In , we performed a full bioinformatic analysis of the SNCA
5′UTR demonstrating by computer-mediated predictions [50
] that the 5′UTR of the SNCA
transcript folds into a unique RNA stem loop resembling an iron-responsive element (IRE) RNA structure that is related to, but distinct from, the H-ferritin and APP 5′UTR-specific IREs [33
]. We are currently testing the capacity of intracellular iron chelation with desferrioxamine to repress neuralα-
synuclein translation acting via the IRE in the 5′UTR of its transcript, as we reported for the APP and ferritin mRNAs [42
Figure 1 The 5′untranslated region (5′UTR) of the Parkinson's disease alpha-synuclein (SNCA) transcript is homologous to the iron-responsive element (IRE) in H-ferritin mRNA. (a) Top panel: The SNCA 5′UTR is encoded by exon-1 and exon-2 (more ...)
Previously, the RNA-directed anticholinesterase drug phenserine, together with its cholinergically inert chiral (+) enantiomer, posiphen, which are both in clinical development for AD, was shown to therapeutically limit brain Aβ
levels in wild-type and AD mouse models [46
]. This action was mediated, in whole or in part, by lowering the rate of synthesis of APP, from which Aβ
is proteolytically cleaved. Here, we sought to demonstrate that phenserine and posiphen, likewise, blockedα-
synuclein expression via their related 5′UTRs, encoding variant versions of the iron-responsive element RNA elements that potentially bind iron-regulatory proteins.
To elucidate whether our defined target would translate across species, we provide the results of an evolutionary evaluation of the conservation of IRE RNA stem loops in SNCA
mRNA as shown in . This is the RNA secondary structure sequence, together with the APP 5′UTR, that is targeted by phenserine and posiphen, as shown in Figures –. The alpha-synuclein-specific IRE stem loop was formed at the splice junction of the first two exons in SNCA
]. We also compared the predicted structure of the SNCA
mRNA IRE with the canonical IREs in the H-ferritin and APP transcripts, which are transcribed from the single first exon of their genes confirming the uniqueness of translational repression of SNCA
mRNA via its 5′UTR.
The anticholinesterase phenserine and its (+) enantiomer, posiphen, are proven APP 5′UTR mediated drugs with known pharmacokinetics in rodents and humans and identified target concentrations [46
]. Since we anticipated that both agents would be active in limitingα-
synuclein translation via its 5′UTR, we had spiked an FDA library with posiphen and phenserine when we formerly screened against the SNCA
5′UTR RNA target [34
]. Here, we confirm that both posiphen and phenserine repressed the SNCA
5′UTR-driven translation of a luciferase reporter in stable cells lines. Their capacity to inhibit SNCA
mRNA translation is similar to that of certain other defined FDA drug leads, including three glycosides and an immunosuppressant, mycophenolic acid (secondary Fe chelator), as we previously reported [34
In , the 5′UTRs of both APP and SNCA
showed 40% homology to each other and also 50% homology to the IRE in H-ferritin mRNA. Multiple Western blot experiments were hence conducted to determine the impact of phenserine compared to posiphen to limitα-
syn compared to APP expression. In this regard, SH-SY5Y cells were treated with both compounds for 48 hours with concentrations ranging from 0 to 10μ
M. Viability studies determined that this range was well tolerated, in accordance with prior studies. In general and as shown in , posiphen (in addition to but potentially slightly more potently than phenserine) decreased levels ofα-
syn in a dose-dependent manner in cultured neural cells (SH-SY5Y) as previously reported for APP. Whether this higher potency would translate to primary neurons and in vivo
is a focus of future studies. In this paper, this was achieved with a preliminary determined IC50
in the order of 5μ
M, in the absence of toxicity.
Multiple transient transfections of SH-SY5Y cells were performed with the constructs that either translated luciferase driven by the 48 base SNCA 5′UTR (SNCA-5′UTR-pGL3) or the empty pGL3 expression vector (Figures and ). As shown (Figures and ), posiphen 50% repressed SNCA 5′UTR-conferred translation of a luciferase reporter transcript. In this regard, posiphen proved a highly selective inhibitor of SNCA 5′UTR driven activity since this chirally pure compound inhibited SNCA 5′UTR-driven luciferase expression in H2A neural cells (i.e., SNCA 5′UTR-positive stably transfected neural cells). By contrast, phenserine and the known APP translation blocker, compound number 9 (included as a comparator), did not suppress SNCA 5′UTR conferred translation in H2A cells. Indeed, phenserine and compound number 9 elevated SNCA 5′UTR-conferred translation.
These data support the mechanism-of-action of posiphen as a highly selective blocker of SNCA
5′UTR activity. However, phenserine—with the identical chemical structure but in a different three-dimensional (chiral) configuration—that has previously been shown to effectively inhibit translation driven by the APP 5′UTR clearly has different actions to posiphen at the SNCA
5′UTR. From this, we can deduce that the element of the SNCA
5′UTR targeted by posiphen has a stereospecific component. Additionally, since phenserine lowersα-
syn levels (), further SNCA
RNA sequences are likely involved in controlling this pathway ofα-
syn translational regulation.
Extending beyond transformed neuronal cell lines, primary neurons from wild-type mice
and PAC SNCA
transgenic mice (PAC-Tg SNCA
) were evaluated for the capacity of posiphen to repressα-
syn expression, as shown in (tested at 100
nM drug concentration). Posiphen proved not only active and well tolerated in SH-SY5Y cells but consistently reduced humanα-
synuclein expression in primary neurons (E18) at doses as low as 1
uM (75% reduction, not shown) without toxicity. This margin appeared to be greater than its capacity to lower APP production (20%) (utilized as a positive control) in these same cells (data not shown).
Following oral administration of posiphen to rodents, dogs, and human, the compound is subjected to metabolic processes and generates the same metabolic profile across these species. Specifically, via a phase 1 metabolism, posiphen undergoes N-demethylation in both the N1 and N8 positions to generate the respective primary metabolites, N1-norposiphen and N8-norposiphen (). Each then undergoes further N-demethylation to generate the common metabolite, N1, N8-bisnorposiphen. Unlike phenserine, posiphen is devoid of cholinesterase inhibitory activity and, therefore, can be advantageously administered at higher clinical doses (in the order of 5, to 8-fold greater).
Figure 5 Metabolic analogs of posiphen and their respective anticholinesterase activities . Posiphen is devoid of anticholinesterase activity. However, its phase 1 metabolites, N8 demethylated, N1 demethylated, and di-demethylated N1, N8-bisnorposiphen showed (more ...)
As shown in , specific N-demethylated metabolites (in particular the N1-nor and N1,N8-bisnorposiphen) possess potentially clinically relevant IC50
values to inhibit acetylcholinesterase (AChE). Such activity is less than phenserine and would be expected to have a slow onset in line with the time-dependent metabolism of posiphen to generate its metabolites. However, with regard to actions onα-
synuclein expression, activity of the metabolites at this target, where posiphen is potent, could usefully increase and extend the drug's in vivo
efficacy. To elucidate this and aid planning for future animal studies, we characterized the relative capacities of posiphen's metabolic analogs to impactα
-syn expression ().
shows a Western Blot analysis representative of three experiments that compared the efficacy of posiphen with that of its three primary metabolites to limitα
-synuclein expression ex vivo
using E18 primary cortical neurons from human SNCA
PAC mice. We consistently measured that N1-norposiphen
(possessing AChE activity) proved most potent to limitα-
synuclein expression (by 50%), and, likewise, APP expression was also reduced. As assessed at 100
nM in , the other metabolites proved to be less active at these targets, indicating that relatively small structural changes (i.e., the loss of a methyl moiety) have significant impact. An assessment of dose-response for each metabolite is a focus of future studies, particularly within the achievable clinical range of the agents in human and in vivo
range in animal models (Dr. Maria Maccecchini and Dr. Robert Nussbaum, personal communications).