A more integrated approach is needed for genetic instrument selection in Mendelian randomization (MR) studies of type 2 diabetes and HbA1c, argue Garfield et al. (p. 175). According to the authors, these studies have become particularly popular in recent years, yet instrument selection strategies are often suboptimal and, on occasion, poorly reported. Underlining this, they write that choosing the most appropriate instrument is one of the most important decisions when designing such studies. This is because inappropriate selection can lead to misleading and conflicting findings, a very real issue with certain MR studies in the diabetes space. On that basis, they provide a set of five recommendations for researchers who wish to use genetic instruments in such studies. Using a literature review, they found far more MR studies that use diabetes as a genetic instrument (as an exposure; n = 48) than HbA1c (n = 22). They also used data from just under 350,000 individuals in the UK Biobank to calculate two instrument strength metrics, the F statistic for average strength and R2 for total strength. They found that the instrument for HbA1c, which was based on 51 single nucleotide polymorphisms (SNPs), was strong both in terms of average and total strength. This was irrespective of the method used to calculate the metrics and irrespective of whether HbA1c was portioned by function (i.e., glycemic or erythrocytic). The equivalent instrument for diabetes had good average and total strengths, although both were substantially smaller than those for HbA1c, despite including 157 SNPs. Among the recommendations they make, they highlight that larger instruments that include a greater number of SNPs will not necessarily give stronger results than smaller instruments. They also stress the importance of calculating and reporting the instrument strength metrics. Commenting further, author Victoria Garfield said, “We put a lot of work into this methodological review, and we really hope that it will become the guide for diabetes researchers who wish to conduct robust MR studies, particularly from an instrument selection perspective.”

Criteria for instrument selection in Mendelian randomization studies in type 2 diabetes and HbA1c.

Criteria for instrument selection in Mendelian randomization studies in type 2 diabetes and HbA1c.

Close modal

Garfield et al. A guide for selection of genetic instruments in Mendelian randomization studies of type 2 diabetes and HbA1c: toward an integrated approach. Diabetes 2023;72:175–183

Patients with type 2 diabetes with and without diabetic neuropathy (DN) appear to have diametrical correlations between HbA1c and peripheral nerve capillary permeability, according to Mooshage et al. (p. 290). Specifically, the constant of peripheral nerve capillary permeability (Ktrans; obtained with dynamic contrast-enhanced magnetic resonance imaging) positively correlated with HbA1c in patients with DN but negatively correlated with HbA1c in patients without DN. According to the authors, the findings raise the prospect of glucose control still having significant benefits in the case of no DN, but the same cannot (yet) be said in the context of type 2 diabetes and DN. At this stage, the mechanisms underlying the positive relationship between capillary permeability and HbA1c in DN remain undetermined. On that basis, the authors call for future longitudinal studies on the relationship and that studies should distinguish between patients with and without DN. The findings come from a cross-sectional, single-center study involving 58 individuals with type 2 diabetes and either DN (n = 20) or no DN (n = 38) who underwent dynamic contrast-enhanced magnetic resonance neurography (measuring a series of in vivo parameters of microvascular perfusion). Clinical, electrophysiologic, and serologic testing were also included. While there were clear differences in capillary permeability between the groups, other measures of volume fraction of plasma and extracapillary/cellular space were not different. There were also further differences between the groups in terms of electrophysiological parameters. The authors also present a series of correlation analyses exploring potential relationships between the various measures. “The results indicate that the impact of glucose levels on the perfusion of peripheral nerves differs fundamentally between those individuals with type 2 diabetes with diabetic neuropathy and those without,” said author Johann M.E. Jende. “This may explain why clinical studies addressing the preventive effect of glucose control on the development of diabetic neuropathy in type 2 diabetes have come to controversial results.”

Coregistered image of T2-weighted (red) and VIBE sequence (green) of sciatic nerve, taken with magnetic resonance neurography.

Coregistered image of T2-weighted (red) and VIBE sequence (green) of sciatic nerve, taken with magnetic resonance neurography.

Close modal

Mooshage et al. Diametrical effects of glucose levels on microvascular permeability of peripheral nerves in patients with type 2 diabetes with and without diabetic neuropathy. Diabetes 2023;72:290–298

GC-globulin (alternatively named vitamin D-binding protein) may have a central regulatory role in glucagon secretion and is required for pancreatic α-cell adaptation during metabolic stress, according to Viloria et al. (p. 275). The findings, which extend previous research, provide the first hints that GC-globulin might be targeted during type 2 diabetes to maintain pancreatic α-cell function (i.e., maintain glucagon secretion). Such an effect would, according to the authors, then restrain pancreatic β-cell proliferation and the hyperinsulinemia known to drive insulin resistance in type 2 diabetes. While they also note that GC-globulin may be reduced in type 2 diabetes, they caution that careful preclinical studies will still be needed to take the idea forward. Based on studies of a mouse model with GC-globulin knocked out, they found that such mice fed a high-fat diet for 12 weeks had basal hyperglucagonemia associated with decreased α-cell size, impaired glucagon secretion and Ca2+ fluxes, and alterations in F-actin abundance. Crucially, the impairments could be rescued with exogenous GC-globulin. This restored GC-globulin levels in pancreatic α-cells, increased glucagon granule area, restored the F-actin cytoskeleton, and normalized glucagon secretion. The authors point out that it is unusual that decreases in expression of a key signature gene can be offset by supplementation with the gene’s protein product. For this reason, they suggest studies of the uptake mechanisms are warranted along with studies on potentially wider effects of such supplementation. Commenting further, lead author Katrina Viloria said, “The present studies confirm that GC-globulin is essential for properly regulated glucagon release and extend findings to models of obesity as well as pancreas samples from patients living with type 2 diabetes. Since GC-globulin supplementation is well tolerated and pancreatic α-cells can take up the protein, we are now working to understand how we can leverage our findings to normalize glucagon release during diabetes.”

Representative images from nondiabetic (ND) and type 2 diabetes (T2D) donors showing a large reduction in GC-globulin expression.

Representative images from nondiabetic (ND) and type 2 diabetes (T2D) donors showing a large reduction in GC-globulin expression.

Close modal

Viloria et al. GC-globulin/vitamin D-binding protein is required for pancreatic α-cell adaptation to metabolic stress. Diabetes 2023;72:275–289

The stability of the protein adipose triglyceride lipase, or ATGL, appears to be regulated by two ubiquitin ligases, UBR1 and UBR2, according to Xu et al. (p. 210). More specifically, it appears this stability is essential for maintaining health, particularly from the cellular lipid accumulation perspective, and thus potentially has relevance to the treatment of obesity. The authors caution, however, that the level and the site of ATGL regulation will be crucial in determining whether lipid levels can be manipulated successfully via this route. Using a combination of cell and mouse models, the authors show that manipulation of ATGL levels results in changes in lipid levels at the cellular level. Previous studies already demonstrated ATGL as a rate-limiting lipase critical to lipolysis (i.e., the breakdown of lipids). While control, transcription, localization, and activation of the ATGL gene have been widely studied, the stability of the ATGL protein is much less understood, which motivated the study. According to the authors, they identified two N-end rule E3 ligases, UBR1 and UBR2, as reducing the level of ATGL, resulting in the promotion of lipid storage. They point out that ubiquitin ligases recognize and bind to proteins that have destabilizing N-terminal residues, resulting in ubiquitination and degradation. Crucially, ATGL has a destabilizing N-end phenylalanine residue. This allowed the authors to generate an N-end rule–resistant mutant where the N-end phenylalanine was swapped out for alanine. This resulted in increased protein stability and enhanced lipolysis, and when the change was knocked into mice they were protected against high-fat-diet–induced obesity, hepatic steatosis, and insulin resistance. Commenting further, author Xun Huang said, “Our study demonstrates N-end rule–mediated proteasomal regulation of ATGL, a finding which could potentially be beneficial in the treatment of obesity. Spatial and temporal fine-tuning of the action of ATGL will be key in the treatment of metabolic diseases.”

Body weight gain over 16 weeks of high-fat diet in Atgl+/+ (blue) and AtglF2A/F2A knock-in (red) mice.

Body weight gain over 16 weeks of high-fat diet in Atgl+/+ (blue) and AtglF2A/F2A knock-in (red) mice.

Close modal

Xu et al. N-end rule–mediated proteasomal degradation of ATGL promotes lipid storage. Diabetes 2023;72:210–222

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/journals/pages/license.