Tjernberg peptide: a double edged sword in Alzheimer’s disease

Abstract

Alzheimer’s disease is a neurodegenerative disorder affecting millions of people worldwide, clinically manifested by the presence of amyloid plaques and neurofibrillary tangles. Senile plaques are composed of amyloid beta protein, while neurofibrillary tangles are formed by the hyperphosphorylation of tau protein. A plethora of reports on the anti-oxidant and pro-oxidant properties of amyloid beta peptides are available. However the molecular candidates involved in bringing about this therapeutic behaviour have not been explored. To investigate this phenomenon we have used a pentapeptide sequence, KLVFF, derived from the core-recognition motif of amyloid beta peptides to study the altered signaling cascade in neuronal cells. Our data showed the unique dual behaviour of the KLVFF peptide as a pro-oxidant and toxicant based on the dosage concentration. The peptide’s inherent ability to scavenge free radicals at low concentrations <100 μM to offset oxidative stress was proved by the down-regulation of SOD1, AP-1 and FoxO3a genes. However at concentrations >100 μM, the cytotoxic effect of the peptide dominates, leading to apoptosis through the activation of p53, ERK1 and p38 in a caspase-dependent mechanism accompanied by mitochondrial membrane depolarization. The therapeutic role of the KLVFF peptide stems out from the regulation of the SOD1 gene by AP-1 and NF-κB and the Nrf2 gene to regulate the intracellular glutathione levels. Collectively our experimental findings revealed a threshold concentration of 100 μM, beyond which the KLVFF peptide mimics the amyloid beta of senile plaques, which can be used as a model system to understand the pathological role of amyloid beta peptides. Concentrations below 100 μM may be actively employed for therapeutic applications to prevent the further aggregation of amyloid beta.

---

1. Introduction

The extracellular deposition of amyloid beta peptide 39–43 aminoacid residues, derived from the amyloid precursor protein (APP) is a characteristic feature of the amyloidosis condition during Alzheimer’s disease.¹ The formation of insoluble toxic amyloid beta occurs through an amyloidogenic pathway, where APP undergoes sequential cleavage by β and γ secretases.² APP is an integral transmembrane protein present in neuronal cells, which upon cleavage by α and γ secretases yields a non-toxic soluble fragment in healthy brain through the non-amyloidogenic pathway.² Mutations in presenilin genes 1 and 2 result in altered proteolytic processing of the APP protein through the amyloidogenic pathway.² The presence of beta sheet structures in amyloid beta peptides results in aggregation with each other, leading to plaque formation.

Peptide based therapeutic systems have emerged as a therapeutic solution to disrupt the amyloid plaques or prevent the aggregation of plaques. The pentapeptides LPFFD (Soto peptide) and KLVFF (Tjernberg peptide) have been used as beta sheet breakers to disrupt amyloid protein aggregation.³ The KLVFF peptide was derived from the amyloid beta peptide residues 16–20 and hence possesses the core recognition motif that enables it to selectively interact with the amyloid beta peptide, thereby preventing further aggregation.³ The LPFFD peptide has been extensively used to disrupt plaques and improve cognitive deficits in vivo.⁴ Modified KLVFF derivatives with either positively charged amino acid residues (KLVFFKKKK) or negatively charged amino acid residues (KLVFFEEE) in the C-terminal have been shown to prevent protein aggregation efficiently in vitro.⁵ In silico studies have suggested that KLVFF can act as a therapeutic peptide by selective binding to the core-recognition motif of amyloid beta peptides, thus preventing its further aggregation.⁶ Surprisingly Hamley et al., have reported that the KLVFF peptide and its PEGylated analogue (KLVFF-PEG) tend to self-assemble into higher order fibril structures in phosphate buffered saline.⁷ The mechanism of self-assembly followed by the KLVFF peptide mimics the in vivo amyloid plaque formation and hence can serve as an effective model system to understand the pathogenesis of amyloid beta aggregation in vitro. Thus it appears that the KLVFF peptide possesses a dual ability to block as well as promote aggregation of amyloid peptides under critical concentration gradient. The mechanism underlying the transition behaviour of peptides, along with the molecular targets involved, remains unexplored in the literature, which forms the major focus of the present study.

Literature reports suggest that the amyloid beta peptide possesses a “Jekyll and Hyde” property with a neuroprotective effect at concentrations in the nanomolar range and a neurotoxic effect at concentrations beyond the nanomolar range.⁸ Many conflicting reports exist for the pro-oxidant and anti-oxidant properties of amyloid beta peptides.⁹ However, the key players involved in the neuronal signaling pathway that are responsible for these complementary effects have not been elucidated thus far. In the present study, we have used the KLVFF peptide, derived from the core-recognition motif of amyloid beta, to understand its concentration-dependent effects on neuronal cells. Since the KLVFF motif in amyloid beta peptides has been identified to be crucial for beta sheet formation, the results obtained in this study can be extrapolated to the effects caused by amyloid beta peptides. Earlier, our group showed that KLVFF at low concentrations can chelate metal ions, thereby overcoming metal ion-induced cytotoxicity, while at high concentrations it aggravated the toxic effects of the metal ions [unpublished work]. However the signaling cascade involved in cell survival or death, induced by the KLVFF peptide in the absence of external stress, has not been explored thus far and this forms the major crux of the present study.

2. Materials and methods

2.1. Materials

Copper chloride (CuCl₂), iron(III) chloride (FeCl₃) and zinc chloride (ZnCl₂) were purchased from Merck, India and were used without further purification. H₂N–KLVFF–CONH₂, a pentapeptide with >95% purity, was purchased from Bioconcept Labs Pvt. Ltd (Gurgaon, India). 2′,7′-Dichlorofluorescein diacetate (DCFDA) of >97% purity used for reactive oxygen species estimation, 5′,5′-dithiobis(2-nitrobenzoic acid), also known as Elman’s reagent (>98% purity), used for the measurement of intracellular glutathione, 2-thiobarbituric acid and trichloroacetic acid used for the TBARS assay, and 2,2-diphenyl-1-picrylhydrazyl of >95% purity used for the DPPH scavenging assay were procured from M/s. Sigma-Aldrich, USA. Fetal Bovine Serum (FBS), Dulbecco’s Modified Eagles Medium (DMEM), phosphate buffered saline (PBS), and antibiotics (penicillin/streptomycin (P/S)) were purchased from Gibco, USA.

2.2. Methods

2.2.1. Cell culture & seeding. IMR-32 human neuroblastoma cells procured from the National Centre for Cell Sciences (NCCS), Pune, India, were cultured in DMEM (Gibco®, USA) supplemented with 10% FBS (Gibco®, USA) and 1% penicillin/streptomycin (Gibco®, USA). The culture was maintained at 37 °C in a 5% CO₂ incubator.

2.2.2. Lactate dehydrogenase assay. Ten thousand cells were seeded in a 96 well plate followed by incubation with the desired concentration of peptide (50, 100, 200 and 400 μM) solution for 48 hours. The concentrations of the peptide used in this study were determined based on previous experiments from our group.¹⁰ One hundred microliters of cell culture supernatant treated with various concentrations of peptide and untreated sample was incubated with an equal amount of LDH reagent (CytoTox-One™ Homogeneous membrane integrity assay, Promega, USA) for about 20 min at room temperature. The reaction was then terminated by the addition of 25 μL stop solution to each well. The fluorescence intensity was measured at 590 nm using a multimode reader (Infinite 200M, Tecan®, Austria) after excitation at 560 nm.

2.2.3. Reactive oxygen species determination (ROS) assay. 10 000 cells were seeded in a 96 well plate followed by incubation with a specific concentration of peptide (50, 100, 200 and 400 μM) for 48 hours. After incubation, the medium was removed and the cells were washed with PBS to remove any non-adherent cells. Then 10 μM DCFDA solution was added and incubated for 45 minutes. After diffusion into the cell, the DCFDA gets deacetylated by the cellular esterases to form a non-fluorescent compound, which is later oxidized by ROS into 2′,7′-dichlorofluorescein, which can be detected with excitation and emission maxima of 495 nm and 530 nm respectively.

2.2.4. Determination of intracellular glutathione levels. 100 000 cells were seeded in a 24-well plate followed by incubation with the desired concentration of peptide (50, 100, 200 & 400 μM) for 48 hours. After incubation, the medium was removed and the cells were washed with PBS to remove any non-adherent cells. Cell lysis buffer was added and incubated for 20 minutes to disrupt the cells and the lysate was collected. The proteins in the cell lysate were removed by adding 10% trichloroacetic acid and centrifuged at 4500 rpm for 15 min to precipitate the proteins. DTNB reagent was added to the supernatant and analyzed using a multimode reader (Infinite 200M, Tecan®, USA). The values were normalized with respect to the protein content present in the cells and were compared with the untreated cells.

2.2.5. Thiobarbituric acid reactive substances (TBARS) assay. 100 000 cells were seeded in a 24-well plate and followed by incubation with a specific concentration of peptide (50, 100, 200 & 400 μM) for 48 hours. After incubation, the medium was removed and the cells were washed with PBS to remove any non-adherent cells. Cell lysis buffer was added and incubated for 20 minutes to disrupt the cells and the lysate was collected. The proteins in the cell lysate were removed by adding 20% trichloroacetic acid and centrifuged at 4500 rpm for 15 minutes to precipitate the proteins. 0.6% thiobarbituric acid was added to the supernatant and heated to 95 °C for one hour in a water bath. The absorbance at 532 nm was analyzed using a multimode reader (Infinite 200M, Tecan®, Austria). Values were normalized with respect to the protein content present in the cells and were compared with the untreated cells.

2.2.6. Imaging of cells. 100 000 cells were seeded in a 24 well plate followed by incubation with a specific concentration of peptide (50, 100, 200 & 400 μM) for 48 hours. The medium was removed and the cells were washed with PBS to remove any non-adherent cells. The cells were then imaged using a phase contrast microscope (Carl Zeiss Axiovert A1, Germany).

2.2.7. Gene expression studies. The expression of genes for various transcription factors associated with the oxidative stress, namely Nrf2, SOD1, p53, Akt, FoxO-3a, mTOR, GSK3β, AP-1 NF-kB, and certain MAP kinase genes, namely ERK-1 and p38, were evaluated in the presence of low and high concentrations of peptides (50 & 400 μM) using real time RT-PCR. The total RNA was isolated using Trizol (Invitrogen, USA) following the procedure described by the manufacturer. In brief, 1 mL Trizol was added to the samples and kept for 30 min at room temperature. The solution was collected and RNA was extracted with 0.2 mL of chloroform (Merck, India). The solution was centrifuged at 12 000 rpm for 15 min at 4 °C. The extracted RNA was stabilized using 70% ethanol prepared with nuclease-free water (Qiagen, USA). The RNA was centrifuged using a QIA shredder spin column (Qiagen) and dissolved in RNAse-free water (Qiagen, USA). The cDNA obtained after a two-step reaction was subjected to a real-time RT-PCR (Eppendorf AG22331, Germany). The primers used in this study are shown in Table 1. Quantitative values were determined by a delta–delta method and normalized with the house keeping gene GAPDH and the control.

Table 1 Primer sequences employed to determine the gene expression levels

Genes	Sequences	Base no.
Nrf2	CTGCTTTCATAGCTGAGCCC	20
	CCTGAGATGGTGACAAGGGT	20
p53	CCCAAGCAATGGATGATTTGA	21
	GGCATTCTGGGAGCTTCATCT	21
ERK-1	CTGGATCAGCTCAACCACATT	21
	AGAGACTGTAGGTAGTTTCGGG	22
p38	TCGACTTGCTGGAGAAGATGCTTGT	25
	CAGGACTCCATCTCTTCTTGGTCAA	25
Akt	CTCACAGCCCTGAAGTACTCTTTCCA	26
	TCCAGCATGAGGTTCTCCAGCTTGA	25
FoxO-3a	TCTACGAGTGGATGGTGCGTT	21
	CGACTATGCAGTGACAGGTTGTG	23
mTOR	TCGCTGAAGTCACACAGACC	20
	CTTTGGCATATGCTCGGCAC	20
GSK3β	ATTTCACCTCAGGAGTGCGG	20
	AAGAGTGCAGGTGTGTCTCG	20
SOD1	ACAAAGATGGTGTGGCCGAT	20
	AACGACTTCCAGCGTTTCCT	20
NF-kB	CGCTTAGGAGGGAGAGCCC	19
	TATGGGCCATCTGTTGGCAGTG	22
AP-1	TCCTGCCCAGTGTTGTTTGT	20
	GACTTCTCAGTGGGCTGTCC	20
GAPDH	ACCACAGTCCATGCCATCAC	20
	TCCACCACCCTGTTGCTGTA	20

---

2.2.8. Superoxide dismutase (SOD) assay. The levels of SOD protein were estimated using a standard SOD assay kit (Promega, USA). In brief, the untreated cells and those treated with peptides (50 & 400 μM) were washed with PBS, scrapped and collected in lysis buffer (20 mM HEPES, 1 mM EDTA, 210 mM mannitol and 70 mM sucrose). The cell suspension was centrifuged at 1500g for 5 minutes at 4 °C. The supernatant was then used for the assay. Diluted radical detector (200 μL) was mixed with 10 μL of the sample and then 20 μL of diluted xanthine oxidase was added. The plate was shaken continuously for a few minutes and incubated for 20 minutes at room temperature. Absorbance was measured between 440–460 nm using a multimode reader (Infinite 200M, Tecan®, Austria).

2.2.9. Caspase 3/7 activity assay. The involvement of caspases in causing apoptosis due to peptide exposure was determined by measuring caspase 3/7 activity (Apo-One® Homogeneous caspase 3/7 assay kit, Promega, USA). In brief, the cells treated with peptides (50 & 400 μM) were incubated with the caspase 3/7 reagent for about 3 hours at room temperature. The fluorescence intensity was measured at 527 nm using a multimode reader (Infinite 200M, Tecan®, Austria) after excitation at 485 nm.

2.2.10. MitoTracker® Red staining. The untreated and peptide treated (50 & 400 μM) cells were washed with PBS followed by the addition of 200 nM MitoTracker® Red dye (Molecular Probes, USA). After incubation for 45 minutes at 37 °C in a 5% CO₂ incubator, the media with dye was gently removed and rinsed with PBS. The cells were incubated with 3.7% paraformaldehyde at 37 °C in a 5% CO₂ incubator for 15 minutes. After gentle PBS washing, the Hoechst 33342 dye (Molecular Probes, USA) was added and incubated for 15 minutes. The excess dye was removed from the wells followed by PBS washing. Later the cells were imaged using laser scanning confocal microscopy (FV1000, Olympus, Japan).

2.2.11. JC-1 mitochondrial membrane potential assay. The measurement of mitochondrial membrane potential (ΔΨ_m) in untreated and peptide treated (50 & 400 μM) cells were carried out using a JC-1 assay. JC-1 is a lipophilic cationic dye, which shows two emission peaks based on the two different conformations adopted by the dye. In mitochondria of healthy cells that possess high mitochondrial potential, J-aggregates are formed, which show emission maxima at 590 nm. However, in the case of dead cells or in cells undergoing apoptosis, due to a decrease in mitochondrial potential, the JC-1 dye exists in the monomeric form and shows a characteristic emission at 530 nm. The cells were washed with PBS followed by the addition of staining solution (JC-1 dye at a concentration of 2 μM) and incubated for one hour in a 5% CO₂ incubator. The dye was removed and washed with PBS followed by the addition of PBS with 5% BSA. After 5 minutes incubation, the BSA containing PBS was removed and fresh PBS was added before taking measurements. The fluorescence emission intensity was recorded at two wavelengths, namely 530 and 590 nm, after excitation at 514 nm using a multimode reader (Infinite 200M, Tecan®, Austria).

2.3. Statistical analysis

The analysis of variance (one-way ANOVA) was performed to determine the statistical significance (p < 0.05) for LDH, ROS, glutathione, the TBARS assay (n = 4), the RT-PCR, caspase 3/7, and the JC-1 mitochondrial membrane permeability assay (n = 5).

3. Results and discussion

3.1. Influence of the KLVFF peptide concentration on cell viability

The results of the LDH assay show that the viability of cells remains unaltered until 100 μM concentration of the peptide, beyond which a significant reduction in cell viability is observed (Fig. 1). Our results suggest that the KLVFF peptide at low concentrations does not exert any cytotoxic effect as it does not affect the cell viability. The drastic reduction in cell viability at high concentrations, namely 200 and 400 μM of the KLVFF peptide, arises due to the tendency of the peptide to form self-assembled aggregates.¹⁰ The molecular mechanism behind the cytotoxicity of the KLVFF peptide needs to be explored in-depth to devise possible strategies to counter its toxic effects and exploit its therapeutic potential.

Fig. 1 Viability assessment of the IMR-32 human neuroblastoma cell line at 50, 100, 200 and 400 μM concentrations of the KLVFF peptide. Results are shown as the mean ± SD (n = 4; *p < 0.05).

3.2. Influence of the KLVFF peptide concentration on the reactive oxygen species (ROS) levels

As the concentration of Tjernberg peptide increases, the ability of the KLVFF peptide to scavenge ROS increases, as shown in Fig. 2A. The results were in correlation with the LDH assay, where the viability reduced at KLVFF peptide concentrations of 200 and 400 μM. However at peptide concentrations ≤100 μM, the cell viability remains unaffected, while the ROS levels declined by about 35% when compared with the control cells. This indicates that the peptide exhibits an inherent radical scavenging effect until 100 μM, beyond which a sudden transition to a toxic form occurs, that appears to cause lethality in a ROS-independent manner.

Fig. 2 Biochemical analysis of the KLVFF peptide at 50, 100, 200 and 400 μM: [A] ROS assay, [B] DPPH assay, [C] glutathione assay and [D] TBARS assay. Results are shown as the mean ± SD (n = 4; *p < 0.05).

A reduction in the ROS levels was observed in all peptide concentrations tested from 50–400 μM, which is in contrast with a recent report on Aβ_1-42 peptide toxicity in glioma and neuroblastoma cells.¹¹ Our results indicate that KLVFF displays contrasting properties with Aβ_1-42 depending on its concentration. Therefore, it is evident that KLVFF can serve as an Aβ_1-42 mimic only under certain conditions. The results of the ROS assay suggest that the KLVFF peptide’s ability to scavenge free radicals arises due to the nature of amino acid residues present in it. It is now recognized that the aromatic amino acids tyrosine (Y) and phenylalanine (F) have the ability to donate protons to electron deficient centres and the resultant aromatic carbanion will be stabilized through resonance.¹² The hydrophobic amino acids are known to have strong hydroxyl radical scavenging activities.¹³ The protein hydrolysates containing higher concentrations of phenylalanine (F), isoleucine (I), leucine (L) and valine (V) possess strong superoxide radical scavenging activities.¹⁴ The amino acid lysine (K) is reported to exhibit a radical scavenging property by virtue of its –NH₃⁺ group.¹⁵ Since the Tjernberg peptide possesses the amino acid sequence KLVFF, the free radical scavenging ability of the individual amino acids could have synergistically contributed to the observed results. In addition, peptides are also known to chelate metal ions via their carboxyl and amino groups,¹⁶ which may further prevent formation of hydroxyl radicals by the metal ions.

3.3. Scavenging ability of the KLVFF peptide using a DPPH assay

The inherent free radical scavenging ability of the KLVFF peptide in the absence of cells was quantified at various concentrations using a DPPH assay and the results are shown in Fig. 2B. A concentration-dependent increase in the free radical scavenging ability of the peptide was observed. However, the maximum radical scavenging activity of the peptide was 5%, indicating that the Tjernberg peptide possesses a weak anti-oxidant nature. The anti-oxidant property of the Tjernberg peptide arose from the aromatic amino acid phenylalanine, that acts as a proton donor to DPPH. The Tjernberg peptide, with a molecular weight of 651 Da, falls in the lower molecular weight category, which may also have contributed to its radical scavenging property.¹⁷ The presence of the hydrophobic amino acids leucine, valine and phenylalanine in high proportions confers the radical scavenging property by enhancing its solubility in a non-polar environment, thereby facilitating better interaction with the free radicals.

3.4. Influence of KLVFF concentration on glutathione levels

The alterations in the intracellular glutathione levels due to the addition of various concentrations of the KLVFF peptide are presented in Fig. 2C. A marginal increase in the glutathione levels was observed at concentrations of 50 and 100 μM when compared with the control cells. However, the intracellular glutathione levels were found to exhibit a concentration-dependent decrease with increasing peptide concentrations. The reduced glutathione levels at higher concentrations, namely 200 and 400 μM, were in correlation with the reduced cell viability, owing to the toxic effects of the peptide. The small increase in the glutathione levels at concentrations ≤100 μM suggests that the peptide recruits anti-oxidant machinery to scavenge free radicals.

3.5. Influence of the KLVFF peptide concentration on lipid peroxidation

A TBARS assay was performed in cells treated with various concentrations of the KLVFF peptide to quantify the extent of lipid peroxidation and the results are shown in Fig. 2D. A marginal decrease in the malondialdehyde (MDA) levels was observed in all cases when compared with the untreated control cells. Our group had earlier demonstrated the localization of the KLVFF peptide in the lipid bilayer.¹⁸ The presence of hydrophobic leucine and valine in the peptide sequence increases its ability to penetrate into the lipid phase of the membrane, thereby facilitating better interaction with free radicals. It has also been reported that hydrophobic amino acids tend to protect linoleic acid from degradation by donating protons to the hydrophobic peroxyl radicals.¹³ Thus, from the results of the ROS and TBARS assays, it may be hypothesized that the KLVFF peptide can act as a lipid soluble anti-oxidant, which can localize into the lipid bilayer and scavenge free radicals. However, beyond 100 μM, no apparent changes in the MDA levels are observed, indicating that toxicity due to the peptide does not act via lipid peroxidation.

3.6. Influence of the KLVFF peptide on the cell morphology

The phase contrast microscopic images of IMR-32 cells treated with different concentrations of the Tjernberg peptide is shown in Fig. 3. The results show that at peptide concentrations of 50 μM and 100 μM, the cells exhibit their native elongated morphology with few dead cells. However at high concentrations i.e., 200 and 400 μM, a significant distortion in the cell morphology is observed with more floating cells. They also lack cell–cell contacts followed by a reduction in the cell number. This suggests that at high concentrations of 200 and 400 μM, the KLVFF peptide triggers the cell death machinery by activating apoptotic pathways directly or indirectly. Therefore, gene expression studies were carried out to decipher the possible pathways triggered by the peptide at a low concentration (50 μM) where it exerts a non-toxic effect and at a high concentration (400 μM), where its effect is cytotoxic.

Fig. 3 Phase contrast microscopic images of IMR-32 human neuroblastoma cell lines treated with various concentrations of the KLVFF peptide: [a] 50 μM, [b] 100 μM, [c] 200 μM and [d] 400 μM peptide. Magnification: ×100.

3.7. Gene expression studies

The expression profiles of genes responsible for cell survival, apoptosis and autophagy, and those that are induced in response to genotoxic or oxidative stress, were evaluated in cells exposed to either a non-toxic concentration (50 μM) or a toxic concentration (400 μM) of KLVFF. Fig. 4 summarizes the relative fold change of various genes in IMR32 cells treated with 50 or 400 μM of the peptide. A distinct difference in the expression profiles of the genes was observed in cells treated with the different peptide concentrations. Cells exposed to 50 μM of the Tjernberg peptide exhibited an up-regulation of Nrf2, Akt, mTOR and GSK3β genes, while the SOD1, p53, ERK-1, p38, FoxO3a, NF-κB and AP-1 genes were down-regulated when compared with the untreated control cells (Fig. 4a). In contrast, cells exposed to 400 μM of the peptide were found to over-express all genes except Nrf2 when compared with the untreated control cells (Fig. 4b).

Fig. 4 Gene expression analysis in IMR-32 human neuroblastoma cell lines treated with various concentrations of the KLVFF peptide: [a] 50 μM and [b] 400 μM peptide.

Interestingly, the mTOR gene expression levels were elevated in cells treated with 50 μM and 400 μM of the KLVFF peptide. mTOR exhibits multiple roles among which the sensing of nutrients and thereby enhancing cell proliferation plays a crucial role.¹⁹ For the mTOR gene, in connection with autophagy, it has been demonstrated that inhibition of mTOR by rapamycin improves cognitive deficits in vivo.²⁰ Recent evidence demonstrates that excessive activation of the autophagy pathway results in cell death due to autophagy of the cell’s own cytosolic contents.²¹ Elevated mTOR levels have also been found to activate the pro-apoptotic nuclear transcription factor p53.²² It has been suggested that mTOR translocates from the cytosol to the nuclear compartment, where it phosphorylates a serine residue in p53, resulting in its activation.²² Our data reveals that cells treated with 400 μM of peptide exhibit up-regulation of p53 in contrast to cells treated with 50 μM of the peptide. This indicates that high concentrations of the peptide trigger the translocation of mTOR and the subsequent phosphorylation and activation of p53, leading to a halt in the cell cycle and the initiation of apoptosis.

The upstream activator of mTOR is Akt, also known as protein kinase B.²³ The activation of Akt is seen in cells treated with the Tjernberg peptide irrespective of the concentration. The up-regulation of Akt suppresses the activation of pro-apoptotic mitochondrial proteins and inhibits caspase activation and the JNK pathway.²⁴ This has been suggested as a potential therapeutic strategy to combat amyloid beta 1-42 induced apoptosis. Therefore, our results indicate that a 50 μM concentration of KLVFF peptide up-regulates Akt while causing no cytotoxicity, indicating that its therapeutic activity may be due to the activation of the Akt pathway. Interestingly the Akt levels are also up-regulated in cells treated with 400 μM of the Tjernberg peptide. Like many other proteins that regulate cell survival and proliferation, a dual role for Akt has been identified recently.²⁵ Over-expression of Akt has been implicated in inducing cell death, though the exact mechanism has not been deciphered yet. It has been suggested that Akt might translocate to the nucleus where it may interact with anti-apoptotic factors like Epb1, an inhibitor of DNA fragmentation, thereby altering its function.²⁶ Thus, the treatment with high concentrations of the Tjernberg peptide is likely to promote translocation of Akt and mTOR to the nucleus, thereby initiating autophagy and apoptotic signals leading to cell death.

It is reported that amyloid beta (Aβ) accumulation induces neuronal apoptosis via transcriptional activation of death-associated genes.²⁷ The FoxO3a is a forkhead transcriptional factor implicated in inducing apoptosis. Its transcription activity is modulated by various post-translational modifications especially phosphorylation by various kinases.²⁸ The phosphorylation of FoxO3a by survival kinases like Akt in the cytosol, prevents its entry in to the nucleus and promotes degradation through ubiquitination.²⁸ In the present study, it was found that in cells treated with a low concentration (50 μM) of the Tjernberg peptide, FoxO3a gene levels are down-regulated, while up-regulated in cells treated with a high concentration (400 μM) of the peptide. This indicates that the presence of low concentrations of the peptide promotes phosphorylation of FoxO3a by Akt, leading to its subsequent degradation. However in cells treated with high concentrations of the peptide, a translocation of FoxO3a to the nucleus might occur, thereby preventing its phosphorylation by Akt and hence escapes degradation. The nuclear sequestration of FoxO3a results in the activation of CDKI p27, a cell cycle inhibitor.²⁹ In addition the nuclear FoxO3a has been found to activate several apoptotic and autophagy factors like TRAIL (TNF-related apoptosis-inducing ligand), FasL (Fas ligand or CD95L), PUMA (p53 upregulated modulator of apoptosis), PTEN (phosphatase and tensin homolog deleted on chromosome 10) etc.³⁰ These events are likely to be the reason for the observed cell death in cells treated with high concentrations of the Tjernberg peptide. Recent evidence has highlighted the existence of cross-talk between the FoxO3a and p53 signaling pathways as they act on common molecular targets.³¹ Therefore the nuclear translocation of FoxO3a may synergistically act with p53 in inducing apoptosis of cells treated with a high concentration of the Tjernberg peptide.

Glycogen synthase kinase 3 beta (GSK3β) levels were found to be up-regulated in cells treated with 50 μM and 400 μM of the Tjernberg peptide. GSK3β is involved in the phosphorylation of different protein substrates in cells. Its over-expression has been implicated in the hyperphosphorylation of tau protein, leading to the formation of neurofibrillary tangles and neuronal apoptosis.³² The GSK3β levels are regulated by the activation of Akt.³³ Our results suggest that in cells treated with a high concentration (400 μM) of the Tjernberg peptide, there was an up-regulation of GSK3β leading to the reduction in cell viability. Interestingly, both Akt and GSK3β levels are also up-regulated in cells treated with a low concentration (50 μM) of the peptide. Apparently the intracellular localization of Akt and GSK3β may be responsible for the difference in the cell viability at this concentration. Like Akt, GSK3β is found to sequester free radicals in the cytosol and nucleus, though the factors promoting this sequestration and their implications on the GSK3β activity are unknown. Another aspect that may have implications on the differential effects of GSK3β expression on cell survival could be the nature of its phosphorylated site. It has been found that GSK3β exists in an inactive state, when the serine residue in the 9^th position is phosphorylated and transforms into an active state if the phosphorylation occurred in the tyrosine residue at the 216^th position.³⁴ It has now been demonstrated that alpha 5 beta 1 integrin triggers resistance to apoptosis in cells through GSK3β signaling.³⁵ This introduces another facet to the complex interplay of signaling factors in the cells treated with a low concentration of the Tjernberg peptide. An earlier report has suggested that alpha 5 beta 1 integrin could rescue cells from amyloid beta induced apoptosis.³⁵ Therefore the up-regulation of GSK3β in cells treated with a low concentration of the peptide could promote cell survival and may have implications in its therapeutic potential, through a difference in GSK3β localization, or through different phosphorylation sites, or through the activation of alpha 5 beta 1 integrin or a combination of all these factors.

It can be observed from Fig. 7 that the ERK1 and p38 genes are down-regulated in cells treated with a low concentration of the Tjernberg peptide, while cells exposed to high concentrations of the peptide over-express both ERK1 and p38. Both ERK1 and p38 are MAP kinases related to stress and are the downstream effectors in the anti-oxidant response of cells.³⁶ The up-regulation of the p38 gene has been implicated in early Alzheimer’s disease, leading to neuronal death.³⁷ Our results indicate that at a high concentration of the Tjernberg peptide, p38 activation is induced leading to stimulation of pro-apoptotic factors. Low concentrations of the peptide do not activate p38. ERK1 activation has been implicated in the initiation of apoptosis through activation of caspase 3.³⁸ The up-regulation of ERK1 in cells treated with 400 μM of the Tjernberg peptide suggests the probable activation of caspase-dependent apoptosis in these cells. The expression levels of ERK1 and p38 are also connected with the anti-oxidant levels in the cells. It has been reported that GSH depletion leads to the activation of p38.³⁹ The glutathione assay in the present study reveals that depletion of glutathione is observed only in cells treated with high concentrations of the Tjernberg peptide and the p38 levels are up-regulated in these cells, suggesting a direct correlation. GSH possesses binding sites for various transcription factors, namely AP-1, AP-2 and NF-κB.⁴⁰ Intriguingly even though the GSH level is enhanced in cells treated with 50 μM peptide, neither AP-1 nor NF-κB is up-regulated. However, Nrf2 levels were up-regulated in these cells and hence it is evident that glutathione levels in cells exposed to the Tjernberg peptide are regulated by Nrf2 and not through AP-1 or NF-κB. This is further substantiated from the high expression levels of AP-1 and NF-κB in cells treated with a high concentration of the peptide, but the intracellular glutathione levels in these cells were depleted (Fig. 4). These cells express low levels of Nrf2 indicating that this factor is commonly involved in the transcriptional regulation of genes encoding the antioxidant proteins and regulates the glutathione levels in cells treated with the Tjernberg peptide. The Nrf2 expression levels may be associated with Akt activation in cells treated with a low concentration of the peptide. The implications of the ERK/MAP kinase pathway in Alzheimer’s disease have been correlated with the induction of synaptic plasticity, increased tau phosphorylation and the development of cytoskeletal abnormalities.⁴¹

The phase contrast images of cells treated with a high concentration of the peptide reveal dystrophic neurites that may be attributed to elevated levels of ERK. Such phenotypic changes were absent in cells treated with a low concentration of the peptide, which correlates well with the low levels of ERK1 expression in these cells. The gene expression level of the major intracellular anti-oxidant system SOD1 was found to be up-regulated only in cells treated with 400 μM of the peptide, while it was down-regulated in cells treated with a low concentration of the Tjernberg peptide. The upstream factors of the SOD1 gene are NF-κB, Nrf2, and AP-1 and its downstream factors are Akt, GSK3β, mTOR, p53, FoxO3a, ERK1 and p38. In cells exposed to a low concentration of the peptide, the SOD1 gene is down-regulated in spite of over-expression of its upstream transcription factor Nrf2. This indicates that the activation of the SOD1 gene by the peptide requires other transcription factors such as NF-κB and AP-1, that are down-regulated in cells exposed to 50 μM of the peptide. All downstream factors of SOD1 except Akt and mTOR are consequently under-expressed. The Akt protein, an upstream factor of mTOR, enhances SOD1 expression through activation of NF-κB.⁴² However in cells treated with a low concentration of the peptide, the NF-κB remains down-regulated and this is reflected in the low levels of expression of the SOD1 gene. In contrast, the cells exposed to a high concentration of the peptide were found to over-express the SOD1 gene, which may have implications in increased oxidative stress and neurodegeneration. The over-expression of SOD1 can be correlated with the up-regulation of the transcription factors AP-1 and NF-κB. In addition both Akt and NF-κB are up-regulated, which is also reflected in the SOD1 expression levels. The major signaling pathways activated by the low (50 μM) and high (400 μM) concentrations of the peptide are shown in Fig. 5a and b.

Fig. 5 Major cell signaling molecules and pathways triggered in neuronal cells due to treatment with [a] 50 μM and [b] 400 μM of the KLVFF peptide.

3.8. Influence of the KLVFF peptide concentration on SOD activity

Fig. 6a shows the SOD levels of cells treated with low and high concentrations of the Tjernberg peptide. It was observed that when cells were treated with a low concentration of the peptide, no significant increase in the SOD levels were observed when compared with the untreated control cells. This is concurrent with the gene expression results, where the SOD1 gene was down-regulated in cells treated with the same concentration of the peptide. The SOD levels of cells treated with a high concentration of the peptide are significantly high in line with the gene expression data. SOD has been hypothesized to promote neurodegeneration by inducing abnormal free-radical metabolism, causing the production of hydroxyl radicals, toxic derivatives of peroxynitrite, metal toxicity and abnormal protein aggregation. Thus, high concentrations of the Tjernberg peptide, increase SOD levels and promote protein aggregation and apoptosis.

Fig. 6 Intracellular alterations in the IMR32 cells due to low (50 μM) and high (400 μM) doses of the KLVFF peptide: [a] SOD assay, [b] caspase 3/7 assay and [c] mitochondrial membrane potential assay.

3.9. Influence of the KLVFF peptide concentration on caspase 3/7 activity

The biochemical assays and gene expression studies have indicated that a high concentration of the Tjernberg peptide induces cell death. In order to understand whether this cell death follows a caspase-dependent mechanism, caspase 3/7 activity was assessed and the results are shown in Fig. 6b. The caspase levels were found to be significantly elevated in the cells treated with a high concentration of the peptide, while they were comparable to the untreated control in cells treated with low concentration of the peptide. This shows that the Tjernberg peptide at a high concentration induces apoptosis through the activation of caspases. This may be due to the activation of the ERK1 gene that activates the caspase 3 pathway. The activation of the p53 gene in these cells may also have contributed to the apoptosis through the transcription of pro-apoptotic genes such as Bcl-2, Bax, Noxa and PUMA. This behaviour of the Tjernberg peptide is akin to amyloid beta peptide, which has also been reported to cause apoptosis through a caspase-dependent mechanism.⁴³ Therefore at high concentrations, the Tjernberg peptide mimics amyloid beta peptides and can serve as a model system to study Alzheimer’s disease.

3.10. JC-1 mitochondrial membrane potential measurements

The measurement of mitochondrial membrane potential were carried out using a JC-1 probe to understand the influence of the peptide treatment on mitochondria and the results are presented in Fig. 6c. The treatment with a low concentration of peptide does not cause significant change in the mitochondrial membrane potential (ΔΨ_m) when compared with the untreated control cells. However a 50% reduction in the membrane potential (ΔΨ_m) is observed in cells treated with a high concentration of the Tjernberg peptide. This indicates that the peptide at high concentrations can depolarize the mitochondrial membrane, triggering the intrinsic pathway of apoptosis. A similar observation was reported for amyloid beta protein by Moreira et al., who demonstrated mitochondrial dysfunction as the prime contributor to the cytotoxicity of amyloid beta and the associated alterations in energy metabolism observed in Alzheimer’s disease.⁴⁴

3.11. MitoTracker® Red staining

To investigate the influence on the mitochondria during peptide dosing, the MitoTracker® Red staining of cells treated with different concentrations of the Tjernberg peptide was performed and the results are shown in Fig. 7. The cells treated with the peptide exhibit a higher intensity when compared with the untreated control cells, suggesting the presence of a greater number of active mitochondria. However, the cells treated with a higher concentration of the peptide displayed reduced viability, indicating the toxic nature of the peptide at this concentration.

Fig. 7 MitoTracker® Red staining of cells treated with 50 μM and 400 μM of the KLVFF peptide.

4. Conclusion

Our experimental results reveal the concentration dependent dual nature of the KLVFF peptide. At concentrations below 100 μM, the peptide exhibits a free radical scavenging property and does not affect the cell viability. It also activates the cell proliferation genes Akt and mTOR that augur well for cell survival. However at higher concentrations, the peptide transforms into a toxic form by promoting cell death through a caspase-dependent mechanism. The peptide depolarizes the mitochondrial membrane at high concentrations and mediates the translocation of signals to the nucleus, leading to the activation of pro-apoptotic genes. Activation of the MAP kinases ERK1, p38 and p53 may be postulated to be key events contributing to the cytotoxicity that strongly resemble the toxic manifestations reported for amyloid beta peptide. The present study unveils a “Jekyll and Hyde” nature of KLVFF based on its concentration. While it exhibits beneficial effects to the cells at concentrations below 100 μM, it could serve as an Aβ_1-42 mimic at concentrations beyond 200 μM. The glutathione levels in the KLVFF peptide treated cells are regulated by the expression of the transcriptional factor Nrf2, while SOD levels are regulated through the factors AP-1 and NF-κB. The molecular mechanisms involved during the transition of the KLVFF peptide from its therapeutic to toxic role were elucidated in the present study, which hold immense potential in designing KLVFF peptide based therapeutic strategies in the future.

Acknowledgements

The first author Priyadharshini Kumaraswamy wishes to thank the Department of Science & Technology for the INSPIRE Fellowship grant (DST/INSPIRE Fellowship/2011/74). The authors also wish to acknowledge the Prof. TRR fund, SASTRA University, PG Teaching Programme (No. SR/NM/PG-16/2007) of the Nano Mission Council and the FIST grant (SR/FST/LSI-453/2010), Department of Science & Technology, New Delhi for the financial and infrastructural support. This article is dedicated to the memory of the first author Dr Priyadharshini Kumaraswamy who died in a tragic road accident at the age of 27 years.

References

1. H. Yamaguchi, M. L. Maat-Schieman, S. G. van Duinen, F. A. Prins, P. Neeskens, R. Natte and R. A. Roos, J. Neuropathol. Exp. Neurol., 2000, 59, 723–732.
2. V. J. R. Paula, F. M. Guimaraes, B. S. Diniz and O. V. Forlenza, Dement. Neuropsychol., 2009, 3, 188–194.
3. C. Hetenyi, Z. Szabo, E. Klement, Z. Datki, T. Kortvelyesi, M. Zarandi and B. Penke, Biochem. Biophys. Res. Commun., 2002, 292, 931–936.
4. M. A. Chacon, M. I. Barria, C. Soto and N. C. Inestrosa, Mol. Psychiatry, 2004, 9, 953–961.
5. T. L. Lowe, A. Strzelec, L. L. Kiessling and R. M. Murphy, Biochemistry, 2001, 40, 7882–7889.
6. M. H. Viet, S. T. Ngo, N. S. Lam and M. S. Li, J. Phys. Chem. B, 2011, 115, 7433–7446.
7. I. W. Hamley and M. J. Krysmann, Langmuir, 2008, 24, 8210–8214.
8. J. E. Morley, S. A. Farr, W. A. Banks, S. N. Johnson, K. A. Yamada and L. Xu, J. Alzheimer’s Dis., 2010, 19, 441–449.
9. A. Kontush, Free Radical Biol. Med., 2001, 31, 1120–1131.
10. P. Kumaraswamy, S. Sethuraman and U. M. Krishnan, Soft Matter, 2013, 9, 2684–2694.
11. H. Wang, J. Ma, Y. Tan, Z. Wang, C. Sheng, S. Chen and J. Ding, J. Alzheimer’s Dis., 2010, 21, 597–610.
12. B. Sarmadi, A. Ismail and M. Hamid, Food Res. Int., 2011, 44, 290–296.
13. H. Sun, T. Ye, Y. Wang, L. Wang, Y. Chen and B. Li, Food Sci. Nutr., 2014, 2, 58–64.
14. X. He, W. Cao, Z. Zhao and C. Zhang, J. Food Nutr. Res., 2013, 1, 30–36.
15. A. Amiri, M. Memarpoor-Yazdi, M. Shanbedi and H. Eshghi, J. Biomed. Mater. Res., Part A, 2013, 101, 2219–2228.
16. A. E. Fisher and D. P. Naughton, Biomed. Pharmacother., 2005, 59, 158–162.
17. S. Jha, P. Mandal, P. Bhattacharyya and A. Ghosh, J. Anim. Sci. Biotechnol., 2014, 5, 16–24.
18. P. Kumaraswamy, S. Sethuraman and U. M. Krishnan, J. Agric. Food Chem., 2013, 61, 3278–3285.
19. G. A. Soliman, Nutrients, 2013, 5, 2231–2257.
20. P. Spilman, N. Podlutskaya, M. J. Hart, J. Debnath, O. Gorostiza, D. Bredesen, A. Richardson, R. Strong and V. Galvan, PLoS One, 2010, 5, e9979.
21. N. Mizushima, Genes Dev., 2007, 21, 2861–2873.
22. M. V. Poyurovsky and C. Prives, Aging, 2010, 2, 380–382.
23. N. Hay and N. Sonenberg, Genes Dev., 2004, 18, 1926–1945.
24. S.-O. Yoon, M.-M. Kim, S.-J. Park, D. Kim, J. Chung and A.-S. Chung, FASEB J., 2002, 16, 111–113.
25. S. Maddika, S. R. Ande, E. Wiechec, L. L. Hansen, S. Wesselborg and M. Los, J. Cell Sci., 2008, 121, 979–988.
26. J. Y. Ahn, X. Liu, Z. Liu, L. Pereira, D. Cheng, J. Peng, P. A. Wade, A. W. Hamburger and K. Ye, EMBO J., 2006, 25, 2083–2095.
27. M. S. Yu, K. C. Suen, N. S. Kwok, K. F. So, J. Hugon and R. C. Chang, Apoptosis, 2006, 11, 687–690.
28. P. Sanphui and S. C. Biswas, Cell Death Dis., 2013, 4, e625.
29. A. Chatterjee, U. Chatterjee and M. K. Ghosh, Cell Death Dis., 2013, 4, e543.
30. H. Luo, Y. Yang, J. Duan, P. Wu, Q. Jiang and C. Xu, Cell Death Dis., 2013, 4, e481.
31. H. You and T. W. Mak, Cell Cycle, 2005, 4, 37–38.
32. S. Jayapalan and J. Natarajan, Bioinformation, 2013, 9, 1023–1030.
33. Z. Fryberg, S. J. Ferrando and J. A. Javitch, Am. J. Psychiatry, 2010, 167, 388–396.
34. C. A. Grimes and R. S. Jope, Prog. Neurobiol., 2001, 65, 391–426.
35. F. D. Toni-Costes, M. Despeaux, J. Bertrand, E. Bouogaa, L. Ysebaret, B. Payrastre and C. Racaud-Sultan, PLoS One, 2010, 5, e9807.
36. J. H. Limon-Pacheco, N. A. Hernandex, M. L. Fanjul-Moles and M. E. Gonsebatt, Free Radical Biol. Med., 2007, 43, 1335–1347.
37. S. A. L. Correa and K. L. Eales, J. Signal Transduction, 2012, 2012, 649079.
38. S. Zhuang and R. G. Schnellmann, J. Pharmacol., 2006, 319, 991–997.
39. X. M. Hu, B. Yuan, S. Tanaka, Q. Zhou, K. Onda, H. Toyoda and T. Hirano, Leuk. Lymphoma, 2014, 55, 392–404.
40. C. K. Sen and L. Packer, FASEB J., 1996, 10, 709–720.
41. G. Drewes, B. Lichtenberg-Kraag, F. Doring, E. M. Mandelkow, J. Biernat, J. Goris, M. Doree and E. Mandelkow, EMBO J., 1992, 11, 2131–2138.
42. A. I. Rojo, M. Salinas, D. Martin, R. Perona and A. Cuadrado, J. Neurosci., 2004, 24, 7324–7334.
43. L. Zhao, Z. M. Qian, C. Zhang, H. Y. Wing, F. Du and K. Ya, Aging Cell, 2008, 7, 47–57.
44. P. I. Moreira, Eur. Neurol. Rev., 2010, 5, 17–21.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra10746f

---