Anti‐tumor effects of an antagonistic mAb against the ASCT2 amino acid transporter on KRAS‐mutated human colorectal cancer cells

Abstract KRAS mutations are detected in numerous human cancers, but there are few effective drugs for KRAS‐mutated cancers. Transporters for amino acids and glucose are highly expressed on cancer cells, possibly to maintain rapid cell growth and metabolism. Alanine‐serine‐cysteine transporter 2 (ASCT2) is a primary transporter for glutamine in cancer cells. In this study, we developed a novel monoclonal antibody (mAb) recognizing the extracellular domain of human ASCT2, and investigated whether ASCT2 can be a therapeutic target for KRAS‐mutated cancers. Rats were immunized with RH7777 rat hepatoma cells expressing human ASCT2 fused to green fluorescent protein (GFP). Splenocytes from the immunized rats were fused with P3X63Ag8.653 mouse myeloma cells, and selected and cloned hybridoma cells secreting Ab3‐8 mAb were established. This mAb reacted with RH7777 transfectants expressing ASCT2‐GFP proteins in a GFP intensity‐dependent manner. Ab3‐8 reacted with various human cancer cells, but not with non‐cancer breast epithelial cells or ASCT2‐knocked out HEK293 and SW1116 cells. In SW1116 and HCT116 human colon cancer cells with KRAS mutations, treatment with Ab3‐8 reduced intracellular glutamine transport, phosphorylation of AKT and ERK, and inhibited in vivo tumor growth of these cells in athymic mice. Inhibition of in vivo tumor growth by Ab3‐8 was not observed in HT29 colon and HeLa uterus cancer cells with wild‐type KRAS. These results suggest that ASCT2 is an excellent therapeutic target for KRAS‐mutated cancers.

are used in CRC. Cetuximab and panitumumab, therapeutic monoclonal antibodies (mAb) against epidermal growth factor receptor (HER1/EGFR), are frequently used, but are less effective in a subset of cancers with KRAS mutations. 6 In clinical trials in patients with KRAS-mutated CRC, multiple combination therapies such as AKT inhibitor-MEK inhibitor, 7 MEK inhibitor-FOLFIRI, 8 and cetuximab-FOLFOX 9 have shown disappointing results. Similar to strategies to suppress KRAS signaling pathway and function, developing attempts at direct targeting KRAS is anticipated to be long and difficult way. 10 As mentioned above, there are no effective therapies for KRAS-mutated CRC, and furthermore, the exact molecular pathology of CRC with KRAS mutations is still unsolved.
Glutamine is considered as a conditional essential amino acid for proliferation and survival of cancer cells. [11][12][13] In particular, KRAS-mutated cancer cells are reported to be prone to increase the demand of glutamine. 14 Indeed, several studies demonstrated the effects of oncogenic KRAS on glutamine metabolism. Utilization of glutamine for anabolic synthesis and the expression of genes associated with glutamine metabolism are upregulated in NIH3T3 cells expressing KRAS-mutated proteins and in KRAS-mutated human breast cancers. 15 In pancreatic cancers, oncogenic KRAS plays an important role in the reprogramming of glutamine metabolism. 16 Furthermore, glutamine catabolism in the TCA cycle is necessary for KRAS-induced anchorage-independent growth. 17 Taken together, glutamine plays an essential role in the growth of KRAS-mutated cells.
ASCT2 (SLC1A5) is a multi-pass transmembrane protein that transports neutral amino acids. In many cancer cells, ASCT2 is highly expressed and regarded as a primary transporter for glutamine. 18 Several studies have reported that treatment with ASCT2 inhibitor or knockdown of ASCT2 causes marked decreases in intracellular glutamine concentration and cell viability in various human cancer cells. [19][20][21][22][23][24] In human tumor tissues harboring KRAS mutation, ASCT2 proteins were highly expressed as compared with KRAS wild-type tumors. 25 Although the relationship between KRAS mutations and ASCT2 remains unclear, inhibition of ASCT2 function may be a promising method to treat KRAS-mutated cancers considering the importance of glutamine in cells with KRAS mutation. In this context, we developed specific mAb recognizing the extracellular domain of human ASCT2 and examined the effects of mAb on in vitro and in vivo growth of KRAS-mutated human colon cancer cells. (#CRL-10317). ABC-1 (#JCRB0815) was purchased from Japanese Collection of Research Bioresources (JCRB). Tagawa was provided by Dr Ohuchi N, Second Department of Surgery, School of Medicine, Tohoku University. Cells were cultured in equal volumes of RPMI-1640 medium and Dulbecco's modified Eagle's medium (RD medium; #05918 and #05919, Nissui Pharmaceutical Co., Ltd) supplemented with 7% fetal bovine serum (FBS; #10270-106, Thermo Fisher Scientific Inc) in humidified CO 2 incubators. RH7777 rat hepatoma (#CRL-1601; provided by Dr Chiba, Tanabe Mitsubishi Pharm) and HEK293F human embryonic kidney cells (#R79007, Invitrogen) expressing human ASCT2 fused to green fluorescent protein (GFP) (ASCT2-GFP) were cultured in RD medium with 7% FBS and 0.4 mg/mL of G418 (#09380-44, Nacalai Tesque).

| Animals
Six-week-old female F344/N rats and 6-week-old male KSN athymic (nude) mice were purchased from SLC Inc (Hamamatsu, Japan). They were housed in specific pathogen-free conditions, kept individually in cages under a standard light/dark cycle (12-hr light cycle starting at 7:00) at a constant temperature of 23 ± 1°C, and had ad libitum access to food and water. Animal experiments in this study were approved by the Committee for the Care and Use of Laboratory Animals at Kindai University, and performed following the institutional guidelines and the United States National Institutes of Health Guide for the Care and Use of Laboratory Animals.

| Rat mAb against human ASCT2
Production of the anti-human ASCT2 mAb was performed according to previous reports. 28,29 In brief, RH7777 cells expressing human ASCT2-GFP were injected three times into F344/N rats every 2 weeks. Three days after the final immunization, cell fusion was carried out by mixing the splenocytes of immunized rats with P3X63Ag8.653 mouse myeloma cells (#CRL-1580, ATCC) using 50% polyethylene glycol (#10783641001, Roche). Hybridomas were chosen for their binding capacity of antibodies in culture supernatant to transfectants expressing ASCT2. Selected and cloned hybridoma cells were injected into athymic mice pretreated with 2,6,10,14-tetramethylpentadecane (Pristane; #161-27483, Wako). Ascites was harvested and purified using Protein G sepharose (#17061801, GE Healthcare). The isotype of mAb was determined with the Rapid Monoclonal Antibody Isotyping Kit (#ISO-M6-20, Antagen Pharmaceuticals, Inc).

| Internalization of ASCT2 proteins
Microscopy study of direct internalization: HEK293 and RH7777 cells overexpressing human ASCT2-GFP were used. Cells were incubated with Ab3-8 (10 μg/mL) for 1 hour at 37°C. Fluorescence images were obtained using a BioZero microscope (Keyence) with a 20 × Plan Fluor objective lens. We observed at least 20 cells in each cell line. Quantification analysis using FCM: Cells (2 × 10 6 ) were reacted with Ab3-8 (10 μg/mL) at 4°C for 30 minutes. After that, the cells were divided into two groups; one was incubated at 4°C, and the other did at 37°C for 1 hour. Cells were reacted with PElabeled anti-rat IgG at 4°C for 1 hour, followed by FCM analysis.

| Glutamine-dependent cell proliferation assay
Cells (2 × 10 3 ) were seeded onto each well of 96-well plates in serum-and glutamine-free DMEM medium (#A1443001, Thermo Fisher Scientific) containing 7% dialyzed FBS, 1 mg/mL of D-glucose, and 0.11 mg/mL of sodium pyruvate. The cells were cultured for 3 days in the absence or presence of 2 mmol/L glutamine, and were measured for cell proliferation every 24 hours. For the evaluation of cell proliferation, WST-8-based Cell Count Reagent SF (#07553-44, Nacalai | 305 Tesque) was added to each well, and incubated for 3 hours at 37°C. After the incubation, the absorbance was read at 450 nm using a FilterMax F3 microplate reader (Molecular Devices, LLC). The assay was performed in duplicate.

| Glutamine uptake
Cells (2 × 10 6 ) were cultured in 35-mm culture dishes overnight. After cells were incubated for 2 hours with Ab3-8 (30 μg/mL) in PBS containing 1 g/L of D-glucose and 0.11 mg/ mL of sodium pyruvate, L-glutamine (2 mmol/L) was added for 10 minutes. Concentrations of intracellular glutamine were measured in cell lysates using the glutamine/glutamate-Glo Assay kit (#J8021, Promega Co.) according to the manufacturer's instructions. Luminescence was measured using a FilterMax F3 microplate reader. The data were normalized to total protein levels and this assay was performed in duplicate.

| In vivo tumor formation
The in vivo tumor study was performed as previously described. 28,29 Tumor-bearing mice were prepared by subcutaneous inoculation of 5 × 10 6 of human colon cancer cells or 1.4 × 10 6 of HeLa cells in the left or right flank. The day on which the tumor was engrafted was defined as day 1. Tumor volumes were measured every third day using digital calipers, and were quantified using the formula: volume [mm 3 ] = (length [mm]) × (width [mm]) 2 × 0.5. Ab3-8 (100 μg/mouse) was administered intraperitoneally on days 1 and 9.

| Statistical analysis
Data are presented as mean ± standard deviation (SD). The significance of the differences was determined by the Mann-Whitney U test using GraphPad Prism 7 for Windows (GraphPad Software). The criterion for significance was P < .05.

ASCT2 mAb
We have previously established a screening method for mAbs using flow cytometry based on the reactivity with transfectants stably expressing GFP-tagged target protein. 30,31 We examined the reactivity of rat antibodies in supernatants of hybridoma cultures with cells expressing human ASCT2-GFP. The Ab3-8 (IgG2a/κ) rat mAb definitely reacted with HEK293 cells expressing human ASCT2-GFP, and with RH7777 cells expressing ASCT2-GFP in a GFP expression level-dependent manner ( Figure 1A). Furthermore, immunoprecipitation analysis revealed that Ab3-8 immunoprecipitated glycosylated 75-and non-glycosylated 50-kDa proteins in extracts prepared from three human colon cancer cell lines ( Figure 1B). As anti-ASCT2 rabbit mAb (D7C12) is available for biochemical experiments, we evaluated the specificity of Ab3-8 by comparing staining pattern of Ab3-8 with that of D7C12. On two-color immunostaining of human colon cancer cell lines (SW1116, HCT116, and HT29), the components stained by Ab3-8 were merged with components stained by the anti-ASCT2 rabbit mAb ( Figure 1C). In FCM, the Ab3-8-positive populations in these colon cancer cells were also stained by the anti-ASCT2 rabbit mAb ( Figure 1D). Next, we examined the reactivity of Ab3-8 with ASCT2-KO cells using the CRISPR-Cas9 system. Ab3-8 did not react with ASCT2-KO-HEK293 and SW1116 cells ( Figure 1E). Furthermore, Ab3-8 reacted with various human F I G U R E 1 Specificity of the Ab3-8 anti-human ASCT2 rat mAb. A, FCM analysis of Ab3-8 against HEK293 (upper) and RH7777 (lower) cells expressing ASCT2-GFP. PE, phycoerythrin. B, Immunoprecipitation. Lysates of SW1116, HCT116 and HT29 cells were immunoprecipitated with Ab3-8, and then detected using a commercial anti-ASCT2 rabbit mAb. Arrowhead, position of human ASCT2 proteins. C, PFA-fixed cells were stained with a combination of Ab3-8 and anti-ASCT2 rabbit mAb, followed by Alexa Fluor 488-conjugated anti-rat IgG and Alexa Fluor 647-conjugated anti-rabbit IgG. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI). Images were taken at 60 × magnification. Scale bar, 10 μm. D, PFA-fixed cells were stained with the same antibodies as described above. E, ASCT2-KO HEK293 and SW1116 cells were reacted with Ab3-8, followed by PE-conjugated anti-rat IgG. The reactivity of Ab3-8 with these cells was analyzed by flow cytometry. F, FCM analysis of cell surface expression of ASCT2 proteins in various human cell lines. The indicated cells were stained with Ab3-8, followed by PEconjugated anti-rat IgG. From values of the mean fluorescence intensity (MFI), the ratio (+mAb/ −mAb) of MFI (rMFI) was calculated. cancer cells, but its reactivity with non-cancer cells was weak ( Figure 1F). Among these cell lines, SW1116, HCT116, DLD-1, and HCT15 cells harbor a KRAS mutation, whereas other cell lines have wild-type KRAS. 32-37

| Functional activity of Ab3-8
We investigated the functional activity of Ab3-8. One of the anti-cancer mechanisms mediated by mAb is internalization; that is, cell surface targets are transferred into the intracellular part after the binding of antibodies. [38][39][40] Membrane localization of green fluorescence was observed in untreated HEK293 (Figure 2A, left) and RH7777 (data not shown) cells expressing ASCT2-GFP, whereas intracellular localization of green fluorescence was observed in cells treated with Ab3-8 (Figure 2A, middle and right). Furthermore, we carried out FCM analysis to quantitate cell surface expression of ASCT2 proteins. In SW1116 and HCT116 cells, which harbor KRAS mutation, treatment with Ab3-8 at 37°C reduced surface expression of ASCT2 compared with at 4°C. On the other hand, treatment with Ab3-8 had little effect on ASCT2 surface expression levels in HT29 with no KRAS mutation ( Figure 2B).
Next, we examined whether growth of KRAS-mutated and wild-type KRAS colon cancer cells is dependent on glutamine ( Figure 2C). All cell lines did not grow under glutamine-depleted conditions. However, SW1116 cells markedly proliferated in the presence of glutamine. Although HCT116 and HT29 cells proliferated in the presence of glutamine, the proliferation rate of these cells was very small as compared with that of SW1116 cells.
Since ASCT2 is a major glutamine transporter in cancer cells, 18 we examined whether Ab3-8 inhibited intracellular glutamine transport. As shown in Figure 2D, Ab3-8 markedly inhibited glutamine uptake into SW1116 and HCT116 cells. On the other hand, treatment with Ab3-8 had no effect on intracellular glutamine concentration in HT29 cells.
AKT and ERK signaling pathways have been reported to be involved in growth and survival of cancer cells. [41][42][43] As shown in Figure 2E, treatment with Ab3-8 suppressed the phosphor-(p-) AKT and p-ERK levels in SW1116 cells. However, consistent with the glutamine uptake assay, Ab3-8 did not change the p-AKT and p-ERK levels in HT29 cells.

| Effects of Ab3-8 on in vivo tumor growth of colon cancer cells
We have next evaluated whether Ab3-8 inhibits tumor growth of KRAS-mutated colon cancer cells ( Figure 3A). Treatment with Ab3-8 slowed tumor growth of both SW1116 and HCT116 ( Figure 3B). Similar to the in vitro cultured colon cancer cells ( Figure 1F), ASCT2 proteins were highly expressed in these in vivo colon cancer tumors ( Figure 3C).
In addition to inhibitory effects on tumor growth, Ab3-8 suppressed the expression of p-AKT, p-ERK, and Ki67 proteins ( Figure 3D). However, Ab3-8 had no effect on the tumor growth of HT29 and HeLa cells, both of which do not have a KRAS mutation ( Figure 3B).

| DISCUSSION
KRAS mutations are observed in many human malignancies, such as non-small-cell lung cancer, 44,45 CRC, 3,46 and pancreatic ductal adenocarcinoma 47 and are related to their oncogenesis. However, there are no effective methods to treat KRAS-mutated cancers.
Cancer cells increase the demand for amino acids compared with normal cells because of their rapid metabolism and proliferation. To fulfill this demand, cancer cells upregulate the expression of amino acid transporters on the cell membrane. 11 In particular, glutamine is one of the most important nutrients in cancer cells because of enhanced glutaminolysis. [11][12][13] In this context, glutamine is estimated as a conditional essential amino acid. Previous studies have revealed that ASCT2 is a major glutamine transporter, and is upregulated in multiple cancer cell types such as breast, gastric, prostate, and CRC. 18,48 Furthermore, expression of ASCT2 is correlated with malignant features. 18 Regarding CRC, several studies have suggested that glutamine metabolism plays an essential role in cellular growth. 49 Pharmacological blockade and genetic silencing of ASCT2 can suppress cancer cell proliferation and tumor growth through the inhibition of glutamine influx. [19][20][21][22][23][24]50 Taken together, these findings suggest that ASCT2 is a promising target for cancer therapy.
Recently, production of anti-human ASCT2 mAbs was reported, and these mAbs suppressed glutamine-dependent cancer cell growth. 51 However, it remains unclear whether these mAbs have in vivo anti-tumor activity against KRAS-mutated cancers. Here, we successfully established a novel hybridoma clone secreting the Ab3-8 mAb against human ASCT2. Ab3-8 reacted with transfectants expressing human ASCT2-GFP in a GFP-expression-dependent manner. Furthermore, immunoprecipitation assay revealed that Ab3-8 detected unglycosylated 50-kDa and glycosylated mature 75-kDa ASCT2 proteins in human colon cancer cells. These results indicate that Ab3-8 specifically reacts with human ASCT2 proteins.
We attempted to directly observe Ab3-8-mediated internalization of ASCT2 proteins by means of a fluorescent microscopy. Treatment with Ab3-8 resulted in increased green fluorescent dots in the intracellular region. The dots are assumed to be endocytic vesicles. However, it remains unclear whether Ab3-8 is released into the cytoplasm or is degrade in the lysosome after the internalization. To confirm this, it needs to further stain lysosome using antibody against the lysosome marker or dye as other groups have done. 52,53 We found that Ab3-8 inhibited the tumor growth of SW1116 and HCT116 colon cancer cells. Furthermore, treatment with Ab3-8 reduced intracellular glutamine concentrations in both cell lines. On the other hand, Ab3-8 had no effects on intracellular glutamine concentration and in vivo tumor growth in HT29 cells. We assume that cause of these results would be attributed to difference in Ab3-8-mediated internalization of ASCT2 proteins. FCM analysis revealed that treatment with Ab3-8-induced internalization of ASCT2 in both SW1116 and HCT116 cells, but not in HT29 cells. Ma et al 54 had reported that treatment with L-γ-glutamyl-p-nitroanilide, an ASCT2 inhibitor, decreased glutamine (100 μg/mouse) was intraperitoneally administered on days 1 and 9. Each dot represents the mean of 4-9 mice. Data are expressed as the mean ± SD. Tumor volume of individual mice shows in Figure  S3. C,Tumor sections from control mice were subjected to immunoperoxidase staining using Ab3-8. Scale bar, 20 μm. D, Tumor sections were subjected to immunoperoxidase staining using the indicated antibodies. Representative images of the SW1116-and HCT116-derived tumor sections, which immune labeled with p-AKT, p-ERK and Ki67. Images were obtained at 4 × magnification. Scale bar, 20 μm. Insets, higher magnification of the boxes. Images were obtained at 20 × magnification. Scale bar, 20 μm consumptions in wild-type KRAS Caco-2 cells as well as in KRAS-mutated SW480 cells. Taken together, Ab3-8 does not directly inhibit function of ASCT2, but suppresses glutamine uptake through internalization of ASCT2 expressing on cell surface. Although Ab3-8 binds to ASCT2 proteins expressed on membrane of HT29 cells, the reason why Ab3-8 did not induce internalization in HT29 cells remains unclear.
HT29 cell line does not harbor KRAS mutation, whereas the cell line harbors BRAF and PIK3CA, the most commonly altered genes in human CRC, mutations. 32 For this reason, we cannot accurately judge the effects of Ab3-8 on tumor growth of KRAS-mutated cancer. Therefore, we used HeLa cells because the cell line harbors wild-type KRAS, BRAF, and PIK3CA genes. 35,55,56 Similar to HT29 cells, treatment with Ab3-8 did not affect tumor growth of HeLa cells, although we found that the expression of ASCT2 proteins in HT29 and HeLa cells was higher than that in SW1116 and HCT116 cells. Many studies have reported the importance of glutamine metabolism in KRAS-mutated cancers. [15][16][17] For this reason, glutamine is likely more important for the survival of SW1116 and HCT116 cells than for HT29 and HeLa cells. In fact, we found that SW1116 cells effectively grew under the glutamine-enriched conditions as compared with HT29 cells. Accordingly, we speculate that inhibition of ASCT2 function is more effective for KRAS-mutated cancer than for wild-type KRAS cancer regardless of the ASCT2 expression level. However, we found unexpectedly that glutamine-dependent growth rate of HCT116 cells was nearly identical to that of HT29 cells. It would be of great interest to elucidate why Ab3-8 had no effect on KRAS wild-type tumors, and this could lead to the development of novel drugs for KRAS-mutated cancers.
ASCT2 is also expressed in normal mouse tissues such as lung, skeletal muscle, and large intestine. 57 However, Ab3-8 did not recognize mouse ASCT2 protein (data not shown); therefore, we speculated that Ab3-8 specifically bound to the inoculated tumor expressing human ASCT2 proteins. It was recently reported that Slc1a5 knockout mice exhibit normal growth, survival, and development of adaptive immune cells. 58,59 This suggests that ASCT2-targeted cancer therapy will be well tolerated with few adverse effects.
Although the importance of glutamine in KRAS-mutated cancers has been investigated, the exact effects of the KRAS mutation status on ASCT2 function are unclear. KRAS regulates the expression of amino acid transporters including ASCT2, and glutamine uptake in KRAS-mutated lung cancer cells. 60 Furthermore, ASCT2 was upregulated in human CRC cells expressing mutant KRAS as compared with cells expressing wild-type KRAS. 61 Analysis of human CRC samples revealed a significant positive correlation between the KRAS mutation status and ASCT2 protein expression, and an association between tumor depth and ASCT2 expression. 25 These findings suggest that regulation of the expression and function of ASCT2 is closely related to the KRAS mutation status. Further studies are needed to clarify the precise relationship between KRAS mutation and ASCT2, particularly how KRAS regulates ASCT2 expression.
ASCT2 is often expressed together with L-type amino acid transporter 1 (LAT1)/CD98hc (SLC7A5/SLC3A2), a heterodimeric antiporter that exchanges large neutral amino acids. 48,62,63 Both transporters have been implicated in cancer growth and mTOR signaling in many studies. 63,64 ASCT2 was proposed to take up glutamine, which then acts as an exchange substrate to accumulate leucine via LAT1/ CD98hc. We have recently reported the anti-tumor effects of a mAb against LAT1, which preferentially transports many essential amino acids. 29 As tumor growth highly depends on the nutritional condition, maintaining a low-nutrient environment is likely to induce tumor regression. Taken together, simultaneous inhibition of ASCT2 and LAT1 may be more effective than their individual inhibition for treating cancer.
In conclusion, this study demonstrated that the Ab3-8 anti-ASCT2 mAb inhibits the tumor growth of colon cancer cells with KRAS mutation by inhibiting glutamine uptake. Together with the finding that the growth of KRAS-mutated cancers, including pancreatic and lung cancers, is strongly dependent on glutamine metabolism, 16,65 this study suggests inhibition of glutamine uptake as a potential therapeutic target in KRAS-mutated various human malignancies.