Genetic variations in amino acid metabolism-related genes are associated with risk of papillary thyroid carcinoma: a case-control study

Abstract

Background

Abnormal amino acid metabolic pathways, especially those of glutamine, serine and proline, are crucial to tumourigenesis and the development of papillary thyroid carcinoma (PTC). However, genetic variants in key genes regulating these metabolic pathways remain poorly characterized in PTC.

Methods

Seven SNPs in the SLC1A5, SLC1A3, SHMT1, and PRODH genes were genotyped in 620 patients with PTC and 620 controls using a flight mass spectrometry platform.

Results

The frequency of the minor allele A of SLC1A5-rs2070246 was significantly greater in the PTC group than in the control group, thereby increasing the risk of PTC by 1.587 times (p < 0.0001). Similarly, the minor alleles C, T and A of SLC1A3-rs16903247, SHMT1-rs4925166 and PRODH-rs372055 increased susceptibility to PTC by 2.584, 1.346 and 1.349 times, respectively (p_rs16903247 < 0.0001, p_rs4925166 = 0.00024, and p_rs372055= 0.001, respectively). Moreover, carriers of the rs2070246-GA/AA genotype had a 1.77- and 2.35-fold increased risk of PTC, and the rs16903247-PTC/CC genotype was associated with a 2.42- and 9.46-fold increased risk of PTC (p < 0.0001). In addition, carriers of the TG or TT genotypes of rs4925166 and the AA genotype of rs372055 all presented a greater risk of PTC (p_rs4925166 = 0.0005, p_rs372055= 0.0021). Genetic model data further confirmed that the above four SNPs indeed increased the individual’s sensitivity to PTC, as these SNPs were all associated with an elevated disease risk under different models (Bonferroni p < 0.0014).

Conclusion

Our results revealed significant associations between amino acid metabolism gene polymorphisms and PTC risk, suggesting potential biomarkers for PTC susceptibility.

Supplementary information

The online version contains supplementary material available at 10.1186/s12902-025-02034-8.

Introduction

As the most common malignant endocrine tumour, thyroid cancer accounts for approximately 95% of all endocrine tumours, especially papillary thyroid carcinoma (PTC) [1]. Its rapid growth trend in recent years has attracted widespread attention [2, 3]. Most PTC patients can be effectively managed with surgery, endocrine suppression therapy and iodine radiotherapy, with a good outcome [4]. However, some patients have early progressive metastasis, postoperative recurrence, insensitivity to treatment, and even death [5]. Therefore, the current difficulties and challenges are early identification of PTC, reliable screening of patients at high risk for recurrence and metastasis, and better disease management; thus, new diagnostic markers and treatment targets urgently need to be explored.

Amino acids, as one of the main nutritional substrates in the human body, are essential structural elements for cell growth, differentiation and function and critically modulate the tumour microenvironment (TME) [6, 7]. Amino acids can not only serve as carbon-nitrogen donors to provide raw materials for the synthesis of biological macromolecules required for the rapid proliferation of tumour cells but also can be taken up by immune cells in the TME to promote their proliferation and activation, thereby exerting antitumour effects [8–10]. Therefore, genetic variants related to amino acid metabolism and the resulting abnormal amino acid metabolism are the key to tumourigenesis and tumour development.

Glutamine is the most heavily utilized amino acid in PTC cells because it serves as a key substrate in protein and lipid synthesis [11]. The metabolic pathways of serine have also attracted much attention due to its important role in redox balance [12]. Proline is the substrate for collagen synthesis and thus plays an important role in tumour migration [13]. Furthermore, recent multiomics analyses have shown that branched-chain amino acid metabolism also varies among different types of PTC [14]. Certain key genes are essential for the metabolic pathways of these amino acids [15]. Solute carrier family 1 member 5 (SLC1A5) and member 3 (SLC1A3) transport glutamine/glutamic acid from the TME into cells and are highly expressed in PTC cells [16]. Serine hydromethyl transferase (SHMT) is also differentially expressed in different types of PTC [17], and proline dehydrogenase (PRODH) is the key enzyme involved in proline degradation [18]. An increasing number of studies have been conducted on the treatment of PTC through targeting amino acid metabolic pathways. However, few studies have focused on whether amino acid metabolism-related genetic variants can be used to assess the risk of PTC.

Single-nucleotide polymorphism (SNP) genotyping is widely used in early disease screening and personalized treatment due to its low detection cost and high sample throughput. Here, we selected seven SNPs associated with genes involved in amino acid metabolism through comprehensive literature screening and consideration of East Asian-specific minor allele frequencies (MAFs) and conducted genotyping in patients with PTC and healthy controls to analyse whether these SNPs could be used for PTC risk assessment.

Materials and methods

Participants

First, we calculated the expected sample size using G*Power 3.1.9.7 software. With an effect size of 0.2, a significance level (α) of 0.05 and a statistical power of 0.95, we needed to collect at least 542 patients and 542 controls.

Subject data were collected at Shanxi Provincial People’s Hospital. A total of 620 patients from the Department of Thyroid Surgery with papillary thyroid carcinoma who underwent strict histopathological diagnosis were selected as the PTC group. The pathological features of PTC include characteristic papillary structures with fibrovascular cores lined by a single layer of neoplastic cells, distinctive nuclear features (ground-glass appearance, nuclear grooves, nuclear overlapping, and intranuclear pseudoinclusions), the presence of psammoma bodies (calcified foci), and invasive growth patterns manifested by local tissue infiltration or lymph node metastasis [19]. At the physical examination centre, another 620 healthy individuals with no evidence of cancer or thyroid-related diseases were selected. All the research subjects were Han individuals residing in Shanxi Province with no blood relationship to one another. The PTC group included 201 males and 419 females, with an average age of 57.69 years. The control group included 196 males and 424 females, with an average age of 56.88 years (Table 1). The two groups had comparable sex and age distributions (p > 0.05).

Table 1.

Basic information on the participants

Characteristics	Case (n = 620)	Control (n = 620)	χ2/t	p
Sex (%)
Male	201 (32.4)	196 (31.6)	0.093	0.761
Female	419 (67.6)	424 (68.4)
Age			1.452	0.146
mean ± SD	57.69 ± 9.77	56.88 ± 9.95

Genotyping

Seven SNPs in SLC1A5, SLC1A3, SHMT1 and PRODH with an MAF > 5% in the East Asian population from the 1000 Genomes Project database and previously reported to be associated with cancers or other diseases were selected for genotyping [20–24]. Each participant contributed 5 millilitres of whole blood. After extracting DNA in batches using a PureLink™ Pro 96 Kit (Invitrogen), we stored the DNA samples at −80 ℃. After accumulating a sufficient sample size, we used a flight mass spectrometry platform (Agena Bioscience) to perform SNP genotyping. The specific operation steps are described in the literature [25]. The Genotype data is listed in Supplementary Table 1.

Statistical analysis

The chi-square and Student’s t tests were used to compare sex and age, respectively, between the groups. The MAFs of all the SNPs were evaluated for Hardy–Weinberg equilibrium (HWE). The genotyping data were analysed using SNPstats (https://www.snpstats.net/start.htm). Statistical significance was established when p < 0.05. All P values were Bonferroni corrected, and statistical significance was set at p < 0.0014 (0.05/35).

Results

In this study, a total of seven SNPs in four genes were investigated, namely, SLC1A5, SLC1A3, SHMT1 and PRODH. The chromosomal positions, alleles and variant types of these SNPs are listed in Table 2. Among these SNPs, four were intron variants, including SLC1A5-rs3826793, SLC1A3-rs3776578, rs16903247, and SHMT1-rs4925166. Two synonymous variants, SLC1A5-rs2070246 and PRODH-rs372055, and one missense variant, SHMT1-rs1979277, were also included. The potential functions of these SNPs were annotated using the HaploReg database, which revealed that they are involved in regulating the promoters, enhancers, DNases and/or motifs of the genes. In addition, four SNPs are known expression quantitative trait loci (eQTLs), including rs2070246, rs3776578, rs4925166 and rs372055.

Table 2.

Basic information on the SNPs

SNP	Gene	Position	Allele	Type	Predicted Functions
rs3826793	SLC1A5	chr19:46787394	G > T	Intron Variant	Promoter/Enhancer histone marks, DNAse, motifs changed
rs2070246	SLC1A5	chr19:46778905	G > A	Synonymous Variant	Promoter/Enhancer histone marks, DNAse, motifs changed, eQTL hits
rs3776578	SLC1A3	chr5:36651884	T > C	Intron Variant	Enhancer histone marks, DNAse, motifs changed, eQTL hits
rs16903247	SLC1A3	chr5:36653353	T > C	Intron Variant	Enhancer histone marks, DNAse, motifs changed
rs1979277	SHMT1	chr17:18328782	G > A	Missense Variant	L (Leu) > F (Phe)
rs4925166	SHMT1	chr17:18307496	G > T	Intron Variant	Motifs changed, eQTL hits
rs372055	PRODH	chr22:18913237	G > A	Synonymous Variant	Enhancer histone marks, DNAse, motifs changed, eQTL hits

SNP, single-nucleotide polymorphism; eQTL, expression quantitative trait locus

After processing the SNP genotyping data, we calculated the MAFs of each SNP in the PTC and the control groups (Table 3). After comparing the MAFs of all SNPs, we observed that the frequency of the minor allele A of SLC1A5-rs2070246 was significantly greater in the PTC group (38%) than in the control group (28%), thereby increasing the risk of PTC by 1.587 times (p < 0.0001). Similarly, the frequencies of the minor alleles C, T and A of SLC1A3-rs16903247, SHMT1-rs4925166 and PRODH-rs372055, respectively, were all higher in the PTC group than in the control group and increased the susceptibility to PTC by 2.584, 1.346 and 1.349 times, respectively (p_rs16903247 < 0.0001, p_rs4925166 = 0.00024, and p_rs372055= 0.001, respectively). SLC1A5-rs2070246, SLC1A3-rs16903247 and SHMT1-rs4925166 remained significant after Bonferroni correction (p < 0.0014). However, PRODH-rs372055 was not significant after Bonferroni correction.

Table 3.

MAF and HWE of the SNPs in the PTC and control groups

SNP	Gene	Allele	MAF-case	MAF-control	HWE p	OR (95% CI)	p
rs3826793	SLC1A5	G > T	0.21	0.19	0.36	1.134(0.932–1.381)	0.210
rs2070246	SLC1A5	G > A	0.38	0.28	0.69	1.587(1.340–1.881)	< 0.0001*
rs3776578	SLC1A3	T > C	0.08	0.08	0.57	1.057(0.790–1.414)	0.710
rs16903247	SLC1A3	T > C	0.18	0.08	0.99	2.584(2.006–3.327)	< 0.0001*
rs1979277	SHMT1	G > A	0.07	0.07	0.99	1.118(0.821–1.524)	0.478
rs4925166	SHMT1	G > T	0.48	0.41	0.21	1.346(1.149–1.578)	0.00024*
rs372055	PRODH	G > A	0.27	0.22	0.23	1.349(1.122–1.622)	0.001*

SNP, single-nucleotide polymorphism; MAF, minor allele frequency; HWE, Hardy–Weinberg equilibrium

*p < 0.05; Bonferroni multiple adjustment was applied, with p < 0.0014 (0.05/35)

We further compared the differences in the genotype frequency distributions of all SNPs between the patients with PTC and healthy controls (Table 4; Fig. 1). Based on the wild genotype of each SNP, we calculated the risk ratio of disease for individuals carrying the mutant heterozygous or homozygous genotype. We observed that carriers of the GA and AA genotypes of rs2070246 had an increased risk of PTC by 1.77- and 2.35-fold, respectively, and the TC and CC genotypes of rs16903247 were associated with a 2.42- and 9.46-fold higher risk of PTC, respectively (p < 0.0001). In addition, carriers of the TG or TT genotypes of rs4925166 and the AA genotype of rs372055 all exhibited a higher risk of PTC (p_rs4925166 = 0.0005, p_rs372055= 0.0021). However, after Bonferroni correction, no association was observed between PRODH-rs372055 and the risk of PTC.

Table 4.

Genotype frequencies of SNPs in the PTC and control groups

SNP	Genotype	Control	Case	OR (95% CI)	p
rs3826793	GG	410 (66.1%)	391 (63.1%)	1	0.48
	GT	184 (29.7%)	197 (31.8%)	1.09 (0.85–1.39)
	TT	26 (4.2%)	32 (5.2%)	1.35 (0.79–2.31)
rs2070246	GG	328 (52.9%)	234 (37.7%)	1	< 0.0001*
	GA	243 (39.2%)	306 (49.4%)	1.77 (1.40–2.25)
	AA	49 (7.9%)	80 (12.9%)	2.35 (1.58–3.49)
rs3776578	TT	526 (84.8%)	522 (84.2%)	1	0.87
	PTC	92 (14.8%)	95 (15.3%)	1.02 (0.75–1.40)
	CC	2 (0.3%)	3 (0.5%)	1.59 (0.26–9.59)
rs16903247	TT	526 (84.8%)	419 (67.6%)	1	< 0.0001*
	PTC	91 (14.7%)	179 (28.9%)	2.42 (1.81–3.23)
	CC	3 (0.5%)	22 (3.5%)	9.46 (2.80-31.93)
rs1979277	GG	540 (87.1%)	533 (86%)	1	0.66
	GA	78 (12.6%)	83 (13.4%)	1.07 (0.76–1.49)
	AA	2 (0.3%)	4 (0.6%)	2.05 (0.37–11.23)
rs4925166	GG	207 (33.4%)	157 (25.3%)	1	0.0005*
	TG	316 (51%)	325 (52.4%)	1.41 (1.08–1.84)
	TT	97 (15.7%)	138 (22.3%)	1.95 (1.39–2.75)
rs372055	GG	387 (62.4%)	349 (56.3%)	1	0.0021*
	GA	199 (32.1%)	207 (33.4%)	1.16 (0.91–1.47)
	AA	34 (5.5%)	64 (10.3%)	2.15 (1.38–3.35)

SNP, single-nucleotide polymorphism; OR, odds ratio; CI, confidence interval

*p < 0.05; Bonferroni multiple adjustment was applied, with p < 0.0014 (0.05/35)

Fig. 1.

Genotype distributions of SNPs between the PTC and control groups

Genetic models can better demonstrate the impact of different genotypes on disease risk. Therefore, we evaluated the influence of amino acid metabolism-related gene polymorphisms on the risk of PTC in three different genetic models (Table 5). These data further confirmed that the genetic variants of rs2070246, rs16903247, rs4925166 and rs372055 indeed increased the individual’s sensitivity to PTC, as these SNPs were all associated with an elevated disease risk under different genetic models. The degree of its increased risk varies from 1.32 to 8.35 times. All four SNPs remained significant after Bonferroni correction (p < 0.0014).

Table 5.

Evaluation of the effects of SNPs on the development of PTC by genetic models

SNP	Model	Genotype	Control	Case	OR (95% CI)	p
rs3826793	Dominant	GG	410 (66.1%)	391 (63.1%)	1	0.34
		GT-TT	210 (33.9%)	229 (36.9%)	1.12 (0.89–1.42)
	Recessive	GG-GT	594 (95.8%)	588 (94.8%)	1	0.31
		TT	26 (4.2%)	32 (5.2%)	1.32 (0.77–2.25)
	Log-additive	---	---	---	1.12 (0.92–1.36)	0.25
rs2070246	Dominant	GG	328 (52.9%)	234 (37.7%)	1	< 0.0001*
		GA-AA	292 (47.1%)	386 (62.3%)	1.87 (1.49–2.34)
	Recessive	GG-GA	571 (92.1%)	540 (87.1%)	1	0.0025*
		AA	49 (7.9%)	80 (12.9%)	1.77 (1.22–2.58)
	Log-additive	---	---	---	1.62 (1.36–1.93)	< 0.0001*
rs3776578	Dominant	TT	526 (84.8%)	522 (84.2%)	1	0.82
		TC-CC	94 (15.2%)	98 (15.8%)	1.04 (0.76–1.41)
	Recessive	TT-TC	618 (99.7%)	617 (99.5%)	1	0.61
		CC	2 (0.3%)	3 (0.5%)	1.58 (0.26–9.56)
	Log-additive	---	---	---	1.05 (0.78–1.41)	0.76
rs16903247	Dominant	TT	526 (84.8%)	419 (67.6%)	1	< 0.0001*
		TC-CC	94 (15.2%)	201 (32.4%)	2.66 (2.01–3.53)
	Recessive	TT-TC	617 (99.5%)	598 (96.5%)	1	< 0.0001*
		CC	3 (0.5%)	22 (3.5%)	8.35 (2.47–28.22)
	Log-additive	---	---	---	2.54 (1.96–3.29)	< 0.0001*
rs1979277	Dominant	GG	540 (87.1%)	533 (86%)	1	0.61
		GA-AA	80 (12.9%)	87 (14%)	1.09 (0.79–1.51)
	Recessive	GG-GA	618 (99.7%)	616 (99.3%)	1	0.4
		AA	2 (0.3%)	4 (0.6%)	2.03 (0.37–11.13)
	Log-additive	---	---	---	1.11 (0.81–1.51)	0.52
rs4925166	Dominant	GG	207 (33.4%)	157 (25.3%)	1	0.0011*
		TG-TT	413 (66.6%)	463 (74.7%)	1.53 (1.18–1.98)
	Recessive	GG-TG	523 (84.3%)	482 (77.7%)	1	0.003*
		TT	97 (15.7%)	138 (22.3%)	1.55 (1.16–2.07)
	Log-additive	---	---	---	1.40 (1.18–1.66)	0.0001*
rs372055	Dominant	GG	387 (62.4%)	349 (56.3%)	1	0.024*
		GA-AA	233 (37.6%)	271 (43.7%)	1.30 (1.03–1.63)
	Recessive	GG-GA	586 (94.5%)	556 (89.7%)	1	0.0009*
		AA	34 (5.5%)	64 (10.3%)	2.04 (1.32–3.15)
	Log-additive	---	---	---	1.32 (1.11–1.58)	0.0018*

SNP single-nucleotide polymorphism, OR odds ratio, CI confidence interval

*p < 0.05; Bonferroni multiple adjustment was applied, with p < 0.0014 (0.05/35)

Given that women are more prone to PTC, we further explored the influence of the above four SNPs on the risk of PTC by sex (Table 6). Overall, rs2070246, rs16903247 and rs372055 were associated with an elevated risk of PTC in both males and females, but the increased hazard ratios varied by sex and different genetic models, ranging from 1.3 times to 9.57 times. Notably, after subgroup analysis, rs4925166 was only associated with PTC in females, and no correlation was detected in the male population. However, after applying the Bonferroni correction, no significant associations were observed for rs372055 and PTC in females.

Table 6.

Confirmation of the influence of four SNPs on PTC risk in male and female subgroups

SNP	Model	Genotype	Males				Females
			Control	Case	OR (95% CI)	p	Control	Case	OR (95% CI)	p
rs2070246	Dominant	GG	117 (59.7%)	75 (37.3%)	1	< 0.0001*	211 (49.8%)	159 (38%)	1	0.0004*
		GA-AA	79 (40.3%)	126 (62.7%)	2.47 (1.65–3.70)		213 (50.2%)	260 (62%)	1.63 (1.24–2.15)
	Recessive	GG-GA	181 (92.3%)	167 (83.1%)	1	0.0035*	390 (92%)	373 (89%)	1	0.12
		AA	15 (7.7%)	34 (16.9%)	2.52 (1.32–4.82)		34 (8%)	46 (11%)	1.45 (0.91–2.32)
	Log-additive	---	---	---	2.00 (1.47–2.71)	< 0.0001*	---	---	1.45 (1.17–1.79)	0.0006*
rs16903247	Dominant	TT	174 (88.8%)	123 (61.2%)	1	< 0.0001*	352 (83%)	296 (70.6%)	1	< 0.0001*
		TC-CC	22 (11.2%)	78 (38.8%)	5.05 (2.98–8.56)		72 (17%)	123 (29.4%)	2.04 (1.45–2.86)
	Recessive	TT-TC	195 (99.5%)	195 (97%)	1	0.054	422 (99.5%)	403 (96.2%)	1	0.0002*
		CC	1 (0.5%)	6 (3%)	5.78 (0.69–48.57)		2 (0.5%)	16 (3.8%)	9.57 (2.16–42.38)
	Log-additive	---	---	---	4.53 (2.73–7.53)	< 0.0001*	---	---	2.02 (1.49–2.74)	< 0.0001*
rs4925166	Dominant	GG	48 (24.5%)	56 (27.9%)	1	0.35	190 (44.8%)	127 (30.3%)	1	< 0.0001*
		TG-TT	148 (75.5%)	145 (72.1%)	0.81 (0.51–1.27)		234 (55.2%)	292 (69.7%)	1.89 (1.42–2.51)
	Recessive	GG-TG	179 (91.3%)	171 (85.1%)	1	0.065	375 (88.4%)	337 (80.4%)	1	0.0015*
		TT	17 (8.7%)	30 (14.9%)	1.80 (0.95–3.38)		49 (11.6%)	82 (19.6%)	1.85 (1.26–2.71)
	Log-additive	---	---	---	1.05 (0.76–1.47)	0.76	---	---	1.61 (1.32–1.97)	< 0.0001*
rs372055	Dominant	GG	116 (59.2%)	112 (55.7%)	1	0.46	271 (63.9%)	237 (56.6%)	1	0.026*
		GA-AA	80 (40.8%)	89 (44.3%)	1.16 (0.78–1.73)		153 (36.1%)	182 (43.4%)	1.37 (1.04–1.81)
	Recessive	GG-GA	192 (98%)	181 (90%)	1	0.0008*	394 (92.9%)	375 (89.5%)	1	0.058
		AA	4 (2%)	20 (9.9%)	5.10 (1.70-15.25)		30 (7.1%)	44 (10.5%)	1.60 (0.98–2.61)
	Log-additive	---	---	---	1.36 (0.98–1.89)	0.064	---	---	1.30 (1.06–1.61)	0.012*

SNP single-nucleotide polymorphism, OR odds ratio, CI confidence interval

*p < 0.05; Bonferroni multiple adjustment was applied, with p < 0.0014 (0.05/35)

Discussion

At present, ultrasound-guided fine needle aspiration biopsy is the gold standard for diagnosing PTC [26]. However, it cannot effectively distinguish between benign and malignant lesions at the cytological level, resulting in a diagnostic uncertainty of 10–25% [27]. In addition, some molecular biomarkers, including BRAF, TERT, RAS and RET translocations, have been reported for the diagnosis of PTC [28]. However, none of the above methods can diagnose PTC early and accurately. Moreover, these methods have disadvantages, such as low sensitivity, high cost and invasive examination methods, resulting in unsatisfactory clinical application effects. Therefore, an urgent need exists to develop new preoperative methods, especially biomarker methods, to effectively distinguish PTC-susceptible populations, thereby improving the accuracy and sensitivity of clinical diagnosis. Here, we identified several SNPs in amino acid metabolism-related genes, SLC1A5, SLC1A3, SHMT1, and PRODH, that are associated with PTC risk and promising for use in clinical practice to screen high-risk populations for the prevention or early diagnosis of the disease.

The rapid growth and proliferation characteristics of tumour cells cause their demand for glutamine to far exceed the normal supply; thus, the uptake of glutamine from the extracellular environment becomes crucial. SLC1A5 is the main transport vector of glutamine in cancer cells [29]. Studies have shown that SLC1A5 can transport a large amount of glutamine into head and neck squamous cell carcinoma cells and that silencing SLC1A5 significantly leads to a decrease in the levels of glutamine and glutathione within cancer cells and in cell growth [30]. SLC1A5 can activate the downstream mTOR pathway to regulate the rapid proliferation of tumour cells [31]. Furthermore, under conditions of hypoxia or amino acid deficiency, SLC1A5 can be regulated by ATF4, which is synthesized in large quantities in the stressed tumour microenvironment and then translocated to the cell nucleus [32]. At present, many studies have revealed that negatively regulated miRNA and inhibitors of SLC1A5 have antitumour effects, which further indicates the potential of SLC1A5 as a tumour therapeutic target [33]. For polymorphisms in SLC1A5, little information is available. rs3826793 and rs2070246 in SLC1A5 have been investigated in patients with hepatocellular carcinoma, and carriers of the rs2070246-T allele exhibited better outcomes [20]. We first detected the MAFs of rs3826793 and rs2070246 in PTC patients and found that the MAF of rs2070246 was significantly higher in the PTC group than in the control group, thereby increasing the risk of the disease. Although the detailed mechanism needs to be further explored, the results suggest that individuals with SLC1A5-rs2070246 T > C should pay attention to their thyroid health.

SLC1A3 is an important glutamate-aspartate transporter located on the cell membrane and in mitochondria and lysosomes and is mainly responsible for the transport of glutamic acid [34]. Glutamate transporters are crucial for the termination of excitatory neural transmission and the glutamate metabolic process [35]. The dysregulation of SLC1A3 contributes to the pathogenesis of various neurological diseases, such as Alzheimer’s disease, depression and schizophrenia [36–38]. SLC1A3 is related to tumour metabolism and contributes to the progression of a series of tumours, such as glioblastoma [39], gastric cancer [40] and thyroid cancer [41]. Sun et al. demonstrated that the overexpression of SLC1A3 contributes to the characteristics of cell proliferation in solid tumours [42]. However, the regulatory role and underlying molecular mechanism of SLC1A3 in the process of tumourigenesis and tumour development have not been fully revealed. In recent reports, the functional role of SLC1A3 has been demonstrated in tumours such as thyroid cancer and gastric cancer. Wang et al. reported that SLC1A3 was highly expressed in CD133⁺ thyroid cancer cells, resulting in an increase in glutamate levels in PTC cells and thereby promoting the proliferation of cancer cells [41]. rs3776578 and rs16903247 in SLC1A3 were related to a reduced risk of migraine in a previous study [21]. The present study revealed that SLC1A3-rs16903247 T > C could lead to susceptibility to PTC. Whether this genetic variant affects the expression of SLC1A3 synergistically with other factors and thereby influences the progression of PTC remains to be further explored. However, the current results show that the risk of PTC is increased in the population carrying this genetic variant.

One-carbon metabolism provides the necessary precursor substances for growth and maintains the homeostasis of cancer cells by regulating antioxidants and methylation [43]. Enzymes including cytoplasmic SHMT1 and mitochondrial SHMT2 are the key enzymes involved in one-carbon metabolism in tumour cells [44]. SHMT1 affects the production of NADH/NADPH and the status of DNA methylation by supplying the one-carbon units required for nucleotide synthesis [45]. In addition, SHMT1 is implicated in driving tumour development, facilitating metastasis, and preventing apoptosis in diverse human cancers. Nevertheless, the precise biological roles and regulatory mechanism of SHMT1 in PTC remain unclear. SHMT1-rs1979277 is a missense variant G > A and has been investigated in patients with laryngeal tumours [22]; however, we observed no correlation between it and PTC risk. The TT genotype of SHMT1-rs4925166 is related to multiple sclerosis risk [23], and we also found that carriers of the TG or TT genotype of rs4925166 presented an increased risk of PTC. In addition, previous studies have identified several SNPs related to the risk of PTC in women. Song et al. reported that women with certain cell differentiation- and inflammation-related SNPs could be more susceptible to PTC [46]. Ren et al. reported that VAV3-rs7521681 was related to an elevated risk of thyroid cancer in women [47]. In this study, considering the high MAF of rs4925166 and its significant association with PTC in females, we suggested that its detection combined with that of other genetic variants could be used to assess the risk of PTC in women.

Most of the current evidence linking proline metabolism to cancer development centres around PRODH, which catalyses the first step of proline catabolism [48]. PRODH is usually expressed at low levels in tumours and is regarded as a tumour suppressor [49]. It has important regulatory functions in apoptosis, autophagy, osmotic regulation, oxidative homeostasis and tumour metastasis [50, 51]. Existing studies have shown that PRODH variants and the resulting abnormal proline metabolism are associated with hyperprolinemia, neuropsychiatric disorders and tumour development [18]. PRODH-rs372055 G > A is a synonymous variant and could lead to an increased risk of schizophrenia [24]. We also determined that the MAF of rs372055 in the PTC group was greater than that in the control group, indicating that it may increase susceptibility to PTC.

This study has several limitations that should be acknowledged. First, the lack of detailed information on tumour characteristics (including tumour size, stage, and metastasis status) in some patients precluded stratified analyses based on these clinically important parameters. Second, the absence of a benign thyroid tumour control group limits our ability to determine whether the identified susceptibility SNPs are specifically associated with malignant PTC development rather than thyroid tumourigenesis in general. Third, as our study population was predominantly recruited from Shanxi Province, the findings may primarily reflect genetic associations in northern Chinese Han populations, necessitating further validation in more diverse ethnic cohorts to confirm the associations. These limitations highlight important considerations for the interpretation of our results and the design of future studies.

In conclusion, the present study focused on amino acid metabolism-related gene polymorphisms and identified four SNPs, namely SLC1A5-rs2070246, SLC1A3-rs16903247, SHMT1-rs4925166 and PRODH-rs372055, that are associated with the risk of PTC. While this association study revealed SNPs significantly associated with PTC susceptibility, their potential utility in clinical risk stratification necessitates additional replication studies across independent populations to ensure generalizability.

Abbreviations

PTC

papillary thyroid carcinoma

TME

tumour microenvironment

SLC1A5

solute carrier family 1 member 5

SLC1A3

solute carrier family 1 member 3

SHMT

serine hydromethyl transferase

PRODH

proline dehydrogenase

SNP

single-nucleotide polymorphism

HWE

Hardy–Weinberg equilibrium

MAF

minor allele frequency

OR

odds ratio

CI

confidence interval

Author contributions

HM: Investigation, Formal analysis, Writing- Original draft preparation. ZX: Investigation, Formal analysis, Writing- Review & Editing. QW: Formal analysis, Writing- Review & Editing. DL: Formal analysis, Writing- Review & Editing. ZL: Validation, Supervision, Writing- Review & Editing. All authors read and approved the final manuscript.

Funding

This study was supported by the Youth Project of Shanxi Province Health Commission (No. 2024036) and the Key Research Project of Statistical Science in Shanxi Province (No. 2024Z026).

Data availability

The datasets generated and/or analysed during the current study are available in the Genome Variation Map repository, [Accession number GVM001134].

Declarations

Ethics approval and consent to participate

All participants provided informed consent, demonstrating their voluntary participation and understanding of the study’s objectives and potential implications. The study was approved by the Ethics Department of Shanxi Provincial People’s Hospital (No. 2018-035) and was conducted in accordance with the World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects.

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.

Clinical trial number

Not applicable.