Research Progress of Natural Small-Molecule Compounds Related to Tumor Differentiation

Tumor differentiation is a therapeutic strategy aimed at reactivating the endogenous differentiation program of cancer cells and inducing cancer cells to mature and differentiate into other types of cells. It has been found that a variety of natural small-molecule drugs can induce tumor cell differentiation both in vitro and in vivo. Relevant molecules involved in the differentiation process may be potential therapeutic targets for tumor cells. Compared with synthetic drugs, natural small-molecule antitumor compounds have the characteristics of wide sources, structural diversity and low toxicity. In addition, natural drugs with structural modification and transformation have relatively concentrated targets and enhanced efficacy. Therefore, using natural small-molecule compounds to induce malignant cell differentiation represents a more targeted and potential low-toxicity means of tumor treatment. In this review, we focus on natural small-molecule compounds that induce differentiation of myeloid leukemia cells, osteoblasts and other malignant cells into functional cells by regulating signaling pathways and the expression of specific genes. We provide a reference for the subsequent development of natural small molecules for antitumor applications and promote the development of differentiation therapy.


Introduction
Cancer is one of the major diseases threatening human health in the 21st century. Statistical studies have shown that 19.3 million new cancer patients were diagnosed, and 10 million died from cancer in 2020. If the national incidence rate stays the same, there will be 28.4 million cancer patients worldwide in 2040-47% higher than in 2020 [1]. A tumor is a new organism formed by local tissue and cell proliferation under the action of various tumorigenic factors. Because the new organism mostly presents space-occupying massive processes, tumors are also known as vegetation. In addition to uncontrolled cell proliferation and survival, the continuous loss of normal differentiation is also a feature of most malignant cells [2]. Malignant tumors include carcinoma, sarcoma, hematopoietic and lymphoid malignancies, etc. The conventional treatments of tumors include surgery, radiotherapy, chemotherapy, targeted therapy, immunotherapy, traditional Chinese medicine therapy, etc., which are aimed at killing tumor cells. However, killing tumor cells inevitably damages normal cells, and tumor cells cannot be eliminated due to their migratory and invasive nature. Tumor differentiation is a method aimed at reactivating the endogenous differentiation program of cancer cells and inducing cancer cells to stop proliferation and restore normal cell characteristics [3]. The molecular mechanisms involved in the differentiation process are potential therapeutic targets for tumor cells, and the induction of differentiation is an ideal outcome for both chemopreventive and therapeutic agents.
Based on the fact that a tumor is a disease with disturbed cell differentiation and impaired maturation, attempts to induce cancer cells to redifferentiate towards maturity into structurally and functionally normal cells have gradually started to move from the laboratory to the clinic. In 1961, Pierce first found that mouse testicular teratoma cells could spontaneously differentiate into benign normal cells, which verified that malignant tumors could be cured by inducing tumor cell differentiation, as proposed in the 1950s, and research was initiated to induce tumor cell differentiation [4]. Then in 1975, Friend reported that dimethyl sulfoxide induced erythrocyte differentiation and stimulated hemoglobin synthesis in mouse leukemia cells, which confirmed the possibility of inducing differentiation in vitro [5]. In addition, Wang Zhengyi found that symptoms of acute promyelocytic leukemia (APL) were completely relieved by all-trans retinoic acid (ATRA) in 72% of patients in 1985, creating a precedent for the clinical application of differentiation-inducing agents in the treatment of leukemia [6]. Now, APL can be cured by the combined treatment of retinoic acid (RA) and arsenic.
According to the mechanisms of tumor differentiation, differentiation inducers can be divided into endogenous and exogenous agents. The former refers to the chemicals with the differentiation-inducing effect produced by tumor or host cells, such as colonystimulating factors and glucocorticoids; and the latter refers to the differentiation inducers that cannot be produced by tumor or host cells but must rely on external supply, including vitamins, nucleosides and their analogs, organic compounds, etc. At present, a variety of compounds inducing the differentiation of cancer cells have been reported, such as theophylline [7], trimeric glutaryl pentane [8] and 5,7-dimethoxycoumarin [9], which can induce the differentiation of blood tumor cells or solid tumor cells by arresting their proliferation through diverse mechanisms. More research shows that a large number of differentiation-inducing agents have been found and synthesized all over the world.
However, the above agents also have certain limitations, such as causing adverse reactions and drug resistance. It has been found that natural small-molecule compounds widely present in plants, such as polyphenols, flavonoids, carotenoids, alkaloids, etc., have a broad safety profile and therapeutic effects on a wide range of diseases; therefore, some of them have entered clinical research, and many strategies have been proposed for their improvement [2,10,11]. Myeloid leukemia cells are particularly sensitive to most of these natural small-molecule compounds, which are induced to differentiate into morphologically and functionally mature cells. The use of natural small-molecule compounds to induce malignant cell differentiation represents a more targeted and potentially less toxic treatment of oncological disease, which could provide hope for improved treatment of malignancies by avoiding the significant side effects seen with conventional therapies.
This review focuses on natural small-molecule compounds that induce myeloid leukemia cells, osteoblasts, and some other solid tumor cells to advanced stages of differentiation by regulating signaling pathways and the expression of specific genes. In particular, we aim to summarize the current status of research on natural small-molecule compounds that show differentiation induction of malignant cells and their mechanisms (several natural compounds with a detailed mechanism of action are shown in Figure 1) and to provide basic knowledge to facilitate the development of differentiation therapies. We used "small molecules", "natural compounds", "tumors", "hematologic tumors" or "solid tumors" and "differentiation" or "differentiation therapy" or other relevant items as keywords in a search of the PubMed database. In addition, the Clinicaltrials.gov database was searched for relevant diseases to obtain clinical trials of natural compounds. No restrictions were set on the date of publication of the literature.

Hematologic Tumors
Hematologic tumor is a general term for a large group of malignant tumors that originate in the hematopoietic system. Common hematological tumors include various types of leukemia, multiple myeloma and malignant lymphoma. Hematological malignancies are a group of heterogeneous hematological tumors usually characterized by abnormal production of blood cells (hematopoiesis) [12]. Although the clinical management of hematologic malignancies has improved significantly over the past few years, the socioeconomic costs are a concern. Meanwhile, many key challenges remain, such as recurrence, refractory lesions, and morbidity and mortality associated with malignancy. If tumor cells can be induced to differentiate and prevented from proliferating, their malignant potential could be controlled. Therefore, differentiation therapy is a promising treatment method. A successful example of this method is all-trans retinoic acid (ATRA) against acute myeloid leukemia, which promotes the implementation of differentiation therapy. It is worth noting that many natural small molecule compounds have shown significantly therapeutic potential in promoting differentiation in myeloid leukemia (Table 1).

Myeloid Leukemia
Leukemia is a kind of malignant clonal disease of hematopoietic stem cells. Its incidence rate is high (about 27,600 per 100,000 people) because of its diverse pathogenic factors [13]. According to the data of global cancer statistics in 2020, leukemia ranks 15th among all types of cancer, with 474,519 new cases (2020) and 311,594 deaths [1]. The classification of leukemias is more complex, and it is often clinically classified into myeloid and lymphocytic leukemias based on different series of leukodystrophies. Myeloid leukemias can in turn be divided into acute myeloid leukemia (AML) and chronic myeloid leukemia (CML) based on the rapidity of the condition and the maturation of leukemic differentiation, of which AML is the most common, accounting for more than 50%.

Hematologic Tumors
Hematologic tumor is a general term for a large group of malignant tumors that originate in the hematopoietic system. Common hematological tumors include various types of leukemia, multiple myeloma and malignant lymphoma. Hematological malignancies are a group of heterogeneous hematological tumors usually characterized by abnormal production of blood cells (hematopoiesis) [12]. Although the clinical management of hematologic malignancies has improved significantly over the past few years, the socioeconomic costs are a concern. Meanwhile, many key challenges remain, such as recurrence, refractory lesions, and morbidity and mortality associated with malignancy. If tumor cells can be induced to differentiate and prevented from proliferating, their malignant potential could be controlled. Therefore, differentiation therapy is a promising treatment method. A successful example of this method is all-trans retinoic acid (ATRA) against acute myeloid leukemia, which promotes the implementation of differentiation therapy. It is worth noting that many natural small molecule compounds have shown significantly therapeutic potential in promoting differentiation in myeloid leukemia (Table 1).

Myeloid Leukemia
Leukemia is a kind of malignant clonal disease of hematopoietic stem cells. Its incidence rate is high (about 27,600 per 100,000 people) because of its diverse pathogenic factors [13]. According to the data of global cancer statistics in 2020, leukemia ranks 15th among all types of cancer, with 474,519 new cases (2020) and 311,594 deaths [1]. The classification of leukemias is more complex, and it is often clinically classified into myeloid and lymphocytic leukemias based on different series of leukodystrophies. Myeloid leukemias can in turn be divided into acute myeloid leukemia (AML) and chronic myeloid leukemia (CML) based on the rapidity of the condition and the maturation of leukemic differentiation, of which AML is the most common, accounting for more than 50%. Acute myeloid leukemia (AML) is mainly caused by chromosome translocation, which produces fusion proteins with abnormal activity, resulting in cell-cycle imbalance and failure of hematopoietic differentiation. The consequence of the replacement of normal hematopoietic stem cells with tumor cells is a decrease in the number of blood cells, anemia, thrombocytopenia and neutropenia. Although current treatment strategies provide a reasonable possibility for the majority of patients to achieve complete remission, the recurrence rate and subsequent disease-related mortality remain high [2]. In 2020, there were 19,940 newly diagnosed AML cases, and 11,180 patients died of AML in the United States [14]. The improvement of chemotherapy and maintenance therapy increased the overall survival rate of young patients (less than 15 years old) to more than 60%; however, in the end, 40% of people relapse and need salvage therapy.
Induction of cell differentiation has also become an important choice in cancer therapy due to the cyclic imbalance and abnormal hematopoietic differentiation characteristics of AML. The potential of AML differentiation agents has also been demonstrated through the use of all-trans retinoic acid (ATRA), which has achieved significant clinical success in a small number of AML patients with the acute promyelocytic leukemia (APL) subtype [15,16]. In addition to synthetic drugs, a variety of plant-derived compounds (phytochemicals) induce cell differentiation through a variety of mechanisms and show antitumor activity. Many natural small molecules contribute to mediation of apoptosis, cell cycle and differentiation through oxidative stress pathways. Reactive oxygen species (ROS) are one of the key markers of differentiation; in this case, they are considered potential therapeutic agents for the treatment of AML. Securinine is an alkaloid from the root of the plant securinega suffruticosa. Securinine activates receptor tyrosine kinase (nRTK) to induce HL-60 differentiation towards monocytes, with increased expression of CD11b and CD14. Not only does securinine decrease the expression of transcription factors, such as CCAAT/enhancer-binding proteins α (C/EBPα), C/EBPε, GATA-1 and cellularmyelocytomatosis viral oncogene (c-myc), but it also initiates ROS-induced DNA damage by activating the JNK/ERK pathway, which induces ATM/ATR and Chk1-dependent cell differentiation. More importantly, securinine has also been reported to induce AML tumor differentiation in primary leukemia patients, as well as in nude mice, showing potent differentiation activity and potential for clinical application. At present, securinine has been used for clinical treatment of neurological disorders, although it occasionally induces epilepsy due to GABA. GABA activation does not adversely affect securinine's activity to induce AML differentiation. Therefore, this alkaloid could be considered as a potential therapeutic agent of AML [17][18][19][20]. Coincidentally, diallyl disulfide (DADS), a major anticancer active ingredient derived from garlic, can also induce leukemia cell differentiation via ROS pathways in which CRT was downregulated and translocated, resulting in the release of the creatine transporter (CRT) and C/EBP α mRNA interactions that promote C/EBP α protein expression. It was previously believed that DADS exerted anticancer activity by defending against carcinogen invasion and inhibiting proliferation of various cancer cells, but the latest study showed that, apart from ROS pathways, DADS could also negatively regulate the rac1-rock1-limk1-cofilin1 axis to induce leukemia cell differentiation to play an anticancer role. In addition, DADS significantly inhibited the growth and induced the differentiation of HL-60 cells in SCID mice [21][22][23][24]. Small-molecule natural compounds that induce cancer cell differentiation via ROS pathways also include shikonin and ethyl acetate extract (CAE). Shikonin, a major component of the herb comfrey, may interact with biological targets through the formation of covalent bonds or as an electron transfer agent in redox reactions to regulate redox homeostasis between cells via the Nrf2/ARE pathway, thereby promoting differentiation and apoptosis in HL-60 cells. Low doses (<100 ng/mL) of paclitaxel shift cells from proliferation to differentiation; however, higher concentrations lead to cell death [25][26][27]. Another small-molecule natural compound is ethyl acetate extract (CAE), an ethyl acetate extract from Caesalpinia sappan L, which differentially inhibits AML cells in vitro in a concentration-dependent manner without toxic effects on normal cells and induces mitochondrial apoptosis, G2/M-phase cell cycle arrest and late cell differentiation through reactive oxygen species [28].
Studies have shown that orphan nuclear receptor Nur77 is an important tumor suppressor gene in AML and that deletion of Nur77 expression is a common feature in patients with AML. Deletion of Nur77 leads to impaired differentiation of hematopoietic stem cells and bone marrow progenitor cells, which contributes to development of AML. Targeted activation of Nur77 expression has been shown to be a potential new intervention approach in the treatment of AML. Ginsenoside 20(s)-Rh2 is a protopanaxadiol-type ginsenoside and also one of the typical components of red ginseng. Ginsenoside 20(s)-Rh2 [29][30][31] promotes Nur77 translocation from the nucleus to the mitochondria by activating the Nur77-mediated death receptor pathway (Fas and DR5), followed by Nur77 interaction with Bcl-2 to cause apoptosis. 20(S)-Rh2 induces AML cell differentiation via the Nur77mediated transcription factors c-Jun and JunB [32,33]. Like 20(s)-Rh2, CTD also induces apoptosis and differentiation through a Nur77-mediated signaling pathway. Cantharidin (CTD) is the main bioactive component of cantharidin. Recent studies have found that CTD and its derivatives, such as Cantharidic acid and norcantharidin, inhibit leukemia cell proliferation, induce apoptosis, cause cell cycle arrest and enhance the inhibitory effect of chemotherapeutic drugs [34][35][36]. Notopterol, a coumarin, is an active monomer extracted from N. incisum with antipyretic, analgesic, and anti-inflammatory properties. Notopterol [37][38][39] induces apoptosis, differentiation and G0/G1 cell cycle arrest in human AML HL-60 cells. Notopterol markedly induces protein expression of c-Jun and JunB and decreases c-myc, thereby inducing differentiation [40]. DT-13 is a new compound isolated from Liriope muscari (Decne) Baily, which has strong cytotoxicity to a variety of solid tumors. Similar to 20(S)-Rh2, Liriodendron chinense saponin C (DT-13) [41][42][43] induces apoptosis in HL-60 and Kasumi-1 cells by upregulating Fas, FasL, DR5 and TRAIL through the expression of Kruppel-like factor 2 (KLF2). The restoration of KLF2 by DT-13 is highly correlated with AMPK-related histone acetylation mechanisms. DT-13 significantly enriches Ace-H3 (AH3-120) in the promoter region of the KLF2 gene. In addition, KLF2 promotes differentiation by upregulating the expression levels of the differentiation markers CD11b and CD14 and the transcription factors C/EBPα and C/EBPβ. Therefore DT-13 may mediate apoptosis and differentiation through AMPL-KLF2 [44]. DT-13 is classified by the Chinese Ministry of Health as an herbal medicine due to its high efficiency and safety.

Chronic Myeloid Leukemia (CML)
Chronic myeloid leukemia (CML) is characterized by the abnormal accumulation of immature leukemia mother cells in blood, bone marrow and spleen, which prevents the terminal differentiation of myeloid cells but promotes the expression of BCR-ABL fusion oncoprotein. BCR-ABL fusion protein plays a role in the imbalance of tyrosine kinase activity, which activates proliferation and antiapoptotic signal pathways and leads to the malignant expansion of pluripotent stem cells in bone marrow. Furthermore, BCR-ABL fusion protein mediates downregulation of C/EBPα expression, which triggers a block in the myeloid differentiation of CML cells. By inducing terminal differentiation, leukemic cells may lose their ability to proliferate and form mature, functional cells. Therefore, induction of differentiation is an ideal therapeutic strategy.
Unlike AML, CML cells are particularly sensitive to erythrocyte differentiation. Many natural compounds, including apigenin and fagaronine or other anthracycline antibiotics and HSP90 inhibitors, have been proved to induce CML cell differentiation into erythroid lineage in vitro. These inducers work through different mechanisms, including direct or indirect inhibition of BCR-ABL. Almost all these compounds activate the key erythroblast transcription factor GATA-1. Apigenin [45,46] is considered to be a bioactive flavonoid with anti-inflammatory and antioxidant effects. It upregulates the mRNA expression of the erythroid transcription factor GATA-1 and alters the nuclear and cytoplasmic DTP distribution, thereby affecting the nuclear translocation and localization of signaling molecules, such as GATA-1. Long-term treatment with apigenin induced erythroid-like differentiation of K562 cells. Analysis of the structure-function relationships of the flavonoids showed that apigenin-induced differentiation of K562 cells was due to the 2-3 double bonds and hydroxyl groups in their structure. Apigenin mainly acts as a glucoside conjugate due to its unstable pure form, and its conjugated form may be an important determinant of its bioavailability and absorption [47,48]. Fagaronine, a benzo phenanthrene alkaloid extracted from Fagara zanthoxyloides Lam, exerts its differentiation activity by specifically activating genes involved in the regulation of GATA-1 in the region of erythroid phenotypic expression [49]. Wogonine [50][51][52] is a flavonoid extracted from Scutellaria baicalensis Georgi that has a good curative effect in the treatment of hematological malignancies. Wogonine inhibited BCR-ABL-mediated phosphorylation of MEK and ERK in K562 cells and increased the expression of erythroblast differentiation genes by upregulating transcription factor GATA-1 and enhancing the combination between GATA-1 and friend of GATA-1 (FOG-1). Differentiation was accompanied by cell cycle arrest in the G0/G1 phase, as well as concomitant upregulation of p21 and downregulation of CDK4 and cyclin D1 [53].
Small-molecule tyrosine kinase inhibitors (TKIs) [54], such as imatinib, dasatinib, nilotinib and ponatinib, have been developed to treat CML by blocking the kinase structural domain of the BCR-ABL oncoprotein. However, there are a variety of adverse reactions of these compounds, such as off-target and metabolic toxicity, due to the resistance of BCR-ABL mutation to TKI treatment and the progression to advanced diseases. It is necessary to seek new treatment strategies. Gambogic acid (GA) is a small molecule derived from the Chinese herb Garcinia cambogia. It promotes apoptosis and differentiation in imatinib-resistant chronic granulocytic leukemia cells through induction of proteasome inhibition and caspase-dependent BCR-ABL downregulation. Studies on GA derivatives found that GA derivatives lacking reactive α, β-unsaturated carbonyl groups were ineffective, suggesting that α, β-unsaturated carbonyl groups are important for the exertion of GA activity [55][56][57]. Andrographolide (Andro), a major component of the medicinal plant Andrographis paniculata, promotes the differentiation of imatinib-sensitive and imatinib-resistant CML cells through the production of ROS, which prompts BCR-ABL downregulation and Hsp90 cleavage. The aliphatic ester enhances the cytotoxic activity of Andro and the double bond between C-8 and C-17, in addition to the lactone and its conjugated double bond, maybe the central structure in which it acts [58]. In general, Andro and GA provide an alternative and complementary strategy for imatinib-resistant CML by inhibiting BCR-ABL function through a different mechanism than that of imatinib [59][60][61]. 8-hydroxydaidzein (8-OHD, 7,8,4 -trihydroxyisoflavone) is a hydroxylated derivative of daidzein isolated from fermented soybean products. The o-dihydroxy group in the 8-OHD structure has superior antioxidant and free radical scavenging activity and may be the key moiety for its action. 8-OHD activates the mitogen-activated protein kinase (MAPK) and nuclear clearance factor-κB (NF-κB) signaling pathways, thereby inducing autophagy, apoptosis and cycle arrest. In addition, 8-OHD induces degradation of BCR-ABL oncoprotein and promotes early growth response 1 (EGR1)-mediated megakaryocyte differentiation, which could be a potential compound for the treatment of CML [62,63]. Atractylodes macrocephala (also known as white bamboo) is an important component of several traditional Chinese medicine prescriptions used to treat abdominal pain and gastrointestinal diseases for thousands of years. The main active compound, atractylenolide I (ATL-I), induces apoptosis and differentiation of human leukemia cells [64].

Multiple Myeloma (MM) and Osteosarcoma
Multiple myeloma (MM) is a malignant tumor in which plasma cells proliferate abnormally. MM can cause organ damage, including osteolytic lesions, anemia, renal failure or hypercalcemia (clonal expansion of malignant plasma cells). To date, multiple myeloma is still considered incurable, accounting for about 10% of all hematological malignancies and 2% of all cancers (Cancer Research UK, 2018). In 2016, there were 138,509 incident cases worldwide (Cowan, 2018). Between 1990 and 2016, the incidence rate of the whole world increased by 126% and was closely related to age (Cancer Research UK 2018; Cowan 2018) [65].
In MM disease, clonal malignant plasma cells accumulate in bone marrow, reduce osteoblast formation and stimulate osteoclasts to destroy the bone. Osteoblasts build bones by forming connective cell populations, and osteoclasts are large multinucleated cells that destroy bones. The functional balance of the two cell types is essential for bone maintenance and repair. Therefore, in addition to targeting myeloma cells to treat MM patients, osteoclasts and osteoblasts are also regarded as potential targets. At present, it has been proposed that drugs that induce osteoblast differentiation can be used as an alternative therapy for MM treatment. Osteoblast differentiation requires the activity of transcription factor 2 (Runx2) and Osterix transcription factors, as well as growth factors.
In MM therapy, natural small-molecule compounds act as differentiation agents for osteoblasts ( Table 2). This study shows that resveratrol, silibinin, acerogenin A and baicalein can regulate the expression and activity of specific markers, including bone morphogenetic proteins (BMPs) and key transcription factors Osterix and Runx2, to induce differentiation. Bone morphogenetic proteins (BMPs), members of the TGF-β superfamily, are often the key regulators of bone formation and remodeling, and BMPs also play an important role in the differentiation of hBMSCs into osteoblast-like cells [66]. Silibinin is a strong antioxidant with a very strong hydrogen bonding of the 5-OH group to the adjacent oxo group, which is conjugated to the aromatic ring and acts as a free electron pair donor for the hydrogen bonding to the 5-OH group. Silibinin enhances osteoblast differentiation of human bone marrow stromal cells by inducing the expression of BMPs and activating BMP and Runx2 pathways through hydrogen bonding [67,68]. Acerogenin A (ACE) is a natural compound isolated from Acer nikoense Maxim. In MC3T3-E1 and RD-C6 cells and primary osteoblasts, ACE stimulates osteoblast differentiation via BMP action, which is mediated by Runx2-dependent and Runx2-independent pathways. In addition, ACE increases the expression of osteocalcin mRNA in which the hydroxyl groups of C-9 and C-11 may play a role [69,70]. Baicalein is a coumarin-like derivative extracted from Chinese herbs that induces early osteoblast differentiation through activation of the MAP kinase/NF-κB signaling pathway, and this activation is associated with increased expression of osteoblast differentiation markers. In addition, baicalin induced the differentiation of MC3T3-E1 in mouse osteoblasts [71][72][73][74][75]. Resveratrol (RSV) is a natural compound present in various plant species that reduces the growth of myeloma cell lines (RPMI 8226 and OPM-2) in a dose-dependent manner and inhibits ligand-receptor activator (RANKL)-induced osteoclast differentiation. In addition, RSV induced the expression of the osteoblast markers osteocalcin and osteoprotegerin in immortalized human bone marrow mesenchymal stem cells (HMSC-TERT), leading to osteoblast differentiation [76][77][78][79]. Several studies on RSV and its natural or synthetic analogues have highlighted the importance of 3,4-dihydroxyl groups in the expression of cytotoxic and proapoptotic activity, suggesting that hydroxyl groups are important for the expression of RSV activity. The RSV derivative STR50 is currently being tested in phase II clinical trials for the treatment of MM. Icaritin is the main constituent of Herba Epimedii. It has obvious inhibitory effects on a variety of blood cancer cells, including acute myeloid leukemia (AML), chronic myeloid leukemia (CML), multiple myeloma (MM) and lymphoma. By inhibiting the JAK-STAT pathway, icaritin induces cell differentiation, inhibits tumor cell migration and inhibits the growth of cancer stem/progenitor cells. Besides, icaritin has an estrogen-like chemical structure, which can induce MC3T3-E1 cell differentiation through estrogen receptor-mediated activation of ERK1/2 and p38 signaling [80][81][82][83][84]. Similarly, osteoblast differentiation and mineralization induced by the estrogen receptor (ER) pathway can also be used to treat osteosarcoma. The ER functions primarily as a DNA-binding transcription factor that regulates gene expression.
Osteosarcoma is the most common non-hematological skeletal malignancy in both children and adults [85]. Although modern treatment schemes have combined multiple approaches, such as chemotherapy, surgery and radiotherapy, the five-year survival rate of osteosarcoma patients has been maintained at 60-70% since the 1970s [86,87]. Osteosarcoma has a high rate of metastasis and chemoresistance and therefore has a poor prognosis. Although surgical techniques and implants have shown a steady progress, current chemotherapeutic drugs associated with toxic side effects, including cardiotoxicity, infertility and renal insufficiency, seem to be similar to those used 40 years ago [88]. This highlights the urgent need for new therapies and drugs. Osteosarcoma cells share many similar characteristics with undifferentiated osteogenic progenitor cells, including high proliferative capacity, resistance to apoptosis and similar expression profiles of many osteogenic markers, such as connective tissue growth factor, Runx2, alkaline phosphatase (ALP) and Osterix. Because more than 80% of osteosarcomas have histopathological differentiation differences, new therapies based on non-cytotoxic induction of cell differentiation response pathways may have good prospects.

Other Tumors
Many natural small-molecule compounds have antitumor and differentiation-promoting effects in a variety of tumors (Table 3). For example, kaempferol and genistein promote not only osteoblast differentiation but also melanoma differentiation. Melanoma

Other Tumors
Many natural small-molecule compounds have antitumor and differentiation-promoting effects in a variety of tumors (Table 3). For example, kaempferol and genistein promote not only osteoblast differentiation but also melanoma differentiation. Melanoma is a type of cancer caused by the malignant transformation of melanocytes. Melanin production is thought to be a major marker of differentiation in melanoma models. For melanoma, many cell lines considered to represent the "blocked" stage of melanoma differentiation have been isolated and studied. Finding new cancer-specific differentiation inducers and clarifying the basis of cancer cell differentiation provides support for the development of improved melanoma therapy. An early study showed that theophylline (1,3-dimethylxanthine; Theo) induced the maturation and differentiation of B16 melanoma cells by activating cAMP-dependent protein kinase A and inducing phosphorylation of cAMP response element-binding proprotein (CREB) [121][122][123]. Isoliquiritigenin (ISL) from licorice can induce the differentiation of mouse melanoma B16F10 cells [3,[124][125][126]. Recent studies have shown that three isopentenyl flavonoids in GF-1, GF-4 and GF-9 in licorice roots can significantly induce the differentiation of B16-F10 melanoma cells. In addition, kaempferol [127][128][129][130][131], genistein [132][133][134] and 3 3-diindolylmethane (DIM) [135,136] inhibit the viability of melanoma cells and induce apoptosis and differentiation of malignant melanoma cells mediated by reactive oxygen species and endoplasmic reticulum stress [137].
Neuroblastoma is the most common extracranial solid tumor in children, accounting for about 15% of all cancer-related deaths in children [138]. The prognosis is related to patient age, MYCN oncogene amplification and tumor differentiation. Retinoic acid (RA) is used as a differentiation agent for maintenance therapy of high-risk neuroblastoma (NB), but its use is limited by its related toxicity. Natural compounds can avoid the limitations of toxicity and have great potential. Kaempferol (KFL) induces differentiation of neuroblastoma cells via the IRE1α-XBP1 pathway [139,140]. Melatonin (N-acetyl-5methoxytrimethylamine) is an endocrine hormone mainly released by the pineal gland. Melatonin promotes the differentiation of neuroblastoma cells by activating cell phagocytosis induced by hyaluronate synthase 3 (HAS3) [141][142][143][144][145]. Tinospora cordifolia (TCE) has a strong ability to resist metastasis and induce differentiation in neuroblasts, and IMR-32 cells can be induced by treatment with higher concentrations of TCE. Studies have shown that TCE or phytochemicals derived from this plant can be used as safe pharmacological agents together with conventional drugs [146,147]. Chlorogenic acid (CA) is a plant compound isolated from Eucommia ulmoides, honeysuckle and other plants. In glioma cells, CA increases the expression of specific differentiation biomarkers Tuj1 and GFAP to induce differentiation. The therapeutic effect of CA in glioma cells is equivalent to that of temozolomide [148][149][150].
Glioblastoma is the most common and invasive primary brain tumor. It is the most malignant brain tumor, with a very poor prognosis. The current standard of care for patients with GBM includes tumor resection, followed by temozolomide radiotherapy and chemotherapy. However, due to the highly invasive nature of the tumor, it is almost impossible to remove completely. In addition, stem-like tumor cells resistant to chemotherapy and radiotherapy were found to regroup in the tumor cavity, resulting in tumor recurrence and drug resistance. The prognosis of patients with GBM remains very poor, with a median survival of 9.9 to 15 months.
Differentiation therapy has been proposed as a promising strategy for GBM treatment because after differentiation, GBM cells lose tumorigenicity and become sensitive to chemotherapy and radiotherapy [151]. Resveratrol promotes differentiation and inhibits the tumorigenic and self-renewal capacity of glioma stem cells (GSCs). Studies have demonstrated that Nanog is a key factor in the maintenance of GSCs and that resveratrol induces phosphorylation-dependent activation of p53, leading to Nanog degradation in GSCs, which may overcome treatment resistance in GSCs [152][153][154][155]. Xanthone-rich extract from Gentiana dinarica transformed roots, and its active component, norswertianin, induce autophagy and ROS-dependent differentiation of the human glioblastoma cell line [156,157].
Similarly, curcumin [158,159] can significantly induce differentiation of glioma-initiating cells (GICs) in vivo and in vitro through the induction of autophagy. Thus, curcumin can be used as an adjuvant to conventional chemotherapy in the treatment of GBM [160]. Falcarindiol (FAD) is a polyacetylene found in many dietary plants and has shown a variety of physiological activities, such as anti-inflammatory, antibacterial and hepatotoxic inhibition. FAD has an anticancer effect on glioblastoma cells by inducing the differentiation of glioblastoma stem cells and activating the apoptosis pathway of glioblastoma cells [161]. Differentiation therapy has been proposed as a promising strategy for GBM treatment because after differentiation, GBM cells lose tumorigenicity and become sensitive to chemotherapy and radiotherapy [151]. Resveratrol promotes differentiation and inhibits the tumorigenic and self-renewal capacity of glioma stem cells (GSCs). Studies have demonstrated that Nanog is a key factor in the maintenance of GSCs and that resveratrol induces phosphorylation-dependent activation of p53, leading to Nanog degradation in GSCs, which may overcome treatment resistance in GSCs [152][153][154][155]. Xanthone-rich extract from Gentiana dinarica transformed roots, and its active component, norswertianin, induce autophagy and ROS-dependent differentiation of the human glioblastoma cell line [156,157]. Similarly, curcumin [158,159] can significantly induce differentiation of gliomainitiating cells (GICs) in vivo and in vitro through the induction of autophagy. Thus, curcumin can be used as an adjuvant to conventional chemotherapy in the treatment of GBM [160]. Falcarindiol (FAD) is a polyacetylene found in many dietary plants and has shown a variety of physiological activities, such as anti-inflammatory, antibacterial and hepatotoxic inhibition. FAD has an anticancer effect on glioblastoma cells by inducing the differentiation of glioblastoma stem cells and activating the apoptosis pathway of glioblastoma cells [161]. Differentiation therapy has been proposed as a promising strategy for GBM treatment because after differentiation, GBM cells lose tumorigenicity and become sensitive to chemotherapy and radiotherapy [151]. Resveratrol promotes differentiation and inhibits the tumorigenic and self-renewal capacity of glioma stem cells (GSCs). Studies have demonstrated that Nanog is a key factor in the maintenance of GSCs and that resveratrol induces phosphorylation-dependent activation of p53, leading to Nanog degradation in GSCs, which may overcome treatment resistance in GSCs [152][153][154][155]. Xanthone-rich extract from Gentiana dinarica transformed roots, and its active component, norswertianin, induce autophagy and ROS-dependent differentiation of the human glioblastoma cell line [156,157]. Similarly, curcumin [158,159] can significantly induce differentiation of gliomainitiating cells (GICs) in vivo and in vitro through the induction of autophagy. Thus, curcumin can be used as an adjuvant to conventional chemotherapy in the treatment of GBM [160]. Falcarindiol (FAD) is a polyacetylene found in many dietary plants and has shown a variety of physiological activities, such as anti-inflammatory, antibacterial and hepatotoxic inhibition. FAD has an anticancer effect on glioblastoma cells by inducing the differentiation of glioblastoma stem cells and activating the apoptosis pathway of glioblastoma cells [161]. Differentiation therapy has been proposed as a promising strategy for GBM treatment because after differentiation, GBM cells lose tumorigenicity and become sensitive to chemotherapy and radiotherapy [151]. Resveratrol promotes differentiation and inhibits the tumorigenic and self-renewal capacity of glioma stem cells (GSCs). Studies have demonstrated that Nanog is a key factor in the maintenance of GSCs and that resveratrol induces phosphorylation-dependent activation of p53, leading to Nanog degradation in GSCs, which may overcome treatment resistance in GSCs [152][153][154][155]. Xanthone-rich extract from Gentiana dinarica transformed roots, and its active component, norswertianin, induce autophagy and ROS-dependent differentiation of the human glioblastoma cell line [156,157]. Similarly, curcumin [158,159] can significantly induce differentiation of gliomainitiating cells (GICs) in vivo and in vitro through the induction of autophagy. Thus, curcumin can be used as an adjuvant to conventional chemotherapy in the treatment of GBM [160]. Falcarindiol (FAD) is a polyacetylene found in many dietary plants and has shown a variety of physiological activities, such as anti-inflammatory, antibacterial and hepatotoxic inhibition. FAD has an anticancer effect on glioblastoma cells by inducing the differentiation of glioblastoma stem cells and activating the apoptosis pathway of glioblastoma cells [161]. Differentiation therapy has been proposed as a promising strategy for GBM treatment because after differentiation, GBM cells lose tumorigenicity and become sensitive to chemotherapy and radiotherapy [151]. Resveratrol promotes differentiation and inhibits the tumorigenic and self-renewal capacity of glioma stem cells (GSCs). Studies have demonstrated that Nanog is a key factor in the maintenance of GSCs and that resveratrol induces phosphorylation-dependent activation of p53, leading to Nanog degradation in GSCs, which may overcome treatment resistance in GSCs [152][153][154][155]. Xanthone-rich extract from Gentiana dinarica transformed roots, and its active component, norswertianin, induce autophagy and ROS-dependent differentiation of the human glioblastoma cell line [156,157]. Similarly, curcumin [158,159] can significantly induce differentiation of gliomainitiating cells (GICs) in vivo and in vitro through the induction of autophagy. Thus, curcumin can be used as an adjuvant to conventional chemotherapy in the treatment of GBM [160]. Falcarindiol (FAD) is a polyacetylene found in many dietary plants and has shown a variety of physiological activities, such as anti-inflammatory, antibacterial and hepatotoxic inhibition. FAD has an anticancer effect on glioblastoma cells by inducing the differentiation of glioblastoma stem cells and activating the apoptosis pathway of glioblastoma cells [161]. Differentiation therapy has been proposed as a promising strategy for GBM treatment because after differentiation, GBM cells lose tumorigenicity and become sensitive to chemotherapy and radiotherapy [151]. Resveratrol promotes differentiation and inhibits the tumorigenic and self-renewal capacity of glioma stem cells (GSCs). Studies have demonstrated that Nanog is a key factor in the maintenance of GSCs and that resveratrol induces phosphorylation-dependent activation of p53, leading to Nanog degradation in GSCs, which may overcome treatment resistance in GSCs [152][153][154][155]. Xanthone-rich extract from Gentiana dinarica transformed roots, and its active component, norswertianin, induce autophagy and ROS-dependent differentiation of the human glioblastoma cell line [156,157]. Similarly, curcumin [158,159] can significantly induce differentiation of gliomainitiating cells (GICs) in vivo and in vitro through the induction of autophagy. Thus, curcumin can be used as an adjuvant to conventional chemotherapy in the treatment of GBM [160]. Falcarindiol (FAD) is a polyacetylene found in many dietary plants and has shown a variety of physiological activities, such as anti-inflammatory, antibacterial and hepatotoxic inhibition. FAD has an anticancer effect on glioblastoma cells by inducing the differentiation of glioblastoma stem cells and activating the apoptosis pathway of glioblastoma cells [161].

Discussion
One of the most important characteristics of cancer is the loss of differentiation; therefore, cancer is considered to be a disease of cell differentiation. Differentiation-related preclinical models were developed as early as the 1980s [162]. The concept of inducing cancer cell differentiation and cell proliferation arrest has become an alternative to cytotoxic chemotherapy. The purpose of this therapy is not to eliminate cells through cytotoxic and non-selective drug activity but to regulate the expression of signaling pathways and specific genes to guide cancer cells into higher stages of differentiation and reverse growth/differentiation, thereby reprogramming malignant cells into functional cells with different subtoxic doses.

Discussion
One of the most important characteristics of cancer is the loss of differentiation; therefore, cancer is considered to be a disease of cell differentiation. Differentiation-related preclinical models were developed as early as the 1980s [162]. The concept of inducing cancer cell differentiation and cell proliferation arrest has become an alternative to cytotoxic chemotherapy. The purpose of this therapy is not to eliminate cells through cytotoxic and non-selective drug activity but to regulate the expression of signaling pathways and specific genes to guide cancer cells into higher stages of differentiation and reverse growth/differentiation, thereby reprogramming malignant cells into functional cells with different subtoxic doses.

Discussion
One of the most important characteristics of cancer is the loss of differentiation; therefore, cancer is considered to be a disease of cell differentiation. Differentiation-related preclinical models were developed as early as the 1980s [162]. The concept of inducing cancer cell differentiation and cell proliferation arrest has become an alternative to cytotoxic chemotherapy. The purpose of this therapy is not to eliminate cells through cytotoxic and non-selective drug activity but to regulate the expression of signaling pathways and specific genes to guide cancer cells into higher stages of differentiation and reverse growth/differentiation, thereby reprogramming malignant cells into functional cells with different subtoxic doses.

Discussion
One of the most important characteristics of cancer is the loss of differentiation; therefore, cancer is considered to be a disease of cell differentiation. Differentiation-related preclinical models were developed as early as the 1980s [162]. The concept of inducing cancer cell differentiation and cell proliferation arrest has become an alternative to cytotoxic chemotherapy. The purpose of this therapy is not to eliminate cells through cytotoxic and non-selective drug activity but to regulate the expression of signaling pathways and specific genes to guide cancer cells into higher stages of differentiation and reverse growth/differentiation, thereby reprogramming malignant cells into functional cells with different subtoxic doses.

Discussion
One of the most important characteristics of cancer is the loss of differentiation; therefore, cancer is considered to be a disease of cell differentiation. Differentiation-related preclinical models were developed as early as the 1980s [162]. The concept of inducing cancer cell differentiation and cell proliferation arrest has become an alternative to cytotoxic chemotherapy. The purpose of this therapy is not to eliminate cells through cytotoxic and non-selective drug activity but to regulate the expression of signaling pathways and specific genes to guide cancer cells into higher stages of differentiation and reverse growth/differentiation, thereby reprogramming malignant cells into functional cells with different subtoxic doses.
In recent years, a large number of natural compounds have been reported, with good activity in tumor prevention and treatment, inducing tumor cell differentiation and apoptosis through a variety of mechanisms and inhibiting tumor cell cycle and proliferation ( Figure 2). The mechanisms by which natural antitumor compounds work against tumors vary, but in general, they directly induce the expression of proteins associated with tumor cell differentiation or differentiation of cells with blocked differentiation, leading to antitumor effects. The natural small-molecule compounds used for the treatment of AML, such as securinine, CAE, DADS, etc., induce apoptosis and differentiation by inducing intracellular oxidative stress. In addition, 20(s)-Rh2, CTD, DT-13 and notopterol inhibit tumor cell proliferation and induce apoptosis and differentiation through cell cycle arrest. Compared to AML, natural small-molecule compounds exhibit different molecular mechanisms in CML cells, especially sensitivity to erythroid differentiation. Fagaronine, apigenin, and wogonin induce erythroid differentiation or indirectly inhibit BCL-ABL, activating the key erythroid factor GATA-1 and increasing the expression of erythroid differentiation genes. There are also many natural small-molecule compounds, such as GA, andro, 8-OHD and ATL-I, which play an antitumor role by inducing apoptosis and differentiation of CML cells. In MM disease, the accumulation of clonal malignant plasma cells in the bone marrow reduces osteoblast formation and stimulates osteoclasts to destroy bone. RSV, ACE, silibinin and baicalein have been shown to enhance osteoblast differentiation of human bone marrow stromal cells by regulating the expression and activity of specific markers, including bone morphogenetic proteins (BMPs) and other key transcription factors. Induction of osteoblast differentiation can therefore be used as an alternative therapy to MM treatment. Meanwhile, induction of osteoblast differentiation is also an effective treatment for osteosarcoma. In the treatment of osteosarcoma, several flavonoids, including ugonin K, icaritin, genistein and KFL, promote osteoblast differentiation through their estrogenic effects, whereas galangin and hyperoside induce osteogenic differentiation by targeting TGF-β1. Furthermore, hyperoside and vitamin D3 induce osteogenic differentiation by inhibiting the entire cell cycle process, inducing apoptosis in osteosarcoma cells and driving the cells to a more differentiated phenotype. A significant proportion of natural small-molecule compounds that induce differentiation of tumor cells are also found in other types of tumors. Studies have shown that KFL, genistein and DIM induce apoptosis and differentiation of malignant melanoma cells mediated by reactive oxygen species and endoplasmic reticulum stress through their inhibitory effects on melanoma cell viability. In addition, theophylline and ISL induce B16 melanoma cell maturation and differentiation. CA, KFL, melatonin and TCE promote neuroblastoma cell differentiation, and norswertianin, FAD and curcumin promote GBM cell differentiation. Among the 36 natural antitumor compounds mentioned in this article, 25 have in vivo anticancer effects, 15 are in clinical trials and one is FDA-approved for use. In general, although many methods and techniques such as surgery, radiotherapy, chemotherapy, targeted therapy, immunotherapy and Chinese medicine therapy have been used for tumor treatment, most patients still face problems, such as disease recurrence and drug resistance. The development and research of natural small-molecule compounds provide a good solution for prolonging the remission period and improving the survival rate. Natural compounds are an important resource for modern drug development. From 1981 to 2019, 59% of newly approved small-molecule drugs (1123 in total) were derived from natural compounds and their derivatives [163,164]. Natural-molecule medicine and its preparations used under the guidance of modern medical theory have their unique advantages in treatment. Firstly, natural drugs come from nature, with wide sources and few toxic side effects. The material basis and in vivo mechanism of action of natural drugs are relatively clear, quality-controlled and safe, and natural drugs have shown high antitumor activity [165]. In addition, under the guidance of modern drug theory, natural drugs modified and transformed in structure have relatively concentrated targets, enhanced efficacy and richer forms of clinical application. For example, the combined use of modified and transformed natural molecule medicine with conventional anticancer drugs can provide relevant advantages of therapeutic efficacy by sensitizing malignant cells to drugs and overcoming drug-induced resistance in cancer.
Nevertheless, natural small molecules have several shortcomings. In many preclinical studies, pharmacokinetics and pharmacodynamics have not been well addressed, especially when natural compounds are combined with conventional drug therapy. This is because the heterogeneity and often combinatorial nature of natural compounds make it difficult to distinguish between the cause and the exact mechanism of action and exact information on the target in the numerous experimental studies, which once hampered research into the development of natural antitumor drugs. Most phytochemicals exhibit In general, although many methods and techniques such as surgery, radiotherapy, chemotherapy, targeted therapy, immunotherapy and Chinese medicine therapy have been used for tumor treatment, most patients still face problems, such as disease recurrence and drug resistance. The development and research of natural small-molecule compounds provide a good solution for prolonging the remission period and improving the survival rate. Natural compounds are an important resource for modern drug development. From 1981 to 2019, 59% of newly approved small-molecule drugs (1123 in total) were derived from natural compounds and their derivatives [163,164]. Natural-molecule medicine and its preparations used under the guidance of modern medical theory have their unique advantages in treatment. Firstly, natural drugs come from nature, with wide sources and few toxic side effects. The material basis and in vivo mechanism of action of natural drugs are relatively clear, quality-controlled and safe, and natural drugs have shown high antitumor activity [165]. In addition, under the guidance of modern drug theory, natural drugs modified and transformed in structure have relatively concentrated targets, enhanced efficacy and richer forms of clinical application. For example, the combined use of modified and transformed natural molecule medicine with conventional anticancer drugs can provide relevant advantages of therapeutic efficacy by sensitizing malignant cells to drugs and overcoming drug-induced resistance in cancer.
Nevertheless, natural small molecules have several shortcomings. In many preclinical studies, pharmacokinetics and pharmacodynamics have not been well addressed, especially when natural compounds are combined with conventional drug therapy. This is because the heterogeneity and often combinatorial nature of natural compounds make it difficult to distinguish between the cause and the exact mechanism of action and exact information on the target in the numerous experimental studies, which once hampered research into the development of natural antitumor drugs. Most phytochemicals exhibit poor stability in their natural form. Some natural small-molecule compounds may also cause side effects, including headache, dizziness, nausea, diarrhea and liver dysfunction. The most important barrier to the clinical use of natural small molecules is that most natural small molecules are fat-soluble, and their low solubility in water and poor internal absorption and rapid metabolism contribute to their low bioavailability. Despite this, many scientists are still actively involved in the development and research of natural small-molecule drugs and have provided some current and future strategies for the development of effective and safe natural antitumor drugs.

1.
Using medicinal chemistry, natural small-molecule compounds can be structurally modified and optimized to obtain new chemical entities with high drug resistance and low toxic side effects.

2.
Using computer simulations of the drug screening process to predict the likely activity of compounds allows multiple biologically relevant pathways to be probed in a targetagnostic manner, which allows for targeted screening and greatly reduces the number of compounds to be experimentally screened, thereby shortening development cycles and saving money. 3.
The development of innovative target-prediction tools to help identify biomolecular targets or potential off-target effects of drugs may help to determine the biological activity of natural products and guide the biochemical screening of natural products to reduce the number of experiments and save valuable resources.

4.
Establishing sound testing methods and experimental approaches to promote drug development, improving methods for purification of active ingredients in natural compounds and extracting key components for subsequent experiments.

5.
The search for or development of lead compounds for novel drugs with improved pharmacokinetic profiles through the study of natural product molecules with specific active backbones, active moieties and excellent biological activity. 6.
Diverse and well-established animal or neural models are established for screening, development and validation.
Although much research has been conducted to treat cancer, the incidence and mortality of cancer continue to rise, and effective treatment remains a formidable challenge. The drugs currently used for treatment have very strong adverse side effects, such as vomiting, bone marrow suppression and liver dysfunction. Inducing the differentiation of malignant cells into functional cells is an effective alternative approach to treating tumors, where natural small-molecule compounds play an important role in inducing tumor cell differentiation due to their wide availability and low toxicity. However, problems encountered in clinical studies need to be addressed to improve the use of natural compounds in effective cancer treatment. Overall, natural small-molecule compounds, as new anticancer properties and mechanisms continue to be revealed, hold outstanding promise for future cancer differentiation therapy.

Conclusions
In this paper, we summarized various natural small-molecule compounds inducing differentiation in myeloid leukemia cells, osteoblasts and some solid tumor cells. This may promote the research of differentiation therapy, improve therapeutic efficiency and reduce drug resistance and cytotoxic side effects, etc.

Conflicts of Interest:
The authors declare no conflict of interest.