14 May 2021

Part 2 of two

Polyamines, the nitrogen-packed aliphatic molecules produced by our own cells and found in the foods we eat, have been cast by some as anti-aging dynamos. Studies suggest the compounds, most notably spermidine, counter cognitive slippage in animal models by revving up autophagy and mitochondrial function in the brain. One pilot trial hinted that a daily spermidine supplement could impart similar benefits in people with subjective cognitive complaint (see Part 1 of this two-part story). Could it also ward off Aβ and tau shenanigans? That’s unknown, but it might be time to take a closer look at these ubiquitous molecules. Several new studies reported that, in mouse and cell culture models, polyamines promote degradation of soluble Aβ, suppress tau aggregation, and temper neuroinflammation. What’s more, postmortem studies indicate that polyamine metabolism is cranked into high gear in the AD brain, a sign of a chronic activation of the so-called polyamine stress response. At the moment, the consequences of further upping polyamines in the AD brain are unclear.

Polyamines are perhaps best known as activators of autophagy, one of the molecular digestion pathways whereby cells metabolize and recycle old and unwanted components. Autophagy wanes with age and even more so in AD, and some researchers believe this may contribute to the accumulation of toxic forms of Aβ and tau that are deposited as amyloid plaques and neurofibrillary tangles, respectively (Mai et al., 2019; Zare-Shahabadi et al., 2015; Nixon and Yang, 2011). Spermidine boosts autophagy and a specialized version of it, mitophagy, which recycles dysfunctional mitochondria. In so doing, spermidine bolsters overall mitochondrial function in the brains of aging mice and flies (see Part 1). 

Polyamines and AD Pathology
Might this autophagic boost also promote riddance of the proteopathic proteins that pile up in neurodegenerative disease? Researchers led by Marina Jendrach and Frank Heppner addressed this question in a manuscript they posted to bioRxiv last December. First author Kiara Freitag and colleagues asked how spermidine supplementation would influence Aβ deposition in APP/PS1 mice. When the animals were 1 month old, prior to noticeable Aβ deposition, the researchers added spermidine to the drinking water to a final concentration of 3 mM. They found that at 4 and 9.5 months old, mice imbibing the polyamine had less soluble Aβ40 and Aβ42 in the brain than did untreated mice. Spermidine did not influence levels of insoluble Aβ, nor the number or size of Aβ plaques. This suggested that spermidine had somehow promoted the clearance of only soluble Aβ species.

Despite the plaques’ persistence, spermidine treatment dampened neuroinflammation. In the brains of treated 9.5-month-old APP/PS1 male mice, the scientists measured lower levels of pro-inflammatory cytokines typically released by microglia, including IL-6, TNF-a, and IL-12. They have yet to test female mice. Added to cell culture, the polyamine prevented astrocyte and microglial reactions to generic inflammatory stimuli including lipopolysaccharide and poly I:C. Spermidine quieted inflammation at the transcriptional level, by dampening the activity of the pro-inflammatory transcription factor NF-kB. Spermidine also calmed microglia that had previously been riled by LPS or Poly I:C.

When added to these activated cells, spermidine threw a wrench into the assembly of the inflammasome, an IL-1β- and IL-18-producing apparatus that takes inflammation to the next level and exacerbates neuropathology (Dec 2012 news; Nov 2019 news). 

Notably, the researchers found that autophagy was required for spermidine’s anti-inflammatory effect. When they knocked down Beclin-1, a protein required for autophagy, spermidine failed to rein in the glial release of pro-inflammatory cytokines. Likewise, the polyamine offered no further anti-inflammatory benefit if the authors had first starved the cells—a classic trigger of autophagy.

Heppner and colleagues propose that spermidine’s induction of autophagy both calms neuroinflammation and promotes clearance of potentially neurotoxic soluble Aβ species. The findings mesh with a report that spermidine quelled expression of IL-1β and IL-12 in macrophages in the experimental autoimmune encephalomyelitis mouse model of multiple sclerosis (Yang et al., 2016). 

Curiously, endogenous polyamine levels rise in response to Aβ accumulation, as recently reported in a study of APP knock-in mice. In those animals, which carry copies of APP containing the Swedish, Arctic, and Austrian mutations that cause familial AD, plaque-associated microglia were stuffed with spermine (Feb 2021 news). The authors of that study, led by Pascal Sanchez at Denali Therapeutics in San Francisco, interpreted the massing of polyamine as a sign of microglial stress amid amyloidosis.

Polyamine Stress Overload in Alzheimer's
In fact, a so-called polyamine stress response (PSR) is triggered in the brain in response to different stressors, including traumatic brain injury, neurological disorders, and emotional distress (Zahedi et al., 2010; Gross and Turecki, 2013). The PSR brings with it rising polyamine production, which may initially serve a protective role. However, some researchers claim that, as with any stress response, chronic activation has the potential to become harmful (Gilad and Gilad, 2003). 

Is the PSR active in AD and, if so, does it influence pathogenesis? Researchers led by Daniel Lee at the University of Kentucky, Lexington, addressed these questions in a study published February 15 in the Journal of Clinical Investigation. First author Leslie Sandusky-Beltran and colleagues started by comparing expression of polyamine metabolism genes in 10 people with AD versus eight controls. They measured levels of 26 transcripts in hippocampal slices taken postmortem.

What did they find? For you to appreciate the answer, we can no longer delay the inevitable—a quick, hopefully painless, overview of polyamine metabolism. No chemical formulas required. Here goes … It all starts with arginine, which the enzyme arginase converts to ornithine (see image at right). In what is considered the rate-limiting step of polyamine production, ornithine decarboxylase then turns ornithine into putrescine. This diamine morphs sequentially into the polyamines spermidine and spermine, courtesy of the enzymes spermidine synthase (SRM) and spermine synthase (SMS), respectively. In short, starting with arginine, cells can make spermidine and spermine.

However, the supply is finite. These polyamines rarely accumulate, because other enzymes join in and either acetylate them, or “back-convert” them into putrescine. What’s more, all these steps are in delicate balance. This is where the biochemistry gets fiendishly complicated. Ornithine decarboxylase can be blocked by any of three antizymes, called ornithine antizyme 1, 2, or 3 (OAZ1-3). Antizymes are a class of small inhibitory proteins. They not only block ornithine decarboxylase activity but promote its degradation by the cell’s proteasome. Antizymes have also been linked to degradation of other targets (for a review see Mangold, 2005; Kahana et al., 2009). In turn, OAZs 1-3 are themselves held in check by two antizyme inhibitors (AZIN1/2), which block the binding of OAZs to ornithine decarboxylase (Wu et al., 2015). 

Still with us? Essentially, Lee and colleagues discovered a top-to-bottom perturbation of the PSR pathway in people with AD. Most genes involved were upregulated relative to controls. Expression of both ornithine decarboxylase and AZIN2 more than doubled in AD compared to controls, suggesting that the polyamine stress response was cranked into high gear. The researchers also found more AZIN2 protein in AD brains. Since AZIN2 blocks the ornithine decarboxylase antizymes, the upshot would be ramped-up production of putrescine, followed by spermidine and spermine. Alas, Sandusky-Beltran and colleagues did not measure polyamine levels in the brain samples. Therefore the net result of these gene expression changes remains unclear, especially since genes encoding enzymes that acetylate or back-convert polyamines were also elevated in AD brain.

Even so, these findings suggest that the polyamine train may be running off the rails in AD. What’s more, they dovetail with a recent analysis from the Baltimore Longitudinal Study of Aging (Mahajan et al., 2020). Researchers led by Madhav Thambisetty at the NIA, Bethesda, Maryland, compared levels of metabolites in the brains of 17 people with AD, 13 asymptomatic people with AD neuropathology, and 13 controls. They found more spermidine in the brains of people with AD, at levels that correlated with the case’s burden of neuritic plaques and neurofibrillary tangles.

In this study, too, none of this seems straightforward. Using gene expression analysis, Thambisetty and colleagues found that enzymes that acetylate or back-convert polyamines were also elevated in the AD brain. By contrast, SRM, which converts putrescine to spermidine, was down. This hints that any rise in polyamine levels is likely greeted by an uptick in their acetylation, rendering them inactive.

The findings raise the question of whether simply giving people with AD polyamine supplements is an effective therapeutic strategy, Lee and other researchers acknowledged. They do, however, introduce a metabolic regulatory network that could be more deeply explored in search of a druggable target.

Do Polyamines Influence Tau?
To investigate the relationship between polyamine dysregulation and tau pathology, Lee and colleagues used an adeno-associated virus to overexpress the antizyme inhibitor AZIN2 in the hippocampi and cortices of PS19 mice. These animals express human tau bearing the P301L mutation that causes frontotemporal dementia. Compared to PS19 mice with normal AZIN2 expression, those churning out the enzyme had roughly double the amount of phospho-tau in their hippocampi and cortices.

Upping AZIN2 expression had no effect on wild-type mice, but it exacerbated deficits in memory and worsened anxiety in the 8-month-old tauopathy mice.

In both wild-type and PS19 mice, AZIN2 overload triggered an uptick in putrescine and spermine, though not spermidine. Acetylspermidine, however, ramped up in control mice and even more so in PS19s. In summary, the findings suggested that overproduction of polyamines and their acetylated counterparts exacerbated tau pathology and behavioral symptoms in PS19 mice.

But how? Polyamines have positive charges evenly distributed along their length, Lee explained. This enables them to interact with a wide array of negatively charged molecules, including nucleic acids, phospholipids, acidic proteins, ion channels, polysaccharides—maybe even tau. In its native state, the microtubule binding protein is positively charged, but when hyperphosphorylated, its net charge turns negative, and it is this form of tau that tends to aggregate. In multiple in vitro assays and in a tau biosensor cell line, the researchers found that spermidine and spermine prevented tau oligomerization and fibrillization, while acetylated versions of the polyamines—which had increased in the AD brain—either had no effect, or exacerbated aggregation. The findings go hand in hand with a previous report from Lee’s lab. He found that knocking out spermidine/spermine acetyltransferase, the enzyme that acetylates polyamines, ameliorated tau pathology and its detrimental consequences in the rTg4510 mouse model of tauopathy (Sandusky-Beltran et al., 2019). 

“Lee’s work confirms the causal link between polyamine metabolism and neuropathology in AD, and reinforces the importance of thoroughly studying all the mechanisms involved in synthesis and degradation, and how they can be manipulated in order to use them as drugs for AD and other neurodegenerative diseases,” commented Elvira Leonibus of Telethon Foundation, Pozzuoli, Italy, who recently reported that spermidine supplementation slows cognitive decline in mice by enhancing autophagy (see De Risi et al., 2020). However, a recent computational study suggested polyamines promote the fibrillization of hyperphosphorylated tau (Ivanov et al., 2020). 

Overall, the growing collection of studies underscores the complexity of the polyamine metabolism and its relationship to neuropathology, Lee noted.

Polyamines for AD?
A handful of studies suggest supplementation with spermidine might counter cognitive decline with age (see Part 1) but not everyone is convinced that this is a sound therapeutic strategy for AD. Baruh Polis of Bar-Ilan University in Safed, Israel, noted that while brain polyamines levels drop during healthy aging, they rise in the AD brain. Polis takes the view that a chronically activated polyamine stress response is central to AD pathogenesis (Polis et al., 2021). He noted that arginase—the first enzyme in the polyamine pathway—is elevated in the AD brain, and favors using arginase inhibitors, such as norvaline, to temper this enzyme. His studies suggest that arginase inhibitors have beneficial effects in mouse models of AD (Polis et al., 2019). 

Running counter to this idea, another study from Lee’s lab found that knocking out one copy of the arginase 1 gene in myeloid cells exacerbated Aβ deposition, neuroinflammation, and behavioral impairment in a mouse model of amyloidosis (Ma et al., 2021; Ma et al., 2021). Lee also noted that arginase 1 activity, and production of putrescine, are important for macrophages to phagocytose apoptotic cells (Yurdagul et al., 2020). He wondered if the arginase/polyamine pathway could influence microglial phagocytosis, which was recently reported to be instrumental in the building of Aβ plaques (Apr 2021 news). 

More detailed analyses—ideally at the single-cell level—are needed to sort out how polyamines influence neuroinflammation and neurodegeneration, wrote Jendrach, Freitag, and Heppner in a joint comment to Alzforum. Specifically, studies could explore how different cell types, such as neurons and glia, respond to spermidine supplementation in the AD brain, they wrote.

Polyamines Playing the Field?

In addition to physically interacting with tau, polyamines also bolster mitophagy, and in so doing, cull the mitochondrial herd to enhance metabolism (see Part 1). Curiously, pathological tau species have been implicated in blocking mitophagy (Feb 2019 news). “Knowing that pathological tau impairs not only mitophagy, but also mitochondrial transport, mitochondrial dynamics, and oxidative phosphorylation, it would be great to determine which of these aspects spermidine can rescue,” wrote Jurgen Götz of University of Queensland in Brisbane, Australia, who led one of the studies implicating tau as a mitophagy blocker.

Last but not least, there is also some evidence that polyamines play a hand in other neurodegenerative proteinopathies. Maj-Linda Selenica, who co-leads a lab with Lee, recently reported that eIF5α, a protein-elongation factor that requires spermidine to function properly (see Part 1), binds to TDP-43 and regulates its localization and aggregation (Smeltzer et al., 2021). Suppression of the related translational initiation factor eIF2α has been linked to neurodegeneration in AD and frontotemporal dementia/ALS models, hinting at the importance of protein synthesis responses in neuronal health (Feb 2021 news; Dec 2013 news).—Jessica Shugart