In the present study we have endeavored to understand the molecular basis for the onset of insulin deficiency in the syndrome of M
iabetes of Y
. Proinsulin mutants causing the MIDY
syndrome are distinct both from insulinopathies previously described as associated with adult onset-associated diabetes 
and from recessive alleles that result in neonatal diabetes through reduced insulin expression 
. Unequivocally, MIDY
mutations cause misfolding of mutant proinsulin, blocking its progression to insulin 
(this report, Figure S1
), and activating ER stress response pathways 
(this report, ).
A prevailing hypothesis is that ER stress-mediated death of pancreatic beta cells with a resultant loss of beta cell mass triggers diabetes onset in MIDY 
. Indeed, in Akita
mice, there is little dispute that ultimately, after diabetes progresses, there is a significant loss of beta cell mass 
; presumably this is the case in human MIDY
, just as is the case in type 2 diabetes 
. However, decreased insulin production and diabetes in Akita
mice, which is linked to inadequate wild-type proinsulin delivery to secretory granules 
, may occur before there is any loss of pancreatic beta-cell mass 
. We now find that misfolding of the wild-type gene product is induced by the presence of the mutant gene product. Additionally, regardless of whether such mutations result in the loss, or creation, of an extra unpaired cysteine within proinsulin, we find that Cys residues play a critical role in the dominant-negative blockade of insulin production from the wild-type allele. Furthermore, while each of the MIDY
mutants does cause ER stress and ER stress response, we find that ER stress and ER stress response alone cannot efficiently block insulin production from wild-type proinsulin and thus do not appear sufficient to account for the initiation of insulin deficiency.
Whereas a previous study using a two-dimensional non-reducing/reducing gel system failed to obtain evidence that the Akita
mutant proinsulin promoted formation of aberrant disulfide bonds 
, we believe that a simpler, single-dimensional system of analysis results in a more robust assay 
; demonstrating that MIDY
mutants are predisposed to form aberrant disulfide-linked protein complexes () and that MIDY
mutants exhibit selective perturbation of the intracellular trafficking of co-expressed wild-type proinsulin (, ). This dominant-negative action precedes impairment of cellular ATP production or cell viability as it occurs in living cells that are translating and secreting proteins. Thus, we conclude that the following events are among the earliest in the molecular pathogenesis of MIDY
. Specifically, shortly after expressing any of the MIDY
mutants, the general ER export pathway for secretory proteins remains functional () even as wild-type proinsulin begins to be recruited in abnormal disulfide-linked protein complexes (), impairing its delivery to secretory granules () and thereby impairing wild-type insulin secretion (). This series of defects appears sufficient to account for insulin deficiency that is already in evidence on postnatal day 1 in Akita
mice (for review, see 
Our evidence favoring increased recruitment of wild-type proinsulin into aberrant disulfide-linked proinsulin-containing protein complexes, which serves to entrap wild-type proinsulin within the ER, is highly reminiscent of results from recent studies perturbing endogenous level of ERO1-beta (the endocrine pancreas-specific disulfide oxidase) 
. In these knockout mice, a diabetic phenotype is linked to abnormally-increased proinsulin recruitment into interchain disulfide-linked adducts within the ER of pancreatic beta cells, accompanied by deficient proinsulin delivery to secretory granules for production of mature insulin 
. These findings point strongly to the idea that perturbation of the proinsulin folding pathway, with or without mutations in the proinsulin coding sequence, is sufficient to trigger diabetes onset.
While it may be easy to dismiss as trivial the structural basis for aberrant intermolecular thiol attack by MIDY mutants that add or remove a cysteine (), it is more challenging to obtain data providing structural evidence for why non-cysteinyl proinsulin mutants are also predisposed to form aberrant disulfide-linked protein complexes in vivo (). In one particular example studied in detail, we examined insulin-G(B23)V chemically-synthesized from an insulin scaffold with native disulfide bonds already in place. Remarkably, insulin-G(B23)V has a largely native structure () with essentially normal thermodynamic stability — even as the isolated B-chain bearing G(B23)V was essentially completely blocked in assembly with the A-chain to yield insulin. Since chain alignment leading to interchain disulfide bond formation is the crux of the chain-assembly assay, the data strongly suggest that there is an inability to create interchain disulfide pairing, and this kinetic blockade in the folding of the insulin chains (rather than instability of the native state) leads to enhanced cysteine thiol exposure that can promote intermolecular attack. Thus far, structural analyses of non-cysteinyl MIDY mutants remain quite limited. Nevertheless, from the experiments performed to date, we posit that MIDY mutations act as kinetic blocks in the folding pathway to proinsulin disulfide bond formation, resulting in free thiol availability that is a key to the molecular pathogenesis of MIDY.
In support of this hypothesis, we find that proinsulin-DelCys — despite being the most misfolded of all proinsulin mutants and completely blocked within the secretory pathway — cannot induce recruitment of co-expressed wild-type proinsulin into aberrant protein complexes (), cannot efficiently impair intracellular transport of co-expressed wild-type proinsulin () and thus cannot block insulin production or secretion (). These findings emphasize that at least one Cys residues is required for efficient dominant-negative blockade of insulin production from the wild-type allele, leading to an unfavorable chain of molecular events that results in progressive ER stress and beta cell failure. Unmistakably, ER stress and ER stress response are important consequences of proinsulin retention in the ER (). However, ER stress and ER stress response alone cannot efficiently block production of insulin from wild-type proinsulin (). Thus it is not clear that ER stress, ER stress response, and loss of beta cell mass can account for initial pancreatic insulin deficiency that leads to the onset of diabetes in Akita mice, a model of MIDY.
, the severity of phenotypes may be linked to the degree of folding (and secretion) disturbance 
. Of the mutations classically associated with adult-onset diabetes 
, curiously, proinsulin-F(B24)S exhibits a more perturbed distal B-chain structure (Figure S3
), a twofold decrease in insulin yield from the chain-assembly assay, a partial defect for secretion (Figure S1A
), partial engagement in disulfide-linked protein complexes in the ER (), a partial dominant-negative effect on insulin production (), partial recruitment of wild-type proinsulin into disulfide-linked complexes (), and partial activation of ER stress response (). Loss of F(B24) may de-stabilize the native-like cluster of hydrophobic side chains near C(B19) and C(A20), decreasing the efficiency of disulfide pairing 
. These findings appear consistent with a spectrum in the molecular pathogenesis of early-onset and late-onset diabetes caused by autosomal dominant INS
gene mutations, ranging all the way to proinsulin-G(C28)R 
which ultimately generates perfect human insulin lacking any mutation, does not use the MIDY
mechanism (this report) and instead operates either through novel mechanisms involving the mutant C-peptide 
or is coincidental to the pathogenesis of diabetes.
In summary, we have defined the molecular pathogenesis of MIDY as a syndrome in which mutant proinsulins use unpaired cysteine residues to recruit nonmutant proinsulins into disulfide-linked complexes, blocking insulin production that leads to insulin-deficiency, beta cell ER stress, and diabetes. Uncovering the earliest events in the molecular mechanism of the disease may help in identifying therapies designed to rescue proinsulin folding in the ER of pancreatic beta-cells.