Article activity feed

  1. Author Response

    Reviewer 3

    This papers builds on a previous publication from the same group that showed compartmentalisation model of beta-cell fuel metabolism in which plasma membrane-localized pyruvate kinase is sufficient to close KATP channels required for insulin secretion. In this current manuscript the authors identified the PK isoforms involved in this process using tissue specific KO mouse models. Using excised patch-clamp experiments, they demonstrated that although redundant in their function both the constitutively active PKm1 and allosterically PKm2 are associated with the PM and locally regulate KATP channel closure. Further, the authors showed that the mitochondrial PEP carboxylase (PCK2) is essential for amino acids to promote an increase in cytosolic ATP/ADP and closure of KATP channels. Therefore, this study very nicely demonstrates that he distinct response of PK isoforms to the mitochondrial and glycolytic sources of PEP impacts beta cell nutrient preference and affects the oscillatory cycle regulating secretion. These findings do provide new mechanistic information about the control of the regulated secretory pathway and will be of interest to broader audience.

    Strength
    The major strength of the study is the use of tissue/isoform specific KO mouse models. Although limited by constitutive KOs with compensatory increase in other isoforms, the authors have achieved what they were set out to do i.e identify the PK isoform involved in the regulation of PM ATP generation and regulation of KATP channel closure. Their experimental rigorosity including the ability to perform the excised patch clamp experiments and use of PKa to show the specific effect of the allosterically regulated PKM2 are also strength.

    Weakness
    It is not clear from the manuscript what the "littermate controls" are used in all the experiments. Given the limitations of the cre lox system, it is really important to clearly show what controls have been used and their phenotypes (and the rationale for pooling the different controls if that is what is done here).

    Response: We apologize that this was unclear. Littermate Ins1-Cre controls were used for the PKm1-βKO and PKm2-βKO models, whereas floxed controls (i.e. Pck2f/f) were used for the PCK2-βKO. This is described in the first paragraph of the results section as well as the methods.

    The data adds to our understanding of the role of PM localised PK on the regulated exocytosis pathway however the claim that these findings question the canonical mitochondrial ATP coupled to KATP channel closure is not fully supported by the data especially given glucose induced insulin secretion is not affected by any of the KO models.

    Response: Had we performed experiments that fully block flux through pyruvate kinase, we expect that glucose-stimulated insulin secretion would be impaired if not eliminated. However, as this experiment would prevent both glycolytic and mitochondrial ATP production, it would not address whether a mitochondrially-derived increase in cytosolic ATP/ADP is the primary mechanism of KATP channel closure, as proposed by the canonical model.

    Due to the redundant response of PKm1 and PKm2 to glucose, isoform-specific deletion allowed us to test whether a fuel-stimulated rise in cytosolic ATP/ADP is sufficient to close KATP channels in healthy β-cells (i.e. in the absence of glucose intolerance) using a protocol of low glucose and amino acid stimulation. Our findings show that raising cytosolic ATP/ADP with amino acids is insufficient to close KATP channels in the absence of PK activity or PCK2, indicating that KATP channels are regulated primarily by PEP that provides ATP via plasma membrane-associated PK rather than mitochondrially-derived ATP. In these experiments, insulin secretion shows the expected reduction in both the β-cell PCK2 and PKm1 knockout models.

    In the discussion, we point out that mitochondrially-derived ATP is likely important for buffering this plasma membrane-compartmentalized KATP closure by pyruvate kinase. However, on balance, our data argue that glycolysis preferentially closes KATP channels, as previously shown in cardiac myocytes (Lamp and Weiss, Science 1987).

    Was this evaluation helpful?
  2. **Evaluation Summary:
    **
    This manuscript employs in vitro studies and elegant mouse models to detail how specific pyruvate kinase isoforms impact pancreatic beta-cell ATP/ADP levels, ATP-sensitive K+ channel (KATP channel) activity, calcium handling, and insulin secretion. This is an important study that challenges the current paradigms of KATP-channel regulation, the major signaling mechanism through which pancreatic beta cells couple blood glucose levels to insulin release. Future studies will be necessary to determine whether similar mechanisms are used in human pancreatic beta cells.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #2 and Reviewer #3 agreed to share their names with the authors.)

    Was this evaluation helpful?
  3. Reviewer #1 (Public review):

    The manuscript by Foster et al. details how PEP cycling and specific pyruvate kinase isoforms impact beta-cell ATP/ADP levels, KATP activity, calcium handling, and insulin secretion. The manuscript clearly illuminates the beta-cell specific roles of PKm1, PKm2, and mitochondrial PEP carboxykinase. The manuscript finds that beta-cell PEP production leads to KATP inhibition via ATP produced by PKm1 and PKm2. The manuscript also finds that amino acid induced closure of KATP channels depends on mitochondrial PEP production but not elevations in cytoplasmic ATP/ADP. Finally, the manuscript suggests that the PEP cycle is also involved in KATP activation, but the mechanism remains to be determined. The manuscript is well written and easy to follow. Overall, this is an excellent manuscript that will be very useful to the diabetes research community.

    Was this evaluation helpful?
  4. Reviewer #2 (Public review):

    The present studies by Foster and colleagues use mouse genetics to show that pyruvate kinase 1 and 2 (PKM1 and PKM2) regulate ATP-sensitive K+ channel activity (KATP channel) through mitochondrial PEP-dependent cytoplasmic ATP/ADP increases, leading to first phase insulin secretion. During the second phase of insulin secretion, when ATP hydrolysis is maximal, oxidative phosphorylation is engaged to sustain ATP/ADP ratios and KATP channel closure. As such, the work challenges the consensus view of KATP channel activity, which states that ATP derived from oxidative phosphorylation in the mitochondrial matrix increases cytoplasmic ATP/ADP ratio, thus closing KATP channels and increasing Ca2+ fluxes.

    Strengths of the study include: 1) careful experimental design and execution; 2) use of comprehensive mouse genetics to pinpoint roles of PKM1, PKM2 and phosphoenolpyruvate carboxykinase 2 (which produces PEP from oxoaloacetic acid); and 3) multiple lines of corroboratory evidence that the PEP-PKM1/2 system influences KATP channel activity and downstream signaling, via changes in non-mitochondrial ATP/ADP.

    Weaknesses include: 1) lack of in vivo data to support a role of PKM1/PKM2 in determining glucose levels; and 2) over-reliance on mouse models, meaning that translational relevance to human biology is unclear.

    Nonetheless, on balance, the authors have achieved their aims of showing that PEP and PKM1/PKM2 are critical regulators of KATP channel activity, Ca2+ fluxes and insulin secretion.

    Overall, this is a potentially important study, which updates the textbook view of KATP-channel regulation, the major signaling mechanism through which pancreatic beta cells couple blood glucose levels to insulin release.

    Was this evaluation helpful?
  5. Reviewer #3 (Public review):

    This papers builds on a previous publication from the same group that showed compartmentalisation model of beta-cell fuel metabolism in which plasma membrane-localized pyruvate kinase is sufficient to close KATP channels required for insulin secretion. In this current manuscript the authors identified the PK isoforms involved in this process using tissue specific KO mouse models. Using excised patch-clamp experiments, they demonstrated that although redundant in their function both the constitutively active PKm1 and allosterically PKm2 are associated with the PM and locally regulate KATP channel closure. Further, the authors showed that the mitochondrial PEP carboxylase (PCK2) is essential for amino acids to promote an increase in cytosolic ATP/ADP and closure of KATP channels. Therefore, this study very nicely demonstrates that he distinct response of PK isoforms to the mitochondrial and glycolytic sources of PEP impacts beta cell nutrient preference and affects the oscillatory cycle regulating secretion. These findings do provide new mechanistic information about the control of the regulated secretory pathway and will be of interest to broader audience.

    Strength
    The major strength of the study is the use of tissue/isoform specific KO mouse models. Although limited by constitutive KOs with compensatory increase in other isoforms, the authors have achieved what they were set out to do i.e identify the PK isoform involved in the regulation of PM ATP generation and regulation of KATP channel closure. Their experimental rigorosity including the ability to perform the excised patch clamp experiments and use of PKa to show the specific effect of the allosterically regulated PKM2 are also strength.

    Weakness
    It is not clear from the manuscript what the "littermate controls" are used in all the experiments. Given the limitations of the cre lox system, it is really important to clearly show what controls have been used and their phenotypes (and the rationale for pooling the different controls if that is what is done here).

    The data adds to our understanding of the role of PM localised PK on the regulated exocytosis pathway however the claim that these findings question the canonical mitochondrial ATP coupled to KATP channel closure is not fully supported by the data especially given glucose induced insulin secretion is not affected by any of the KO models.

    Was this evaluation helpful?