-53). Exposure of pregnant rats to exogenous glucocorticoids mimics one feature of experimental models that causes low birth weight and program glucose intolerance in adult life. Such model has already been used to this end, as also stated in the revised manuscript (please, see lines 54-56).

 

• Specific details regarding: the amount of dexamethasone in circulation in mothers, quantification of water intake, and importance of dose and window of exposure is critical.

Answer: Unfortunately, we do not have data on the circulating concentration of DEX in mothers. However, we have added data on maternal liquid intake (please see new supplementary figure 1) and commented on the window of exposure to glucocorticoid (please, see lines 58-59).

 

• Does the dose used correlate to levels seen in the clinic?

Answer: Not actually. As mentioned above, the dose was chosen based on previous studies reporting that this amount of DEX was able to mimic low-birth weight and glucose intolerance already described in the offspring born to undernourished mothers.

 

• What is the percentage of women using glucocorticoids/dexamethasone in pregnancy?

Answer: High-income countries have antenatal corticotherapy coverage of approximately 90% of mothers at risk of preterm delivery. On the other hand, low- and middle-income countries have variable coverage (27-58%) (PMID: 26390927). As the aim of the experimental model used in our study was not to simulate the clinical use of antenatal corticotherapy, we feel that this information did not fit in the introduction.

 

• Is the litter size affected between both groups?

Answer: Litter size is not affected by this treatment. This data was added in supplementary

figure 1

 

• Why is the exposure period only between embryonic day 14-19 and not include lactation?

Answer: We believe that this point was addressed previously in this letter. In utero exposure to DEX during the third week of pregnancy is sufficient to cause low-birth weight and the metabolic consequences of it (please, see PMID: 9593773).

 

• In the discussion, can the authors provide insight as to whether there is an opportunity to intervene postnatally to prevent this enhanced hepatic lipogenesis and gluconeogenesis in dexamethasone-exposed offspring?

Answer: This is a very interesting point. There are reports showing that it is possible to prevent/revert behavioral features of the phenotype programmed by antenatal DEX. For instance, the noradrenaline reuptake inhibitor desipramine (DMI) could prevent the onset of depression-like behavior seen in mice born to DEX-treated mothers (PMID: 31624238; PMID: 28916331). Similarly, swimming exercise during adult life ameliorates depression-like behaviors induced by prenatal exposure to glucocorticoids in rats (PMID: 22813978).

            We could not find reports of postnatal interventions that are able to change the enhanced hepatic lipogenesis and gluconeogenesis seen in the offspring exposed to antenatal DEX. On the other hand, it was already described that leptin treatment during the first 21 days of life can revert steatosis and insulin resistance seen in the offspring born to mothers subjected to experimental models hallmarked by increased endogenous glucocorticoid production (PMID: 28028317). Detailed information about this study was added in the revised manuscript (please, see lines 613-617).

 

• Throughout the paper, bar graphs quantifying data do not have an x-axis, including figures 1-4, 6-9.

Answer: We have found missing labels under the bars in the figure 2. We have now corrected this figure. The remaining bars throughout the figures are labeled as “CTL” or “DEX”.

 

• Although whole body insulin tolerance is not affected, can the authors comments on whether fasting insulin levels differ between groups?

Answer: This is a relevant issue. We have previously described that insulin levels are not modulated in 21-day old offspring born to DEX-treated mothers (PMID: 32473248). Hence, it is plausible to assume that, rather than reduced insulin levels, transcriptional changes in the liver are the putative causes of increased gluconeogenesis. This argument was added to the discussion (lines 538-543).

 

• In addition, is there any evidence to suggest there is hepatic insulin resistance (assessed via Western blot)? This data will help contribute to our understanding that in utero exposure to dexamethasone mainly enhances gluconeogenesis and complement the functional metabolic tests performed (specifically the pyruvate tolerance test).

Answer: Yes, there is a study showing impaired insulin signaling (IRS-1 association with PI3K) in the liver of the progeny born to DEX-treated mothers (PMID: 18828053). We have now referenced this study in the revised text (please, see lines 65-66). This information is in accordance with the reference originally mentioned in the introduction acknowledging that an acute challenge with insulin ineffectively inhibits PEPCK expression in the liver of rat exposed in utero to DEX.

 

 

• Can the authors provide a rationale as to why male pups were only used in this study (not female)?

Answer: We actually have assessed some parameters in the female littermates. We have data on body weight, glucose tolerance tests and insulin tolerance tests. These data were now added in supplementary figure 2. In this initial analysis, we have found that female and male are equally affected by antenatal exposure to DEX. Thus, we decided to perform the additional experiments only in males.

 

 

• In the discussion the authors provide a clear explanation regarding the importance of most mitochondrial measures included in the paper, except for citrate synthase. Could the authors please include a statement to discuss the importance of enhanced citrate synthase in DEX-exposed offspring?

Answer: We have added a sentence discussing this topic. Increased CS activity favor the proposition that higher citrate availability might feed fatty acid synthesis and hepatic steatosis in the liver of the offspring born to DEX-treated mothers (please, see lines 610-612).

 

Minors:

 

Line 430: replace comma with period for “ATP/ADP ratio was 2.9 times higher…”

Answer: This correction was performed. Please see line 455.

 

Line 497: Replace “already” with “previously”

Answer: This correction was performed. Please see line 524.

 

Line 505: Remove “little” with “unexplored”

Answer: This correction was performed. Please see line 537.

 

Line 513: Define “GR” acronym – “glucocorticoid receptor”?

Answer: This correction was performed. Please see line 545

 

Line 559-560: Sentence slightly awkward – rephrase

Answer: This correction was performed. Please see lines 595-596.

Reviewer 2 Report

During gestation, some maternal hormones cross the placental barrier and interact with the fetus in essential and adaptive ways. Glucocorticoids (mostly corticosterone in rats and mice) are one such hormone with many kind of effects on both the fetus and the mother during pregnancy. In the present experimental study the authors have investigated in more detail the glucocorticoid (dexametasone, DEX)) effects on fasting offsprings of rat mothers that were using DEX from the 14th to the 19th day of pregnancy as described in their previous reports (refs#3,12). Offsprings were studied on the 1st, the 8th and the 21st days of life. The major focus was on PGC-1α, one of the major metabolic programmers of newborn rats that were challenged in utero to DEX. The major findings showed that conversion of pyruvate into glucose was increased concomitant with the key enzymes PEPCK and G6pase, mitochondrial oxidative phosphorylation increased, there was higher ATP/ADP ratio as well as enhanced mitochondria biogenesis. At day 21 progeny of DEX-treated mothers had elevated hepatic TGs and lower CPT-1 activity indicating attenuated fatty acid ß-oxidation. Interestingly in mechanistic terms increased PGC-1α was coupled to increased association with HNF-4α and NRF1 according to immunoprecipitation experiments. As a whole the manuscript is well and clearly written. The methods used are relevant to test the aims of the study. However, there remains several issues that need further discussions.

MAJOR COMMENTS

1. The study was concentrated only on male offspring. What was the reason not to include other gender, females. The results would be relevant to show in gender-wise.
2. PGC-1α protein level was highest at day 21 whereas Ppargc1a mRNA expression was highest at day 1 and lowest at day 21, i.e. mRNA expression and protein level are shown in opposite directions. Is this explained with enhanced degradation of Ppargc1a mRNA with a more stable accumulating  PGC-1α protein or with the epigenetic methylation only?
3. Elevated DNA methylation at cytosine at the promoter region of Ppargc1a gene via the function of DNMT3a was observed in livers of  DEX mothers' offspring.  Did the authors study any of the SIRT functions?
4. Assumably there are changes facilitated via DEX in adipose tissue. Did the authors study whether adipose tissue is affected and whether adipose lipid droplets are increased? This would be an interesting issue regarding the whole body homeostasis and especially the cross-talk between liver and adipose tissue.
5. Hepatic TG level was clearly elevated in DEX-offsprings as well as serum TG level. Typically elevation on serum TG is associated inversely to serum HDL levels. What were HDL levels in these DEX-offsprings and in controls? In addition, when hepatic TG level is elevated this is in many times associated with elevated inflammation. Inflammation data would be relevant to include in the manuscript both data from the liver and serum.
6. On page 17, lines 563-565 the authors cite that adult rats born to DEX-mothers display elevated hepatic TG when exposed to long-term fasting or fructose feeding. What about long-term saturated fat feeding?
7. What was the response of DEX treatment of mothers on corticosteroid-binding globulin (CBG) in DEX-offsprings? Based on previous reports total plasma corticosterone increases and CBG decreases in late pregnancy,implying an increase in free corticosterone. But what about the offspring of DEX-mothers? 
8. During pregnancy a tight balance between glucocorticoids and progesterone may take place. When DEX was used at the level of 0.2 mg/kg body mass could there be imbalance between total glucocorticoids and progesterone that would affect newborns?
9. Could the authors comment possible glucocorticoid mediated adverse neurodevelopmental outcomes of prenatal stress exposure in DEX-offsprings?

 

 

Author Response

Dear reviewer, thank you for your valuable comments. Please, find below the detailed answers to your concerns.

 

1. The study was concentrated only on male offspring. What was the reason not to include other gender, females. The results would be relevant to show in gender-wise.

Answer: We have performed the initial analysis in both genders. As the changes in body weight, GTT and ITT in females were similar to those in males, we decided to perform the following experiments only in males. We have now added the data on female offspring in the supplementary figure 2.

 

2. PGC-1α protein level was highest at day 21 whereas Ppargc1a mRNA expression was highest at day 1 and lowest at day 21, i.e. mRNA expression and protein level are shown in opposite directions. Is this explained with enhanced degradation of Ppargc1a mRNA with a more stable accumulating PGC-1α protein or with the epigenetic methylation only?

Answer: This an interesting issue. We do not have data that allow us to speculate about Ppargc1a mRNA degradation or PGC-1α stability. Our data allow us to suggest that the (i) low Ppargc1a mRNA can be explained by in Ppargc1a promoter cytosine methylation and (ii) high PGC-1α protein content be explained by reduced miR-29c. These two mechanisms presently described certainly do exclude other regulatory pathways. Indeed, there are papers showing that PGC-1α is degraded via the ubiquitin proteasome system (PMID: 20713359) and such mechanism is increased in the adipose tissue of obese individuals (PMID: 29678181). These arguments were added to the revised text (Please, see lines 582-586).

            In regard to a possible modulation of Ppargc1a mRNA degradation at L21, we could not find reports in the literature that sustain such hypothesis.

 

3. Elevated DNA methylation at cytosine at the promoter region of Ppargc1a gene via the function of DNMT3a was observed in livers of DEX mothers' offspring. Did the authors study any of the SIRT functions?

Answer: We acknowledge that PGC-1α deacetylation by SIRT is a relevant mechanism that controls its activity and the expression of gluconeogenic genes. However, changes in PGC-1α acetylation had no impact on mitochondrial genes (PMID: 15744310). Therefore, we explored other mechanisms by with PGC-1α could modulate mitochondrial genes, such as its association with NRF1.

 

4. Assumably there are changes facilitated via DEX in adipose tissue. Did the authors study whether adipose tissue is affected and whether adipose lipid droplets are increased? This would be an interesting issue regarding the whole body homeostasis and especially the cross-talk between liver and adipose tissue.

Answer: Programming of adiposity by antenatal exposure to DEX has been previously addressed by our and others groups in the adult offspring. Unfortunately, we do not have data on adipose tissue parameters between birth and 21 days of life. We believe that the body of data presented herein is sufficient to provide new and relevant information about the programming of energy metabolism in the offspring born to DEX-treated mothers. Whether these changes are synchronized with responses in the adipose tissue, is an aspect that certainly deserves further investigation

 

5. Hepatic TG level was clearly elevated in DEXoffsprings as well as serum TG level. Typically elevation on serum TG is associated inversely to serum HDL levels. What were HDL levels in these DEXoffsprings and in controls? In addition, when hepatic TG level is elevated this is in many times associated with elevated inflammation. Inflammation data would be relevant to include in the manuscript both data from the liver and serum.

Answer: We do not have data on HDL, but have added data on total cholesterol (please, see new Figure 9B). These new data shows that total cholesterol is not modulated in the offspring of DEX-treated mothers.

            We do agree with the reviewer that inflammatory parameter is potentially relevant to the present data. However, the manuscript has already a great body of data that allow new and relevant conclusions. For sure, further investigations will continue to explore new mechanisms related to the phenomena presently described.

 

6. On page 17, lines 563-565 the authors cite that adult rats born to DEX-mothers display elevated hepatic TG when exposed to long-term fasting or fructose feeding. What about long-term saturated fat feeding?

Answer: Yes, hepatic steatosis induced in adult life by HFD is exacerbated by antenatal exposure to DEX (PMID: 22067322; PMID: 20133452). This information was added to the revised text, please see line 610.

 

7. What was the response of DEX treatment of mothers on corticosteroid-binding globulin (CBG) in DEXoffsprings? Based on previous reports total plasma corticosterone increases and CBG decreases in late pregnancy,implying an increase in free corticosterone. But what about the offspring of DEX-mothers?

Answer: As the reviewer acknowledged, we have previously demonstrated that CBG levels are modulated in undisturbed pregnant rats (PMID: 22700767). Unfortunately, we do not have data to speculate on maternal CBG levels in DEX-treated pregnant rats and their offspring.

 

8. During pregnancy a tight balance between glucocorticoids and progesterone may take place. When DEX was used at the level of 0.2 mg/kg body mass could there be imbalance between total glucocorticoids and progesterone that would affect newborns?

Answer: In fact, treatment of pseudopregnant rats with DEX was reported to reduce progesterone levels (PMID: 9284513). However, we do not have data on progesterone levels in DEX-treated pregnant rats, thus we feel that we cannot speculate on this topic.

 

9. Could the authors comment possible glucocorticoid mediated adverse neurodevelopmental outcomes of prenatal stress exposure in DEX-offsprings?

Answer: We acknowledge that the neurodevelopmental impacts of antenatal glucocorticoids are widely investigated. However, we understand that such information do not fit in the scope of the manuscript and the journal.

Round 2

Reviewer 2 Report

I do not have any further queries to the authors.