Overexpression of L-Type Amino Acid Transporter 1 (LAT1) and 2 (LAT2): Novel Markers of Neuroendocrine Tumors

Background 6-18F-fluoro-L-3,4-dihydroxyphenylalanine (18F-FDOPA) PET is a useful tool in the clinical management of pheochromocytoma (PHEO) and medullary thyroid carcinoma (MTC). 18F-FDOPA is a large neutral amino acid biochemically resembling endogenous L-DOPA and taken up by the L-type amino acid transporters (LAT1 and LAT2). This study was conducted to examine the expression of the LAT system in PHEO and MTC. Methods Real-time PCR and Western blot analyses were used to assess LAT1 and LAT2 gene and protein expression in 32 PHEO, 38 MTC, 16 normal adrenal medulla and 15 normal thyroid tissue samples. Immunohistochemistry method was applied to identify the proteins’ subcellular localization. Results LAT1 and LAT2 were overexpressed in both PHEO and MTC by comparison with normal tissues. LAT1 presented a stronger induction than LAT2, and their greater expression was more evident in PHEO (15.1- and 4.1-fold increases, respectively) than in MTC (9.9- and 4.1-fold increases, respectively). Furthermore we found a good correlation between LAT1/2 and GLUT1 expression levels. A positive correlation was also found between urinary noradrenaline and adrenaline levels and LAT1 gene expression in PHEO. The increased expression of LAT1 is also confirmed at the protein level, in both PHEO and MTC, with a strong cytoplasmic localization. Conclusions The present study is the first to provide experimental evidence of the overexpression in some NET cancers (such as PHEO or MTC) of L-type amino acid transporters, and the LAT1 isoform in particular, giving the molecular basis to explain the increase of the DOPA uptake seen in such tumor cells.


Results
LAT1 and LAT2 were overexpressed in both PHEO and MTC by comparison with normal tissues. LAT1 presented a stronger induction than LAT2, and their greater expression was more evident in PHEO (15.1-and 4.1-fold increases, respectively) than in MTC (9.9-and 4.1-fold increases, respectively). Furthermore we found a good correlation between LAT1/2 and GLUT1 expression levels. A positive correlation was also found between urinary noradrenaline and adrenaline levels and LAT1 gene expression in PHEO. The increased expression of LAT1 is also confirmed at the protein level, in both PHEO and MTC, with a strong cytoplasmic localization.
The aims of this study were: 1) to study the LAT system in tissues obtained from patients with PHEO or MTC with a view to elucidating the molecular grounds for using 18 F-FDOPA in the diagnostic/clinical work-up; 2) to correlate the expression of LAT with that of GLUT1 transporters in the same tissues; and 3) to identify any relationships between LAT expression and patients' clinical features.

Patients
The study concerned a consecutive series of 32 patients with PHEO, all but 4 of them sporadic, and all but 2 benign (10 men and 22 women; median age 55, range 25-74 years), and 38 patients with MTC, all but 3 of them sporadic (18 men and 20 women; median age 60, range 10-81 years). Only one of the patients showed both PHEO and MTC.
Normal adrenal medulla tissue samples (n = 16) were obtained during surgical procedures for the purposes of renal transplantation. Normal thyroid tissue samples (n = 15) were obtained from the contralateral thyroid lobe in patients undergoing thyroid surgery for unifocal differentiated thyroid cancers.
All studies were performed in accordance with the guidelines proposed in the Declaration of Helsinki: the local ethical committee (Ethical Committee for the Clinical Experimentation of the Hospital of Padua) approved our study protocol (Ref. 3388) and all patients (including the parent/guardian on behalf of the minor) gave their written informed consent.

Mutation screening by direct sequencing
At germinal level, all exons of succinate dehydrogenase complex B (SDHB), SDHD, and VHL, and exons 5, 8, 10, 11 and 13-16 of the RET, and MAX and TMEM127, and exons 2-10 of MEN-1 were examined by direct sequencing. Somatic mutations of N-K-H RAS and RET were also analyzed in cases of sporadic MTC.

RNA extraction and reverse transcription
Each surgical specimen was snap-frozen in liquid nitrogen within 15 minutes of collection and stored at -80°C pending RNA recovery. Total RNA was extracted using the TRIzol reagent lysis buffer (Invitrogen, Life Technologies, Carlsbad, CA) according to the manufacturer's protocol.

qRT-PCR
A real-time quantitative PCR (qRT-PCR) was performed in an ABI-PRISM 7900HT Sequence Detector (Applied Biosystems, Milan, Italy) using the relative quantification method (2 -ΔΔCt method) as previously described [13]. The genes were analyzed using the following TaqMan assays: SLC7A5 (Hs00185826_m1); SLC7A8 (Hs00794796_m1); and GLUT1 (Hs00892681_m1), all from Applied Biosystems. Data were analyzed with the Sequence Detection Software rel. 2.4 (Applied Biosystems), adopting an automatically-set baseline and a fluorescence threshold adjusted to measure quantification cycle (Ct) values. Validation experiments performed using the standard curve method with five serial dilutions of genomic DNA from control subjects showed identical amplification efficiencies (100% ± 10%) calculated according to the formula E = 10 1/-slope -1 for all assays. Using the 2 -ΔΔCt method the data were presented as the fold-change in gene expression normalized by a reference gene and relative to a calibrator sample. As the reference gene in this study we used β-actin (Hs99999903_m1), one of the most commonly used housekeeping genes. A pool of cDNA derived from mixed normal human thyroid and adrenal medulla tissues was used as the calibrator source in our study.

Western blot analysis
Tumor samples were collected after surgery, frozen immediately in liquid nitrogen, and stored at −80°C. Immunoblot analysis was performed on normal and cancerous frozen thyroid tissue segments, as described elsewhere [14]. Briefly, proteins were separated with SDS/PAGE under reducing condition, in the presence of the S-S reducing agent dithiothreitol (DTT), electroblotted onto nitrocellulose membranes and saturated in 5% non-fat dry milk. Membranes were incubated overnight with the primary antibodies [anti-SLC7A5 (1:1000), Abcam (Cambridge, UK) and anti-β-actin, Sigma-Aldrich (1:5000)] and then incubated with HRP-conjugated secondary antibodies (Jackson ImmunoResearch, Europe). Blots were developed using Pierce ECL Substrate and exposed to CL-XPosure Film (Thermo Scientific, Rockford, US). Films were scanned and band intensity was quantified with ImageJ software 1.44p. To validate the anti-SLC7A5 antibody stain, LI-COR Odyssey Imaging Systems with the specific infrared fluorescence IRDye 1 secondary antibodies were used (S1 Fig) [15]. All experiments were performed in duplicate.

Immunohistochemistry
To further validate the immunofluorescence data, immunohistochemistry was performed on formalin-fixed, paraffin-embedded tissue sections 4-6 μm thick from 5 PHEO and 5 MTC cases, using the same anti-LAT1 antibody (1:500). Appropriate positive and negative controls were run concurrently. Two different pathologists (G.P. and F.G.) blindly assessed the findings, describing the intensity of staining as weak, moderate or strong. The subcellular localization of staining was also considered.

Statistical analysis
Proportions and rates were calculated for categorical variables; means ± standard deviations, or medians and ranges for parametric or non-parametric variables. In the qRT-PCR experiments, groups were compared with the Mann-Whitney or Wilcoxon tests for quantitative variables not distributed normally. Spearman's rank correlation and regression analyses were used to test the association between mRNA expression and different genes. To correlate the clinical/ pathological features and LAT1/2 gene expression in PHEO/MTC, the PHEO/MTC patients in our series were divided into two groups (PHEO/MTC strongly or weakly expressing LAT) based on median receptor expression values, using the chi-square and Fisher's exact tests as appropriate. The MedCalc for Windows, version 14.7 (MedCalc software, Ostend, Belgium) was used to manage the dataset on our patients and for the statistical analyses. The level of significance was set at p < 0.05 for all tests.

LAT1, LAT2 and GLUT1 gene expression in PHEO
PHEO specimens showed a significant increase in LAT1 (p<0.0001) and LAT2 (p = 0.0005) mRNA levels by comparison with normal adrenal medulla tissues. No statistically significant difference was found in GLUT1 (p = 0.26) mRNA expression levels between pathological and normal tissues. In particular, the median upregulation was 15.1-fold for LAT1 (95% CI: 12.54-34.92) and 4.1-fold for LAT2 (95% CI: 2.43-4.16) (Fig 1A). LAT1 mRNA was more overexpressed that of LAT2 in PHEO samples (p = 0.0001), while no differences were found between the two transporters' expression levels in normal adrenal medulla tissues.
A statistically significant link also emerged between LAT1 overexpression and high urinary levels of noradrenaline and adrenaline (p = 0.04 and p = 0.03, respectively, Fig 1D and 1E), while this was not apparent for LAT2 (Fig 1F). Patients' clinical characteristics are summarized in Table 1.
We then focused on LAT1 because it was expressed more strongly. Western blot experiments testing LAT1 expression in four representative normal/tumoral PHEO paired samples confirmed that the tumoral part expressed more protein than the non-tumoral tissue (Fig 2A). A stronger expression of LAT1 in tumors than in normal tissues was thus confirmed, and this trend was the same that we found in the mRNA expression analysis (Fig 2C).
The differences in mRNA expression levels were then considered in relation to clinical aspects: only tumor size greater than 10 mm correlated with a higher than median LAT1 expression level (p = 0.01, Fig 3D). Patients' clinical characteristics are summarized in Table 2.
Western blot experiments testing LAT1 expression in four representative paired normal/ tumoral MTC samples showed that the tumoral part expressed more protein than the nontumoral tissue in the most part of assessed cases (Fig 2B) showing that the expression at mRNA level reflects the amount of LAT1 protein (Fig 2D).
In one MEN2A patient, we were able to analyze both PHEO and MTC tumor tissues simultaneously for the three genes: for PHEO, LAT1 expression was 41 times higher than in normal tissues, while LAT2 and GLUT1 genes were expressed on much the same levels as in normal tissue; for MTC, on the other hand, LAT1 expression was 6.5 times higher while that of LAT2 was 3.1 times and that of GLUT1 was 5 times higher than in normal thyroids.

LAT1 immunohistochemistry
LAT1 protein expression was then assessed in paired normal/tumoral PHEO and MTC samples using immunohistochemistry (IHC). As previously shown in other tissues [16], LAT1 protein again displayed a strong cytoplasmic expression with membranous enhancement, in both PHEO and MTC, by comparison with normal tissues in all assessed tissues (Fig 4). The LAT1 reactivity was comparable with the Western blot findings for tumoral samples investigated and mainly revealed a cytoplasmic distribution.

Discussion
While initially developed for the neurological scanning of the basal ganglia in Parkinson's disease and in patients with psychiatric disorders, 18 F-FDOPA was shown to accumulate in PHEO, MTC, and gastroenteropancreatic NET-due to such tumors' ability to take up, store, and decarboxylate L-DOPA. Applications of this metabolic tracer in PET/CT scanning have proved more effective than anatomical imaging methods not only for the purpose of identifying the tumor burden (particularly in patients with recurrent disease), but also in providing further functional evidence to support patient prognosis and treatment planning. In metastatic PHEO, several studies have demonstrated that the diagnostic accuracy of 18 F-FDOPA-PET is clearly superior to that of 123 I-MIBG or CT/MRI, and that 18 F-FDOPA--PET can also be performed in the presence of drugs interfering with MIBG uptake [17][18][19]. 18 F-FDOPA scanning is now considered the first-line imaging tool for detecting glomus tumors at diagnosis or localizing extra-adrenal PHEO, and in patients with hereditary PHEO syndromes [20,21].
It has been also demonstrated that 18 F-FDOPA PET/CT produces better results than other functional/anatomical imaging procedures in patients with persistent or recurrent MTC, regional metastases, and basal calcitonin over 60 pg/ml, in which case scanning with 18 F-FDOPA seems to be the most useful functional imaging method for detecting regional lymph node disease and identifying candidates for surgery with a curative intent [20,22].
While it has been demonstrated that VMAT1 protein expression is the molecular prerequisite for functional imaging of PHEO with 123 I-MIBG scintigraphy [20], no experimental data on the molecular mechanism responsible for 18 F-FDOPA uptake have been published to date. To our knowledge, this is the first study in the literature to analyze the levels of expression (both mRNA and protein) of the two main LAT isoforms (LAT1 and LAT2) responsible for 18 F-FDOPA uptake in PHEO and MTC. The expression of these two transporters was also correlated with that of GLUT1 at mRNA level.
Using real-time PCR, we demonstrated that LAT1 and LAT2 are overexpressed in both PHEO and MTC samples by comparison with normal adrenal medulla and thyroid tissue. LAT1 was markedly more overexpressed than LAT2, and more so in PHEO (with 15.1-and 4.1-fold increases, respectively) than in MTC (with 9.9-and 4.1-fold increases, respectively). To further validate our results concerning LAT1 mRNA expression, we ran parallel Western blot experiments to test LAT1 protein expression in normal/tumoral samples: we These data were also confirmed with immunohistochemical analysis; in fact LAT1 protein showed a strong cytoplasmic expression with membranous enhancement, in both PHEO and MTC, by comparison with normal tissues. There are many evidences in literature that LAT1 immunostaining, especially in tumor tissue, is not only on plasma membrane, as we can expect about a transporter protein, but also diffuse in cytoplasm and/or in granules within the cytoplasm. Considering that after translation on ribosomes the membrane proteins are packaged into transport vesicles in the cytosolic compartment, we can suppose that the overexpression of LAT1 in the tumors involves an accumulation of vesicles in the cytoplasm. However, there were described some cases in which neoplastic cell showed only a strong cytoplasmic positivity [23].
Concerning PHEO patients, the statistically significant link between their LAT1 overexpression and their high levels of urinary catecholamines strongly suggests that this amino acid transporter may be a leading molecular step not only in the higher 18 F-FDOPA uptake, but also in the increased catecholamine secretion. However no significant correlation between  LAT2 expression levels and the urinary levels of catecholamine was found. This may depend on at least two causes: the first, the amount of LAT1 is higher than that of LAT2 in cancer cells as shown in Shennan et al. 2003 [24]; additionally LAT1 transports large neutral amino acids with higher affinity than LAT2 [7]. LAT1 has recently been associated closely with cancerous and/or proliferative cells, and previous studies found it strongly expressed in proliferating tissues, in many tumor cell lines, and in primary human tumors [25,26]. From a basic oncological viewpoint, cancer cells need plenty of nourishment for rapid growth and cell division, and LAT1 expression has been described as a significant indicator of a poor outcome in various human cancers, including lung [27], pancreas [28] or breast [29], and hepatocellular cancer [30]. Cancer cell growth is also supported by an increased glucose metabolism: this phenomenon corresponds to an increased glucose uptake across the plasma membrane by the glucose transporter proteins (GLUT). Several studies have demonstrated that the expression of glucose transporters, and of GLUT1 in particular, increases in a variety of malignancies. GLUT1 overexpression has also been found associated with tumor progression and poor overall survival in various malignant tumors [31]. From a clinical viewpoint, previous findings showed that 18 F-FDG-PET (which visualizes glucose metabolism and glucose transport, mainly via the GLUT1 transporter) performs better than other functional imaging procedures with more specific tracers such as 18 F-FDOPA, particularly in cases of malignant PHEO, and especially in patients carrying SDHB mutations. Our molecular PHEO findings reveal some degree of discrepancy on this issue, since we demonstrated a good correlation between LAT1/2 and GLUT1 expression levels, but the levels of the latter were comparable in tumoral and normal adrenal medulla cells. Such apparently conflicting data might be explained by the fact that the majority of PHEO patients in our series had benign disease; only 2 cases showed a malignant behavior according to the presence of distant metastases; finally, none of them carried SDHB mutations. Our MTC patients revealed a good correlation between LAT1/2 and GLUT1 expression levels, together with an increase in GLUT1 gene expression, though less marked than that of 18 F-FDOPA transporters. This confirms that functional imaging with 18 F-FDG PET could have a complementary role in the radiological work-up of a subset of MTC patients, particularly those with recurrent disease and low calcitonin doubling times, as a surrogate of aggressive disease, as recently suggested [32].
As concerns tumor size, we demonstrated a correlation between LAT1 overexpression and tumors larger than 1 cm in size, but only in MTC, not in PHEO. The presence of necrotic tissue and degenerative rearrangement (a frequent finding in larger PHEO) can cause variations in tumor size, and might explain the lack of correlation in PHEO, compared to MTC finding.
In conclusion, the present study is the first to provide experimental evidence of the overexpression in some NET cancers (such as PHEO or MTC) of L-type amino acid transporters, and the LAT1 isoform in particular, giving the molecular basis to explain the increase of the DOPA uptake seen in such tumor cells. Further studies are needed to define if the LAT1/LAT2 overexpression is also the molecular basis for justifying the 18 F-FDOPA use in functional imaging of NET cancers.