Within the cell, ASNS catalyzes the synthesis of Asn in an ATP-dependent reaction, using glutamine (Gln) and Asp as substrates. This intracellular production of Asn is somewhat dependent on the functioning of the mitochondrial electron transport chain (ETC) as inhibition of the ETC using metformin impedes the conversion of Asp from Gln, and therefore affects de novo synthesis of Asn (Davidsen and Sullivan, 2020; Krall et al., 2021). ASNS-catalyzed synthesis of Asn is also dependent on somatic mutation patterns within the cancer cells, since ASNS expression is regulated by KRAS, SOX12 and TP53, which we discuss further here.

The proteasome is essential in tumor cells for maintaining the decomposition of intracellular proteins such as p53 that aids DNA repair, and p21 and p27 that impede cancer cell growth, resulting in the promotion of tumor cell survival and proliferation (Almond and Cohen, 2002). Consequently, proteasomal inhibition is a major therapeutic target for cancer chemotherapy (Almond and Cohen, 2002). However, proteasome inhibitors are typically accompanied by resistance, which could be addressed by targeting upstream regulators (Manasanch and Orlowski, 2017). One example of this is restricting the actions of the upstream regulator glycogen synthase kinase 3 (GSK3). GSK3-dependent protein ubiquitination and consequent destruction at the proteasome can generate sources of free Asn. The actions of GSK3 can be restricted by WNT-dependent stabilization of proteins (WNT/STOP) to mitigate resistance to ASNase in CRC (Hinze et al., 2020). When WNT/STOP is off, GSK3-dependent protein degradation remains unperturbed and produces catabolic free Asn which induces ASNase resistance in CRC; whereas CRCs with R-spondin (RSPO) fusions such as RSPO3 have activated WNT/STOP which inhibits the proteasome, limiting Asn availability and thus increasing vulnerability to ASNase (Hinze et al., 2020). Moreover, tumor cells that have KRAS/TP53 mutations, are forced to survive on Asn recycled from protein degradation in the proteasome as autophagy is inhibited and not available as an alternative protein recycling pathway (Hinze et al., 2020).

As described above, tumor cells are sensitized to Asn depletion through actions of ASNase which subsequently increases Asn levels through proteasomal release (Davidsen and Sullivan, 2020). Under normal cellular quiescent conditions, WNT signaling is silenced through proteasomal degradation of β-catenin mediated by the APC/Axin/GSK-3 destruction complex (Fearon, 2011). However, tumor formation, mediated by elevated WNT signaling leads to the accumulation of cytosolic and nuclear β-catenin which promotes cell proliferation. WNT activity, therefore, prevents proteasomal degradation, an event that leads to the recycling of protein into free Asn and consequently survival of tumor cells (Davidsen and Sullivan, 2020).

Therefore, proteasomal degradation that recycles protein into Asn is associated with ASNase resistance. Activation of WNT/STOP signaling may improve the therapeutic effect of ASNase in CRC.

Dietary intake of foods high in Asn, and dietary patterns that might influence ASNS expression mediate free Asn levels. In this context, other amino acids that are involved in Asn synthesis and are acquired exogenously may also serve as indirect regulators of Asn bioavailability. However, it is difficult to investigate the exact dietary sources and consumption of Asn (and other amino acids), due to the lack of sufficient and reliable nutrient databases that quantify micronutrients such as Asn in foods. Most existing nutrient databases do not provide information specifically on Asn (Lenders et al., 2009). This may be due to the difficultly in assessing levels of Asn in foods via common laboratory methods. Since most of the common methods of assessing dietary intake in populations, such as food frequency questionnaires (FFQs), rely on nutrient databases, most large epidemiologic studies cannot quantify Asn levels within the diet. Therefore, to our knowledge, the direct association between self-reported dietary intake of Asn and CRC has not been studied.

An alternative approach would be to examine Asn abundance in foods that are observed to be associated with the risk of CRC. For example, the consumption of red meat is associated with a higher risk of CRC development (Thanikachalam and Khan, 2019), and short-term red meat consumption is positively associated with higher serum levels of Asn (Guan et al., 2021), however, data on Asn levels in red meat compared to other foods is not available. Even though Asn may be higher in the serum of those consuming red meat, the role of dietary Asn from red meat in CRC risk would be difficult to tease apart as there are other components of red meat that are linked to CRC development such as N-nitroso compounds, and those that are formed when red meat is cooked at high temperatures such as heterocyclic amines and polycyclic aromatic hydrocarbons (Aykan, 2015). It is still too premature to conclude a link between red meat, Asn levels, and CRC, especially since vegetables which are often associated with reduced risk of CRC are also sources of Asn and would thus confound these associations.

Food additives could also be linked to cancer. A recent study hypothesized that consumption of animals fed with ractopamine might be pro-tumoral for CRC patients as ractopamine could increase ASNS expression in the skeletal muscle of pigs (Fan, 2022).

Some dietary patterns, such as the vegan diet (not consuming any animal products), and vegetarian diets (not consuming any meat or fish) (Clarys et al., 2014), have been shown to increase levels of intestinal ASNase (assessed in fecal samples) in individuals consuming these diets as reported via FFQs (De Angelis et al., 2020). Whereas other diets which have dietary protein restriction, or an imbalanced dietary amino acid composition have shown that the ASNS gene can be activated through amino acid response (AAR) that limits single essential amino acids (Balasubramanian et al., 2013).

The consumption of specific dietary components has been shown to modulate proliferation, gene expression, and metabolism of gut microorganisms, leading to changes in protein synthesis (Figure 1), which could further affect Asn levels. For example, dietary consumption of sourdough bread might bring a benefit to host health by increasing healthy gut microorganisms in the colon (Da Ros et al., 2021), such as boosting the growth of Bacteroides that express ASNase (Abbondio et al., 2019).

Overall, with regards to dietary sources of Asn, laboratory studies that accurately assess Asn levels in foods are needed, so that Asn can be accurately added to nutrient databases. If dietary Asn could be accurately quantified, then associations between dietary Asn and CRC risk could be evaluated in existing and future large epidemiologic populations.

Experimental models have shown that limiting dietary intake of Asn can decrease cancer growth. A study found that decreasing Asn in chow that was fed to mice with a breast cancer xenograft had a protective effect on metastatic cancer progression. Furthermore, enhancing Asn in the diet resulted in more invasive tumors (Knott et al., 2018). Dietary restriction to exogenous non-essential amino acids, including Asn, also hindered the growth of lung cancers that harbored Kelch-like ECH-associated protein 1 (KEAP1)/NRF2 mutations in vivo, wherein the oxidative stress response was activated (LeBoeuf et al., 2020). In a wider variety of cancer types, inhibition of the ETC was shown to impair mitochondrial respiration by reducing ATP and Asp levels, thus limiting Asn synthesis. With ETC impairment, sufficient exogenous Asn supply rescued ATF4 protein levels and ASNS expression, and restored mTORC1 activation; restricted Asn dietary intake together with ETC inhibition further suppressed proliferation (Krall et al., 2021).

There are trillions of microbiota residing in the human gut that play a critical role in health (Lozupone et al., 2012). These microbes are known to aid in the production of metabolites including short-chain fatty acids, vitamins, indoles and amino acids that ultimately impact the host metabolism (Sharon et al., 2014). The gut microbiome has been associated with cancer progression, and alterations in the diversity of gut microbiome can also affect the course of the treatment (Gopalakrishnan et al., 2018). The gut microbiome has been shown to mediate ASNS and ASNase activity which can alter Asn levels, and could also ultimately impact the efficacy of L-ASNase therapy (Amer et al., 2013; Hillman et al., 2020; Dunn et al., 2021). The efficacy of ASNase treatment is determined by its ability to maintain the depleted levels of Asn for a prolonged period of time (Pieters et al., 2011; Covini et al., 2012). Dunn et al. reported that pediatric ALL patients treated with ASNase therapy showed distinctive gut microbial composition that correlated with serum ASNase activity (Dunn et al., 2021). Patients with decreased serum ASNase activity had enrichment of Escherichia, whereas those with increased serum ASNase activity has an enrichment of Bacteroides and Streptococcus (Dunn et al., 2021).

Microbial species express the ASNS gene and two ASNases, ansA and ansB. Therefore, the differential taxonomic distribution of these genes could underlie the major reasons for such kind of microbial community distinction based on increased or decreased serum ASNase activity. For example, it had been found that E. coli contains all three genes in its genome, Blautia [Ruminococcus] gnavus and Bacteroides vulgatus contains ASNS and ansB, while Bacteroides ovatus, Bacteroides stercoris, Bacteroides uniformis, Faecalibacterium prausnitzii, Roseburia intestinalis contains ASNS and ansA (Dunn et al., 2021). There are some microbial species that contain only one of these genes. For example, Bifidobacterium longum, Collinsella aerofaciens contains only ansA, Blautia obeum, Blautia torques contains only ASNS, while Clostridium bartlettii, Veillonella parvula contains ansB (Dunn et al., 2021). There has been a limited number of studies that showed the association between microbiome and ASNase treatment in cancer. Therefore, future studies to establish the predictive microbial biomarkers that could potentially impact ASNase levels during treatment are warranted. The manipulation of these microbial species and their associated genes (ASNS, ansA and ansB) could aid in optimizing the patient’s response to ASNase therapy.

Han et al. designed a systematic metabolomics-based pipeline for studying the functional microbiome and discovered that Asn and Gln can be used as sole nitrogen sources by various bacteria such as Bacteroidetes (Han et al., 2021). When grown on media supplemented with Asn, Bacteroidetes taxa growth was strongly correlated with an L-asparaginase II homologue (ansB; >59% identity), which was essential for growth when Asn was the only nitrogen source (Han et al., 2021). Using their data explorer pipeline, levels of Asn in feces and cecal contents was significantly lower in Bacteroides thetaiotaomicron (Bt.) inoculated mice compared to germ-free control (Han, 2020). It was shown that some microbes that were grown in mega media had decreased levels of Asn compared to media controls (i.e. with no bacteria), such as E. coli and Bacteroides species including Bt. and Bacteroides caccae, Bifidobacterium dentium, and Blautia producta., indicating that these bacteria used Asn as a growth source, and thus could affect Asn levels in the gut and perhaps of CRC patients. Whereas there are side effects associated with ASNase administration; E. coli ASNase can bind and inhibit the concanavalin A receptor, which accounts for non-immunological side effects and toxicities by ASNase (Van Trimpont et al., 2022). Moreover, the Rhodospirillum rubrum mutant L-ASNase is unique to other bacterial ASNases as it can penetrate cancer cells via a clathrin receptor-dependent manner that suppressed telomerase activity and led to asparagine-independent cytotoxicity (Zhdanov et al., 2017).

As previously discussed, Asn can be generated through the breakdown of proteins in the proteasome which is regulated by β-catenin. Sex-specific differences in the expression of voltage-gated K+ (Kv) channels such as those in the KCNQ family (KCNQ1 and KCNE3) can affect this process; the sex-steroid hormone 17β-estradiol inhibits the KCNQ1:KCNE3 channel complex, resulting in a relatively higher proportion of free homomeric KCNQ1 that inhibits WNT/β-catenin signaling pathway (Abancens et al., 2020). Therefore, estrogens regulate Kv channel expression and might indirectly decrease Asn levels by inhibiting WNT/β-catenin signaling, this would thus associate with a better outcome.

The hormone estradiol has been hypothesized to decrease the overall incidence of CRC. Males produce a basal level of estradiol (10–40 pg/ml), which is at a 10-fold lower concentration than in females (Icahn School of Medicine at Mount Sinai, 2021). The protective effects of estradiol are through activation of the nuclear estrogen receptor (ERβ) in the colon mucosal epithelium. Activation of ERβ causes pro-apoptotic signaling, inhibition of inflammatory signals, and modulation of the tumor microenvironment, including inactivation of IL-6 and increased DNA repair capacity (Edvardsson et al., 2011). ERβ expression is significantly reduced in tumor tissues compared to normal tissues, and in advanced stages of colon cancer, ERβ expression can be completely eliminated. ERα is not expressed in the colonic epithelium, therefore understanding the protective role of ERβ is of primary importance in CRC (Caiazza et al., 2015).

Loss of ERβ expression in CRC has been associated with more advanced stages of cancer and poorer prognosis compared to patients with high ERβ expression (Rudolph et al., 2012). In CRC cells, ERβ could upregulate PTEN and downregulate PI3K/AKT/mTOR signaling pathways, thus promoting autophagy (Wei et al., 2019). In Figure 2, we propose that an inverse relationship exists between ERβ and PTEN/PI3K/AKT which modulates ASNS expression and influences CRC progression; PTEN can inhibit PI3K/AKT signaling, and loss of PTEN expression co-occurs with KRAS and PIK3CA mutations (De Roock et al., 2011), enhanced PI3K/AKT signaling can induce ERβ ubiquitination and degradation (Zhou and Liu, 2020). We propose that when ERβ expression is lost, ASNS expression increases, driving Asn production (Figure 2). Activation of PI3K/AKT signaling by mutant KRAS can cause ERβ degradation and disables its tumor suppressing effect. When estrogen levels are low, there is decreased activation of ERβ, therefore PTEN can no longer inhibit PI3K/AKT. As activation of PI3K/AKT induces ASNS expression, this cascade causes increased Asn synthesis, which predisposes female patients to a poorer prognosis.

The GPER is a membrane estrogen receptor encoded by the GPER1 gene, which exercises the functions of an estrogen receptor in the absence of nuclear ERs (Gaudet et al., 2015). In CRC, GPER differentially regulates the effects of 17β-estradiol on CRC cell proliferation depending on the tumor microenvironment: 17β-estradiol and a GPER agonist suppressed migration and proliferation in a normoxic tumor microenvironment, while they enhanced migration and proliferation under hypoxia, and these effects were lost in GPER-silenced cells (Bustos et al., 2017). Hypoxia is a decrease in the normal level of tissue oxygen tension (Muz et al., 2015) and typically occurs in advanced cancers (Petrova et al., 2018). Cells respond to hypoxic conditions through hypoxia-inducible transcription factor 1 (HIF-1) which activates the expression of the pro-angiogenic growth factor vascular endothelial growth factor (VEGF), a hypoxia-response gene (Harris, 2002). Specifically in CRC, the expression of HIF-1α and VEGFA (an important angiogenesis regulator in VEGF family) were associated with poorer CRC prognosis (Cao et al., 2009). Under normoxia, 17β-estradiol suppressed VEGFA expression, and its expression was lost in GPER-silenced cells, whereas 17β-estradiol increased VEGFA expression (Bustos et al., 2017).

Ataxia telangiectasia mutated (ATM) is a key DNA damage repair gene, and loss of ATM expression was associated with adverse prognosis in CRC (Beggs et al., 2012). ATM expression was decreased in both hypoxia and normoxia environments via GPER signaling (Bustos et al., 2017). These findings suggest a synergistic effect of estrogens and hypoxia in ERβ-negative CRC that represses ATM expression to promote cancer cell proliferation and migration, which is induced by hypoxia via GPER signaling, activating HIF-1α and VEGFA (Bustos et al., 2017). Furthermore, a sex and stage-specific association has been observed, in which GPER expression was only associated with poor survival in CRC stages III-IV females but not in stage-matched males (Bustos et al., 2017). Lack of ERβ was associated with higher cancer stages of CRC and greater tumor extent (Rudolph et al., 2012). Thus, it might be assumed that GPER could be the primary estrogenic signaling pathway if ERβ expression is low or lost at later stages in females, resulting in poorer clinical outcomes than in earlier stages.

There are links between GPER signaling and downstream targets of KRAS (Chaturantabut et al., 2019; Xu et al., 2020; Shi et al., 2022), wherein activation of GPER increases PI3K/AKT via cyclic adenosine monophosphate (cAMP) (Yu et al., 2017; Chuang et al., 2020). Increased expression of genes linked to cAMP was also observed in CRC tumors from female patients (Sun et al., 2020). PI3K/AKT is a downstream target of KRAS which regulates ASNS via NRF2 and ATF4. Thus, GPER is linked to ASNS regulation via its regulatory effects on PI3K/AKT through cAMP (Figure 2).

Therefore, sex interacts with Asn synthesis and CRC through GPER activation; nutrient supply influences the function of estrogen and GPER1 in KRAS-mutant CRC by upregulating ASNS expression, Asn synthesis, PI3K/AKT and HIF-1α-mediated SOX12 overexpression (Figure 2). Due to the multifaceted roles of nuclear receptors in cellular homeostasis, the links between GPER1 signaling and ASNS expression require further mechanistic investigation.

5′AMP-activated protein kinase (AMPK) and mTORC1 play important roles in nutrient signaling and balancing cellular energy homeostasis by sensing cellular ATP and nutrients such as amino acids (Chun and Kim, 2021). We previously discovered downregulation of AMPK alpha subunit (AMPKα) and enriched mTORC1 signaling pathway in females with right-sided colon cancer (RCC) compared to left-sided colon cancer under a nutrient deplete environment (Sun et al., 2020). We also observed enhanced Asn synthesis and amino acids uptake in RCC tissues from females (Cai et al., 2020), which was not observed in RCC tissues from males. Asn and Gln can promote mTORC1 activity (Meng et al., 2020). Moreover, Gln can activate mTORC1 at the surface of the lysosome through the ASNS-mediated production of ATP by inhibiting AMPK (Bodineau et al., 2021). Therefore, enhancement of de novo Asn synthesis that upregulates mTORC1 by downregulating AMPK to promote CRC growth is hypothesized in tumors from females with RCC.

Diabetes mellitus and CRC have similar deregulation of cellular and molecular pathways in their disease pathogenesis, such as PI3K/AKT/mTOR signaling, WNT/β-catenin pathway, and the AMPK pathway (Yu et al., 2022). Both diseases also have common risk factors including obesity, inflammation, and poor diets (Gonzalez et al., 2017). Metformin, a biguanide compound commonly used for treating diabetes, has shown great potential in CRC treatment (Higurashi and Nakajima, 2018), as suggested by its capability to activate AMPK and inhibit TGF-β/PI3K/AKT signaling transduction (Xiao et al., 2022). Moreover, by functioning as a type of ETC inhibitor that impairs mitochondrial respiration to cause depletion of TCA-cycle-derived Asp, metformin in combination with ASNase or dietary Asn restriction is able to halt tumor growth by limiting de novo Asn synthesis (Krall et al., 2021) (Figure 3). Therefore, metformin plays a comprehensive role in Asn metabolism by regulating its upstream substrate Asp, downstream mTORC1 inhibition (Bordini et al., 2017), and AMPK activation which induces p53 phosphorylation and activation (Yi et al., 2019) (Figure 3). Furthermore, compared with KRAS wild-type metastatic CRC (mCRC), KRAS-mutant mCRC had a greater benefit from metformin administration which inactivated AKT/mTOR signaling and cell proliferation (Xie et al., 2020).

Studies have shown that there are possible sex-related factors that could alter the outcomes of metformin treatment on diabetic CRC. One study using two large cohorts discovered that type 2 diabetes (T2D) was associated with higher CRC risk in males, but not in females (Ma et al., 2018). Diabetic female CRC patients who took metformin had significantly lower CRC-specific mortality than males, showing a sex-specific difference in the effect of metformin on diabetic CRC patients (Park et al., 2017). Such sex differences could be attributable to sex hormone levels, as metformin usage had no significant impact on CRC-specific survival in postmenopausal female CRC patients with diabetes (Cossor et al., 2013). In human endometrial cancer cells, metformin increased ERβ expression and AMPK phosphorylation (Zhang et al., 2017). As estrogens bind to the ERs to exert their functions (Fuentes and Silveyra, 2019), and postmenopausal and diabetic female CRC patients have lower levels of estrogens, it is possible that this underlies the lack of benefit from metformin in those women with CRC (Figure 3). However, the role of metformin with sex interactions in CRC warrants further investigation.