
This paper discusses the robustness of neutral-type Cohen-Grossberg neural networks with time delays and stochastic disturbances. And the problem is whether the Cohen-Grossberg neural networks, which originally maintain exponential stability, still achieves exponential stability when subjected to three simultaneous disturbances, namely, time delays, stochastic perturbations, and neutral terms. First, the width of the time delays, the strength of the stochastic disturbances, and the neutral term preset parameter size are derived through the Bellman-Gronwall Lemma, the Itô formula, and the properties of integrals. Next, the values of the three perturbation factors of time delay, stochastic disturbance, and neutral term are obtained by solving a multivariate privacy transcendental equation, which allows the Cohen-Grossberg neural networks to remain exponentially stable after being disturbed. Finally, the numerical example is provided to validate the results of this brief.
Citation: Yijia Zhang, Tao Xie, Yunlong Ma. Robustness analysis of exponential stability of Cohen-Grossberg neural network with neutral terms[J]. AIMS Mathematics, 2025, 10(3): 4938-4954. doi: 10.3934/math.2025226
[1] | K. Saritha, K. Thilagavathi . Differential subordination, superordination results associated with Pascal distribution. AIMS Mathematics, 2023, 8(4): 7856-7864. doi: 10.3934/math.2023395 |
[2] | Rabha W. Ibrahim, Dumitru Baleanu . Fractional operators on the bounded symmetric domains of the Bergman spaces. AIMS Mathematics, 2024, 9(2): 3810-3835. doi: 10.3934/math.2024188 |
[3] | Madan Mohan Soren, Luminiţa-Ioana Cotîrlǎ . Fuzzy differential subordination and superordination results for the Mittag-Leffler type Pascal distribution. AIMS Mathematics, 2024, 9(8): 21053-21078. doi: 10.3934/math.20241023 |
[4] | Rabha W. Ibrahim, Jay M. Jahangiri . Conformable differential operator generalizes the Briot-Bouquet differential equation in a complex domain. AIMS Mathematics, 2019, 4(6): 1582-1595. doi: 10.3934/math.2019.6.1582 |
[5] | Ekram E. Ali, Rabha M. El-Ashwah, R. Sidaoui . Application of subordination and superordination for multivalent analytic functions associated with differintegral operator. AIMS Mathematics, 2023, 8(5): 11440-11459. doi: 10.3934/math.2023579 |
[6] | Huo Tang, Muhammad Arif, Khalil Ullah, Nazar Khan, Bilal Khan . Majorization results for non vanishing analytic functions in different domains. AIMS Mathematics, 2022, 7(11): 19727-19738. doi: 10.3934/math.20221081 |
[7] | Alina Alb Lupaş, Shujaat Ali Shah, Loredana Florentina Iambor . Fuzzy differential subordination and superordination results for q -analogue of multiplier transformation. AIMS Mathematics, 2023, 8(7): 15569-15584. doi: 10.3934/math.2023794 |
[8] | Ekram E. Ali, Nicoleta Breaz, Rabha M. El-Ashwah . Subordinations and superordinations studies using q-difference operator. AIMS Mathematics, 2024, 9(7): 18143-18162. doi: 10.3934/math.2024886 |
[9] | Ebrahim Amini, Mojtaba Fardi, Shrideh Al-Omari, Rania Saadeh . Certain differential subordination results for univalent functions associated with q-Salagean operators. AIMS Mathematics, 2023, 8(7): 15892-15906. doi: 10.3934/math.2023811 |
[10] | Gangadharan Murugusundaramoorthy, Luminiţa-Ioana Cotîrlă . Bi-univalent functions of complex order defined by Hohlov operator associated with legendrae polynomial. AIMS Mathematics, 2022, 7(5): 8733-8750. doi: 10.3934/math.2022488 |
This paper discusses the robustness of neutral-type Cohen-Grossberg neural networks with time delays and stochastic disturbances. And the problem is whether the Cohen-Grossberg neural networks, which originally maintain exponential stability, still achieves exponential stability when subjected to three simultaneous disturbances, namely, time delays, stochastic perturbations, and neutral terms. First, the width of the time delays, the strength of the stochastic disturbances, and the neutral term preset parameter size are derived through the Bellman-Gronwall Lemma, the Itô formula, and the properties of integrals. Next, the values of the three perturbation factors of time delay, stochastic disturbance, and neutral term are obtained by solving a multivariate privacy transcendental equation, which allows the Cohen-Grossberg neural networks to remain exponentially stable after being disturbed. Finally, the numerical example is provided to validate the results of this brief.
From current studies on human whole genomes, we have already learned that “cancer is a genetic disease” after understanding the genomic/genetic differences between cancer and normal cells through next-generation sequencing [1],[2]. At present, whole-genome sequencing data can predict the incidence of specific diseases, including cancer, in healthy individuals [3]. However, recent metabolome studies suggest that metabolites could be biomarkers for cancer and its development [4], suggesting as well that “cancer is a metabolic disease”. It is well known that cancer mainly uses glucose to produce adenosine triphosphate (ATP) by glycolysis. The “Warburg effect”, or abnormal metabolism, which was described by Dr. Otto Warburg over 60 years ago, is widely known as one of the most essential characteristics in cancer cells [5]–[7]. Importantly, it is frequently observed that both dysregulation of the tricarboxylic acid (TCA) cycle (Krebs/Citrate cycle) and oxidative phosphorylation occur in cancer cells, suggesting that mitochondrial dysfunction precedes the metabolic change [8], indicating that “cancer is a mitochondrial disease”. Mitochondria have their own DNA or mitochondrial DNA (mtDNA), suggesting that they have evolved from the symbiosis of aerobic bacteria (α-proteobacteria) into ancestral eukaryotic cells. However, it should be noted that most (99%) of the genes that encode mitochondrial function-associated proteins are contained in nuclear genomes [9]. Therefore, mitochondrial function-associated genes are commonly regulated by transcription factors (TFs) that regulate other cellular protein-encoding genes.
In this regard, scientific interest should be focused on causative factors that lead to mitochondrial dysfunction by asking whether “cancer could be a transcriptional disease”. In this chapter, through a discussion mainly about biological functions and regulations of nicotinamide adenine dinucleotide (NAD+), which is not only required as a co-enzyme for various biologically significant reactions but also as a substrate for poly (ADP-ribose) polymerase (PARP), we propose a novel anti-cancer therapy that converts or forces cancerous cells to return to a non-proliferating or differentiated state.
Before discussing the biological roles and functions of the NAD+ molecule, we should reconfirm the characteristics of cancer cells, which frequently show differences in metabolism and transcriptional state from normal healthy cells. Cancer cells very frequently show disruption in metabolism compared with normal differentiated cells. First, proliferating mammalian cells exhibit anaerobic metabolism [6], and the energy they require is thought to be supplied mainly from glycolysis. Moreover, in cancer cells, abnormalities in structures and functions of mitochondria are observed [8]. Hence, deficiencies in carbohydrate metabolism, including the TCA cycle and oxidative phosphorylation, which are carried out in the mitochondria, are frequently accompanied by malignancies of cancer cells [7]. Second is the difference in amino acid metabolism. It is well known that cancer cells require high amounts of glutamine to survive and proliferate [10]–[12]. Expression of the GS and GDH genes, which encode glutamine synthetase and glutamate dehydrogenase, respectively, is significantly increased in breast cancer cells, and accumulated ammonia is efficiently used for glutamine and glutamate synthesis without engaging the urea (ornithine) cycle [13]. As described in many textbooks on biological chemistry, the TCA cycle is overlapped by the urea cycle, in which ammonia is processed to urea and is safely eliminated. However, in cancer cells, the urea cycle does not work efficiently because of insufficient progression of the TCA cycle. In addition, serine and one-carbon metabolism is also dysregulated in cancer, which might also affect the rate of synthesis of thymidine and purine nucleotides [14]. If mitochondrial functions are disrupted, serine synthesis, which is catalyzed by SHMT2 with the coenzyme tetrahydrofolate, would be insufficient in the mitochondria [14]. Consequently, cancer cells might exhibit dependence on higher amounts of serine to produce more glycine for use in the de novo synthesis of purines and pyrimidines. Generally, de novo synthesis of nucleotides is hyperactivated in proliferating cells. Moreover, malignant tumors usually utilize greater amounts of uracil compared with the normal surrounding cells [15]. Thus, the nucleotide metabolism in cancer cells is significantly different from that in normal cells [16],[17]. In proliferating malignant cancer cells, both higher telomerase reverse transcriptase activities and longer telomeric regions of chromosomes than in normal differentiated cells are observed [18]. Nucleotides that were synthesized in excess could be incorporated into the terminal regions of chromosomes that will result in telomere elongation. This may be consistent with the observations that communications between mitochondria and nuclear DNAs/telomeres play roles in the regulation of longevity and the incidence of cancer [19],[20].
It is widely known that cancer cells carry numerous genomic mutations in chromosomal DNAs. Recent studies on human genomic DNA sequences revealed several cancer predisposition genes [2]. In this context, “cancer is a genetic disease” [21]. However, cancer cells have many different characteristics of metabolism from those of normal differentiated cells, including those related to carbohydrates, amino acids, and nucleotides. Because most of these biological reactions depend on mitochondrial functions, “cancer is a metabolic/mitochondrial disease” [8]. Although mutations in mtDNA cause mitochondrial deficiencies that could lead to cancer development [22], most of the mitochondrial proteins, including the TCA cycle/urea cycle enzymes, are encoded on nuclear DNAs [9]. Not only cancer but also neurodegeneration is thought to be caused by mitochondrial deficiencies [23]. To better understand molecular pathogenesis in the development of cancer and aging-related diseases, transcription mechanisms that control the expression of mitochondrial function-associated genes should first be elucidated.
Niacin is classified as vitamin B3, whose depletion causes pellagra. As shown in Figure 1, nicotinic acid (NA) and nicotinamide, which provide the biological activity of niacin, are generated from the metabolism of tryptophan, giving an essential group for constructing an NAD+ molecule [24]. NAD+ is known as a molecule that is involved in energy metabolism, substrate oxidation, and antioxidation [25]–[27]. In glycolysis, NAD+ is reduced to NADH when glyceraldehyde 3-phosphate is oxidized to bisphosphoglycerate. Accepting H-from NADH, pyruvate is reduced to lactate. Through this simple oxidization/reduction cycle of NAD+/NADH, one glucose molecule can produce two ATP and two lactate molecules independent of mitochondrial function or the respiration process. In this way, the Warburg effect would be observed specifically in most cancer cells. In normal cells, pyruvate is oxidized by pyruvate dehydrogenase [28] and converted to acetyl CoA, which can then be completely oxidized to CO2 and H2O through the TCA cycle. This reaction requires NAD+ as an acceptor of H-. Three enzymes, isocitrate dehydrogenase, α-ketoglutarate dehydrogenase, and malate dehydrogenase, which are encoded by the IDH (IDH1, IDH2, IDH3A, IDH3B, and IDH3G), OGDH, and MDH (MDH1, MDH2, and MDH1B) genes, respectively, require NAD+ as a coenzyme for oxidization of each substrate in the TCA cycle [29]. As long as the mitochondrial electron transport chain functions, NADH will be immediately oxidized to NAD+ by complex I. Moreover, in β-oxidation of fatty acids, NAD+ is used as an acceptor of H- when 3-hydroxy-acyl-CoA dehydrogenase works. β-Hydroxybutyrate, which is one of the ketone bodies, is converted to acetoacetate. The reaction is catalyzed by β-hydroxybutyrate dehydrogenase (BDH), which is encoded on the BDH1 and BDH2 genes [24]. In the metabolism of nucleotide bases, NAD+ functions as a coenzyme for inosine 5′-phosphate (IMP) dehydrogenase (IMPDH) and dihydropyrimidine dehydrogenase (DPYD) [30], which are encoded by the IMPDH1/IMPDH2 and DPYD genes, respectively.
It should be noted that the NAD+ molecule is consumed for the synthesis of poly (ADP-ribose) (PAR) by PARP especially when DNA damage occurs [31]. The biological relevance of the NAD+ molecule in the protection of chromosomal DNAs has been reviewed and discussed previously [26],[32]. A recent study showed that nuclear PAR can be used by NUDIX5 to supply ATP molecules, which are required for chromatin remodeling [33]. Moreover, NAD+-dependent deacetylases, sirtuin proteins that regulate both cancer generation and aging [34],[35], possess histone deacetylase activity [36]. This observation suggests that sirtuins play roles in epigenetic regulation.
In summary, NAD+ is not only required for the oxidization process but also for DNA repair, chromatin remodeling, and epigenetic control. Importantly, these biological events that occur in the nucleus corelate with cancer generation/development and the aging process. Therefore, it is important to revisit the biosynthesis and metabolism of NAD+.
Figure 1 summarizes the biosynthesis and metabolism of the NAD+ molecule [26]. NAMPT is a nicotinamide phosphoribosyltransferase that catalyzes the first rate-limiting step of NAD+ synthesis from nicotinamide [37]. Therefore, its activity may modulate biologically important reactions such as the TCA cycle, poly (ADP-ribosyl)ation, and sirtuin-mediated de-acetylation that all depend on NAD+. The reaction product, nicotinamide mononucleotide (NMN), can be bound with ATP to produce NAD+ by nicotinamide nucleotide adenylyltransferases (NMNATs). Depletion of NMNAT2, which localizes in cytosol-facing Golgi apparatus, by siRNA introduction into cultured cells results in a decrease of both cytoplasmic and mitochondrial NAD+, suggesting that an NAD+ transport system is at work in mammalian cells [38]. NMN could be alternatively produced through nicotinamide riboside phosphorylation, which is catalyzed by nicotinamide riboside kinases. NAD+ can also be synthesized from nicotinic acid adenine dinucleotide (NAAD) by NAD synthetase (NADSYN1), and NAAD can be produced by de-phosphorylation of nicotinic acid adenine dinucleotide phosphate (NAADP) [39]. Moreover, it should be noted that PGC-1α, which is a pivotal regulator of mitochondrial biogenesis, up-regulates the de novo synthesis of NAD+ in mice [40].
Expression of the NAMPT gene is induced in glioblastoma cells and upregulates the E2F2/ID pathway [41]. A recent study revealed that the NAMPT/E2F2/SIRT1 axis promotes the proliferation of human melanoma cells [42]. Depletion of NAD+ by NAMPT inhibitor FK866 effectively leads to cell death in Ewing sarcoma [43]. A detailed analysis of metabolites revealed that in cancer cells, NAMPT inhibition suppresses both glycolysis and the TCA cycle [44]. These observations imply that NAMPT could be targeted in therapeutics for the treatment of some specific cancers, including gastric cancer and colorectal cancer, that overexpress the NAMPT (PBEF) gene [45]–[47]. The NAMPT protein was shown to form a complex with beta-actin and POTTE (POTE ankyrin domain family member E), and it might be responsible for the sensitivity to inhibitors such as FK866 [48]. Analyses of cancer genomes have revealed that the NAPRT, encoding nicotinate phosphoribosyltransferase, and NADSYN1 genes are amplified in many cancer types [49]. The NAPRT protein uses nicotinic acid as a substrate to produce nicotinic acid mononucleotide, suggesting that the Preiss-Handler (PH) pathway affects cancer generation. In addition, NAMPT gene transcription could be epigenetically controlled in non-PH-amplified cancer [49]. Another important NAD+ metabolism is the synthesis of S-adenosylhomocysteine (SAH) and a metabolically inert 1-methyl nicotinamide (1-MNA) by a nicotinamide N-methyltransferase (NNMT) that transfers a reactive methyl group from S-adenosylmethionine (AdoMet) to nicotinamide. AdoMet is generally known to be an essential donor of methyl group to histones, DNAs, and RNAs. Thus, NNMT-mediated AdoMet depletion affects gene expression [50]. Recently, it was shown that NNMT is a master metabolic regulator of cancer-associated fibroblasts [51]. In this regard, NAD+ metabolism could be linked with epigenetic regulation. Because nicotinamide, which is an acceptor of the methyl group, could play a role in depletion of AdoMet.
Not only the overproduction of NAD+ but also a decrease in its level can be problematic. NAD+ and its precursor nicotinamide can ameliorate metabolism or mitochondrial functions [52]–[54]. Repletion of NAD+ improves mitochondrial functions to prolong the life span of adult mouse stem cells [55]. Conversely, decreased concentrations of NAD+ will cause aging or aging-related diseases [56]. These observations are consistent with the concept that the NAD+ level correlates with mitohormesis [57] and that nutrient-sensing molecules could control aging [58].
In breast cancer cells, knockdown of the NDUFV1 subunit leads to an aberration in complex I to enhance agammaessiveness or metastasis [59]. Therefore, upregulation of the mitochondrial NAD+ level will improve mitochondrial respiratory function or metabolism and that could suppress oncogenesis. Mitochondrial biogenesis is mainly regulated by PGC-1α, which drives NAD+ biosynthesis and thereby induces stress resistance [39]. PGC-1α suppresses the metastasis of melanoma by affecting the PGC-1α-ID2-TCF-integrin axis through transcriptional control [60]. A loss of mitochondria localizing protein CSB [61] can activate PARP1, which plays essential roles in DNA repair [62], thus consuming the NAD+ molecule. Therefore, dysregulation of the mitochondria or hyperactivation of the PARP enzyme will reduce the NAD+/NADH molecular ratio. Mutations on the IDH1 and IDH2 genes have been identified in human brain cancer cells [21],[63],[64]. In addition, mutation-introduced IDH2 could generate sarcomas [65]. Incomplete functioning of the IDH proteins might retard progression of the TCA cycle, thus leading to mitochondrial dysfunction. It should be noted that 5′-upstream regions of the TCA cycle enzyme-encoding genes, including IDH1, IDH3A, and IDH3B, contain duplicated GGAA motifs (Figure 2), which are recognized by various TFs, such as ETS family proteins [66]. Surveillance of the NCBI database (https://www.ncbi.nlm.nih.gov/gene/) indicated that the duplicated GGAA motifs are contained in the 5′-upstream regions of the TCA cycle enzyme-encoding genes, including PDH, MDH1, and MDH2 (Figure 2). Interestingly, the IDH3G, OGDH, and MDH1B gene promoters have no such elements, but the reactions that are catalyzed by the encoded enzymes require NAD+ as a coenzyme.
Metabolic dysfunction is thought to play a major role in the development of neurodegenerative diseases such as Alzheimer's disease (AD), Parkinson's disease (PD), and Huntington's disease (HD) [67]. Thus, mitochondrial dysfunction is not only thought to be one of the causations of cancer but also of neurodegenerative disease [68]. For example, high levels of deletions in mtDNA have been identified in PD [69]. In addition, an NAD+-dependent deacetylase SIRT3, which localizes in mitochondria, may play a role in the development of neurodegenerative disorders, including AD, PD, and HD [70],[71]. Caloric restriction (CR) attenuated the accumulation of amyloid-β peptide (Aβ) plaques in a transgenic mouse model of AD [72]. It has been reported that CR-induced SIRT1 reduced AD pathology [73]. Interestingly, SIRT2, which is a tubulin deacetylase, plays a role in accelerating Aβ agammaegation [74]. In addition, SIRT6 can protect cells from Aβ-induced DNA damage [75]. Therefore, NAD+ is thought to play a role in the pathogenesis of AD and other neurodegenerative diseases by its effect on sirtuin proteins.
Epigenetic regulation, which is mainly driven by modifications of chromosomal DNAs and histone proteins [76], determines the chromatin compositions and structures [77]. Epigenetic alterations are frequently found in cancer, implying that “cancer is an epigenetic disease” [78]. The correlation between cellular metabolites and gene expression has been proposed as the RNA/enzyme/metabolite (REM) networking system [79]. Because the metabolites NAD+, AdoMet, and acetyl CoA are the substrates for poly (ADP-ribosyl)ation, methylation, and acetylation, respectively [80], they could be involved in epigenetic regulation.
The NAD+ molecule is the substrate of the PARP enzymes, which play pivotal roles in the DNA repair system [26],[31],[81]. The inhibition of PARP1 ameliorates mitochondrial metabolism through the activation of SIRT1 [82]. Conversely, over-activation of PARP1 can lead to mitochondrial dysfunction [83]. PARP1 gene expression was found to be negatively regulated when poly (ADP-ribose) glycohydrolase (PARG) siRNAs were introduced into HeLa S3 cells [84]. Therefore, degradation of the PAR macromolecule might be required for transcription of the PARP1 gene. It has been suggested that PARP1 is involved in the NAD+-sensitive transcription system [85]. Interestingly, duplicated GGAA motifs are commonly contained in the 5′-upstream regions of the human PARP1 and PARG genes [66],[86], suggesting that both genes are regulated in an NAD+-dependent manner. The NAD+/NADH ratio may not only directly affect mitochondrial functions and DNA repair but also the transcription of DNA repair factor-encoding genes.
Presently, it has been known that trans-resveratrol (Rsv)-activated mitochondrial complex I upregulates the NAD+/NADH ratio [87], and that leads to the modification of chromatin-associating proteins [88] and modulation of the gene expression system [89]. The human TP53 gene promoter, which contains a duplicated GGAA motif, is up-regulated in HeLa S3 cells in response to Rsv [90]. Of note, the bidirectional BRCA1/NBR2 promoter region contains a duplicated GGAA motif and is regulated by a metabolic switch depending on the NAD+/NADH ratio, which could be elevated by CR-mimetic drugs [91]. The C-terminal binding protein (CtBP) [92],[93] might have an essential role in this regulation as a metabolic sensor. Recently, it was shown that CtBP could work as an NAD(H)-sensitive transcriptional co-repressor that regulates NF-κB-driven pro-inflammatory transcription in macrophages and microglia [94]. Co-transfection of the CtBP1 or CtBP2 expression vector to CtBP1/CtBP2-knock down cells down-regulated the Luc activities of the NF-κB motif containing Luc reporter plasmid-transfected cells [94]. The NF-κB p65 (RELA) homodimer recognizes two symmetric sequences, 5′-GGAATTTCC-3′ and 5′-GGAATTCCC-3′, which are contained in the IL8 and COL7A1 enhancers, respectively [95]. Therefore, duplicated GGAA motifs might play a role in responding to CtBP that affects NF-κB, which can bind to the duplicated GGAA motifs. Moreover, the GABP TF (nuclear respiratory factor 2), which belongs to the ETS family and directly recognizes GGAA motifs, is required for mitochondrial biogenesis [96]. In summary, the duplicated GGAA (TTCC) motifs, which are very frequently contained in the 5′-upstream regions of a number of genes that encode DNA repair factors and TCA cycle-associating enzymes (Figure 2) [66],[96], may be partly dependent on the level of NAD+ that affects activities of CtBP and GGAA motif-binding TFs. A recent study showed that mitochondrial LACTB could be a potent tumor suppressor [97]. Notably, the duplicated GGAA motifs are present in the 5′-upstream of the LACTB gene.
It is noteworthy that oxidative stress-responding TF Nrf2 recognizes antioxidant response element (ARE) in the promoter region of the NQO1 gene, encoding NAD(P)H:quinone oxidoreductase1 (NOQ1), and upregulates its expression [98]. Interestingly, ARE is contained in the promoter region of the ETS1 gene [99], which encodes TFs that bind to the GGAA motif-containing sequences [100]. Therefore, the Nrf2/ARE pathway may also contribute to regulating the expression of genes encoding proteins that play roles in the redox system or TCA cycle. Another important gene expression controlling system might be epigenetic alteration by sirtuin proteins, which are the NAD+-dependent deacetylases [101],[102], and sirtuin deacetylate TFs such as FOXO, NF-κB, and STAT proteins [103],[104]. Importantly, SIRT1 is involved in mitochondrial biogenesis depending on the metabolic state in muscle [105]. SIRT3, which localizes in mitochondria, can up-regulate functions of mitochondrial proteins, including TCA cycle enzymes [106]. SIRT1, SIRT2, and SIRT3 proteins are involved in the reactive oxygen species defense mechanism co-operatively regulating various TFs, including FOXO, HSF1, PGC-1α, and p53 [34]. Thus, NAD+ may modulate or fine-tune the transcriptional state of both DNA repair- and mitochondrial function-associated genes, which are required for cells to manage various stresses from, for example, DNA damage, nutrient conditions, and viral infections.
It might be paradoxical and controversial that DNA damage-activated PARP1 leads to the depletion of NAD+ from cytoplasm and mitochondria. Several PARP inhibitors (PARPi) have been developed and have successfully completed clinical trials [107],[108]. The favorable anti-cancer effect of PARPi can be explained by the concept of “synthetic lethality” in cancers that carry BRCA1/BRCA2 mutations or double-stranded DNA break repair deficiencies [109]. However, the suppression of cancer development by PARPi could also be explained by the inactivation of PARP1, which consumes NAD+ as a substrate for poly (ADP-ribose) synthesis and increases the NAD+/NADH ratio. Nevertheless, because PARylation is the primarily required response for double-stranded DNA repair synthesis, complete inhibition of the PARP enzyme activity would be unfavorable or harmful for cells. Therefore, advantages and disadvantages of PARPi in the treatment of tumors might have been observed in different clinical settings [110]. Inhibitors of the PAR-degrading enzyme PARG could also be used as anti-cancer drugs to induce cell death or apoptosis. Knockdown of PARG by siRNAs in HeLa S3 cells enhanced staurosporine-induced cell death [84]. Moreover, individual inhibition of either PARP1 or PARG is effective in cancer treatment, whereas combined inhibition of both enzymes does not affect cancer cell death [111]. These observations are consistent with a concept of homeostasis by maintaining the concentration of NAD+ at an appropriate level, which may avoid a pseudohypoxic state that disrupts nuclear-mitochondrial communication [112].
Several drugs that increase the intracellular NAD+ level are expected to contribute to the establishment of novel therapeutics for treating age-related diseases, including cancer [113]. Vitamin B3, which is a precursor of NAD+, could be administered for the treatment of glaucoma by modulating mitochondrial vulnerability [114]. A low dose of Rsv stimulates NADH dehydrogenases and mitochondrial complex I activity to increase the NAD+/NADH ratio [87]. Piceatannol, which is one of the resveratrol analogs with a potent antioxidant effect [115], might have a similar effect on the NAD+/NADH ratio. Rapamycin, which inhibits mTor, rescued the mitochondrial defects in a Ndufs4 knockout mouse model of Leigh syndrome [116]. The anti-diabetic agent metformin could serve as an anti-cancer agent by decreasing glucose oxidation and increasing dependency on reductive glutamine metabolism in the generation of prostate cancer cells [117]. Recently, it was shown that 4-hydroxy-2,2,6,6-tetramethylpiperidine-noxyl (TEMPOL) upregulates the NAD+ level and ameliorates the NAD+/NADH ratio in obese mice [118]. Sulforaphane and curcumin upregulate expression of the NOQ1 protein, which catalyzes the oxidization of NADH to produce NAD+ and is a biomarker for effective stimulation of the hepatic Nrf2/ARE pathway [119],[120]. These compounds could be applied for clinical use if they can ameliorate the inner cellular NAD+ level in vivo.
A recent study showed that the introduction of nicotinamide nucleotide transhydrogenase into glioma-initiating cells up-regulates the mitochondrial NAD+ level and impairs clonogenicity [121]. It should be noted that experiments with Caenorhabditis elegans and a mouse model of ATM deficiency showed that replenishment of NAD+ improves the life span and health span of both animals [122]. The duplicated GGAA motifs are present in numbers of 5′-upstream regions of the human DNA-repair factor encoding genes [66]. Moreover, we have recently reported that NAD+/NADH up-regulating-drugs could lead to up-regulation of promoter activities of human DNA-repair associated genes [123]. In addition to the natural and synthetic drugs, NAD+ up-regulators or GGAA-binding TF-encoding vectors that lead to the restoration of mitochondrial functions and DNA repair systems should be investigated in the establishment of new therapeutics for cancer or age-related diseases.
Presently, almost all established chemotherapies have been strictly based on the concept of killing cancer cells by targeting weaknesses specific to them but not to normal cells, which should be regulated Normal differentiated cells should be under the control not to proliferate unlimitedly. Normal undifferentiated cells, including stem cells or lymphoid cells, should be controlled not to proliferate without specific signals. The anti-cancer drugs can be classified into several groups. For example, they are anti-cancer antibiotics, DNA replication/telomerase inhibitors, and DNA-damage/apoptosis inducers, including alkylating agents and platinum compounds. Molecular targeting drugs have come into clinical application, and they are going to increase in number. In the clinical setting, even the immune receptor targeting drugs such as anti-PD1/-PD-L1 antibodies are included in the list of cancer-killing therapeutics. Moreover, a recent study indicated that inhibitory receptor PD-1 plays an essential role as a suppressor of oncogenic T-cell signaling in mice [124]. Hence, therapy targeting the immunological signaling system will require extra caution. In this chapter, with the understanding that currently used anti-cancer drugs can cause damage to both malignant cancer cells and normal healthy cells, we have asked and discussed whether the most effective cancer therapies might be the control of NAD+ levels in the cytoplasm, nucleus, and mitochondria, by maintaining them in an appropriate state. In a near future, a novel treatment policy will be established for the use of anti-cancer drugs or therapies. The most favorable drugs should neither reduce a patient's quality of life nor kill cancer cells. Rather, they would force malignant cells to return to their normal state with ameliorated mitochondrial functions and DNA repair systems and immunological responses. Instead of natural or chemical compounds, TF(s) could be introduced into cancer cells by specially designed vectors. Specific TF binding sequences, including duplicated GGAA motifs and GC boxes, are contained in numbers of gene promoter regions. We previously reported that the duplicated GGAA-motifs are very frequently present in 5′-regulatory regions of the mitochondrial genes in nucleus [86]. Moreover, the motif can be found in numbers of the 5′-upstream of human DNA-repair factor/tumor suppressor encoding genes [66],[125] and interferon-stimulated genes (ISGs) [126]. In the human TP53 and HELB gene promoters, GGAA motif and GC box elements co-operatively regulate the response to the natural compound Rsv in HeLa cells [90],[127]. However, multiple TFs have been shown to recognize such sequences. Therefore, the molecular mechanisms behind how these gene promoters are up-regulated in response to the NAD+/NADH ratio should be elucidated. Based on this novel concept, the design of anticancer/tumor drugs in combination with gene transfer vector(s) will contribute to the prevention of aging and treatment of its associated diseases, including cancer and neurodegenerative diseases.
[1] |
M. A. Cohen, S. Grossberg, Absolute stability of global pattern formation and parallel memory storage by competitive neural networks, IEEE T. Syst. Man Cy., 1983,815–826. http://dx.doi.org/10.1109/tsmc.1983.6313075 doi: 10.1109/tsmc.1983.6313075
![]() |
[2] |
C. J. Cheng, T. L. Liao, J. J. Yan, C. C. Hwang, Globally asymptotic stability of a class of neutral-type neural networks with delays, IEEE T. Syst. Man Cy. B, 36 (2006), 1191–1195. http://dx.doi.org/10.1109/TSMCB.2006.874677 doi: 10.1109/TSMCB.2006.874677
![]() |
[3] |
P. Kowsalya, S. S. Mohanrasu, A. Kashkynbayev, P. Gokul, R. Rakkiyappan, Fixed-time synchronization of inertial Cohen-Grossberg neural networks with state dependent delayed impulse control and its application to multi-image encryption, Chaos Soliton. Fract., 181 (2024), 114693. http://dx.doi.org/10.1016/j.chaos.2024.114693 doi: 10.1016/j.chaos.2024.114693
![]() |
[4] |
J. Li, Z. Peng, Multi-source image fusion algorithm based on cellular neural networks with genetic algorithm, Optik, 126 (2015), 5230–5236. http://dx.doi.org/10.1016/j.ijleo.2015.09.187 doi: 10.1016/j.ijleo.2015.09.187
![]() |
[5] |
X. Hu, G. Feng, S. Duan, L. Liu, A memristive multilayer cellular neural network with applications to image processing, IEEE T. Neur. Net. Lear., 28 (2016), 1889–1901. http://dx.doi.org/10.1109/TNNLS.2016.2552640 doi: 10.1109/TNNLS.2016.2552640
![]() |
[6] |
X. Yang, Z. Cheng, X. Li, T. Ma, Exponential synchronization of coupled neutral-type neural networks with mixed delays via quantized output control, J. Franklin I., 356 (2019), 8138–8153. http://dx.doi.org/10.1016/j.jfranklin.2019.07.006 doi: 10.1016/j.jfranklin.2019.07.006
![]() |
[7] |
R. Samli, S. Senan, E. Yucel, Z. Orman, Some generalized global stability criteria for delayed Cohen–Grossberg neural networks of neutral-type, Chaos Soliton. Fract., 116 (2019), 198–207. http://dx.doi.org/10.1016/j.neunet.2019.04.023 doi: 10.1016/j.neunet.2019.04.023
![]() |
[8] |
Z. Zhang, X. Zhang, T. Yu, Global exponential stability of neutral-type Cohen–Grossberg neural networks with multiple time-varying neutral and discrete delays, Neurocomputing, 490 (2022), 124–131. http://dx.doi.org/10.1016/j.neucom.2022.03.068 doi: 10.1016/j.neucom.2022.03.068
![]() |
[9] |
F. Kong, Q. Zhu, Antiperiodic dynamical behaviors of discontinuous neutral-type Cohen-Grossberg neural networks with mixed time delays, Comput. Intel., 36 (2020), 698–719. http://dx.doi.org/10.1111/coin.12262 doi: 10.1111/coin.12262
![]() |
[10] |
Y. Shen, J. Wang, Robustness analysis of global exponential stability of non-linear systems with time delays and neutral terms, IET Control Theory A., 7 (2013), 1227–1232. http://dx.doi.org/10.1049/iet-cta.2012.0781 doi: 10.1049/iet-cta.2012.0781
![]() |
[11] |
O. Faydasicok, S. Arik, The combined Lyapunov functionals method for stability analysis of neutral Cohen-Grossberg neural networks with multiple delays, Neural Networks, 180 (2024), 106641. http://dx.doi.org/10.1016/j.neunet.2024.106641 doi: 10.1016/j.neunet.2024.106641
![]() |
[12] |
R. Li, J. Cao, A. Alsaedi, F. Alsaadi, Exponential and fixed-time synchronization of Cohen–Grossberg neural networks with time-varying delays and reaction-diffusion terms, Appl. Math. Comput., 313 (2017), 37–51. http://dx.doi.org/10.1016/j.amc.2017.05.073 doi: 10.1016/j.amc.2017.05.073
![]() |
[13] |
Y. Wan, J. Cao, G. Wen, W. Yu, Robust fixed-time synchronization of delayed Cohen–Grossberg neural networks, Neural Networks, 73 (2016), 86–94. http://dx.doi.org/10.1016/j.neunet.2015.10.009 doi: 10.1016/j.neunet.2015.10.009
![]() |
[14] |
N. Ozcan, New conditions for global stability of neutral-type delayed Cohen–Grossberg neural networks, Neural Networks, 106 (2018), 1–7. http://dx.doi.org/10.1016/j.neunet.2018.06.009 doi: 10.1016/j.neunet.2018.06.009
![]() |
[15] |
H. Zhang, Inequalities and stability of stochastic fuzzy delayed Cohen–Grossberg neural networks, IEEE Access, 2023. http://dx.doi.org/10.1109/ACCESS.2023.3300581 doi: 10.1109/ACCESS.2023.3300581
![]() |
[16] |
X. Z. Liu, K. N. Wu, X. Ding, W. Zhang, Boundary stabilization of stochastic delayed Cohen–Grossberg neural networks with diffusion terms, IEEE T. Neur. Net. Lear., 33 (2021), 3227–3237. http://dx.doi.org/10.1109/TNNLS.2021.3051363 doi: 10.1109/TNNLS.2021.3051363
![]() |
[17] | M. A. Jamal, S. Das, S. Mukhopadhyay, Fixed-time synchronization of delayed inertial Cohen–Grossberg neural networks with desynchronizing impulses, Commun. Nonlinear Sci., 130 (2024), 107772. http://dx.doi.org/0.1016/j.cnsns.2023.107772 |
[18] |
J. Cheng, W. Liu, Stability analysis of anti-periodic solutions for Cohen–Grossberg neural networks with inertial term and time delays, Mathematics, 12 (2024), 198. http://dx.doi.org/10.3390/math12020198 doi: 10.3390/math12020198
![]() |
[19] |
M. A. Jamal, R. Kumar, S. Mukhopadhyay, O. M. Kwon, Fixed-time stability of Cohen-Grossberg BAM neural networks with impulsive perturbations, Neurocomputing, 550 (2023), 126501. http://dx.doi.org/10.1016/j.neucom.2023.126501 doi: 10.1016/j.neucom.2023.126501
![]() |
[20] |
X. Wang, J. Lan, X. Yang, X. Zhang, Global robust exponential synchronization of neutral-type interval Cohen-Grossberg neural networks with mixed time delays, Inform. Sciences, 676 (2024), 120806. http://dx.doi.org/10.1016/j.ins.2024.120806 doi: 10.1016/j.ins.2024.120806
![]() |
[21] |
M. Shen, Z. Wang, S. Zhu, X. Zhao, G. Zong, Q. G. Wang, Neural network adaptive iterative learning control for strict-feedback unknown delay systems against input saturation, IEEE T. Neur. Net. Lear., 2024, 1–10. http://dx.doi.org/10.1109/TNNLS.2024.3452721 doi: 10.1109/TNNLS.2024.3452721
![]() |
[22] |
Q. Meng, Q. Ma, Y. Shi, Adaptive fixed-time stabilization for a class of uncertain nonlinear systems, IEEE T. Automatic Contr., 68 (2023), 6929–6936. http://dx.doi.org/10.1109/TAC.2023.3244151 doi: 10.1109/TAC.2023.3244151
![]() |
[23] |
L. Xing, C. Wen, Dynamic event-triggered adaptive control for a class of uncertain nonlinear systems, Automatica, 158 (2023), 111286. http://dx.doi.org/10.1016/j.automatica.2023.111286 doi: 10.1016/j.automatica.2023.111286
![]() |
[24] |
M. Shen, C. Wang, Q. G. Wang, Y. Sun, G. Zong, Synchronization of fractional reaction-diffusion complex networks with unknown couplings, IEEE T. Netw. Sci. Eng., 11 (2024), 4503–4512. http://dx.doi.org/10.1109/TNSE.2024.3432997 doi: 10.1109/TNSE.2024.3432997
![]() |
[25] |
H. Liu, J. Cheng, J. Cao, I. Katib, Preassigned-time synchronization for complex-valued memristive neural networks with reaction–diffusion terms and Markov parameters, Neural Networks, 169 (2024), 520–531. http://dx.doi.org/10.1016/j.neunet.2023.11.011 doi: 10.1016/j.neunet.2023.11.011
![]() |
[26] |
W. Fang, T. Xie, B. Li, Robustness analysis of fuzzy cellular neural network with deviating argument and stochastic disturbances, IEEE Access, 11 (2023), 3717–3728. http://dx.doi.org/10.1109/ACCESS.2023.3233946 doi: 10.1109/ACCESS.2023.3233946
![]() |
[27] |
S. Majhi, B. Rakshit, A. Sharma, J. Kurths, D. Ghosh, Dynamical robustness of network of oscillators, Phys. Rep., 1082 (2024), 1–46. http://dx.doi.org/10.1016/j.physrep.2024.06.003 doi: 10.1016/j.physrep.2024.06.003
![]() |
[28] |
O. Artime, M. Grassia, M. Domenico, J. P. Gleeson, H. A. Makse, G. Mangioni, Robustness and resilience of complex networks, Nat. Rev. Phys., 6 (2024), 114–131. http://dx.doi.org/10.1038/s42254-023-00676-y doi: 10.1038/s42254-023-00676-y
![]() |
[29] |
W. Si, S. Gao, F. Yan, N. Zhao, H. Zhang, H. Dong, Linkage-constraint criteria for robust exponential stability of nonlinear BAM system with derivative contraction coefficients and piecewise constant arguments, Inform. Sciences, 612 (2022), 926–941. http://dx.doi.org/10.1016/j.ins.2022.08.078 doi: 10.1016/j.ins.2022.08.078
![]() |
[30] |
Y. Zhang, D. W. Gu, S. Xu, Global exponential adaptive synchronization of complex dynamical networks with neutral-type neural network nodes and stochastic disturbances, IEEE T. Circuits I, 60 (2013), 2709–2718. http://dx.doi.org/10.1109/TCSI.2013.2249151 doi: 10.1109/TCSI.2013.2249151
![]() |
[31] |
O. Faydasicok, S. Arik, An analysis of stability of multiple delayed Cohen-Grossberg neural networks of neutral type, 2022 7th International Conference on Mathematics and Computers in Sciences and Industry (MCSI), 2022, 55–60. http://dx.doi.org/10.1109/MCSI55933.2022.00016 doi: 10.1109/MCSI55933.2022.00016
![]() |
[32] |
L. Wan, Q. Zhou, Stability analysis of neutral-type Cohen-Grossberg neural networks with multiple time-varying delays, IEEE Access, 8 (2020), 27618–27623. http://dx.doi.org/10.1109/ACCESS.2020.2971839 doi: 10.1109/ACCESS.2020.2971839
![]() |
[33] |
L. Wan, Q. Zhou, Exponential stability of neutral-type Cohen-Grossberg neural networks with multiple time-varying delays, IEEE Access, 9 (2021), 48914–48922. http://dx.doi.org/10.1109/ACCESS.2021.3068191 doi: 10.1109/ACCESS.2021.3068191
![]() |
[34] |
W. Yao, C. Wang, Y. Sun, C. Zhou, H. Lin, Exponential multistability of memristive Cohen-Grossberg neural networks with stochastic parameter perturbations, Appl. Math. Comput., 386 (2020), 125483. http://dx.doi.org/10.1016/j.amc.2020.125483 doi: 10.1016/j.amc.2020.125483
![]() |
[35] |
Q. Zhu, J. Cao, Robust exponential stability of Markovian jump impulsive stochastic Cohen-Grossberg neural networks with mixed time delays, IEEE T. Neural Networks, 21 (2010), 1314–1325. http://dx.doi.org/10.1109/TNN.2010.2054108 doi: 10.1109/TNN.2010.2054108
![]() |
[36] |
Y. Shen, J. Wang, Robustness of global exponential stability of nonlinear systems with random disturbances and time delays, IEEE T. Syst. Man Cy.-S., 46 (2015), 1157–1166. http://dx.doi.org/10.1109/TSMC.2015.2497208 doi: 10.1109/TSMC.2015.2497208
![]() |
[37] |
N. Ozcan, Stability analysis of Cohen–Grossberg neural networks of neutral-type: Multiple delays case, Neural Networks, 113 (2019), 20–27. http://dx.doi.org/10.1016/j.neunet.2019.01.017 doi: 10.1016/j.neunet.2019.01.017
![]() |
[38] | X. Mao, Stochastic differential equations and applications, 2 Eds., UK: Woodhead Publishing, 2008. |
1. | Fumiaki Uchiumi, 2023, 9780128187876, 33, 10.1016/B978-0-12-818787-6.00012-6 | |
2. | Fumiaki Uchiumi, 2022, Chapter 1, 978-1-80355-621-5, 10.5772/intechopen.103152 | |
3. | Fumiaki Uchiumi, 2023, 9780128187876, 77, 10.1016/B978-0-12-818787-6.00009-6 |