Acute leukemia is characterized by neoplastic proliferation of immature white blood cells, also called blasts, in the bone marrow (BM), which later rapidly disseminate to the blood and other tissues (Colmone et al., 2008; Vardiman et al., 2009; Glait-Santar et al., 2015). Based on the lineage of affected white blood cells, acute leukemia is classified into acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL) (Marks et al., 2009; Lyengar and Shimanovsky, 2021). AML is more common in adults and accounts for about 80% of all AML cases. ALL predominantly occurs in children; it comprises about 80% of childhood and 20% of adult ALLs (Lyengar and Shimanovsky, 2021).

Conventional chemotherapy has successfully decreased the mortality rate of patients with acute leukemia (Marks et al., 2009; Rowe and Tallman, 2010; Rubnitz, 2017). However, patients still suffer from refractory disease or relapse, signifying the need for the development of more effective therapies (Hunger et al., 2012; Winter et al., 2018). One of the devastating features of leukemia is the ability of leukemic cells to colonize at secondary sites for tumorigenesis (Valastyan and Weinberg, 2011). Particularly, infiltration of the central nervous system (CNS) by leukemic cells contributes to an increase in leukemia mortality and treatment failure (Basu et al., 2014; Si et al., 2018).

Growing evidence suggests the crucial role of intrinsic and extrinsic factors in modulating leukemic cell survival (Silva et al., 2011; Giambra et al., 2015; Pitt et al., 2015; Moharram et al., 2017; Ribeiro et al., 2018; van der Zwet et al., 2021). Leukemic cells remain in a quiescent state and highly depend on intrinsic survival factors while circulating in the blood (Guan et al., 2003). When leukemic cells enter the homing tissue, the microenvironmental niche provides multiple signaling ques supporting leukemia survival (Ninomiya et al., 2007). Metabolic reprogramming provides cancer cells with a unique flexibility in adapting to a variety of cell-extrinsic and -intrinsic stimuli. These metabolic adaptations govern tumor transformation, proliferation, invasiveness, and resistance to therapy (Hanahan and Weinberg, 2011; DeBerardinis and Chandel, 2016). Otto Warburg was the first to recognize the aberrant metabolic behavior of tumor cells. He postulated that cancer tissues have higher levels of glucose uptake compared to normal tissues and that cancer cells rely primarily on aerobic glycolysis to produce adenosine triphosphate (ATP) (Warburg, 1925; Warburg, 1956a). Understanding the mechanism underlying metabolic reprogramming of cancer cells could provide a venue for defining novel therapeutic targets (Wolpaw and Dang, 2018).

Acute leukemias commonly display an increase in glucose uptake and aerobic glycolysis (Suganuma et al., 2010; Kishton et al., 2016; Matthijssens et al., 2021). In addition, enhanced mitochondrial respiration (OXPHOS) increases the reactive oxygen species (ROS) levels in leukemic cells (Han et al., 2019). However, leukemic cells can compensate for the harmful effects of elevated ROS levels by enhancing the expression of antioxidants, which ultimately restore redox homeostasis (Sabharwal and Schumacker, 2014; Khan et al., 2016). Importantly, under energy crisis conditions, leukemic cells rely on non-glycolytic resources (Lee et al., 2013) such as fatty acid oxidation, amino acid oxidation (e.g., methionine, cysteine), and glutaminolysis, which all provide essential intermediates to maintain the Krebs cycle (Tabe et al., 2020). In line, growing evidence suggests that major oncogenic drivers, such as PI3K/Akt/mTOR pathway, MYC, FLT3, and RAS, contribute to metabolic rewiring in leukemic cells (Herranz et al., 2015; Rashkovan and Ferrando, 2019).

Cell adhesion plays a key role in cancer progression and metastasis. Adhesion molecules regulate cancer cell survival, differentiation, proliferation, inflammation, and migration. Alterations in cell-cell and cell-matrix adhesion allow malignant cells to increase their motility and degrade the cell-extracellular matrix (ECM) to enter the blood circulation, followed by dissemination to distant sites (Martin et al., 2013). During this multi-step process, cancer cells induce metabolic rewiring to meet distinct metabolic demands (Nepstad et al., 2018; Wei et al., 2020). Recent studies demonstrated that cancer cell adhesion may either induce or be induced by cell signaling pathways associated with metabolic reprogramming (Sousa et al., 2019). In line, several adhesion molecules were identified as critical regulators of leukemia development and chemoresistance (Jacamo et al., 2014; Fonseca et al., 2018; Scharff et al., 2020a; Gutjahr et al., 2021).

The impact of specific genetic lesions (e.g., MLL rearrangements, BCR-ABL), CNS niche, chemokines, cytokines, and growth factors in driving leukemic cells to the CNS and meninges has been extensively reviewed in (Heidari et al., 2016; Gossai and Gordon, 2017; Piovan et al., 2018; Zhou et al., 2019; Lenk et al., 2020; Whiteley et al., 2021). The roles of metabolic reprogramming and cell adhesion in leukemia development and progression have also been discussed elsewhere (Heath et al., 2019; Rashkovan and Ferrando, 2019; Windisch et al., 2019; Härzschel et al., 2020; Scharff et al., 2020a; Di Martino et al., 2021). However, how metabolic reprograming and cell adhesion regulate leukemic cell infiltration to the CNS remains unclear. In this mini review we focus on recent advances toward our understanding of the roles played by metabolic reprogramming and cell adhesion in acute leukemia (ALL, AML) colonization into the CNS.

The CNS is a common extramedullary site for infiltrating ALL cells (Lazarus et al., 2006). CNS involvement is detected either at initial treatment or at relapse. The incidence of CNS in ALL at diagnosis is approximately 5–10%. For ALL patients who have received prophylactic CNS directed chemotherapy, the recurrence of CNS disease is 7–15% (Alakel et al., 2017; Holland et al., 2003; Pinkel and Woo, 1994; Gökbuget and Hoelzer, 1998). Factors associated with CNS-ALL include high white blood cell (WBC) count, hypercellular marrow, and extramedullary infiltration (BLEYER, 1984; Cassileth et al., 1988; Pinkel and Woo, 1994; Cortes, 2001).

CNS infiltration in AML is relatively rare (Galati et al., 2013; Loeb et al., 2016). Approximately 0.6–5% of AML patients present with CNS disease at diagnosis, and 3–15% with CNS relapse (Kouser and Hashmi, 2007; Galati et al., 2013; Alakel et al., 2017). Risk factors involved in AML-CNS include elevated serum lactate dehydrogenase (LDH) levels, increased WBC count, subtypes of myelomonocytic/monoblastic/monocytic leukemias, inversion of chromosome 16, mutations in FLT3 and NPM1, expression of CD56 and 11q23 rearrangements (Holmes et al., 1985; Thompson et al., 1986; Chang et al., 2004; Cheng et al., 2015; Rozovski et al., 2015).

While cytospin-based evaluation of cerebrospinal fluid (CSF) is used for diagnosis of CNS disease/relapse in leukemia (Del Principe et al., 2014), this diagnostic method does not identify patients who have occult CNS involvement (Martínez-Laperche et al., 2013; Bartram et al., 2018; Del Principe et al., 2021). Furthermore, studies showed that CNS prophylaxis (cranial radiation and intrathecal chemotherapy) is associated with various neurological toxicities (Pochedly, 1977). Thus, it is imperative to understand the mechanism underlying CNS involvement to develop more accurate diagnostic tools and novel therapies that will eradicate leukemic cells from the CNS while causing less adverse neurotoxicity.

The CNS serves as a sanctuary site, in which leukemic cells evade the immune response and systemic chemotherapy (Frishman-Levy and Izraeli, 2017). Metabolic adaptation in the CNS niche is a prerequisite for the long-term survival of leukemic cells and the recurrence of the disease (Savino et al., 2020). Leukemic cells depend on cellular metabolism rewiring to survive in nutrient-poor and hypoxic microenvironments (Cairns et al., 2011; Cha and Lee, 2016; DeBerardinis and Chandel, 2016). Thus, metabolic vulnerabilities of leukemic cells could be used for therapeutic purposes (Kuntz et al., 2017; Nachmias and Schimmer, 2018). Despite recent advances in cancer metabolism research, little is known about whether and how cell metabolism affects the migration and adhesion of leukemic cells in the CNS. A better understanding of these metabolic adaptations will advance the development of novel treatment strategies. Below, we discuss major metabolic pathways and their roles in leukemia colonization of the CNS (Figure 1A).

Leukemic cells regulate the expression of adhesion molecules to confer a pro-survival advantage against chemotherapy and to increase their invasiveness to extramedullary sites (Erbani et al., 2020; Barbier et al., 2020; Wang et al., 2018; Gaynes et al., 2017; Akers et al., 2011). Growing evidence suggests that leukemic cells invade and colonize the leptomeningeal microenvironment through specific adhesion and homing mechanisms (Figure 1B). Co-culture of ALL cell lines and primary B- and T-ALL cells with meningeal cells enhanced leukemic cell survival compared to leukemic cells incubated in the CSF suggesting the importance of leukemic and meningeal cell-cell interactions in promoting ALL cell survival (Basile et al., 2020). In fact, Jonart et al., (2020), demonstrated that ALL cells adhere to meningeal cells and that cell-cell adhesion governs leukemic cell dormancy and resistance to chemotherapy. Importantly, disruption of the meningeal ALL adhesion with tMe6TREN (Tris [2-(dimethylamino)ethyl]amine) increased the efficacy of chemotherapy in the CNS in leukemia xenograft murine models (Jonart et al., 2020). Other studies demonstrated that B-ALL cells from children and B-ALL cell lines were adherent to astrocytes, choroid plexus fibroblasts, and epithelial cells, thus promoting leukemic cell survival and chemoresistance (Akers et al., 2011; Fernández-Sevilla et al., 2020).

Integrins and their ligands play a vital role in leukemic cells migration and homing through mediating cell-cell and cell-ECM adhesion (Scharff et al., 2020a). To date, few studies have identified specific adhesion molecules potentially associated with leukemia infiltration in the CNS/meninges. Increased expression of intercellular adhesion molecule 1 (ICAM-1) was correlated with CNS disease in pediatric ALL samples and B-ALL patient derived xenografts (Mielcarek et al., 1997; Holland et al., 2011). Furthermore, the expression of CD56, a neural cell-adhesion molecule (NCAM), was elevated in adult ALL samples with CNS involvement (Ravandi et al., 2002; Hu et al., 2017). NCAM was also associated with intracerebral and leptomeningeal infiltration in adult T cell leukemia (ATL) (Hashiguchi et al., 2002). Thus, NCAM was proposed as a marker for CNS infiltration and poor prognosis in ALL, and ATL. In addition, AML and ALL patients who had elevated levels of matrix metalloproteinase-9 (MMP-9) and soluble vascular cell adhesion molecule 1 (sVCAM-1) in CSF were at risk of CNS involvement (Si et al., 2015).

Elegant studies by Yao et al.(2018), demonstrated that B-ALL cells invade the CNS along emissary vessels passing between vertebral and calvarial BM, and the subarachnoid space. ALL cells expressed integrin subunit alpha 6 (α6), a laminin receptor, which interacted with laminin expressed on the bridging vessels, thus mediating the migration of ALL cells into the meninges (Yao et al., 2018). ALL xenografts treated with specific α6 integrin-neutralizing antibodies showed reduced leukemia burden in the CSF/meninges. High levels of Itga6 mRNA (encodes α6) were also found in leukemic cells in a BCR-ABL1-driven murine model of meningeal leukemia (Yu et al., 2019). On the contrary, recent studies identified ITGA5 (α5) and ITGA9 (α9) expression positively correlated with CSF colonization in primary B-ALL samples (Scharff et al., 2020a; Scharff et al., 2020b).

Functionally, co-culture of B-ALL Nalm6 cells with choroid plexus fibroblasts resulted in upregulation of VLA-4 and LFA-1 in leukemic cells concomitant with increased expression of relevant integrin ligands, VCAM1 and ICAM1, in the tested fibroblasts. The inhibition of VLA-4/VCAM-1 signaling with anti-VLA-4 antibodies sensitized co-cultured leukemic cells to chemotherapy (Fernández-Sevilla et al., 2020). Further studies are required to determine whether targeting VLA-4/VCAM-1 adhesion could be used to eradicate CNS involved leukemia. Other adherent junction proteins such as VE-cadherin and PECAM-1 increased the adhesion and migration of B-ALL cell lines through the human brain-derived microvascular endothelial cells but their role in CNS leukemia has yet to be determined (Akers et al., 2010). In another study, CCR7 was sufficient to drive leukemic cells to the CNS in T-ALL. Interestingly, gene expression analyses identified deregulation in genes encoding integrins and metalloproteases that could potentially interact with CCR7 function to support T-ALL invasion of the CNS (Buonamici et al., 2009). Moreover, recent studies showed that pediatric ALL-derived exosomes contributed to leukemic cell invasion in a model of the blood-cerebrospinal fluid barrier (BCSFB) in vitro. Interestingly, binding/uptake of ALL-derived exosomes was dependent on various exosomal integrins such as αV, α5, β1, and β3 (Erb et al., 2020).

CNS involvement has been emerging as a major challenge in acute leukemia treatment. Patients with CNS infiltration have a low survival rate, particularly those with recurrent or refractory disease. The interaction between leukemic cells and the CNS microenvironment promotes leukemic cell quiescence and subsequently the resistance to chemotherapy. To date, few studies have investigated the roles of chemokine receptors and other molecules in leukemia trafficking to the CNS (Buonamici et al., 2009; Williams et al., 2014; Jost et al., 2016; Williams et al., 2016; Alsadeq et al., 2017; Oruganti et al., 2017; Alsadeq et al., 2018; Prieto et al., 2018; Lenk et al., 2021). A growing body of evidence demonstrates that leukemic cells employ cell-cell and cell-ECM adhesion to reside and survive in the CNS/meninges. Furthermore, we speculate that metabolic reprogramming could provide leukemic cells with sufficient energy to facilitate the invasion and colonization of the nutrient-poor and hypoxic CNS microenvironment. Accordingly, targeting specific cell adhesion molecules and metabolic pathways could potentially increase treatment efficacy and reduce the toxicity of existing therapies. In fact, there are many drugs targeting metabolism or integrin-targeting drugs under clinical evaluation. However, those drugs have not yet been tested for the treatment of CNS involved leukemia. Moreover, the potential cross-talk between metabolic pathways and cell adhesion remains poorly understood, pointing to several questions: Are modifications in leukemic cell adhesion associated with specific energetic demands? Do cell adhesion mechanisms support metabolic rewiring of leukemic cells in the CNS niche? Do metabolic adaptations of leukemic cells lead to changes in cell-cell and cell-ECM adhesion within the CNS microenvironment? What is the role of meningeal fibroblasts and other cells within the CNS microenvironment in facilitating leukemia infiltration? Do they undergo further metabolic adaptations and/or express specific adhesion molecules to create a more permissive microenvironment for leukemia colonization? Further investigation is needed to uncover the specific adhesion molecules and metabolic adaptations underlying CNS disease and CNS relapse in leukemia.