Nucleoside-based anticancer drugs: Mechanism of action and drug resistance

Nucleoside-based drugs, recognized as purine or pyrimidine analogs, have been potent therapeutic agents since their introduction in 1950, deployed widely in the treatment of diverse diseases such as cancers, myelodysplastic syndromes, multiple sclerosis, and viral infections. These antimetabolites establish complex interactions with cellular molecular constituents, primarily via activation of phosphorylation cascades leading to consequential interactions with nucleic acids. However, the therapeutic efficacy of these agents is frequently compromised by the development of drug resistance, a continually emerging challenge in their clinical application. This comprehensive review explores the mechanisms of resistance to nucleoside-based drugs, encompassing a wide spectrum of phenomena from alterations in membrane transporters and activating kinases to changes in drug elimination strategies and DNA damage repair mechanisms. The critical analysis in this review underlines complex interactions of drug and cell and also guides towards novel therapeutic strategies to counteract resistance. The development of targeted therapies, novel nucleoside analogs, and synergistic drug combinations are promising approaches to restore tumor sensitivity and improve patient outcomes.


Introduction
Cancer is a complex disease that affects millions of people around the world.Despite the availability of numerous treatments, curing and eradicating cancer proves challenging, largely due to the development of resistance by cancer cells to treatments.Mechanisms of chemoresistance (MOC) include several cellular alterations that significantly enhance cell survival.The first mechanism represents reduced drug uptake, particularly in the case of nucleoside-based drugs, facilitated by Concentrative Nucleoside Transporters (CNTs/SLC28s) and/ or Equilibrative Nucleoside Transporters (ENTs/SLC29s).MOC-1b refers to an increased drug efflux, which can be often mediated by MDR (multidrug resistance) proteins such as P-glycoprotein (Pgp), Multidrug Resistance-related Proteins (MRPs) or Breast Cancer Resistance Protein (BCRP), MOC-2 is characterized by a decrease in the concentration of active metabolites due to the reduction of pro-drug activation or higher active metabolite elimination.Other mechanisms include cellular changes in drug targets (MOC-3), enhanced DNA-repair mechanisms (MOC-4), decreased activity of pro-apoptotic factors (MOC-5a), and/ or increase of anti-apoptotic factors (MOC-5b), alterations in the tumor cells microenvironment (MOC-6), and activation of epithelial-mesenchymal transition (EMT) (MOC-7) [1,2].Historically, nucleosides, compounds comprising a ribose or 2-deoxyribose sugar ring and purine (adenine and guanine) or pyrimidine (cytosine, thymine, and uracil) bases, were among the first widely used anticancer and antiviral therapeutics [3,4].As anticancer drugs, nucleosides belong to a class of antimetabolites, a category that encompasses purine or pyrimidine analogs (nucleosides), folic acid and amino acid analogs, and ribonucleotide reductase inhibitors.[5] In 1948, the first antimetabolite, aminopterin (4aminopteroyl-glutamic acid), was used to treat acute leukemia in children, where it was shown to induce temporary remission.[6] Nucleosides compete with physiological nucleosides and interact with a plethora of intracellular targets involved in nucleoside metabolism and signaling.The nucleoside 6-mercaptopurine (6-MP), which was synthesized in 1950 for the treatment of acute leukemia, was one of the first nucleosides to be used clinically.A few years later, structurally similar 6-thioguanine (6-TG) was synthesized, which was gradually followed by other compounds such as gemcitabine.[7,8] While the first nucleosides targeted hematological malignancies, later-developed compounds like 5-fluorouracil and gemcitabine demonstrated high efficacy in solid tumors (Fig. 1) [9].
Currently, the primary obstacles in chemotherapy include tumor cell selectivity, systemic side effects, and the development of multidrug resistance due to recurrent treatment cycles, which reduce overall treatment efficacy.[10] This review summarizes the mechanisms underlying resistance of the most prevalently used nucleoside-based drugs, with a focus on drug transporters and clinically relevant resistance mechanisms, and potential strategies to overcome them.

6-Mercaptopurine (6-MP)
6-mercaptopurine is a purine analog widely used for the treatment of acute leukemia.Its immunosuppressive properties also make it a suitable treatment option to treat autoimmune diseases and inflammatory conditions such as rheumatoid arthritis, Crohn's disease, and ulcerative colitis.[11] In clinics, 6-MP is typically administered as a pro-drug  (azathioprine, AZA, Fig. 1), and both compounds are relatively slow acting, limiting their usage in monotherapy.[12] For pediatric patients with acute leukemia, 6-MP is recommended to be combined with methotrexate, resulting in a higher remission rate.[13] Recent research indicates that the addition of a low dose of 6-thioguanine (6-TG) to a 6-MP/methotrexate regimen improves disease-free survival in pediatric and adult patients newly diagnosed with acute lymphoblastic leukemia.[14] Furthermore, this combination therapy can be complemented by prednisone, which induces remission in patients with steroid-dependent inflammatory bowel disease (IBD) and Crohn's disease.[15].
The oral absorption of 6-MP is variable and incomplete, with approximately 50% of the dose entering the systemic circulation and a half-life ranging from 60 to 120 min.[16] In the case of intravenous administration, plasma half-disappearance varies from 21 min in pediatric patients to 47 min in adults.The primary catabolic pathways of 6-MP are methylation and oxidation.Other drugs, such as allopurinol, can inhibit the oxidative degradation of 6-MP by inhibition of xanthine oxidase (XO), which mediates 6-MP oxidation.However, this inhibition leads to an accumulation of 6-MP, resulting in increased toxicity without enhancing therapeutic effects.[17] Despite this, the combined therapy of allopurinol (or alternative XO inhibitor infliximab) and a low dose of thiopurines (AZA and 6-MP) has been suggested as a potential immunosuppressive strategy for treating adult patients with inflammatory bowel disease.[18] The combination of AZA/6-MP and allopurinol was initially considered controversial due to its side effects.Nonetheless, recent studies show that monitoring thiopurine metabolite levels can help adjust the dosage and reduce the adverse effects of the treatment.[19,20].
Resistance to 6-MP can develop through various mechanisms, including alterations in the expression of membrane transporters.Notably, 6-MP acts as a substrate for efflux pumps such as MDR1/Pglycoprotein (Pgp/ABCB1) and multidrug resistance-related proteins 4 and 5 (MRP4/ABCC4; MRP5/ABCC5).Overexpression of these efflux pumps can reduce 6-MP activity and lead to the development of a resistant phenotype.[23][24][25][26] Typically, a decrease in intracellular 6-MP concentration is associated with alterations in the expression of influx/ uptake transporters, such as equilibrative nucleoside transporters (ENTs/SLC29s) and concentrative nucleoside transporters (CNTs/ SLC28s).Approximately half of 6-MP transport is carried out via a sodium-dependent mechanism.[27] Moreover, 6-MP has been identified as a substrate for ENT1/SLC29A1, ENT2/SLC29A2, and CNT3/SLC28A3.[27] Historically, sensitivity to 6-MP has often been associated with a decrease or loss in the activity of enzymes such as inosinic or guanylic acid pyrophosphorylase.These enzymes facilitate the formation of mercaptopurine ribonucleotides in human cancer cells.[28].
An activating mutation in NT5C2 (cytosolic 5 -nucleotidase II) has been recently discovered in approximately 20% of patients diagnosed with acute lymphoblastic leukemia (ALL).NT5C2, as a nucleotidase, dephosphorylates and inactivates HGPRT, effectively inhibiting conversion of 6-MP into its active form and reducing its therapeutic effectiveness.[29] Consequently, leukemia cells with NT5C2 mutations develop resistance to 6-MP.Overcoming this resistance can be achieved either through direct targeting of NT5C2 or by the inhibition of compensatory pathways, which are notably active in cells with NT5C2 mutations.One such pathway involves the enhanced purine biosynthesis due to the excess export of purine nucleosides in cells with NT5C2 mutations.This leads to the activation of allosteric feedback loops, which in turn increases the sensitivity of these cells to the inhibition of IMPDH, a crucial enzyme in purine biosynthesis.By inhibiting IMPDH its possible to partially compensate for the purine nucleotide depletion caused by the NT5C2 mutation, thereby reducing the the resistance of these cells to 6-MP.This approach offers a promising alternative strategy for the treatment of ALL in patients with NT5C2 mutations [30].
Several studies have also identified a relationship between the inhibition of the mTOR pathway, particularly mTOR complex 1, and resistance to 6-MP in cases of B-cell acute lymphoblastic leukemia (B-ALL).[31] Another mechanism contributing to resistance involves the preferential metabolism of AZA into 6-MMP, particularly in patients with inflammatory bowel disease who exhibit a lower response to 6-MP.This effect appears to be dose-dependent; an increase in 6-MP results in a significant elevation in 6-MMP levels and minor changes in 6-thioguanine nucleotide concentrations [32].
Recent research has explored the genes TPMT, NUDT15, ITPA, and APEX1 as potential predictive markers of 6-MP treatment efficacy in pediatric patients diagnosed with ALL.However, these markers have yet to find application in clinical settings.[33,34] Notably, variants of the NUD15 have been associated with the onset of thiopurine-induced leukopenia in Asian patients suffering from Crohn's disease.In particular, the homozygous T/T variant has been linked to the development of early leukopenia in nearly all observed cases.Consequently, the administration of 6-MP and 6-TG treatments is strongly discouraged in these cases [35,36].

6-Thioguanine (6-TG)
6-Thioguanine, a metabolite structurally akin to 6-MP, can often induce cross-resistance due to its similarities.Despite this resemblance, both 6-TG and 6-MP have distinct metabolic pathways and modes of action.In instances where patients with inflammatory bowel disease (IBD) experience adverse effects from 6-MP therapy, 6-TG is often used as an alternative therapeutic agent.[37] In haemato-oncology, 6-TG, when combined with cytarabine, serves to induce remission in patients suffering from acute and chronic myeloid leukemia (ALL, CML).Preclinical studies also suggest its efficacy against prostate and pancreatic cancers.[38][39][40][41] Recent research has explored the potential of a 6-TG and disulfiram/Cu combination in inhibiting the cellular proliferation of triple-negative breast cancer.This combination seemingly disrupts DNA damage checkpoints and enhances DNA damage.[42] Furthermore, this drug combination appears to synergistically inhibit tumor-associated deubiquitinases, ubiquitin-specific protease 2 and 21 (USP2 and USP21).The activity of these enzymes is related to breast and prostate cancers, as well as hepatocellular carcinomas, making them potential targets in cancer therapy [43].
Once orally administered, 6-TG is converted into nucleosidemonophosphate by hypoxanthine-guanine phosphoribosyltransferase (HGRPT) and then phosphorylated into di-(TGDP) and tri-phosphate (TGTP) forms by nucleoside diphosphate kinase (NDPK).[44,45] Subsequently, TGTP is incorporated into the nucleic acid structure, inhibiting DNA and RNA synthesis, and leading to cellular death (Fig. 3).[46] The absorption of 6-TG by patients varies, with an average bioavailability of approximately 30% and a median half-life of 90 min.Unlike 6-MP, 6-TG metabolism does not rely on xanthine oxidase (XO), implying that a combination therapy with XO inhibitors, such as allopurinol, does not impact the activity or toxicity of the drug.6-TG is metabolized into its inactive form, 6-thioinosine, via TPMT.[47] Pancreatic ductal adenocarcinoma cells (PDACs) exhibit low levels of TPMT, which makes 6-TG an ideal candidate for the treatment of this cancer type.The reduced inactivation via TPMT increases the efficacy of 6-TG therapy in such cases [41].
Resistance to 6-TG is often correlated with mutations in HGPRT, as shown in mismatch repair-deficient cell lines.These cell lines exhibit greater resistance compared to those with a fully functional DNA-repair system.[48] Mutations and inactivation of HGPRT impede the metabolism of 6-TG into its active form.Interestingly, such mutations have been detected in circulating human lymphocytes, with the frequency of mutated cells increasing with patient age, indicating a potential relationship between aging and mutagenesis.[49] In vitro experiments with melanoma cells have demonstrated that 6-TG resistant cells express approximately three times more methylguanine-DNA methyltransferase (MGMT), a DNA repair enzyme.Administration of MGMT inhibitors has been shown to enhance the sensitivity of resistant cell lines to 6-TG, suggesting a correlation between MGMT activity and resistance to 6-TG.[50] Resistance to 6-TG can also be linked to variations in the expression levels of membrane transporters.For instance, leukemia cells (MOLT-4) displaying reduced levels of ENT2 and CNT3, transporters critical to drug influx, demonstrate resistance to 6-TG.[51] It is important to note that 6-TG is also a substrate of efflux pumps, such as MRP4 and 5. [52] The relationship between MRP4 levels and 6-TG sensitivity has been observed in Japanese patients with IBD.[53] Since 6-TG is not a substrate of P-glycoprotein (Pgp), it has the potential to be used in tumors with elevated Pgp levels, such as breast and ovarian cancers with BRCA1 mutations.This could potentially overcome resistance to Poly (ADP-ribose) polymerase (PARP) inhibitors [54].

Cytarabine (Ara-C, cytosine arabinose)
Cytarabine, a cytosine analog, is a cornerstone in the treatment of ALL, Acute Non-Lymphocytic Leukemia (ANLL), the blast phase of CML, and Acute Myeloid Leukemia (AML).[55][56][57] The Food and Drug Administration (FDA) gave its approval to cytarabine for cancer treatment in 1969, and it has been administered in combination with other anticancer drugs to augment its efficacy since the 1990s.For example, in AML patients over 60 years, cytarabine is often used in combination with clofarabine.[58] An estimated 20-35% of AML patients harbor a mutation in the DNMT3A gene (DNA methyltransferase 3A), predisposing them to drug resistance (mainly to anthracyclines), a higher incidence of minimal residual disease and an increased risk of clinical relapse, particularly in an advanced-aged cohort.Yet, those carrying the DNMT3A R882 mutation exhibit heightened sensitivity to the replication stress induced by cytarabine.For patients with a wild-type variant of the DNMT3A gene, a combination of cytarabine and PARP inhibitors can enhance treatment efficacy [59].
Furthermore, cytarabine can be co-administered with decitabine, a combination that has shown improved prognosis in elderly patients with high-risk AML.[60] Younger and adult AML patients have also demonstrated favorable responses to the combined regimen of cytarabine, sorafenib, and idarubicine.[61,62]Recently, a combination of venetoclax with low-dose cytarabine has emerged as an effective treatment for adult AML patients who cannot undergo standard intensive chemotherapy (IC).This combined therapy can induce complete remission in 60-80 % of patients aged 16-60 years [63].
While the standard dose of cytarabine is 100-200 mg/m 2 body surface area, a higher dose (3 g/m 2 body surface area) with an elimination half-life of 3-4 h has demonstrated superior efficacy in AML treatment, although the underlying mechanism responsible for this difference remains elusive.[64,65] This high-dose cytarabine therapy is also administered to children and adolescents with ALL, ANLL, CML, and Burkitt's lymphoma.Despite serious side effects, including bone marrow suppression, nausea, vomiting, fever, and diarrhea, the increased cytarabine dose has proven effective in inducing remission in patients with lymphoma resistant to the standard dose [66].
Oral administration of cytarabine is less effective due to its high firstpass metabolism, with less than 20% of the drug being absorbed from the gastrointestinal tract (GIT).The primary metabolism of cytarabine into its inactive form, uracil arabinose (Ara-U), is facilitated by cytidine deaminase (CDA).Overexpression of CDA has been implicated in the development of cytarabine resistance, and it is also recognized as a valuable predictive marker of toxicity in cytarabine treatment.[67,68] Notably, lower CDA activity has been observed in adult patients suffering from severe or lethal toxicity [67].
Cytarabine is typically administered intravenously, intrathecally, or subcutaneously due to its limited absorption through the GIT.Primarily transported into cells via the ENT1 transporter, cytarabine also serves as a substrate for ENT2 CNT1 and CNT3.[69] The efflux of cytarabine is facilitated by MRP4 and 5, and the overexpression of these proteins can contribute to the development of cytarabine resistance [70,71].
Following its transport into cells, mainly via the ENT1 transporter [72], cytarabine is phosphorylated into its active form, cytarabine triphosphate (Ara-CTP).This conversion is facilitated by deoxycytidine kinase (dCK), deoxycytidylate kinase, and nucleoside diphosphate kinase (NDPK) (Fig. 4).The efficacy of cytarabine is directly related to the expression of these proteins.[64] Once activated, Ara-CTP incorporates into the nascent DNA, leading to the induction of single-strand breaks, replication fork stalling, and eventually, cell death [73].
Resistance to cytarabine can primarily emerge due to: a) rapid elimination by CDA, b) restricted intracellular transport associated with reduced ENT1 expression or c) limited activation due to inhibition or reduced expression of activating kinases, notably dCK.[39] Levin et al. (2019) documented the loss of dCK function in cytarabine-resistant acute myeloblastic cells, which exhibited hypersensitivity to a combination of hydroxyurea (HU) and azidothymidine (AZT), an approach that could potentially overcome cytarabine resistance.[74] Recent in vitro studies have highlighted the role of CD157 (ADP-ribosyl cyclase 2) in cytarabine sensitivity in AML.CD157 induces the expression of antiapoptotic proteins, particularly Mcl-1, while its inhibition has been shown to enhance the efficacy of cytarabine treatment in CD157-high AML cells [75].
Cytarabine resistance is also correlated with the expression of cytosolic 5 -nucleotidase II (NT5C2), which inactivates cytarabine by dephosphorylating cytarabine monophosphate into cytarabine.[76] Furthermore, SAMHD1 (SAM and HD domain-containing protein 1), a deoxynucleotide triphosphate triphosphorylase, has been identified as a predictive marker of cytarabine treatment efficacy.SAMHD1 hydrolyzes Ara-CTP, leading to a significant reduction in cytarabine activity.[77] Inhibitors of SAMHD1, such as hydroxyurea, can suppress Ara-CTP hydrolysis and enhance the efficacy of cytarabine [78,79].
An analysis of AML primary cells resistant to cytarabine treatment has revealed a link between high mitochondrial oxidation based on Bcl2 and cytarabine resistance.Consequently, these cells exhibited high sensitivity to the combined therapy of venetoclax and cytarabine.However, adaptive resistance can occur due to changes in oxidative phosphorylation, electron transport chain complex and/or the p53 pathway.Using electron transport chain inhibitors, mitochondrial ClpP protease agonists, or pyruvate dehydrogenase inhibitors could potentially offer ways to overcome this adaptive cytarabine/venetoclax resistance [80].
Recent findings suggest a role for glucose transporters (GLUTs) in determining the sensitivity of AML cells to cytarabine.Specifically, heightened glycolysis appears to diminish the sensitivity of these cells to chemotherapy in AML.Consequently, the inhibiting of glucose uptake may enhance the cells' sensitivity to cytarabine treatment [81].

Fludarabine (F-ara-A)
Fludarabine is a purine analog primarily employed in the treatment of lymphoproliferative malignancies, notably chronic lymphocytic leukemia (CLL).It has also proven to be useful in managing AML.[82,83] Fludarabine can be used alone or in combination with other anticancer drugs, including cytarabine, ibrutinib, bortezomib, cyclophosphamide, rituximab, clofarabine, and idarubicin.[84][85][86] Recent research has revealed fludarabine's potential as an inhibitor of DNA-dependent RNA polymerase in the treating of monkeypox virus [87].
In adults, CML is frequently associated with the overexpression of MDR proteins, such as Pgp, breast cancer resistance protein (BCRP/ ABCG2), multidrug resistance-related proteins (MRPs), and lung resistance-related protein (LRP).Higher expression of these proteins is viewed as a negative prognostic marker.Fludarabine, which is not typically associated with MDR, has shown resistance in cases with heightened expression of Pgp, MRP1, and LRP.[91] As fludarabine is a substrate for BCRP, its treatment efficacy decreases in patients with BCRP-positive CML, which also heightens the risk of relapse [92].
Fludarabine resistance primarily presents in adult patients with CLL.A study of peripheral blood mononuclear cells (PBMCs) from fludarabine-sensitive and resistant CLL patients revealed a significant increase in the expression of Pgp, glucosylceramide synthase (GSC), and CD34, a marker of leukemic stem cells.Higher GSC expression leads to the accumulation of glucosylceramide, which fosters cell proliferation, and survival, inhibits apoptosis, and stimulates resistance development.This fludarabine resistance can be counteracted by inhibiting GSC with 1-phenyl-2-decanoylamino-3-morpholino-1-propanol (PDMP) [93].
In CLL, mitochondrial metabolism plays a key role in fludarabine sensitivity.Beyond the effect of enhanced oxidative phosphorylation and reactive oxygen species (ROS) production, recent studies have drawn a link between mitochondrial abnormalities (such as mtDNA mutations, aberrant mitochondrial biogenesis, respiration, and increased superoxide production) and a negative impact on apoptosis, which results in the development of fludarabine resistance.Moreover, malignancies showing reduced sensitivity to fludarabine seem to rely heavily on ER-Golgi protein transport.Therefore, targeting this process could present a promising therapeutic strategy that may also influence fludarabine efficacy [94][95][96].
CLL cell lines exhibit a more robust endoplasmic reticulum network compared to normal B-lymphocytes.Brefeldin A, an anticancer compound targeting the ER, has demonstrated its ability to induce apoptosis in fludarabine-resistant CLL cells.[94] Additionally, fludarabine resistance in CLL cell lines is linked with increased phosphorylation of ribosomal protein S6 kinase (phospho-p70S6K), activated by the mTORC1 complex implicated in translation.This mTORC1 activation is relatively prevalent across a wide range of cancers.However, mTORC1 inhibitors, including rapamycin and its analogs (rapalogs), often fail to produce the anticipated clinical effect as they only inhibit phosphatidylinositol-3kinase (PI3K) and don't affect Akt.A compensatory surge in Akt activity has been observed, leading to a prosurvival effect.Nevertheless, in fludarabine-resistant and primary CLL cells, rapalogs succeeded in triggering cell death and inhibiting oxidative phosphorylation and glycolysis, both of which are elevated in these fludarabine-resistant cells.Thus, the inhibition of mitochondrial respiration and downregulation of mTORC1 may prove selectively cytotoxic to fludarabineresistant CLL cells, presenting a potential therapeutic target to counteract drug resistance [95].
CLL cells are also found to have higher levels of ROS compared to normal lymphocytes.Trachootham et al. (2008) explored a strategy to eliminate fludarabine-resistant cells using ROS and discovered that fludarabine-sensitive/resistant CLL cell lines were five times more sensitive to the anticancer and chemopreventive agent β-phenethyl isothiocyanate (PEITC) than normal lymphocytes.The exposure of fludarabine-resistant CLL cells to PEITC resulted in ROS accumulation, glutathione depletion, and the oxidation of mitochondrial cardiolipin, causing cellular death [96].

Gemcitabine (GEM, dFdC)
Gemcitabine, a deoxycytidine nucleoside analog, is deployed in the treatment of solid tumors, including breast, non-small cell lung, pancreatic, bladder, and ovarian cancers.[97][98][99] Its more efficient phosphorylation and activation, along with slower elimination compared to other nucleoside drugs, make gemcitabine one of the most effective nucleoside-based anticancer drugs.The half-life of gemcitabine after a brief infusion ranges from 32 to 94 min and is extended to 4 to 10 h with a longer infusion.Innovative strategies are being explored, utilizing gemcitabine in the form of a pro-drug, LY2334737, as its valproic ester [100,101].
Though gemcitabine often prompts an excellent initial response, drug resistance tends to develop over time.This resistance is primarily associated with factors such as increased efflux/reduced influx (primarily via ENT1), alterations in drug metabolism, inhibition of apoptosis, changes in micro-RNA levels, and the presence of cancer stem cells (CSCs) with heightened expression of efflux pumps.[107] CDA is a protein involved in the metabolic deactivation of cytidine analogs, encompassing not only gemcitabine but also cytarabine, azacytidine, and decitabine.The expression level of CDA could serve as a reliable marker for predicting the efficacy of these analog treatments, alongside the expression level of ENT1, the primary transporter for gemcitabine influx and dCK, which mediates gemcitabine activation [108,109].
Various transporters, predominantly ENTs that function as both influx and efflux pumps, as well as ABC transporters, especially MRP5, are involved in gemcitabine efflux.The latter is associated with multidrug resistance.[110] Although the relationship between Pgp expression and gemcitabine resistance remains unclear, over-expression of Pgp seems to heighten the cells' sensitivity to gemcitabine.[111].
Recent studies have suggested an association between alterations in the Erk/Akt/Stat3, Hedgehog (Hh), Wnt and Notch pathways and the development of gemcitabine resistance.[112] Specifically, the Erk/Akt/ Stat3 pathway, activated by cyclin L1 (CCNL1), promotes cell survival, thereby facilitating the development of gemcitabine resistance [113].
Gemcitabine resistance has also been linked to the Hedgehog pathway (Hh), particularly in patients with urothelial carcinoma.In these cases, gemcitabine induces a Gli2-dependent activation of the Hh pathway, which in turn fosters cell migration and invasion, enhancing the resistant phenotype.Consequently, a potential strategy to reduce gemcitabine resistance in these patients could involve targeting the Gli2 pathway.[114] The Hh pathway also plays a significant role in the tumorigenesis of pancreatic adenocarcinoma (PC).However, clinical trials involving patients with PC treated with vismodegib, an Hh pathway inhibitor, alongside gemcitabine, have not demonstrated any significant differences in treatment efficacy when compared to a control group.[115].
The Wnt pathway, featuring β-catenin as a crucial protein in its signaling cascade, also contributes to gemcitabine resistance.Studies focused on the correlation between β-catenin expression and the efficacy of gemcitabine treatment in pancreatic carcinoma have shown that patients with lower β-catenin expression demonstrated enhanced sensitivity to gemcitabine and significantly longer disease-free and overall survival compared to those presenting high β-catenin expression [116].
Gemcitabine resistance has also been associated with dysregulation of the Notch pathway, frequently upregulated in pancreatic and ovarian cancer.[117,118] An increased expression of Notch3 is associated with reduced sensitivity to gemcitabine, which is mediated by the activation of the PI3K/Akt pathway.Consequently, a decrease in Notch3 expression reduces PI3K/Akt activity, thereby inducing apoptosis in resistant cells.[119].
Research has indicated that pancreatic cancer is not only associated with alterations in the Hh, Wnt, and Notch pathways but that gemcitabine-resistant cells also demonstrate an increase in tumor stemness.[107] While the exact mechanism remains somewhat unclear, the induction of pancreatic cell stemness seems to result from the activation of the AKT/Notch1 pathway.Hypoxia further augments the activation of this pathway, thereby promoting the development of a resistant phenotype [120].

5-Fluorouracil (5-FU)
5-fluorouracil ranks among the earliest nucleoside-based anticancer drugs, together with 6-MP and 6-TG.First synthesized in 1957, it has since been employed in the treatment of patients with breast, colon, or head and neck cancers [121] 5-FU, an analog of uracil, is used both as a monotherapy and in combination with other anticancer drugs, such as methotrexate, leucovorin, baicalin, and oxaliplatin in order to enhance treatment efficacy [122][123][124] The combined use of 5-FU and curcumin has also demonstrated promising results in sensitizing 5-FU-resistant tumor cells.[125] Both in vitro and in vivo studies have provided evidence that tazemetostad (an inhibitor of the histone methyltransferase EZH2) can increase sensitivity to 5-FU treatment in cases of colorectal cancer [126].
The limitations of 5-FU therapy include dose-dependent toxicity and the development of multidrug resistance.Several mechanisms have been identified that contribute to altered sensitivity to 5-FU treatment.These include modifications of drug influx/efflux pumps, accelerated degradation or limited activation, and a deficiency in dihydropyrimidine dehydrogenase which can cause fatal 5-FU toxicity and alterations in various genes and molecules.These include the insulin-like growth factor-1 receptor, epidermal growth factor receptor, nuclear factor kappa B, cyclooxygenase-2, activator of transcription 3, the antiapoptotic Bcl-2, and pro-apoptotic Bax proteins.[123,125,133] The expression of thymidylate synthase (TS) has also been recognized as one of the most important predictive markers of 5-FU sensitivity in cancer treatment.Higher TS expression is known to decrease 5-FU sensitivity.[131,134].
The overexpression of Nrf3 (nuclear factor erythroid 2-like 3) and Bcl-2 has been observed in 5-FU-resistant tissues, with Nrf3 found to promote Bcl-2 expression.[135] Another factor contributing to 5-FU resistance is the hypermethylation of the MLH1 gene promoter.This hypermethylation inhibits MLH1 expression, which is involved in DNA mismatch repair (MMR).Cells with hypermethylated MLH1 resist various anticancer drugs, including 5-FU.In vitro experiments have shown that this resistance can be overcome by the re-expression of MLH1 via the demethylation of the promoter region by decitabine.[136].
FOXM1 is another transcription factor involved in 5-FU resistance.FOXM1 is essential for DNA repair, angiogenesis, and metastasis, and the knockout of FOXM1 results in increased sensitivity to 5-FU treatment.[137,138] These studies have also highlighted an association between cancer stem cells (CSCs) and alterations in anticancer drug sensitivity.CSCs, being quiescent and poorly differentiated populations, may evade 5-FU treatment targeting cell proliferation.[137].
The disruption of the p53 pathway is a predictive marker of the efficacy of combination therapy with 5-FU and oxaliplatin in patients with gastric cancer.Defects in this pathway are associated with a poorer response to treatment.[139].
The transport of 5-FU across the cell membrane is mediated by various transporters, including ABC transporters and nucleoside transporters.A study investigating 5-FU-resistant pancreatic carcinoma cells revealed a significant increase in the expression of MRP3, 4, and 5. Furthermore, the silencing of MRP5 was found to increase sensitivity to 5-FU.[140] An association between MRP5 overexpression and 5-FU resistance has been reported in both colon and breast cancers.[141,142] An analysis of small-cell lung cancer cells resistant to 5-FU demonstrated an approximately 25-fold increase in MRP8 expression.Interestingly, the alteration in MRP8 expression was induced by 5-FU treatment.The expression level of MRP8 is directly correlated with active 5-FU efflux and resistance development.[143] Overexpression of both MRP5 and MRP8 has been observed in colorectal cancer patients.[142].
While Pgp does not transport 5-FU, in vitro studies using 5-FU-resistant gastric cancer cell lines have shown that increased Pgp expression can lead to cross-resistance to taxanes such as paclitaxel and docetaxel.[144].
5-FU is a substrate of the breast cancer resistance protein (BCRP), which is primarily associated with drug resistance in breast cancer.Therefore, BCRP expression could potentially serve as another   The primary gemcitabine influx transporter; a higher expression is associated with increased sensitivity to gemcitabine [109] L. Hruba et al.
prognostic marker of 5-FU treatment efficacy.[145] Research has shown that the lung resistance protein (LRP), also known as the human major vault protein (MVP), is involved in the development of resistance to 5-FU.[146] 5-FU uptake is mediated mainly via nucleoside transporters, specifically ENT1, ENT2, and CNT2.The expression levels of these transporters could potentially help predict the efficacy of 5-FU treatment.However, resistance to 5-FU is typically associated with low expression or inhibition of these transporters.[147][148][149].

5-Azacytidine
5-Azacytidine, a pyrimidine analog of cytidine, was firstly synthesized in 1963.Since then, azacytidine has primarily been used in the treatment of AML, chronic myelomonocytic leukemia, and myelodysplastic syndrome (MDS), which transforms into leukemia in approximately 30% of patients.[150,151].
Resistance to azacytidine can arise from various factors, such as the decreased activation of kinases (particularly uridine-cytidine kinase 2 [UCK2]), diminished expression levels of CDA, alterations in the pyrimidine metabolism pathway, or the expression of specific membrane transporters.[156,157] Although azacytidine is transported by ENT1-4 and CNT1-3, CNT3 seems to play a pivotal role.[158] Consequently, a reduction in expression or inhibition of these transporters may diminish the efficacy of azacytidine.While no direct correlation has been reported between azacytidine resistance and Pgp over-expression, cell-based experiments have revealed an enhancement in azacytidine efficacy following the inhibition of ABC transporters (Pgp and MRPs) by erlotinib.[159,160].
Next-generation sequencing of patients with AML and MDS treated with azacytidine has uncovered numerous genetic mutations.However, none of these were significantly associated with treatment response.The only clinical marker linked with shorter survival was a platelet count of less than 50,000/μl.[161] The expression of UCK, which phosphorylates azacytidine into its active form, and CDA, which is responsible for the deactivation and elimination of azacytidine, could potentially serve as predictive markers for azacytidine treatment efficacy.[157].

Decitabine (5-AZA-2 -deoxycytidine, 5-AZA-CdR)
Decitabine is a compound structurally similar to azacytidine, synthesized a year later in 1964.Like azacytidine, decitabine is indicated for the treatment of adult patients diagnosed with AML and MDS.[151,162] Decitabine is used as a monotherapy or in combination with talacotuzumab for AML patients, and with cedazuridine for patients suffering from MDS or CMML.[163].
Our understanding of the mechanism behind decitabine resistance remains incomplete.However, in cases of MDS, non-responders have been observed to exhibit higher expression of CDA compared to responders, leading to increased elimination of decitabine.[165] Decitabine treatment efficacy is also related to the expression of dCK, which plays a role in its activating phosphorylation.Lower expression levels of this kinase may contribute to reduced efficacy of decitabine treatment.[166] Both decitabine and azacytidine resistance have been linked with alterations in the pyrimidine metabolism pathway.[157] Although decitabine is a substrate for BCRP protein, over-expression of this protein has not been reported in decitabine-resistant cell lines.[159,167] Decitabine is a poor Pgp substrate, and in vitro studies show, that decitabine treatment decreases Pgp expression leading to increasing sensitivity to drugs, which are Pgp substrates.[168].
The resistance to cladribine is multifaceted, attributed to factors such as increased activity of 5 -nucleotidase, which dephosphorylates cladribine-monophosphate into its inactive form, decreased activity of the activating dCK, or alterations in the apoptotic pathway.[175,176] Cladribine is a substrate of Pgp and BCRP but not MRPs, and elevated expression of these efflux pumps may adversely impact cladribine's efficacy.[172,173,177].

Current strategies to overcome nucleoside-based drug resistance
The development of drug resistance is a multifactorial process, including cellular changes improving cancer cell viability and proliferation.[10] Drug resistance can be connected with the altered expression of enzymes involved in the metabolic pathways, which activate or eliminate drugs.This is called cross-resistance, which arise among drugs with similar structure or mechanism of action.This kind of resistance is typical, for example, between 6-mercaptopurine and 6-thioguanine, [178] or among cytarabine, gemcitabine, and fludarabine, which are intracellularly activated by the same kinase (dCK).[179] Therapeutic strategy how to over-come this resistance can be choosing a drug with a different structure, and/ or mechanism of action or to employ combinatory therapy, which can enhance treatment efficacy.
Multidrug resistance (MDR) rising between drugs with completely different chemical structures and mechanisms of action is typically associated with altered expression of ABC-transporters (Pgp, MRPs, BCRP).[180,181] Since these proteins were identified as risk factors for MDR development, they became potential therapeutic targets.Until now, four generations of Pgp inhibitors were synthetized, but clinical trials did not demonstrate significantly improved therapeutic effects.[182] Reason for this failure can be the fact that ABC-transporters are very often co-expressed, and their substrate specificity is overlapping, so inhibition of one of them can be compensated by the others.Recently dual Pgp/BCRP inhibitors, such as flavonoids, their precursors and derivatives, curcumin, chalcone, and quinazoline derivatives etc., have been investigated with promising results.[183][184][185][186] In contrast, some MRP inhibitors show clinical relevance by themselves.For example, Ibrutinib, an MRP1 inhibitor, is improving the cytotoxicity of drugs that are substrates of this transporter.[187].
Other cellular mechanisms of chemoresistance are not connected only with the expression of membrane transporters or with the activity of metabolic enzymes.These therapeutic targets include changes at the genomic or proteomic level, which can be heterogenous among patients.Identifying biomarkers is important in finding new therapeutic approaches based on synthesizing small drug molecules, antibiotics, and monoclonal antibodies, which are the main focus of current research.[188] please in this position insert (te'In addition to direct binding to a cellular target and eliciting an immune response, monoclonal antibodies can also be applied as conjugates with anticancer drugs leading to the accumulation of drugs in the tumor tissue or with radionuclides enabling focus cytotoxic effect of radiation only at a specific location.[189,190] Specific binding and reducing the side effect caused by the off-target interactions is the biggest advantage of using monoclonal antibodies in cancer treatment. The first antibody-drug conjugate (ADC), vindesine-anti-CEA, was successfully prepared in 1983, and many others followed.[191] However, finding the optimal combination of drug and antibody is challenging, and not all ADCs have successfully passed the clinical trial.Currently, more than 11 FDA-approved ADCs are utilized in clinical settings, for example, gemtuzumab ozogamicin, which is used in the treatment of acute myeloid leukemia, brentuximab vedotin used in the treatment of Hodgkin lymphoma or ado-trastuzumab emtansine for breast cancer treatment.[192][193].
Nucleoside-based drugs are also being studied in their conjugated form.In this case, aptamer-drug conjugates (ApDCs) are being developed, which utilize nuclei acids instead of monoclonal antibodies.[195] Aptamers have been constructed for many nucleoside drugs.For example, the RNA aptamer (P19) was developed for gemcitabine and 5fluorouracil, and these ApCDs significantly inhibited the proliferantion of gemcitabine-resistant pancreatic cells.[196] In 2022, a new class of drug-constituted DNA-like oligomers, known as "drugtamers", were developed for clofarabine, gemcitabine, floxuridine and ara-guanosine.Compared to ApDCs drugtamers are expected to have a higher loading ratio, active targeting capability, and enzymatically controlled release of active compounds.This represents a promising advancement in the field of drug conjugates.[197].
The ultimate goal of current research is to advance personalized medicine, which can identify the optimal treatment for each individual patient.The identification of predictive markers is a useful tool, enabling the prediction of treatment efficacy and the implementation of the most effective cancer therapy.

Conclusion
Nucleosides, as delineated in Table 1, represent a versatile group of antimetabolites that have been successfully applied in the treatment of a broad spectrum of conditions, including diverse forms of cancer, myelodysplastic syndromes (MDSs), multiple sclerosis, and viral infections.The mechanism of action of this class of drugs typically involves the activation of a phosphorylation cascade, culminating in the formation of triphosphate forms that interact with nucleic acids.However, as these drugs take effect, cancer cells can evolve strategies to resist their impacts.While the pathways leading to resistance are diverse and multifactorial, several common pathways emerge, including a) alterations in membrane transporters, which can be characterized by reduced expression or activity of ENTs, CNTs, and/or overexpression of efflux pumps such as Pgp, MRPs, and BCRP (refer to Table 2); b) reduced activity of activating kinases, which may result in decreased phosphorylation of the drug and its capacity to interact with nucleic acids; and c) accelerated drug elimination through the activation of metabolic processes or increased renal clearance (see Table 1).Alongside these mechanisms, cancer cells may also develop resistance to nucleoside drugs by activating DNA repair mechanisms.This approach enables cancer cells to repair the damage caused by nucleoside drugs, thereby reducing their effectiveness.
Given these complex resistance mechanisms, several strategies for countermeasures have been proposed.These include targeted therapies, the development of innovative nucleoside drugs, and the formulation of unique drug combinations yielding synergistic effects.Such strategies hold the potential to prevent and overcome the resistant phenotype, thereby restoring tumor sensitivity and ultimately enhancing patient survival rates.The development of new strategies to combat resistance to nucleoside drugs continues to be a vibrant area of research, promising further breakthroughs in the future.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig. 1.Chemical structures of the most commonly used nucleoside-based anticancer drugs.

Table 1
Nucleoside-based drugs and their predictive markers.