Current perspective on amyloid aggregation accelerating properties of the artificial butter flavoring,diacetyl

Ashish P.Vartak,Swati S.More

The amyloid—what peptide can resist its entropic bliss? Without kinetic barricades and chaperones,most peptides would simply tumble down that precipice.The amyloid-β (Aβ) peptides are understood to underlie the hallmark pathology of Alzheimer’s disease (AD) and are considered one of the causative factors for neurodegeneration and cognitive impairment.AD affects critical connected structures within the brain that are responsible for memory,language,and social behavior.Various isoforms of Aβ peptides are produced by proteolytic cleavage of the transmembrane amyloid precursor protein (APP) by secretases,which dictates the amyloidogenic fate of the released product.Those released from α-secretase cleavage appear to be non-amyloidogenic,while β-secretase cleavage that releases fragments with intact N-termini is amyloidogenic.Further diversification occurs also through a slew of posttranslational modifications,such as methionine25 oxidation and aspartate racemization,producing fragments with varying amyloidogenic tendencies.Amongst all of these isoforms,peptides Aβ1–42and Aβ1–40show higher propensity for aggregation,ultimately forming insoluble amyloid plaques(also termed senile plaques) present in the diseased brain (Hampel et al.,2021).This global minimum energy peptide assembly is stabilized by backbone-to-backbone in-register hydrogen bonding of peptides in the β-strand conformation.The tertiary structure of amyloid fibrils is biophysically described as β-strands perpendicular to the fibrillar axis and exhibiting the characteristic cross-β X-ray diffraction pattern (Nirmalraj et al.,2020).Quaternary structures include soluble oligomers,higher-order structures that form colloids,and yet higher-order insoluble plaque(Figure 1).Each order of amyloid aggregation is bound to exhibit its own pathological (or as more recently emergent,physiological) relevance.Factors influencing the propensity of amyloid conformation include the identity of residues,local environment,concentration,and local primary structure.The residues Arg and particularly Lys have a low β-strand propensity.Their covalent modification at the side-chain amino or guanidyl functions increases hydrophobicity—one of the promoters of the β-strand conformation.

Figure 1 ｜Proposed mechanism for amyloid aggregation induced by diacetyl due to covalent modification of arginine residue.

Diacetyl (2,3-butandione,DA),an avid electrophile,is used widely to flavor food such as microwave popcorn,coffee,cheese,and e-cigarettes,among many others.An odor receptor,odr-10 which is a seven-transmembrane G-protein coupled receptor,specific for DA has been characterized inCaenorhabditis elegans(Zhang et al.,1997).DA bears a structural resemblance with one of the end-products of glycolysis: methylglyoxal(MG).Over a decade ago,we reported upon the ability of DA to covalently modify guanine in DNA(More et al.,2012a).DA is also known to modify Arg and Lys covalently.We then embarked on the study of the effect of this modification on the conformation of the amyloid peptides Aβ1–42and Aβ1–40that constitute protein aggregates found in AD (More et al.,2012b).Genetic and environmental factors,such as the consumption of fatty acids and dairy products,concerning the aggregation process of the Aβ range,have been deliberated.However,the influence of chemical contaminants and toxins on Aβ progression continues to gain traction as well.MG,a reactive carbonyl species,has been previously implicated in Aβ aggregation and cognitive function.Reactive carbonyl species,generally known for their cytotoxicity,form covalent adducts with proteins termed as advanced glycation end products.Structural similarity between MG and DA attracted our attention toward the potential implications of Aβ1–42modification by DA and resultant conformational changes.

Circular dichroism measurements of Aβ solutions exposed to DA indicated that the modified peptides follow an affected conformational path that involves greater and earlier conformational tilt towards the β-strand.Thioflavin-T bindinginduced fluorescence measurements revealed that DA renders the peptides markedly prone to aggregation.The molecular basis of this phenomenon was traced through mass spectrometry to the adduct formation of DA with Arg5of Aβ1–42(More et al.,2012b).DA was found to easily permeate the blood-brain barrier.It not only was resistant to glyoxalase-I the primary detoxification mechanism for MG,but also irreversibly inhibited the enzyme.This spurred us to examine its toxicity in cell culture.DA was found to potently aggravate the toxicity of the Aβ peptides towards cultured SHSY5Y cells,establishing the need for expanded studies in animal models to ascertain whether our cell culture results are relevant in the intact organism.In 2012,we published an opinion advising due caution before extrapolating our results to the intact animal (Vartak et al.,2012).In 2013,the European Food Safety Authority issued a commentary on our paper (Authority,2013),agreeing that the cell culture results were insufficient to draw conclusions about the occupational toxicity of DA.

The constitutive and colligative properties of amyloids are extremely complex and not yet fully described,while the environment of an intact multicellular organism (much less a vertebrate)is far more dynamic than what can be modeled by cells in culture.Classically,it was the insoluble amyloid plaque that was thought to be one of the primary pathogenic entities of AD as the biophysics of amyloid assembly had not been fully elucidated.Consequently,the view of AD-amyloidosis remained rather simplistic.The past decade saw a gratifying increase in our understanding of amyloidosis.At each order of aggregation,the amyloid structure is polymorphic (Riek,2017;Loquet et al.,2018) at the levels of β-strand intersegment interaction,manner of protofilament packing around the fibrillar core,sense of packing of strands (parallel or antiparallel) or even primary structure.Covalent side-chain modification such as the DA—Arg5(Aβ1–42) adduct formation would affect the area preferred by or accessible to Arg5in the (φ,ψ) Ramachandran space that determines β-strand propensity.Second,this interaction would influence side-chain to side-chain interactions that determine intersegment packing.Both effects of DA would affect tertiary structure,and this would permeate into the interaction of protofibrils with the fibrillar core—affecting in turn the topography of the resulting quaternary entity.It is difficult to conjure a scenario where such changes would be of no significance to the physiologic or pathologic role of the resulting amyloid.Even more unlikely is a notion that such effects could be onedimensional (i.e.,either beneficial or pathologic)given the existence of multiple orders of amyloid assembly.

A few years ago,we began describing the effects of DA on AD in a mouse model (APP-PS1 mouse)that overexpresses the APP and a mutant human presenilin (PS1ΔE9) and reliably develops AD at a predictable age (Xie et al.,2021).A second model was one where AD was induced by direct intracerebroventricular infusion of Aβ peptide into wild-type mice.The transgenic APP/PS1 mice were subjected to Morris water maze for evaluation of spatial reference memory and learning.To our surprise,DA-treated transgenic mice did not show exacerbation of cognitive impairment.While a trend toward improved cognitive function was apparent,intersample variability and possibly the dosing regimen choice prevented any statistical conclusions.The intracerebroventricular infusion of Aβ1–42model was then employed to probe this phenomenon in an expeditious and costeffective manner.Similar to our observations in APP/PS1,the T-maze spontaneous alternation test that examines spatial working memory changes,showed better cognitive performance of DAtreated mice compared to A-treated mice.DA treatment restored cognitive impairment induced by intracerebroventricular injection of Aβ1–42and fostered better retention of spatiotemporal clues.The frontal cortex of DA-treated mice was much more burdened with insoluble plaque than their untreated counterparts.On the other hand,the DA-treated mouse brain contained a lower amount of glial fibrillary acidic protein,a marker of inflammation in the AD-afflicted brain.

We studied the effect of DA on Aβ structural dynamics through immunostaining.DA appears to promote (or induce) fibrillation,but not Aβ oligomerization.Here we will use the definition of “fibrils” as well-ordered β-strands that are stabilized by zipper-like packing: H-bonding,specific hydrophobic interactions,and steric alignment.Oligomers are spherules loosely associated through a hydrophobic core,in other words,a micellar assembly.The fibrils formed in the presence of DA appeared to be unstable to sodium dodecyl sulfate,indicating a greater propensity to break into lower-order fibrillary oligomeric species that can potentially be cleared by the body (Figure 1).Transmission electron microscopy showed that the fibrils formed in the presence of DA were notably narrow and elongated (～91 nm) compared to normal rounded,shorter Aβ fibrils formed in the absence of DA(～33 nm).These results suggest that DA adduct with Arg5 of Aβ1–42drives the formation of fibrils that mature without oligomerization,with the net pharmacological impact of all orders of such an assembly being non-or far less pathogenic than normal Aβ1–42aggregation.Soluble oligomers are increasingly being recognized as the toxic amyloid entities of AD (Tolar et al.,2021),and this may be an explanation for the salubrious effect of DA.

Thein vivostudy was able to clarify Aβ1–42and DA interaction observed in our test tube experiments,as well as formulate a better understanding of the conformational pathway and explain the general gap in the translation ofin vitroAD results in an animal.The neuroprotection offered by DA against amyloid toxicity could be explained by the instability of the DA-modified amyloid as seen in dot blot analysis,the inability of this modified amyloid to propagate a perpetual misfolding cascade,and cellular adaptation mechanisms that lead to the neutralization of the toxin,among many others.Further studies such as clearance of DA-modified Aβ will need to be conducted for complete characterization of this phenomenon.However,we realize that the potential utility of DA’s protection against A cytotoxicity,thus against resultant cognitive deficits,is of limited practical utility due to the inherent toxicity of this chemical.DA has been proven to reduce pulmonary function,as factory workers exposed to diacetyl developed bronchiolitis obliterans,or “popcorn lungs”,a generally lethal occupational hazard among popcorn factory workers.

In conclusion,while the respiratory effects of DA are understood to the extent that OSHA has imposed limits on workers’ exposure,very little is known about long-term consequences on other systems—particularly the central nervous system.Spotlight on DA content in prepackaged foods caused its replacement with other dicarbonyls or equivalents such as aliphatic isomers and homologues;2,3-hexanedione,3,4-hexanedione,and 2,3-heptanedione.Apart from food consumables,the utilization of such additives in e-cigarettes has raised safety concerns for the general population,especially since some of these DA alternatives liberate DA upon heating.Regardless,the observations from ourin vivostudies demonstrate a kinetic address to the basic pathobiology of AD.There should be a focus on new drug design that encourages oligomerfree Aβ1–42fibril formation such as that caused by DA.Indeed,studies with small molecules such as 2,8-bis-(2,4-dihydroxy-phenyl)-7-hydroxyphenoxazin-3-one (O4) and the flavanol ZGM1(Zhang et al.,2019) have shown acceleration of amyloid aggregation by skipping the intermediate toxic oligomers as a rescue strategy for cognitive deficits caused by AD pathology.Therapeutic development targeting Aβ for AD until now has relied on strategies to prevent the formation of Aβ1–42or promote its clearance from the brain such as recently approved immunotherapeutics to clear A plaque.Given the rising importance of A soluble aggregates in neurotoxicity,attempts to limit the residence time of such intermediates in brain tissue could also be valuable.Such strategies could provide a new direction for the future design of AD therapeutics.Given the complex pathobiology of AD,we and others believe that multiple approaches to tackle the emerging AD epidemic are urgently needed.Thusly warranted is the search for entities that are able to chaperone non-pathogenic Aβ aggregation as a novel strategy for the prevention and treatment of AD.

This work was supported by the National Institutes of Health Grant(R01-AG062469),the Grantin-Aid of Research,Artistry,and Scholarship program(GIA,Project 143977)at the University of Minnesota,and funding from the Center for Drug Design(CDD),University of Minnesota(to SSM).

Ashish P.Vartak,Swati S.More*

Center for Drug Design,College of Pharmacy,University of Minnesota,Minneapolis,MN,USA

*Correspondence to:Swati S.More,PhD,morex002@umn.edu.

https://orcid.org/0000-0002-8733-2029(Swati S.More)

Date of submission:October 5,2023

Date of decision:November 25,2023

Date of acceptance:December 7,2023

Date of web publication:January 8,2024

https://doi.org/10.4103/1673-5374.392882

How to cite this article:Vartak AP,More SS(2024)Current perspective on amyloid aggregation accelerating properties of the artificial butter flavoring,diacetyl.Neural Regen Res 19(10):2113-2114.

Open access statement:This is an open access journal,and articles are distributed under the terms of the Creative Commons AttributionNonCommercial-ShareAlike 4.0 License,which allows others to remix,tweak,and build upon the work non-commercially,as long as appropriate credit is given and the new creations are licensed under the identical terms.