Decelerate of Amyloid Fibrillation By The Alkaloids

Naturally occurring phytochemical compounds have received considerable attention as alternative candidates for anti-amyloidogenic agents. This study, utilizing human insulin and amyloid beta peptide as an in vitro model, determined the anti-amyloid effects of alkaloids extracted derived from Stephania venosa. Alkaloids extracts including crebanine, O-methylbulbocapnine, tetrahydropalmatine and N-methyltetrahydropalmatine were used. Inhibition of amyloid protein aggregation were studied by fluorescence spectroscopy. Most alkaloids, except N-methyltetrahydropalmatine, exhibited the inhibitory properties against amyloid fibrillation either insulin or amyloid-beta peptide. Among alkaloids group, crebanine and tetrahydropalmatine showed the potent properties of anti-amyloidogenesis. These results suggest that alkaloids could be used as a natural compound for the development of drugs against amyloid protein aggregation for treatment of amyloid-related diseases.


Introduction
Suppression of amyloid protein aggregation is considered as a promising therapeutic approach to prevent or treat amyloidosis-related disorders. One of the strategies aimed to find the effective compound against amyloidogenic activity is to inhibit the toxic amyloid formation and stabilize its native monomeric form or destabilizing the fibrillated misfold form 1 . The protein aggregation phenomena can be used as a model for studying proteins' properties. The ability of various peptides and proteins undergo self-aggregation that leads to the formation of amyloid fibrils. Human insulin is one protein that was chosen and widely used as a model protein for the study of amyloid formation in vitro. Recently, much attention has been paid to find out an inhibitor of insulin amyloid fibrils.
In recent years, natural products are identified and characterized as an impact class of amyloid inhibitor 1 . Several natural polyphenolic compounds have been well-studied as amyloid inhibitors such as epigallocatechin gallate (EGCG), curcumin and resveratrol. Quinones show different inhibitory effects on insulin oligomerization, especially for 1,4-benzoquinone and 1,4-naphthoquinone 2 . Quercetin dosedependently inhibited amyloid formation of insulin via destabilizing the preformed insulin fibrils and transforming the fibrils into amorphous aggregates 3 .
Alkaloids, a class of nitrogen-containing compounds, are found primarily in plants, especially in flowering plants 4 . Alkaloids have a broad spectrum of pharmacological effects including analgesic, antiasthmatic, antiarrhythmic, anticancer etc. 4,5 . Galantamine, the isoquinoline alkaloid family, inhibits A-aggregation and cytotoxicity 4,6 . However, the common side effects of galantamine can be nausea, vomiting, diarrhea, and anorexia. 7 . The in vitro and in vivo study showed that amyloid beta (A) and A-induced neurotoxicity were reduced by the effect of caffeine 8 . In addition, caffeine also reduces levels of A in neuroblastoma-2a cells stably expressing human Swedish mutant APP and protects cerebellar granule neurons and basal forebrain neurons from neurotoxicity caused by A 8,9 . Furthermore, it should find out an additional alkaloid, either novel compounds or an old one, that could be used for inhibiting A-aggregation.
Stephania venosa (Blume) Spreng., belongs to Menispermaceae family. In South East Asian countries, this plant is traditionally used as tonic drug and medication for various diseases 10 . It was found that alkaloids are the main phytochemical compound of this genus 11 . Their biological activities have been reported including anti-cancer activity 12,13 , chemosensitizer 14 and acetylcholinesterase inhibition 15 . Crebanine, a major component of S. venosa, exhibits anti-proliferative and anti-invasive effects on human cancer cells via the cell cycle arrest and apoptosis induction in K562, K562/adr, GLC4 and reduce the expression of MMP-2, MMP-9, uPA and MT1-MMP 13,[16][17][18] . In additions, there are several natural alkaloids, including Omethylbulbocapnine, tetrahydropalmatine, and N-methyl tetrahydropalmatine have been found in the tubes of S. vernosa 12,13 . Furthermore, tetrahydropalmatine inhibits LPS-induced IL-8 secretion via blocking MAPKs signaling pathway 19 . However, there is no report about the anti-amyloid effects of these alkaloids.
In the present study, the effect of alkaloid molecules on protein aggregation were investigated using the amyloid-forming model. Human insulin and amyloids beta peptide fibril formation were generated by incubating at high temperature and acid environment (pH 2.5 and 80°C) and the effect of alkaloids on this fibrillation was measured by intrinsic Tyrosine (Tyr) fluorescence assay and Thioflavin T assay. Therefore, the prevention capacity of alkaloids for amyloid fibrils conversion can be indicated as a therapeutic potential on protein aggregation diseases

Alkaloids extracts inhibited the kinetic of insulin aggregation
To determine whether alkaloids extracts inhibited insulin fibrillation, tyrosine emission spectra of insulin (0.02 mg.mL -1 ) in the presence of alkaloids (0.002 mg.mL -1 ) was observed over a period of 24 h. In the presence of insulin fibrillation, the fluorescence intensity of Tyr was decreasing. This becomes apparent on plotting the fluorescence intensity at 305 nm as a function of time with the half time value of insulin fibrillation (t 0.5 ins ) equal to 0.55 ± 0.11 hour. However, addition of crebanine (A1), O-methylbulbocapnine (A2), tetrahydropalmatine (A3) and N-methyltetrahydropalmatine (A4) to the insulin did not decreased in intensity of the emission at 306 nm over time of incubation. Therefore, the half-time (t 0.5 ins ) of A1, A2, and A4 significantly increased except A3.

Alkaloids extracts inhibited insulin fibril formation
To confirm the presence of insulin fibril formation. Thioflavin T (ThT) fluorescence assay was performed. It was found that ThT binds specifically to the cross-β sheet structure of amyloid fibers and give more intense once bound. In this experiment, after incubation for 24 hours, 20 mM ThT was added and measured the fluorescence intensity at 488 nm after excited with 420 nm. The fluorescence intensity of ThT of insulin incubated with A1, A2 and A3 after incubation was lower than that of insulin control with the relative ThT fluorescence equal to 0.44 ± 0.06, 0.59 ± 0.05 and 0.57 ± 0.07, respectively (Table 1). Interestingly, ThT fluorescence intensity of A4 incubated with insulin did not change when compared with insulin control. Therefore, all alkaloid extracts, except A4, inhibited insulin fibril formation.

Alkaloids extracts inhibited the insulin fibril formation in dose-dependent manner
The previous data demonstrated that most alkaloids, except A4, had a potential as an inhibitor for insulin fibrillation. We performed further experiment to determine whether alkaloid extracts affect the insulin fibrils formation in dose dependent manner, different concentrations of A1-A4 (0.002-0.01 mg.mL -1 ) were added into insulin (0.02 mg.mL -1 ) prior to warming them to 80 °C for 24 h. After incubation, 20 µM ThT was added and the fluorescence intensity at 488 nm (excitation wavelength at 420 nm) were obtained. Relative ThT fluorescence values were calculated which derived from the ratio of ThT fluorescence intensity of insulin in the presence of alkaloids and the ThT fluorescence intensity of fibrils insulin control. Increased concentration of alkaloid extracts, it was found that A1, A2 and A3 was potently inhibited insulin fibril formation in a dose-dependent manner, while A4 did not. Among alkaloids molecules (at 0.002 mg. mL -1 ), A1 might be the most inhibitor of insulin fibrillation. (Figure 1)

Alkaloids extracts inhibited the Aβ fibrillation
To characterize the process of Aβ fibrillation, the kinetic of fibrillization process of Aβ peptide with different mixing ratios of Aβ40 to Aβ42 were performed. Our model based on the finding that there are two main Aβ peptides of different length which involved in Alzheimer's disease, Aβ40 and Aβ42 residues. It was found that mixing of Aβ40 and Aβ42 is enhanced toxicity in the early onset of some familial Alzheimer diseases 20 . Otherwise, our previous study found that the ratio of Aβ40: Aβ42 (1:4) increased the toxicity in neuroblastoma cell line, SK-N-SH, than treated with Aβ40 or Aβ42 alone (data not shown). To mimic the pathology of AD, three Aβ peptides were prepared in 1 µM Aβ40, 1 µM Aβ42, and a combination of Aβ40:Aβ42 (0.2 µM:0.8 µM). The fibrillization of Aβ was observed by an increase in ThT fluorescence due to the binding of the dye to the fibrils. The representative fibrillization curves were shown in Figure 2. Our results showed that Aβ40, Aβ42 and Aβ40:Aβ42 demonstrated similar fibrillization kinetics which exhibit a sigmoidal appearance. It seems to be that the fibril growth rate which represented by the haft time value of amyloid beta (t 0.5 Aβ ) for Aβ42 was shorter than that of Aβ40, indicating Aβ42 exhibited a fast fibrillation rate than Aβ40. Mixing of Aβ40 to Aβ42 seems to decrease the fibril growth rate of Aβ42 comparing with Aβ42 alone (Figure 2d). In the presence of alkaloid extracts (0.002 mg. mL -1 ), the results showed that different alkaloid extracts showed different effect on the kinetic of amyloid beta fibrillations as indicated by half time (t 0.5 ) and the relative ThT fluorescence value as shown in Table 2. Increased half time and decreased relative ThT fluorescence value represent the decreased of amyloid fibrillation formation. The half-time of amyloid formation for Aβ40 was increased from 6.8 h to 8.8 h and 7.7 h in the presence of A1 and A3 respectively. Accordingly, with half-time value, the relative ThT fluorescence value were decreased in the presence of A1 and A3 compared with Aβ control. Therefore, A1 and A3 inhibited the amyloid fibrillation for Aβ40. It was also found that amyloid formation for Aβ42 was inhibited by A2. However, it seems to be that A2 presumably inhibited amyloid formation for Aβ40 and A1 presumably inhibited amyloid formation for Aβ42 as indicated by the decreased of relative ThT fluorescence value comparing with its control.

Discussion
The incidence of amyloid-related diseases has been growing continuously. Finding an effective treatment became more important. Amyloid fibrillated protein has been being the focus of research for many years 21 .
An ever-growing incidence of amyloid-related diseases has led researchers and clinicians to discover the cure. Hence, the propose of drugs that prevented amyloid accumulation may be a potential treatment. Recent researches have focused on natural products to avoid the side effect of the clinical used 22 . Natural products such as flavonoids, alkaloids and curcuminoids have been extensively researched regarding to reduce the amyloid associated toxicity of Aβ 23,24 . In this study, we have proposed alkaloids as an in-situ inhibitor for amyloid protein fibrillation. Four alkaloids derived from Stephania venosa including crebanine, Omethylbulbocapnine, tetrahydropalmatine and N-methyltetrahydropalmatine were used as interested molecules.
We first studied the interaction of alkaloids with human insulin. The reasons why human insulin was chosen as the model protein in this study are as follows; 1) insulin and Aβ protein share a common characteristic. 2) Under appropriate conditions, they both aggregate into amyloid fibrils. 3) Although the proteins do not share sequence homology, they exhibit similar insoluble filaments and fibrillation responses 25,26 . According to the process of insulin aggregation, it proceeds through the dissociation of oligomeric states into monomers, results in conformational changes. This change leads to a stable state by forming fibrous amyloid aggregates rich in β-sheets 27 . In the present work, the aggregation kinetics of human insulin was studied at low pH and high temperatures. Decreased in Tyr fluorescence intensity, an intrinsic fluorophore, was monitored during insulin aggregation that accompany with insulin fibrillation using Thioflavin T. Our results clearly found that most alkaloids inhibited insulin aggregation and fibril formation as dose dependent manner. Omethylbulbocapnine is isomeric with crebanine with different position of the two methoxyl 13 . Both alkaloid molecules exhibited the similar properties of anti-insulin fibrillation. Therefore, different position of the two methoxyl did not affect to the anti-amyloidogenic properties. Interestingly, it was found that the different properties of anti-insulin fibrillation between N-methyltetrahydropalmatine and tetrahydropalmatine. Nmethyltetrahydropalmatine is an analogue of tetrahydropalmatine. We found that the methyl group on the nitrogen atom of N-methyltetrahydropalmatine decrease the capacity of insulin fibril formations. Therefore, the nitrogen atom on tetrahydropalmatine seems presumably to play a role as active site for an inhibitor of amyloid fibril formation.
We successfully demonstrated the ability of alkaloids to inhibit the kinetics of insulin aggregation. We postulated that a similar strategy could be used to study on amyloid beta peptide. According to the evidence of the major form of the Aβ peptide that found in amyloid plaque that showed Aβ40 and Aβ42 form mixed aggregates 28 . It attempted us to investigate the influence of each Aβ peptide on their aggregation kinetics behavior. Kinetic analysis found that Aβ42 exhibited a fast fibrillation rate than Aβ40. However, mixing of Aβ40 to Aβ42 seems to slow down the fibril growth rate of Aβ42 when comparing with Aβ42 alone. The study from Pauwels et al., used the NMR experiments for visualizing the spontaneous aggregation of mixing Aβ40 to Aβ42. It was showed that Aβ40 slows down the aggregation kinetics of Aβ42 29 .
In conclusion, amyloid fibrillation could be monitored by using intrinsic Tyrosine fluorescence accompany with Thioflavin T assay. Alkaloids have shown some promise against amyloid fibrils both in insulin and amyloid beta peptide. Most alkaloids group, except N-methyltetrahydropalmatine, exhibited potent properties of anti-amyloidogenesis. These results suggest alkaloid can be used as the natural compound for the development of drugs against amyloid protein aggregation for treatment of Alzheimer's disease.

Chemical reagents
Recombinant

Kinetics of insulin fibrillation by intrinsic Tyrosine fluorescence
The fluorescence intensity of tyrosine (Tyr) was used to investigate the insulin fibrillation. The insulin fibrillation was performed by using the thermal-induced fibrillation method 30

Effect of alkaloids on insulin fibrillation
Various concentration of alkaloids (0.002-0.01 mg.mL -1 ) were added into insulin solutions (0.02 mg.mL -1 ) prior to warm at 80 °C for 24 h. After incubation, 20 µM ThT was added and the fluorescence emission spectra (excitation wavelength at 420 nm) were obtained. Relative ThT fluorescence value were calculated from the ratio of ThT fluorescence intensity of insulin in the presence of alkaloids and insulin control.

Kinetic analysis of amyloid fibrillation
The Aβ fibrillation was performed in 1 µM Aβ40, 1 µM Aβ42, and a combination of Aβ40:Aβ42 (0.2 µM:0.8 µM). All samples were added in 250 µL of PBS buffer pH 7.4 with 0.05% SDS containing 10 μM of ThT, and then incubated at 40°C. The fluorescence intensity of ThT was taken using a spectrofluorometer (Perkin Elmer LS55) with the emission wavelength at 488 nm excitation wavelength at 420 nm.

Effect of alkaloids on Aβ fibrillation
The experiments were performed by co-incubating of Aβ40 (1µM), Aβ42 (1µM) or Aβ40:Aβ42 (0.2:0.8 µM) with alkaloids (0.002 mg.mL -1 ) at 40 o C and used 10 μM of ThT for fibrillation analysis. The fluorescence intensity of ThT was measured at 488 nm when excitation at 420 nm using a spectrofluorometer. The kinetics of Aβ fibrillization could be described as sigmoid curves and the aggregation parameters were determined by fitting the plot of fluorescence intensity versus time as indicated in Figure 4b. The fibrillation rate presented in the haft time value of amyloid beta (t 0.5 A ) was used for data analysis. Efficiency of alkaloids to inhibit the

Statistical analysis
All data are expressed as mean ± standard deviation (SD). Statistical significance was determined using Student's t-test between the groups treated and the control. A probability (p) value less than 0.05 was considered statistically significant.      The kinetic of insulin brillation monitoring by the uorescence intensity of tyrosine. (a) and the kinetic of amyloid beta brillation monitoring by the uorescence intensity of Thio avin T. (b)