Neurodegeneration, synaptic loss, and increased neuroinflammation are the common pathological symptoms observed in Alzheimer’s disease (AD). Accumulation of misfolded amyloid beta protein (Aβ) might be the principal cause and is strongly associated with these events. Several experimental, clinical, biochemical, and molecular studies have correlated these pathological symptoms with memory impairment in AD [1
]. Therefore, decreasing neuroinflammation or disaggregating the misfolded Aβ are proposed strategies for preventing neurodegeneration and preserving cognitive function in AD. Because of the complexity of the pathological profile observed in AD, it may not be feasible to reduce major AD symptoms using a single drug. Although the US Food and Drug Administration (FDA) has approved several drugs for treating AD, most of them provide only modest improvements in memory and cognitive function and are unable to prevent progressive neuronal death. Therefore, developing new drugs for the treatment of AD is a challenging task for researchers. One area of drug development receiving renewed attention is the use of boron compounds which, because of their anti-inflammatory properties, are being proposed as potential treatments for reducing AD-induced neuroinflammation and cognitive deficits [3
Boron is a solid, nonmetallic element (atomic number 5) in group 13 of the periodic table. It is one of the most important minerals found in food and in the environment. It has pleiotropic effects, including building strong bones, muscles, treating osteoarthritis, and increasing testosterone levels in humans [7
]. Because of its role in improving cognitive skills and muscle coordination, some people take it as dietary supplement [6
], which has been an impetus for using it in the development of pharmaceutical drugs [7
]. Several boron derivatives have been tested for different diseases [3
]. For example, bortezomib, a proteasome inhibitor, has been used to prevent several malignancies [8
]. Cyclic boron derivatives, such as tavaborole, have been developed as antifungal agents and are FDA-approved topical treatments for onychomycosis [9
]. Similarly, benzoxaborole SCYX-7158 has been tested clinically for the treatment of stage 2 human African trypanosomiasis [11
]. Furthermore, a group of boron compounds have shown to be strong inhibitors of phosphodiesterase 4 enzyme (PDE4) and inflammation-related cytokine release [12
Although studies have demonstrated that boron compounds have strong immunomodulatory functions [13
], only, a few studies have investigated their potential to treat neurodegenerative diseases, such as AD. Recently, Kucukdogru and colleagues reported that boron nitride nanoparticles have neuroprotective effects in the experimental Parkinson’s disease model against MPP+-induced apoptosis [3
]. Similarly, Lu and colleagues discovered a series of boron-containing compounds to be Aβ aggregation inhibitors and antioxidants for the treatment of AD [4
]. In our previous study, we found evidence of the reduction of Aβ plaque-induced neurodegeneration in C. elegans
overexpressed with Aβ42 after treatment with trans-2-phenylvinylboronic acid (TPVA; unpublished data). The TPVA MIDA ester and TBSA are monomeric, free-flowing, crystalline solids which are stable at room temperature. They are mainly used for palladium (Pd)-catalyzed Suzuki-Miyaura coupling reactions, diastereoselective synthesis, and rhodium (Rh)-catalyzed intramolecular amination of aryl azides. They are also used as reagents in preparation of optically active unsaturated amino acids by diastereoselectivity and isomerization. Our preliminary observation in C. elegans
Aβ42-expressing models with boronic compounds show promising results which prompted us to investigate the mechanistic details of these two boronic compounds.
Because of the limited number of neurons (~302 neurons), making it easy to track and manipulate their function, Caenorhabditis elegans
have been used by many researchers to test neuronal vulnerability. To investigate the effect of different anti-amyloid compounds (such as boron) on Aβ toxicity in neurons, researchers used a transgenic C. elegans
model that expressed Aβ 1–42 in glutamatergic neurons, which are particularly vulnerable in AD [15
]. In the present study, we utilized a C. elegans
Aβ42-expressing model to test our hypotheses.
In the present study, we investigated the effects of trans 2-phenyl vinyl-boronic acid MIDA ester (TPVA) and trans-beta-styryl-boronic acid (TBSA) on Aβ aggregation and neurotoxicity in vitro. In addition, survival and morphological changes in a C. elegans model of AD, as well as Aβ plaque loads, neuroinflammatory markers, and behavioral outcomes in 5xFAD mice were explored after treatment with TBSA. We observed that boronic compounds significantly inhibited Aβ aggregation and neurotoxicity in vitro. Increased survival and neuroprotection was observed in the C. elegans model of AD, while decreases in amyloid plaque, neuronal death, and neuroinflammation, along with a partial reduction in memory deficits were also observed in the 5xFAD mouse model of AD after two months of treatment with TBSA.
Disturbances of protein homeostasis and increased neuroinflammation are the hallmark pathologies observed in the brains of AD patients. These pathological changes are strongly linked with neurodegeneration, synaptic loss, as well as cognitive dysfunction. Therefore, reducing neuroinflammation and decreasing misfolded amyloid protein could be a viable approach to mitigate cognitive dysfunction in AD. In the present study, we investigated the effects of chronic administration of a boronic compound, trans-beta-styryl-boronic acid (TBSA) on: (i) survival and neurodegeneration in C. elegans expressing Aβ42; (ii) neuronal morphology in both C. elegans expressing Aβ42 and 5xFAD mice; (iii) amyloid plaque load in 5xFAD mice; (iv) neuroinflammatory markers in 5xFAD mice; and (v) behavioral deficits in 5xFAD mice. We observed that in aged C. elegans expressing Aβ42 (17 days old), survival was improved with TBSA treatment, as was neuronal health. In 5xFAD mice, a significant decrease in neuroinflammatory markers, amyloid plaque load, and prevention of neuronal damage in the cortex and hippocampus, along with modest protection against cognitive dysfunction in 5xFAD mice was observed after treatment with TBSA.
Several anti-amyloid, anti-inflammatory drugs and small molecules have been tested to inhibit amyloid aggregation and neuroinflammation in AD [18
]. Some of these molecules prevented Aβ aggregation and neuroinflammation, but were unable to protect against progressive neurodegeneration, as well as loss of cognitive function in AD [20
]; therefore, none of these compounds proved successful in clinical AD trials. In the present study, we tested the efficacy of two different boron compounds against Aβ-induced neurodegeneration. Boron is a nonmetallic element found in food and the environment, and it has a high affinity to bind to oxygen [21
]. It is naturally electron deficient, thus it is well suited to unique chemical reactions. It readily accepts electrons from nucleophiles like the hydroxyl group in serine, making boronic drugs highly effective serine protease inhibitors [10
]. Although as a trace element, boron has yet to be established as an essential nutrient for humans, recent experimental and clinical studies suggest that it might be an important mineral for cell membrane function [7
]. Because of their anti-inflammatory properties, boron compounds have been used as dietary supplements for the treatment of neuroinflammation and in neurodegeneration [4
]. Boron compounds have been tested to treat a host of disorders, from arthritis to vascular dysfunction [13
]. For example, Bortezomib acts as a potent ubiquitin proteasome inhibitor and is commonly used to treat myeloma [8
]. Similarly, Vaborbactam, a β-lactamase inhibitor has been used to fight bacterial infections, such as treatment for complicated urinary tract infections [23
]. Furthermore, sodium perborate has been used in whitening discolored teeth, disinfecting medical instruments, as an ointment and for the treatment of poison ivy dermatitis [24
However, the use of boron-based compounds has not been tested as a therapy to reduce AD pathology. Penland and colleagues used electrophysiology and cognitive performance tests and reported that dietary boron may play important roles for brain function and cognitive performance [6
]. Previous experimental data suggest that boronic acid may have a high affinity for binding directly to amyloid precursor protein (APP) [25
]. In addition, due to their high pKa value (~9–10), boron compounds remain unionized at physiological pH, which favor their preferential binding to Aβ; thus, they have been utilized in detection of Aβ plaque aggregation in AD [25
Based on the above findings, we were interested in testing the efficacy of different boron compounds in animal models of AD. Boronic compounds have not been explored much in the field of AD research, especially in experimental AD models. In our previous study, we found some evidence of a reduction of neurodegeneration in a C. elegans which genetically overexpressed Aβ42. However, the mechanism in which the boron compounds interact with Aβ remains unknown. In the present study, we investigated whether boron compounds can increase survival and neuronal integrity in a C. elegans, which is genetically modified to overexpress Aβ42 and protect against neuroinflammation, amyloid plaque loads, and cognitive dysfunction in the 5xFAD mouse model of AD.
Initially, we compared the Aβ42 aggregation inhibition capability of trans 2-phenyl vinylboronic acid (TPVA) MIDA ester and trans-beta-styryl-boronic acid (TBSA) using a dot blot assay. To determine the relative efficacy as an Aβ42 aggregation inhibitor between these two boronic compounds, we used a synthesized Aβ42 peptide (at 10 µM) and allowed it to aggregate with or without these compounds (in µg/mL: 1-, 5- and 10 of both TBSA and TPVA). Interestingly, we found that, at low concentrations (5–10 µg, but not 1.0 µg), both TPVA and TBSA were able to inhibit Aβ42 aggregation, suggesting that low doses of boronic compounds would be sufficient to inhibit Aβ aggregation [27
]. We also observed that TBSA was a more potent Aβ42 aggregation inhibitor than TPVA (Figure 1
F). We used 6E10, which binds with monomers and oligomers fibrils (1–16 residues of N-terminal fragments). Although oligomer specific (e.g., A11) or fibril specific antibodies (e.g., OC) are preferable for investigating anti-amyloid properties of boron compounds by dot blot assay, other researchers have also used 6E10 to investigate overall Aβ42 aggregation pattern and to make comparisons with other antibodies, such as A11 or OC to investigate oligomer and fibril formation [21
]. Moreover, to better understand the anti-amyloid properties of boron compounds, we should investigate the changes of secondary structure (random coil to β-pleated sheet) of Aβ42 peptide by circular dichroism spectroscopy (CD spectroscopy), or study the amyloid aggregation size by dynamic light scattering or transmission electron microscopy to investigate the overall morphological changes from monomer to oligomers, protofibril or fibril formation; however, due to limitations in our research facility, we could not use these techniques. The mechanism for this difference is unknown. It is suspected that, being a smaller molecule than TPVA, TBSA may have greater facility for entering the hydrophobic core of the Aβ molecules to prevent its aggregation. Furthermore, the steric hindrance of TPVA might lessen its reaction with Aβ42 molecules more than that of TBSA, although these hypotheses require further investigation.
If these boronic compounds inhibit Aβ42 aggregation, then they might help reduce the toxic effects of Aβ42 and yield increased neuroprotective effects. To test this hypothesis, we cultured mouse neuronal cell lines (N2a) and treated them with Aβ42 as a toxin (10 µM) for 24 h in the presence or absence of either TPVA or TBSA. Interestingly, we observed that TBSA, not by TPVA, protected neuronal death caused by Aβ42 insult (Figure 1
H). The toxicity data paralleled our dot blot assay findings, suggesting TBSA might be a better candidate to inhibit Aβ aggregation and neurotoxicity. However, the mechanistic details of neuroprotection caused by TBSA require further investigation.
The therapeutic efficacy of TBSA was then investigated in vivo using a C. elegans
expressing Aβ42 in its glutamatergic neurons, which has been demonstrated to undergo Aβ-induced neurodegeneration [15
]. Since TBSA has not been previously investigated in vivo, it was important to evaluate whether or not chronic treatment with it had any impact on the viability and survival of the whole animal. The data show that the TBSA treatment did not appear to influence animal survival in control animals or AD animals up to 15 days posthatching (Figure 2
A). At the old age of 17 days, however, TBSA treated AD mice did have a statistically significant improvement in survival compared to the untreated AD animals (Figure 2
A,B). It is possible that this improvement in survival can be explained by the statistically significant improvement in neuronal health seen with treatment with TBSA (Figure 2
C,D). As the data show, examination of the number of live GFP-expressing glutamatergic neurons that persist in the tail region of animals of advanced age provides a robust way to assess neuronal health in Aβ-expressing neurons (Figure 2
C). Healthy young adult to adult animals normally express 5 GFP-positive neurons in their tails. Beyond the difference in the sheer number of living neurons (Figure 2
D), micrographs also reveal that many neurons in the untreated animals, while still present, are more degenerated, as evidenced by their reduced size and the increase in aberrant varicosities (Figure 2
C). Again, the mechanism of action for TBSA to help protect the neurons from degeneration needs further investigation. It is true that the survival of the Peat-4::GFP control animals treated with TBSA had the lowest survival rate, i.e., 17 days of age, but that survival was not statistically different from Peat-4::GFP untreated animals, nor the untreated C. elegans
expressing Aβ42. Hence, only the C. elegans
expressing Aβ42 treated with TBSA had a stronger survival rate by day 17. We reason that if Aβ were actually beneficial, then we would expect the C. elegans
expressing Aβ42 untreated animals to either not be statistically different from the C. elegans
expressing Aβ42-treated and/or to be statistically different from the Peat-4::GFP control worms. We hypothesize that the TBSA drug confers a potent support of health in the Aβ42-expressing animals, and that in the absence of Aβ, the drug may confer a low level of toxicity without this preferred molecular target.
Based on these findings and previous experimental observations [5
], we decided to use TBSA as a potential therapy in 5xFAD mice. After 2 months of oral gavage (0.5 mg/kg), we found a significant decrease in Aβ plaque burden in several brain areas, such as the cortex, and the CA1, CA3, and DG regions of hippocampus (Figure 3
), suggesting that the TBSA crosses the blood brain barrier, permeates into the brain tissue, and inhibits amyloidogenic pathways, either by reducing Aβ production or preventing its aggregation. Although we did not measure the boron levels in the TBSA-treated mice brain tissue, previous studies have shown that boron compounds can cross the blood brain barrier and penetrate the brain tissue [32
] and decrease inflammatory responses [33
]. Combined with previous research showing efficacy of boron compounds in the brain [3
], our findings of the efficacy of TBSA in the present study strongly suggest that TBSA is sufficiently permeable to brain tissue to reduce AD pathology.
We also investigated whether TBSA treatments preserved neuronal morphology in affected areas of the 5xFAD brain. We used cresyl violet (CV) stains in paraffin-embedded tissue sections to investigate overall neuronal morphology and number of pyknotic cells. Our morphometric data suggested that TBSA treatment significantly reduced the number of pyknotic, or tangle-like, cells (Figure 4
) in the cortex and in the CA1 and CA3 subfields of the hippocampus in the 5xFAD mouse brain. These findings are in concert with the findings of decreased Aβ plaque burden in TBSA-treated 5xFAD mice, suggesting TBSA may have an inhibitory role on Aβ production.
Inflammation is an important pathological correlate with Aβ accumulation. Reducing the spread of inflammatory response requires brain immune cells to engulf or sequester the amyloid plaques. Therefore, the activation of astrocytes and microglia is a primary event in AD pathology. Given that boron compounds have a vast array of anti-inflammatory actions, we investigated whether TBSA could prevent neuroinflammation in 5xFAD mice, especially in the most affected brain areas of this AD mouse model. We measured the number of activated astrocytes (GFAP-IR) and microglia (Iba-1-IR) in affected areas. We observed a clear inhibition of astrocytic and microglial activation, as the number of GFAP-IR and Iba-1-IR was significantly decreased in all the areas studied (Figure 5
and Figure 6
). Although we did not categorize which type of microglia were inhibited by TBSA treatment, our findings are in concert with others which demonstrate that boron compounds can attenuate inflammatory responses in different animal models of neurological diseases [14
]. How astrocytes and microglia become inhibited by TBSA treatment is unknown, but in vitro studies have revealed that boron can affect the production of inflammatory cytokines by cartilage cells, which are involved in the inflammatory response [34
]. In addition, boron, as boric acid, can stimulate the synthesis and release of tumor necrosis factor-α (TNF-α) in chick embryo cartilage and fibroblasts [34
], suggesting that it could decrease tau phosphorylation in AD [21
Very few studies have reported that daily intake of boron compounds can have cognitive effects, especially for improving learning, memory, and attention in normal individuals [6
]. In the present study, we attempted to investigate the effects of TBSA on counteracting AD-associated cognitive impairments in 5xFAD mice. Using the NOR task, we measured recognition memory, which is commonly impaired in AD patients. We observed a significant deterioration of both the discrimination index (Figure 8
D) and the exploration index (Figure 8
C) in 12-month-old 5xFAD mice, which was mitigated by TBSA treatments (Figure 8
C,D). This was probably a mnemonic effect and not the result of increased anxiety or activity levels, as measures of fecal boli and movements in the open field did not reveal persistent deficits in the 5xFAD mice.
Impairment of the spatial memory is common in AD patients [35
]. To investigate the role of TBSA on spatial memory tasks in 5xFAD mice, we performed the Morris-water-maze (MWM) task. The significant increase in escape latency (on days 4 and 5) and pathlength (on day 5) to find the hidden platform in the 12-month-old vehicle-treated 5xFAD mice was prevented by the TBSA treatments (Figure 9
). Because the average swim speed was equivalent between both the vehicle- and TBAS-treated 5xFAD mice (Figure 9
E), the differences accurately reflect spatial memory abilities. In addition, during the probe trial, TBSA-treated 5xFAD mice took less time to enter to the target quadrant and kept their mean distance closer to the target quadrant compared to the vehicle-treated 5xFAD mice (Figure 10
), suggesting TBSA protects against spatial memory deficits in 5xFAD mice.
The modest TBSA-induced protection against cognitive deficits in the 5xFAD mice may be due to the TBSA-induced reduction in amyloid plaques, reduced neuroinflammation, and/or reduced neurodegeneration (Figure 11
). Previous reports have suggested that boronic compounds bind directly to amyloid precursor protein (APP) or Aβ [22
]; we found a significant reduction in Aβ aggregation in vitro and decreased plaques levels in 5xFAD mice, which could reduce the overall neuroinflammation and neuronal death, although the extent to which any of these and other possible mechanisms of action can explain the therapeutic effects of TBSA requires further investigation, especially in the context of optimal dose and duration of TBSA treatments. To our knowledge, this is the first study using any boron compound to treat deficits in the 5xFAD mouse model of AD. Extending these findings to other models may provide a stronger basis for the use of boron compounds in clinical trials.
Our findings support the work of Penland, who found that a dietary boron intake increases overall brain function and cognitive performances in humans [6
]. Similarly, Nielsen and Meacham reported that boron supplementation after boron deprivation resulted in improved functioning, including less drowsiness and mental alertness, improved psychomotor skills (e.g., motor speed and dexterity), and improved cognitive processing (e.g., attention and short-term memory) in older men and women as shown by electroencephalograms [6
]. In another study, Nielsen and Penland reported that boron deprivation in rats reduced the number, distance, and time of horizontal movements, front entries, margin distance, and vertical breaks and jumps in assessments of spontaneous activity compared with performance of rats given boron supplements [5
]. Our findings confirm earlier work and extend previous research showing that boron compounds can counteract both impaired recognition and spatial memory dysfunction.