Letter by Zhu et al Regarding Article, “Glucose-sensitive myokine/cardiokine MG53 regulates systemic insulin response and metabolic homeostasis.”

To the editor:

We read with great interest the paper by Wu et al concerning an immuno-approach using MG53 antibody (MG53-mAb) to reduce serum levels to treat diabetes¹. After careful reading, several points were appreciated which may alter the interpretation of the data and its appropriate impact.

Whereas a multi-pronged approach was utilized in attempts to quantify MG53 in serum samples, the combined ELISA and western blot (WB) data are unclear. ELISA demonstrated MG53 levels in the range of ∼200–300 pg/mL (Figure 3c). Assuming loading 1–10 µL serum in the gel, this would result in 0.2–2 pg MG53, far beyond the sensitivity of WB. The sensitivity could raise concerns regarding the accuracy of the WB band for the MG53 signal. MG53 marked in the WB at 53 kD would unlikely be the MG53 band as Lemckert et al have clearly shown that running serum samples shifts target proteins to lower molecular sizes².

Surface plasmon resonance (SPR) assay traces, used to demonstrate that recombinant human MG53 protein (rhMG53) binds to the extracellular domain (ECD) of insulin receptor with a Kd of 8 nM, (Figures 6B and 6C) show different patterns of rhMG53 binding and insulin interaction with ECD. Unlike the rapid association and dissociation of insulin with ECD, rhMG53 shows non-equilibrium, slow association with ECD that does not dissociate during the observed time window. In our opinion, this may represent non-specific binding of rhMG53 to the BIACORE sensor chip.

A Kd of 8 nM represents a concentration of ∼400 ng/mL MG53 in circulation; >1,000-fold higher than the circulating MG53 level determined by ELISA. One would surmise that even if circulating MG53 were doubled or 10-fold higher in diabetics than non-diabetics, the blocking effect of circulating MG53 on the insulin receptor, at that level, would not be achieved in physiological or pathological conditions.

The authors report intravenous injection of MG53-mAb led to a modest reduction of glucose from ∼425 to 375 mg/dL (with a standard error of <10 mg/dL, Figure 5C), marking this as a significant difference (p<0.01). Supplemental data in Figure VI-I reported a blood glucose level >600 mg/dL, which is ∼175 mg/dL higher than data shown in Figure 5C under identical conditions (e.g. db/db mice receiving IgG as control). Our interpretation of these data do not suggest beneficial effects of MG53-mAb in reducing glucose levels.

Methodologically, we read that 1.5 mg/animal of MG53-mAb was administered intravenously. This dose is equivalent to ∼7.5×10⁸ pg/mL, a concentration >10⁶-fold higher than the MG53 protein detected in blood (Figure 3C, ELISA data). Animal safety and efficacy data of repeated MG53-Ab administration with any potential impact on vital organ function is critically important and would be of interest to the community.

Diabetes is a devastating illness with profound impact on well-being and pathobiology. A therapeutic approach involving MG53-mAb is intriguing, though it may not be adequately supported with these current data and analyses. Establishing the safety and efficacy of such an immuno-approach will be an important direction for future studies.