Abstract

1. Introduction

Cancer cells share common distinctive features that dictate their aberrant behavior in the body. Hanahan and Weinberg described six capabilities that tumor cells acquire which enable their growth and proliferation, and named them “the hallmarks of cancer”. These hallmark features consist in the ability of malignant cells to sustain the self-supply of growth signals, escape anti-growth signals, evade programmed cell death, maintain the formation of new blood vessels, possess an infinite replicative potential, and, finally, to invade and metastasize into distant tissues [1]. More than a decade later, the same authors added additional features as emerging hallmarks of cancer. The reprogramming of cellular metabolism is among these “next-generation” hallmarks that ultimately support malignant cell survival [2].

Among the major reprogrammed metabolic processes in cancer cells, there is what is named the Warburg’s effect (or aerobic glycolysis) after the name of its discoverer [3]. Typically, in non-malignant cells, in the presence of oxygen, glucose undergoes glycolysis to produce pyruvate which is then converted to acetyl Co-A and enters the citric acid cycle in the mitochondria to produce CO₂ and energy in the form of adenosine triphosphate (ATP) via oxidative phosphorylation. However, in cancerous cells, glucose metabolism is rewired and, even in the presence of oxygen, glycolysis predominates while oxidative phosphorylation is reduced, leading to the fermentation of pyruvate to lactate [3]. This so-called Warburg effect is closely related to the need that cancer cells have to sustain a higher production of nicotinamide adenine dinucleotide (NAD), a central molecule required not only at the physiological level but also by neoplastic cells to support numerous crucial cellular processes [4].

NAD is a pyridine nucleotide essential coenzyme that plays a major role in the oxidation–reduction reactions taking place inside the cell, and is also responsible for energy production. Through behaving as an electron acceptor/donor that shuttles between oxidized and reduced forms, NAD supports various metabolic pathways such as glycolysis, Krebs cycle (citric acid cycle), oxidative phosphorylation, and fatty acid oxidation [5,6]. Beyond its role in bioenergetics and cellular metabolism, NAD possesses a prominent cell regulatory function. NAD serves as a substrate for NAD-dependent enzymes such as mono (ADP-ribosyl) transferases (MARTs), poly (ADP-ribose) polymerases (PARPs), sirtuins (SIRT1-7), and cyclic ADP-ribose (cADPR) synthases (CD38 and CD157) [7,8,9]. Indeed, NAD is degraded during the poly-ADP ribosylation reactions carried out by PARPs (e.g., PARP1/2), during protein deacetylation reactions through sirtuins (e.g., SIRT1), and by the hydrolase and ADP-ribosyl cyclase activities of CD38 and CD157 [8,9]. These NAD-consuming reactions orchestrate a series of fundamental biological processes including gene expression, transcription, calcium signaling, DNA repair, apoptosis, circadian rhythm, and cell cycle progression [8,10]. Importantly, many of these cellular events were found to be implicated in malignant transformation and cancer cell survival. For instance, cancer cells frequently face DNA-damaging insults and thus, typically rely on elevated PARP activity to carry out their DNA repair, overcome stress signals, and thereupon sustain their survival. As opposed to its fate as a coenzyme, NAD becomes rapidly degraded by the enzymatic reactions that use it as a substrate, potentially leading to a net NAD deficit [8]. As a consequence, a nonstop generation of NAD is required in normal tissues, and more in neoplastic cells, to compensate for the high NAD degradation. In light of the above findings, dysregulating NAD homeostasis through interfering with the NAD biosynthetic machinery and consequently reducing NAD pools within cancer cells has been conceived as a promising strategy for cancer treatment [8].

In this review, we provide an overview of NAD production routes in eukaryotic cells, the milestone findings, and studies in the development of NAD-depleting agents in the context of oncology therapeutics. In addition, the obstacles in the field and the most recent advances, including the ongoing clinical studies of NAD biosynthesis inhibitors, will also be highlighted.

2. NAD Biosynthesis in Mammals

Overall, mammalian cells utilize the following three biosynthetic pathways to generate NAD: the de novo pathway, the Preiss–Handler (PH) pathway, and the salvage pathway (Figure 1) [8,9]. In the de novo pathway, also referred to as the kynurenine pathway, the essential amino acid tryptophan serves as the starting molecule. It undergoes a cascade of enzymatic reactions that yield quinolinic acid (QA), which is then converted by the enzyme quinolinic acid phosphoribosyltransferase (QAPRT/QPRT) to nicotinic acid mononucleotide (NAMN) [11]. NAMN is also generated in the PH pathway by transferring a phosphoribosyl moiety from phosphoribosyl pyrophosphate (PRPP) to nicotinic acid (NA) with the help of another phosphoribosyltransferase enzyme called nicotinic acid phosphoribosyltransferase (NAPRT). NA is thus considered the precursor unit in the PH pathway of NAD synthesis [12,13]. Noticeably, PRPP, the provider of the NAD-sugar moiety, is synthesized from ribose-5-phosphate, which is a central product of the pentose phosphate pathway (PPP). In such a way, PRPP can be regarded as a linking molecule between glucose metabolism and NAD biosynthesis. The third biosynthetic pathway that contributes to maintaining the NAD supply is the salvage pathway, where NAD is generated starting from the end product of NAD-consuming enzymes, nicotinamide (NAM). Another phosphoribosyltransferase enzyme named nicotinamide phosphoribosyltransferase (NAMPT) controls the rate-limiting step in this pathway. NAMPT catalyzes the transfer of a phosphoribosyl group from the co-substrate PRPP to NAM, yielding nicotinamide mononucleotide (NMN) and pyrophosphate as a byproduct. Alternatively, the two mononucleotides NMN and NAMN can also be produced from the phosphorylation of the nucleosides nicotinamide riboside (NR) and nicotinic acid riboside (NAR) by nicotinamide riboside kinases (NMRK1/2) [14,15]. NAMN and NMN are converted to their corresponding dinucleotides nicotinic acid adenine dinucleotide (NAAD) and NAD through reactions governed by nicotinamide/nicotinic acid mononucleotide adenylyltransferases (NMNATs), of which three mammalian isoforms exist (NMNAT 1–3) [16,17]. In the final step of the PH pathway, NAD synthetase (NADSYN) catalyzes the amidation of NAAD into NAD using glutamine as a nitrogen donor.

Open in a separate window

Figure 1

Schematic representation of the NAD biosynthetic pathways. NAD, nicotinamide adenine dinucleotide; NAMPT, nicotinamide phosphoribosyltransferase; NAPRT, nicotinic acid phosphoribosyltransferase; PncA, bacterial nicotinamidase; NMNAT, nicotinamide mononucleotide adenylyltransferase; NMRK, nicotinamide riboside kinase; NADSYN, NAD synthetase; QAPRT, quinolinic acid phosphoribosyltransferase; IDO, indoleamine-2,3-dioxygenase; TDO, tryptophan-2,3-dioxygenase; AFMID, arylformamidase; KMO, kynurenine 3-monooxygenase; KYNU, kynureninase; HAAO, 3-hydroxyanthranilate 3,4-dioxygenase; ACMSD, α-amino-β-carboxymuconate-ε-semialdehyde decarboxylase; MARTs, mono(ADP-ribosyl) transferases; PARPs, poly(ADP-ribose) polymerases; SARM 1, sterile alpha and TIR motif-containing 1; NQO1, NAD(P)H:quinone oxidoreductase; NQO2, NRH:quinone oxidoreductase; AK, adenosine kinase; NNMT, nicotinamide N-methyltransferase; and TCA, tricarboxylic acid.

NA, NAM and NR (collectively referred to as vitamin B3), as well as tryptophan, can all be obtained through diet, e.g., through cow milk in the case of NR [11,18]. As anticipated above, NAM is also largely produced “endogenously” in virtually every tissue as a byproduct of NAD-degrading enzymes, such as PARPs, sirtuins and CD38. NR can also be produced in bodily tissues starting from its reduced form, NRH, via NRH:quinone oxidoreductase 2 (NQO2), which utilizes NRH as an electron donor [19]. NRH was shown to be endogenously present in the liver by metabolomic analysis [20] and it was proposed to come from NADH degradation via NUDIX hydrolases, such as NUDIX5, 12, and 13 [21,22]. Notably, exogenously supplemented NRH can also stimulate NAD production independent of NMRK enzymes, through a mechanism that foresees its phosphorylation to NMNH, the reduced form of NMN, by adenosine kinase followed by the subsequent conversion to NADH and, ultimately, to NAD [20,23,24]. NRH and NMNH have indeed been identified as important NAD precursors both in vitro and in vivo [22,23,24].

Recent work by Shats et al. showed that gut bacteria convert NAM into NA through their nicotinamidase (PncA) and thereby contribute to NAD production in several mouse tissues, as well as in cancer cells growing the mouse, making them resistant to NAMPT inhibition (an effect that was shown to be dependent on NAPRT expression in cancer cells) [25]. These findings indicate that at least some amounts of gut microbiota- and diet-derived NA do reach bodily tissues despite the extensive metabolism NA undergoes in the liver (here, NA undergoes first-pass metabolism, being conjugated with glycine to form nicotinuric acid, which is then excreted in the urine, and being converted to nicotinamide—assumingly, via NA-mediated NAD generation and subsequent degradation with consequent NAM release). Indeed, circulating NA levels in the nanomolar range can be physiologically detected in mammals and such levels are increased by NA infusion [26] as well as by oral, pharmacological NA doses (gram doses of NA have been used for decades as a lipid-lowering approach). Thus, these observations suggest a role for dietary NA intake as well as for NA produced by the intestinal flora in fueling the PH pathway in NAPRT-expressing tissues (including NAPRT-positive tumors).

The role of NADH in the maintenance of the NAD pool should also not be overlooked. In this context, the enzyme NAD(P)H:quinone oxidoreductase (NQO1) oxidizes NADH into NAD, helping to maintain NAD levels [27]. Indeed, pharmacologically activating NQO1 was shown to augment the NAD/NADH ratio in the kidney tissues, meanwhile, NQO1−/− mice demonstrated higher NADH:NAD and NADPH:NADP ratios as compared to the wild-type mice [28,29].

Since many body tissues don’t express the complete set of the kynurenine pathway enzymes, the contribution of tryptophan to the overall NAD pool has long been thought to be less important. Nevertheless, a recent analysis of NAD synthesis and breakdown fluxes revealed that the liver mainly produces NAD from tryptophan, that it consumes NAD, producing NAM, and that the latter gets secreted in large amounts into the circulation to be used by other tissues [26]. This raises the interesting possibility that tryptophan might strongly contribute to the NAD pool of bodily tissues, not so much directly via de novo synthesis but rather indirectly through multistep inter-tissue cooperation. The latter involves (i) liver-dependent de novo synthesis, followed by (ii) the conversion of NAD into NAM (always in the liver), (iii) NAM excretion into the circulation, and (iv) finally, NAM uptake and utilization by the tissues to produce NAD via the canonical salvage pathway.

Recently published work proposes that the enzyme α-amino-β-carboxymuconate-ε-semialdehyde decarboxylase (ACMSD) acts as a metabolic gatekeeper of the de novo NAD biosynthetic route [30]. ACMSD is expressed mainly by the liver and kidney and it catalyzes the decarboxylation of α-amino-β-carboxymuconate-ε-semialdehyde (ACMS) to α-amino-β-muconate-ε-semialdehyde (AMS), thus rerouting tryptophan metabolism to picolinic acid and acetyl Co-A synthesis rather than towards NAD production [5,30]. Indeed, enhancement of de novo NAD synthesis and SIRT1 activity was observed in response to ACMSD inhibition [30]. In line with these findings, another study reported that mice overexpressing ACMSD show reduced blood and tissue NAD levels, as compared to their counterparts with normal ACMSD expression, when both groups were fed with niacin-free diets [31].

Finally, nicotinamide N-methyltransferase (NNMT) is another important modulating enzyme of NAD homeostasis. NNMT catalyzes the methyl transfer from S-adenosyl methionine (SAM) to NAM, yielding 1-methylnicotinamide, which is subsequently metabolized and excreted in urine [10]. NNMT derails NAM away from being recycled into NAD and curbs the NAM accumulation and inhibition of NAD-dependent signaling pathways by excessive NAM [32]. NNMT was found to regulate histone methylation and NAD-dependent SIRT1 signaling in adipose tissue and NNMT knockout was associated with elevated NAMPT, NMNAT2 expression, and NAD levels in adipocytes, but not in the liver [33]. Notably, numerous cancer types and stromal cells upregulate NNMT expression, which epigenetically reprograms gene expression and causes a state of histone hypomethylation by consuming SAM methyl groups which, in turn, depletes SAM and creates a methyl sink [34,35]. Similarly, NNMT was shown to promote DNA hypomethylation in mesenchymal glioblastoma stems cells (GSCs) and to drive alterations in DNA methylation within the promoter regions of several genes in cancer-associated fibroblasts [35,36].

3. Regulation of NAD Production in Cancer Cells

The molecular basis and genetic mechanisms that underly the selection of the NAD biosynthetic pathway in cancer cells have not been fully uncovered. However, a recent and extensive analysis for thousands of tumors and their corresponding normal tissues of origin has revealed that the choice of the NAD-producing pathway in cancer cells is based upon the healthy tissues from which they originate [37]. Accordingly, tumors were classified into the following two broad categories: PH-dependent tumors and salvage-dependent tumors where NAPRT gene amplification and NAMPT enhancer remodeling, respectively, are the key hallmarks. Tissues with high basal NAPRT expression become dependent on NAPRT for survival upon malignant transformation. On the other hand, salvage-dependent tumors originate from tissues that lack NAPRT expression, and consequently, their NAD supply is primarily hinged upon NAMPT. This conclusion was further substantiated by the observation that neither the overexpression of NAPRT in salvage-dependent cancer cells nor the overexpression of NAMPT in PH-amplified cancer cells has reversed their predestined NAD metabolic pathway [37]. Notably, when NAMPT was depleted in salvage-dependent tumors, they were still able to maintain the NAD supply through the alternative salvage NR-NMRK1 pathway, and dual NMRK1 and NAMPT inhibition resulted in more effective NAD reduction and significant tumor suppression in vivo [37]. This finding provides insights into the role of NMRK1 in mediating resistance to NAMPT inhibition. In this section, we describe how the rate-limiting enzyme in each NAD synthetic pathway is regulated in cancer cells (Table 1).

4. Chemical Inhibitors of NAD Biosynthesis

As mentioned earlier, NAMPT is the bottleneck enzyme in the NAD salvage pathway. Elevated NAMPT expression levels were extensively reported in numerous solid and hematological malignancies, including pancreatic cancer [50], gastric cancer [71], prostate cancer [72], breast cancer [73,74], colorectal cancer [75,76,77], gliomas [78], ovarian cancer [79], melanoma [80], thyroid carcinoma [81], sarcomas [82], and lymphomas [83]. Indeed, NAMPT is implicated in driving pro-oncogenic and more aggressive phenotypes, and its overexpression has been associated with poor prognosis in different types of cancer [74,75,76,78,84,85]. In addition to existing as an intracellular protein (intracellular NAMPT, iNAMPT), NAMPT also gets secreted extracellularly and this form of the protein is known as extracellular NAMPT (eNAMPT), but also as visfatin or as pre-B-cell colony-enhancing factor (PBEF). In fact, eNAMPT was originally identified as PBEF, a modulatory cytokine during B-cell development [86]. The eNAMPT also exerts pro-oncogenic effects by modulating the tumor microenvironment, enhancing tumor metabolism, and promoting EMT [87,88]. In melanoma, NAMPT is secreted by the cancer cells and NAMPT silencing was shown to reduce tumor progression and was accompanied by lower levels of circulating eNAMPT [89]. In line with this finding, eNAMPT levels were increased in the mice bearing melanoma cells that acquired resistance to BRAF inhibitors as well as in patients with BRAF-mutated metastatic melanoma, and high eNAMPT levels showed a negative correlation with patients’ overall survival [90]. Patients with invasive prostate cancer also showed higher plasma eNAMPT levels and neutralizing circulating eNAMPT was proven to markedly attenuate cancer invasiveness in prostate cancer mice models [91].

Moreover, a proinflammatory function of eNAMPT has already been established [92]. Consistently, serum levels of eNAMPT were found to be elevated in inflammatory bowel disease (IBD) patients who showed resistance to anti-TNFα therapy, and eNAMPT neutralization, with an anti-eNAMPT monoclonal antibody, ameliorated acute and chronic colitis in experimental mice models [93]. Notably, NAPRT was also found to exist as an extracellular protein (eNAPRT), which mediates inflammation by binding to toll-like receptor 4 (TLR4) and by activating the NF-κB pathway [94]. Consistent with these findings, NAPRT serum levels were strikingly elevated in septic patients [94]. Whether, similar to eNAMPT, eNAPRT also plays a pro-oncogenic role remains to be understood. Nonetheless, primary clues from 312 cancer patients diagnosed with solid or hematological malignancies just indicated mild elevations in their serum eNAPRT levels as compared to healthy donors (median serum eNAPRT was 1.4 ± 0.07ng/mL in healthy donors, 2.4 ± 0.15ng/mL in cancer patients and 27.1 ± 4.9ng/mL in septic individuals) [94]. For a broader overview of the roles of eNAMPT and eNAPRT, we refer the reader to other recent articles [87,95,96].

Given its pleiotropic role in cancer pathogenesis, NAMPT has long been considered an attractive therapeutic target for cancer treatment. In this section, we provide a brief overview of some of the key NAMPT inhibitors that have been discovered so far (Figure 2), as well as of other inhibitors of NAD biosynthesis reported over the last years. The advances in medicinal chemistry and the pharmacology of small molecule NAMPT inhibitors are beyond the scope of this review. They have recently been reviewed by Sampath et al. [97] and by Galli et al. [98].

Open in a separate window

Figure 2

Timeline summary for the development of NAMPT inhibitors and their entry for evaluation in clinical studies. * the trial years are the years in which the study started according to clinicaltrials.gov.

5. Effects of NAD Production Inhibition in Cancer

Deregulating NAD homeostasis in the cancer cells through NAD synthesis inhibitors was shown to result in antiproliferative and cytotoxic effects via different mechanisms. In this section, we report the downstream effects of NAD deprivation in cancer cells and combination strategies that were shown to potentiate the NAMPT inhibitors, eliciting a more pronounced antitumor response (Table 2).

6. In Vivo Studies of NAD Production Inhibitors in Mice

7. Perspectives and Obstacles for Clinical Uses of NAD Biosynthesis Inhibitors

Based on the encouraging preclinical data, the early-generation NAMPT inhibitors, FK866 (APO-866), GMX1778 (CHS-828), and its prodrug GMX1777, were investigated in early-phase clinical trials. Thrombocytopenia was the common dose-limiting toxicity associated with the three NAMPT inhibitors [210,211,212]. Additionally, several gastrointestinal (GI) adverse effects and high intra- and inter-patient pharmacokinetic variability were reported with CHS-828 [106,210,213]. GI hemorrhage and skin rash were other dose-limiting toxicities that accompanied GMX1777 [211,214]. Regrettably, the lack of objective tumor response and unfavorable side effect profile upon treatment with NAMPT inhibitors have halted their further progression in clinical trials so far [106,210,211,212,213,214,215].

The failure of NAMPT inhibitors in clinical trials has fueled research endeavors to overcome the limited clinical outcome of these agents. Indeed, numerous approaches have been sought to reduce their toxicities and re-sensitize cancer cells to NAMPT inhibitors, which can be summarized as the following (Figure 3):

• (i) Developing safer and more effective “next-generation” NAMPT inhibitors;
• (ii) Using NAMPT inhibitors against a subset of cancers that show unique sensitivity to NAMPT inhibitors, such as hematological cancers or IDH-mutant cancers. NA might be administered simultaneously to protect normal tissues (circumventing NAMPT inhibition by the PH pathway to sustain adequate NAD stores). Although NA might alleviate the systemic toxicity and widen the therapeutic index of NAMPT inhibitors, abrogation of antitumor efficacy could represent a caveat to this approach;
• (iii) Combining NAMPT inhibitors with chemotherapy, immunotherapy, or radiation to achieve a synergistic effect;
• (iv) Combining NAMPT inhibitors with NAPRT inhibitors against NAPRT-positive cancer subtypes [62];
• (v) Development of “broad-spectrum” NAMPT inhibitors such as the dual NAMPT-PAK inhibitors, NAMPT-HDAC inhibitors, NAMPT-GLUT1 inhibitors, and recently the NAMPT-EGFR inhibitors;
• (vi) Development of NAMPTi-ADCs that selectively target NAMPT in cancer cells through antibody binding to cancer-specific cell surface markers, thereby sparing the normal cells from systemic NAD depletion.

Open in a separate window

Figure 3

Schematic overview of the different strategies to improve the overall performance of NAMPT inhibitors.

Notably, two compounds of the second wave of NAMPT inhibitors (OT-82 and KPT-9274) are being currently investigated in phase I trials. In a two-stage (dose escalation and dose expansion) phase 1 study, the safety and efficacy of OT-82 are being evaluated in participants with relapsed or refractory lymphoma (NCT03921879). In the PANAMA (NCT02702492) and TEACH (NCT04281420) phase 1 trials, the safety, efficacy, and tolerability of KPT-9274 will be evaluated in patients with advanced solid tumors and malignancies or non-Hodgkin’s lymphoma. In both studies, the experimental arms include KPT-9274 alone or in combination with niacin ER. In the PANAMA trial, additionally, a third arm will also assess the combination of KPT-9274 and the immune checkpoint inhibitor nivolumab (anti-PD1). So far, no results have been disclosed.

8. Future Outlook

In summary, NAD-producing enzymes represent a collection of targets that could be exploited to combat cancer proliferation. Even though NAMPT is the most extensively NAD-producing enzyme targeted by chemical inhibitors, limited objective responses and unacceptable toxicities in the clinical trials have hampered faster progressions in the field. Given its pleiotropic roles in energy metabolism, promoting DNA repair in response to genomic insults, epithelial to mesenchymal transition, oncogenic signaling, and suppressing the immune surveillance within the tumor milieu, NAMPT remains an attractive target in cancer therapy [216]. As mentioned earlier, two NAMPT inhibitors are being currently evaluated in clinical trials. Nevertheless, in light of accumulating evidence that substantiated the significance of NAPRT expression as a parallel regulator of NAD metabolism in cancer and as a biomarker for NAMPT inhibitor therapy [37,62], the development of NAPRT inhibitors also represents a promising therapeutic approach. To our best knowledge, no potent NAPRT inhibitor has been identified so far. As a consequence, an important future challenge will be to develop NAPRT inhibitors with a higher potency than that of the currently available compounds (e.g., 2-hydroxy nicotinic acid) and with optimized drug-like properties. Additionally, given the recently emerging function of the gut microbiota as a key contributor to NAD metabolism, manipulation of the intestinal microbiome, presumably through antibiotics, might be exploited to modulate systemic and tumor NAD biosynthesis [25]. Furthermore, recent studies show that limiting dietary intake of niacin and tryptophan creates NAD deficiency in mouse models that recapitulate human physiology [31]. Consistent with this finding, curbing the content of NAD precursors in diets, together with NAD-depleting agents, might allow to more effectively lower NAD levels in cancer cells and thus, provide enhanced antitumor activity. In addition, processes such as polyunsaturated fatty acid (PUFA) desaturation and lactate fermentation prompt regeneration of NAD from NADH, and permit glycolytic NAD recycling and cell survival when the NAD/NADH ratio decreases in the cytosol [217]. These findings raise the possibility that such processes might be targeted to improve NAMPT inhibitors’ efficacy in malignant cells. Lastly, nicotinamide mononucleotide adenylyltransferase (NMNAT 1–3), nicotinamide riboside kinase (NMRK1/2), and NAD synthetase (NADSYN) stand out as additional understudied enzymes that warrant further exploration as targets in oncology.

Acknowledgments

Author Contributions

Conceptualization, A.N. and M.S.G.; methodology, M.S.G.; software, not applicable; validation, not applicable; formal analysis, not applicable; investigation, not applicable; resources, not applicable; data curation, not applicable; writing—original draft preparation, M.S.G.; writing—review and editing, A.N. and F.M.; visualization, M.S.G.; supervision, A.N.; project administration, not applicable; funding acquisition, A.N. and M.S.G. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported in part by the Associazione Italiana per la Ricerca sul Cancro (AIRC; IG#22098) and by the Italian Ministry of Health grants PE-2016-02362694 and PE-2016-02363073. In addition, this project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement no. 813284.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

References