Postpartum depression (PPD) is a common mental disorder among perinatal women, with a prevalence rate of 13% to 19%
[1] and a rising trend year by year. PPD not only affects the emotional and psychological health of the mother, but also has a profound impact on the growth and development of the baby, which brings a heavy burden to the family and society
[2-4]. Although PPD has gained attention in recent years with the increased awareness of maternal mental health, it still faces many challenges and limitations in prevention and treatment based on existing theoretical hypotheses
[5]. Therefore, in-depth study of the pathogenesis of PPD and search for new research targets are important for early identification of intervention, treatment, and prognosis of PPD.
The pathophysiology of PPD is complex and involves a variety of mechanisms, the pathogenesis of which is not yet fully understood, and the generally recognized etiologies include the monoamine hypothesis, dysregulation of the γ-aminobutyric acid (GABA) system, the hypothalamic-pituitary-adrenal axis (HPA) hypothesis, and others
[6-8]. Currently, the most widely used first-line drugs such as selective serotonin reuptake inhibitors (SSRIs) have delayed effects, with full effects taking 3-5 weeks or more, and even the possibility of aggravating suicidal and homicidal tendencies
[9]. Furthermore, the SSRIs may also be associated with withdrawal syndrome in newborns
[10] and postpartum hemorrhage in mothers
[11], which significantly affects the treatment of PPD.
In recent years, research
[12] has focused on the specific mechanism of action of the tryptophan-kynurenine (TRP-KYN) metabolic pathway in PPD, and an imbalance in the TRP-KYN metabolic pathway has been suggested to be one of the important mechanisms in the pathogenesis of PPD.
Only a very small portion (1%-5%) of TRP entering the bloodstream is converted to 5-hydroxytryptamine (5-HT), melatonin, etc. The rest (about 95%) is metabolized via the KYN pathway. This pathway is initiated by 2 key enzymes: Indoleamine 2,3-dioxygenase (IDO) and tryptophan 2,3-dioxygenase (TDO)
[13]. It subsequently branches into 2 arms: A neurotoxic pathway dominated by TRP-KYN-quinolinic acid (QA) and a TRP-KYN-kynurenic acid (KA) dominated neuroprotective pathway. Savitz, et al
[14] found that patients with major depression had higher levels of QA and significant atrophy of the hippocampus and amygdala compared to healthy subjects; whereas KA and KA/QA ratio were positively correlated with hippocampal and amygdala volumes in the major depressed population. The use of antidepressant means, such as behavioral activation therapy, SSRIs, ketamine (NMDAR antagonist), electroconvulsive therapy (ECT), and exercise can increase the KA/QA ratio
[15]. Thus, the TRP-KYN pathway toward QA neurotoxicity is associated with the development of PPD.
Reducing QA accumulation or antagonizing the neurotoxic effects of QA can theoretically prevent PPD. Current study
[16] on the TRP-KYN metabolic pathway in depression and postpartum depression has focused on the metabolites and enzymes in the middle and upstream of the pathway; quinolinic acid phosphoribosyltransferase (
QPRT), which is located downstream of the pathway, is not a rate-limiting enzyme, but it is capable of converting excitatory neurotoxicity products of QA into nicotinamide adenine dinucleotide (NAD
+), the core of energy metabolism, which is the substrate for DNA repair. NAD
+ plays a key role in neuronal protection and axonal regeneration. Therefore, this study aims to explore the mutation sites with potential functional significance of
QPRT in PPD and the possible regulatory mechanisms, which provide new theoretical references and directions for the prevention and treatment of PPD.
1 Subjects and methods
1.1 Ethics statements
The study protocol was approved by the Ethics Committee of the Third Xiangya Hospital of Central South University (Approval number: P2014/s155) and was registered at the Chinese Clinical Trial Registry (Registration number: ChiCTR2400084073). All participants provided written informed consent prior to enrollment.
1.2 Subjects
In this candidate gene association study, the sample size was estimated to be 242 using the sample size calculator developed by the School of Public Health, Zhejiang University of Traditional Chinese Medicine (α=0.05, β=0.2, P0=0.147, P1=0.296). Accounting for a 20% loss to follow-up, a minimum of 290 patients should be included as study subjects. A total of 374 parturients who were scheduled for and subsequently received cesarean section under lumbar anesthesia were recruited from the Third Xiangya Hospital of Central South University of Hunan Province (n=192) and Hunan Maternal and Child Health Hospital (n=182) between 2024 and 2025. Inclusion criteria: 1) American Society of Anesthesiologists Physical Status Classification System (ASA) II, age ≥18 years, gestational age ≥28 weeks; 2) consent of the women and their families and voluntary participation in this study; 3) elective cesarean delivery under lumbar anesthesia. Exclusion criteria: 1) A history of chronic or psychotic psychiatric disorders (such as schizophrenia or bipolar disorder); 2) combined with heart, lung and other organ dysfunction; 3) difficulty in cooperating with specimen collection and questionnaires; 4) high surgical risk and intraoperative adverse effects. All participants ultimately received the planned cesarean delivery under lumbar anesthesia as per the inclusion criteria. It is critical to clarify that the terms “prenatal depression” and “prenatal self-harm ideation” in this study specifically refer to symptoms identified through prenatal questionnaire screening. These are fundamentally distinct from formal clinical diagnoses of a major psychiatric disorder, the latter of which was an exclusion criterion. A total of 12 parturients were lost to follow-up or only provided incomplete data, leaving a total of 362 participants in this study. A post hoc statistical power analysis was conducted. Given the observed odds ratio of 2.92 for the dominant model of the rs9933310 polymorphism, a risk allele frequency of 0.39, with 29 cases and 333 controls, our study achieved a statistical power of 65.6% at an alpha level of 0.05.
1.3 Experimental procedure
A questionnaire was used to collect general information about the parturients participating in this study, including: Ethnicity, age, height, pre-pregnancy weight, pre-partum weight, week of pregnancy, number of pregnancies, mode of conception, completion of maternity checkups, whether they are twins, whether they are planning to get pregnant, whether their newborns are the same gender as expected, history of birth, history of miscarriage, stress during pregnancy, mood during pregnancy, satisfaction with life status quo, stressful events, domestic violence during pregnancy, relationship during pregnancy, relationship between mother-in-law and daughter-in-law during pregnancy, employment during pregnancy, literacy level, level of monthly family income, and prenatal self-harm ideation. Before anesthesia, 2 mL of maternal peripheral venous blood was collected in a blood collection tube and placed in a -80 ℃ refrigerator for gene polymorphism analysis. On the 42nd day after delivery, the Chinese version of the Edinburgh Postpartum Depression Scale (EPDS) was used to evaluate the degree of maternal depression through telephone follow-up, and the cut-off point of PPD evaluation was a score of EPDS≥10
[17-18]. This cut-off was selected based on the validated Chinese version of the scale to optimize sensitivity and specificity within our study population. For consistency with international literature and to enhance the comparability of our findings, we finally chose an EPDS score of ≥13 as the cut-off value. For the last question of the EPDS, “Do you have thoughts of harming yourself”, the parturients who selected “never” were judged to have no postpartum self-harm thoughts, while those who selected other answers were judged to have postpartum self-harm thoughts.
To investigate the association between common genetic variations in the QPRT gene and PPD, this study employed a tag single nucleotide polymorphism (tagSNP) approach. Four tagSNPs (rs1134700, rs2303255, rs9922666, and rs9933310) were selected using Haploview software based on genetic data from Chinese (Beijing and Southern China) populations, with a minor allele frequency (MAF) threshold of >0.05. This strategy is designed to efficiently capture the majority of common genetic variations in the QPRT gene through linkage disequilibrium. It is important to note that while this approach is powerful for screening common variants, it may not encompass all potential functional variants, particularly rare alleles or those in low linkage disequilibrium with the selected tagSNPs. Genotyping was performed using the Sequenom® MassArray SNP platform.
1.4 Bioinformatics databases and queries related to QPRT SNP loci
To query the existing QPRT SNP loci expression quantitative trait loci (eQTL) analysis in different tissues, this study made a preliminary comparison of QPRT expression before and after the mutation of the relevant loci.
Then we used the ENCODE database, GeneCards database and 3DSNP v2.0 database of UCSC Genome Browser to query the relevant information of the QPRT SNP loci, which mainly includes gene localization, histone markers, locus function prediction, etc. The main information includes gene localization, histone markers, function prediction of the locus, etc.
Finally, we predicted and compared the transcription factors before and after the mutation of the QPRT SNP locus using the JASPAR database.
1.5 Dual-luciferase reporter assay
The
QPRT-W (wild-type) and
QPRT-M (mutant) gene sequences were constructed and cloned into the pGL3 vector, which contains both Firefly and Renilla luciferases. The Supplementary
Table 1 (
https://doi.org/10.57760/sciencedb.xbyxb.00144) displays the precise DNA sequences of
QPRT-W and
QPRT-M genes. The pGL3-Promoter/Enhancer-
QPRT-W and pGL3-Promoter/Enhancer-
QPRT-M were transfected into 293T cells. After 48 hours, the expression of luciferase reporter genes was assayed by the Promega Dual-Luciferase System to verify whether the
QPRT rs9933310 locus has promoter and/or enhancer activity.
Differential transcription factors before and after QPRT rs9933310 mutation were selected. Synthetic plasmids of these transcription factors were prepared. The pGL3 empty vector, wild-type or mutant QPRT rs9933310, and transcription factors mimics were co-transfected into 293T cells, and luciferase activity was measured 48 hours later using the the Promega Dual-Luciferase System to verify the ability of these transcription factors to bind to the promoter before and after the QPRT rs9933310 mutation.
1.6 Statistical analysis
SPSS 26.0 was used to analyze the data. The relationship between the general information of the enrolled mothers and PPD and the association analysis between each gene locus and the incidence of PPD were analyzed using the chi-square test, and the factors that might be associated with PPD were screened out. Comparison of fluorescence expression of control (Control), QPRT-W (wild-type) and QPRT-M (mutant) in the Dual-Luciferase Reporter Gene Assay experiments was analyzed by ANOVA. P<0.05 (two-sided) was considered statistically significant.
2 Results
2.1 General clinical data
A total of 362 cesarean parturients were finally enrolled in this study, 29 of whom were diagnosed with PPD, with an incidence of 8.01%. The results showed that hypertension, thyroid disease, consistency of neonatal sex with desired sex, prenatal self-harm ideation, prenatal depression, poor mood during pregnancy, stressful life events, dissatisfaction with current life, poor couple relationship, poor mother-in-law-daughter-in-law relationship, and domestic violence were risk factors for PPD (all
P<0.05,
Table 1). In addition, with the deterioration of mood during pregnancy and the escalation of stress during pregnancy, the prevalence of PPD risk was significantly higher.
2.2 Association between QPRT gene polymorphisms and PPD
The mutation at the rs9933310 locus of the
QPRT gene was significantly associated with PPD (
P<0.05) and was identified as a risk factor (
OR>1,
Table 2 and
Table 3). Compared to women with the AA genotype (wild-type homozygotes) at the rs9933310 locus, those carrying the AG or GG genotypes exhibited an increased risk of developing PPD (
OR=3.100, 95%
CI 1.210 to 7.930;
OR=2.120, 95%
CI 0.500 to 8.900). No significant associations were observed between mutations at the rs9922666, rs1134700, and rs2303255 loci and PPD.
2.3 Bioinformatics databases and query results for the QPRT rs9933310 polymorphism
The eQTL analysis results for the
QPRT rs9933310 polymorphism were retrieved from the GTEx portal. These findings demonstrate that the expression of
QPRT in various brain regions and other tissues, such as the thyroid, liver, and kidneys, is influenced by the rs9933310 genotype, with the expression pattern following the order of rs9933310 AA>AG>GG (Supplementary
Figure 1,
https://doi.org/10.57760/sciencedb. xbyxb.00144).
Further investigation using the ENCODE database (Human Genome version: GRCh38/hg38; Supplementary Figure 2,
https://doi.org/10.57760/sciencedb.xbyxb.00144) revealed that
QPRT rs9933310 is located in a non-coding region, 373 base pairs downstream of the first exon. This region is enriched with histone marks H3K4me1, H3K4me3, and H3K27ac, all of which are associated with gene activation. Based on these findings, we hypothesize that this region may possess dual promoter and enhancer activities. Similarly, querying the GeneHancer database via GeneCards (chr16: 29679583) suggests that the chromosomal region containing
QPRT rs9933310 could exhibit both promoter and enhancer functions (Supplementary Figure 3,
https://doi.org/10.57760/sciencedb.xbyxb. 00144). Additionally, utilizing the 3DSNP v2.0 database, we performed predictive scoring of the function of the
QPRT rs9933310 locus. This locus is in promoter status in 89 cell types and enhancer status in 23 cell types. Based on the final evaluation scores (99.65 vs 7.82), the
QPRT rs9933310 locus appears more likely to possess promoter activity (Supplementary Figure 4,
https://doi.org/10.57760/sciencedb.xbyxb.00144).
Considering the above database results, it is plausible that the
QPRT rs9933310 polymorphism functions as both a promoter and enhancer. We speculate that the rs9933310 A>G mutation may alter the interaction between transcription factors and the promoter, thereby influencing
QPRT expression. To further explore this hypothesis, transcription factor predictions before and after the rs9933310 mutation were compared using the JASPAR database. The sequences used were: rs9933310-A: TAGCTCGGGAG-AAGCAGAGACAAAGTCCTGTCCCTTCTG and rs9933310-G: TAGCTCGGGAGAAGCAGAGGCAAA-GTCCTGTCCCTTCTG, with a relative profile score threshold set to 80% (default). The results showed that the mutation at the
QPRT rs9933310 site leads to the loss of transcription factors, including
Gata1,
GATA2,
GATA3,
Gata4,
Sox17,
Sox2,
Sox3,
Sox6, and
SRY (Supplementary Tables
2 and
3,
https://doi.org/10.57760/sciencedb.xbyxb. 00144).
2.4 Dual-luciferase reporter assay for validation
In this experiment, compared to the control group, no significant increase in luciferase reporter gene expression was observed following the transfection of pGL3-Promoter-
QPRT-W and pGL3-Promoter-
QPRT-M into 293T cells. Additionally, there was no significant difference in reporter gene expression between the Promoter-
QPRT-M and Promoter-
QPRT-W groups (
Figure 1A). However, following the transfection of pGL3-Enhancer-
QPRT-W and pGL3-Enhancer-
QPRT-M into 293T cells, a marked increase in luciferase reporter gene expression was observed. Notably, the luciferase activity of the pGL3-Enhancer-
QPRT-M group was significantly lower than that of the pGL3-Enhancer-
QPRT-W group (
Figure 1B).
This experiment validated the transcription factor predictions obtained from the JASPAR database by selecting the differential transcription factors before and after the
QPRT rs9933310 mutation:
Gata1,
GATA2,
GATA3,
Gata4,
Sox17,
Sox2,
Sox3,
Sox6, and
SRY. Compared to the
QPRT-W group, the luciferase expression in the
QPRT-M group was significantly reduced, particularly for
SRY,
Gata1,
GATA2,
GATA3,
Gata4,
Sox17, and
Sox2 (all
P<0.05). For the transcription factors
Sox3 and
Sox6, no significant difference in the binding ability to the promoter was observed between the pre- and post-mutation conditions of
QPRT rs9933310 (all
P>0.05,
Figure 1C).
3 Discussion
In this study, analysis of the general clinical data from 362 postpartum women revealed significant correlations between PPD and factors such as hypertension or thyroid disease, discrepancies between expected and actual newborn gender, prenatal depression, self-harm thoughts during pregnancy, domestic violence, poor spousal and mother-in-law relationships, life stress events, dissatisfaction with current life, poor mood, and high stress during pregnancy. In summary, the risk factors for PPD can be categorized into 5 major groups: biological/physiological factors, psychological factors, obstetric/pediatric factors, socio-demographic factors, and cultural factors. Although this study did not find significant associations between maternal age, educational level, economic status, and adverse pregnancy history with PPD, such relationships have been identified and consistently reported in other studies
[19-22].
This study specifically explored the association between
QPRT and PPD, and the results indicated that the
QPRT rs9933310 A>G polymorphism is linked to PPD. Women with AG and GG genotypes exhibited a higher risk of developing PPD compared to those with the AA genotype. Additionally, eQTL analysis revealed that the rs9933310 mutation reduced the expression of
QPRT. On the one hand,
QPRT, as a key enzyme in the neurotoxic QA to NAD
+ metabolic pathway, may alleviate neuronal apoptosis and provide a therapeutic effect in PPD by lowering QA and increasing NAD
+ concentrations. The reduced expression of
QPRT leads to QA accumulation, as evidenced by
QPRT knockout (KO) mice, which showed a 6% conversion rate of QA to NAD
+—half of the activity seen in wild-type mice
[23]. In lipopolysaccharide (LPS)-induced mouse models of depression, brain NAD
+ depletion impaired energy metabolism and neuronal function
[24], while elevating NAD
+ levels improved depressive-like behaviors in rats
[25]. Clinical study
[26] also suggests that supplementing nicotinamide (a precursor of NAD
+) can rapidly improve depressive moods in young individuals with severe subclinical depression.
Based on the genetic risk identified in this study (QPRT rs9933310 A>G polymorphism) and established psychosocial risk factors, we propose a hypothesis regarding gene-environment interaction. On the one hand, this risk allele reduces QPRT expression, impairing its “detoxification” capacity to convert neurotoxic QA into NAD⁺; on the other hand, major psychosocial stress is known to strongly activate IDO/TDO enzymes, thereby substantially increasing QA production. We hypothesize that when these 2 processes—the “weak baseline clearance capacity” of QA and the “sharp increase in production” under stress—occur simultaneously in an individual, they may act synergistically. This could lead to a net accumulation of QA in the brain far exceeding levels achievable by either factor alone, thereby significantly amplifying the risk of PPD onset. It is important to note that due to the limited sample size of the PPD group (case group) in this study, we were unable to conduct a formal test with sufficient statistical power for the gene-environment interaction. Validating this hypothesis in future large-scale prospective cohorts is crucial for identifying the highest-risk populations for PPD and enabling precision prevention strategies.
Supporting a role for
QPRT in modulating apoptosis, it has been noted that depletion of
QPRT leads to increased expression of active caspase-3 (a cell apoptosis protein), thereby inducing neuronal apoptosis; however, substantial
QPRT inhibited this apoptotic process
[27]. Therefore,
QPRT may also exert antidepressant effects by reducing caspase-3 activity and minimizing neuronal damage. Furthermore, the voltage-gated potassium channel
KCNQ3 gene was downregulated in
QPRT KO mice
[28].
KCNQ potassium channels regulate neuronal excitability and activity, serving as mediators of stress recovery. The
KCNQ2/3 channel opener retigabine has shown significant improvements in depressive symptoms in patients with severe depression, highlighting its potential as a new antidepressant
[29].
Moreover, bioinformatics and dual-luciferase reporter assays clarified the regulatory mechanism of the rs9933310 site on
QPRT expression. The rs9933310 polymorphism may function as an enhancer, and its mutation weakens enhancer activity, influencing the interaction between transcription factors and the promoter, thereby reducing
QPRT expression levels. It has been proposed that regulatory elements can function dually as both enhancers and promoters, a concept supported by the observation of significant similarities between their chromatin structure, sequence, and core promoter architecture
[30]. Indeed, within this framework, many enhancers may act as promoters for other genes to initiate local transcription, while promoters themselves can exhibit enhancer activity. However, promoter activity of rs9933310 was not supported in the dual-luciferase reporter assay, possibly due to insufficient promoter strength at this site, which may not be able to significantly activate gene expression. Alternatively, the experimental results may not align with database predictions, suggesting that the rs9933310 site itself may lack promoter activity.
Similar to the resistance of the cytomegalovirus (CMV) promoter in embryonic stem cells, genes activated by the rs9933310 mutation may be silenced in certain cell types
[31]. In transcription factor-promoter binding experiments, significant differences in binding were observed before and after the mutation at the rs9933310 site, particularly with the
GATA and
Sox transcription factor families (including the
SRY gene). The
GATA transcription factor family serves as lineage differentiation factors involved in cellular transdifferentiation and reprogramming
[32-33], while the
Sox family (including
SRY) regulates various early embryonic developmental processes
[34]. Conditional knockout of
Sox2 in GABAergic neurons results in depressive-like behavior in mice. Furthermore, a reduction in c-Fos expression, an early marker of neuronal activity, was observed in
Sox2 conditional knockout mice
[35]. Therefore, the loss of
GATA and
Sox transcription factors following the mutation of rs9933310 may impair neural development.
There are several limitations to this study. First, the impact of the rs9933310 mutation on QPRT expression was inferred from existing results in databases, but further research using gene editing techniques to introduce targeted mutations at rs9933310 and directly measure QPRT expression is warranted. Second, while this study focused on the association between rs9933310 and PPD, further quantitative analysis of QPRT levels in postpartum women would provide more direct and comprehensive evidence for the role of QPRT in the development of PPD. Third, the EPDS score is not a diagnostic tool for PPD but rather a screening instrument for identifying potential cases of PPD. Finally, it should be noted that the statistical power of this study was 65.6%, while sufficient to detect the relatively strong genetic effect observed for the QPRT rs9933310 polymorphism (OR>2.9), may have limited the ability to identify associations with more modest effect sizes. This underscores the value of our positive finding and highlights the necessity for future replication in larger, independent cohorts.
In conclusion, combination of hypertension or thyroid disease, inconsistency of neonatal sex with expectation, prenatal depression, prenatal thoughts of self-harm, domestic violence, poor spousal and mother-in-law relationships, stressful life events, dissatisfaction with life status, poor mood during pregnancy, stressful pregnancy, and mutation of rs9933310 of the QPRT gene are all risk factors for PPD. The QPRT rs9933310 G allele represents an independent risk factor for PPD in women undergoing cesarean section. Its pathogenic mechanism may involve downregulating QPRT expression, thereby disrupting the homeostasis of the TRP-KYN pathway. QPRT may play a potential role in the pathogenesis of PPD and is expected to be a new target for antidepressant action on the TRP-KYN pathway.
the Natural Science Foundation of Hunan Province, China(2018JJ2598)
©Journal of Central South University (Medical Science). All rights reserved.
京公网安备11010202008535号
PDF
(932KB)
