Introduction

Non-syndromic cleft lip with or without cleft palate (NSCL/P) is the most prevalent congenital craniofacial anomaly1, with a global incidence of 1 in 700 people2. NSCL/P can be categorized into cleft lip with cleft palate (CLP), cleft lip only (CLO), and cleft palate only (CPO)3. Therapeutic approaches for NSCL/P are resource-intensive, necessitating a multidisciplinary team. Understanding the pathogenesis of cleft lip/palate is crucial for effective prevention and care. The causes of NSCL/P involve complex interactions between genes and the environment3, with epigenetics, specifically DNA methylation, emerging as a potential etiological factor. Perturbations in DNA methylation via hypermethylation and hypomethylation can induce structural defects in tissues resulting in several abnormalities4. Thus, it is possible that aberrant DNA methylation might play a role in the development of NSCL/P.

Alu elements, the most abundant repeats (~ 45%) in the human genome, are short interspersed repetitive elements (SINEs)5,6. Aberrant Alu methylation has been linked to genomic instability, leading to developmental abnormalities and cellular aging7. Alu hypermethylation has been implicated in the prevention of cell senescence through mitigating DNA damage and enhancing genomic stability8. DNA damage during embryonic craniofacial development, particularly during the formation of lip and palate, resulted in cleft lip and palate9,10. Increased levels of reactive oxygen species (ROS) during embryonic development can result in congenital anomalies, including NSCL/P11,12. ROS act as a teratogens, triggering apoptosis and causing oxidative damage to cellular macromolecules, including double strand break (DSB) in DNA13. ROS impair DNA methylation by diverting homocysteine to glutathione (GSH) synthesis instead of S-adenosylmethionine (SAM) production, leading to SAM depletion and unmethylated status13,14. Oxidative stress is often indicated by advanced glycation end-products (AGEs) and their receptor (RAGEs)15,16. DSB aggravated by ROS activates DNA damage response (DDR), leading to cellular senescence or “stress-induced premature senescence” (SIPS)17, characterized by cell-cycle arrest and inhibition of the transcription18. P16 serves as a marker for senescence and indicates DNA damage-induced SIPS17,19.

Maternal factors such as smoking, advanced age, and nutritional status have been implicated as NSCL/P risk factors20,21. Folate, a crucial component in the methionine cycle22,23, influences epigenetic gene expression24, and deficiency has been linked to orofacial cleft occurrence25,26. Aberrant maternal Alu methylation, influenced by external factors during pregnancy, may be passed on to the developing child4,27. This study aimed to investigate the correlation between Alu methylation, senescence, and severity of orofacial clefts in patients with NSCL/P and in their mothers.

Results

Demographic data

A total of 39 NSCL/P patients (6 CLO, 9 CPO, and 24 CLP) and 33 mothers were recruited. The mothers were divided into three groups in accordance with their children (5 in the CLO, 8 in the CPO, and 20 in the CLP). Maternal age averaged 30 years. Elderly mothers, women who became pregnant at the age of 35 or older28 were found 5 (25%) in CLP and 1 (20%) in CLO. Nearly 40% of mothers had a history of prenatal and perinatal smoking. Only 21% took first-trimester folic acid supplements. None used chemical products such as retinoic acid during the prenatal period or in the first trimester. All demographic data are reported in Table 1.

Table 1 The baseline characteristics of subjects enrolled in this study were compared among groups stratified by the presence of cleft lip and cleft palate.
Full size table

Alu methylation in NSCL/P patients and their mothers

At the first visit, univariate analysis of blood data of patients showed no significant difference in total Alu methylation between each subtype of cleft (Table 2). Blood Alu methylation changes were assessed at the time of cheiloplasty (visit 2) and palatoplasty (visit 3), and total Alu methylation did not differ between groups (Fig. 1). The specific positional analysis of Alu methylation was conducted only at the first visit. Our results showed that, in comparison with CLO, CLP had a 75-fold reduction in Alu hypermethylation (mCmC), along with a 3.07-fold increase in Alu partial methylation (mCuC), whereas CPO had a 4.30-fold increase in Alu partial methylation (mCuC) (p < 0.05, Table 2). We found that there was no difference in blood Alu methylation between the CLP and CPO (Table 2). With regard to the tissue specific Alu methylation, our data showed no significant differences between groups in both lip and palatal tissue (Table 2). These findings suggest that cleft subtype is associated with alterations in blood, rather than with tissue Alu methylation.

Table 2 Univariate analysis between patients and maternal Alu methylation, specific CPG methylation types, aging parameters, and cleft lip or cleft palate status.
Full size table
Fig. 1
figure 1

A: Percentage of Patients’ Blood (WBC) total Alu methylation comparing among cleft lip only (CLO), cleft palate only (CPO), and cleft lip with cleft palate (CLP) with first, second and third visit. The data were represented by mean ± SEM, and the comparison among patients’ groups was conducted by one-way ANOVA followed by Tukey’s multiple comparisons test. B: Percentage of Patients’ Blood (WBC) total Alu methylation comparing among first, second and third visit with cleft lip only (CLO), cleft palate only (CPO), and cleft lip with cleft palate (CLP). The data were represented by mean ± SEM, and the comparison among patients’ groups was conducted by one-way ANOVA followed by Tukey’s multiple comparisons test.

Full size image

The analysis of maternal blood Alu methylation showed that total Alu methylation in CLP mothers was 3.98 times higher than that in CPO mothers, however there were no significant differences in total Alu methylation between CPO vs. CLO or CLP vs. CLO mothers (Table 2). Specific positional analysis of Alu methylation revealed that both mothers of CLP and CPO had decreased Alu partial methylation (uCmC), when compared to CLO (Table 2). Additionally, CLP mothers exhibited lower Alu hypomethylation (uCuC) than CPO mothers (Table 2; Fig. 2). These findings suggest that CLP mothers had the lowest level of Alu methylation.

Fig. 2
figure 2

Percentage of Maternal Alu methylation comparing among cleft lip only (CLO), cleft palate only (CPO), and cleft lip with cleft palate (CLP). The subtypes of maternal Alu methylation including (A) mCmC (hypermethylation), (B) uCmC (partial methylation), (C) mCuC (partial methylation), and (D) uCuC (hypomethylation)). There was no significant difference in percentage of maternal Alu methylation subtypes comparing among groups. The data were represented by Box and whiskers plot. The comparison among patients’ groups was conducted by one-way ANOVA followed by Tukey’s multiple comparisons test.

Full size image

Senescence markers in NSCL/P patients and their mothers

The univariate analysis indicated that CLP had a 0.21-fold higher serum level of p16, when compared to that of CPO and a 2.24-fold higher serum RAGE level when compared to that of CLO (Table 2). Serum RAGE levels of CPO were also higher than those of CLO (Table 2). No statistically significant differences were observed in AGE levels between groups. These data indicated that CLP patients exhibited the highest level of senescence. There were no significant differences in p16, AGE, and RAGE levels among all mothers. All senescence markers are presented in Table 2.

We also analyzed the comparison between Alu methylation and senescence markers. Since our results did not follow a normal distribution, as indicated by the Shapiro-Wilk test, we used a non-parametric test instead of a one-way ANOVA. Specifically, we employed the Kruskal-Wallis test followed by an uncorrected Dunn’s test. The data from Kruskal-Wallis test among groups is shown in the Supplementary Table 1.

Multivariate analysis for potential risk factors for NSCL/P patients

Several factors, including maternal age, smoking, elderly mother status, and folate supplement, were included in the multivariate analysis of blood Alu methylation and senescence. After adjusting for maternal age, CLO patients exhibited a higher Alu hypermethylation, when compared to other groups. We also observed that Alu hypermethylation levels were higher in CLO than in other groups (9.23-fold vs. CPO and 10.62-fold vs. CLP, Table 3). Conversely, CLO had a lower partial methylation (mCuC) than the other groups (6.35-fold vs. CPO and 4.89-fold vs. CLP, Table 3). These results follow the same pattern, when adjusting data with smoking or elderly mother or folate supplement.

Table 3 Multivariate analysis between patients and maternal Alu methylation, specific CPG methylation types, aging parameters, and cleft lip or cleft palate status.
Full size table

With regard to senescence markers, CLO patients had lower RAGE levels than the other groups (2.36-2.60-fold vs. CPO and 2.10-2.17-fold vs. CLP, Table 3). However, no significant differences were observed in p16 levels between all groups (Table 3). This data suggests that CLO patients had high Alu hypermethylation, low partial methylation and low senescence in blood regardless of maternal age, smoking, and folate supplement.

Regarding blood Alu methylation in mothers, after adjusting for maternal age, smoking, or folate supplementation, CLO mothers demonstrated higher Alu partial methylation (uCmC) compared to CPO (8.74- to 9.95-fold) and CLP (8.01- to 8.72-fold) mothers. (Table 3). It should be noted that total Alu methylation in mothers of children with CLP children was 3.89- to 4.17-fold higher than in those with CPO children, after adjusting for smoking and folate supplementation. These findings suggest that maternal Alu methylation might be influenced by smoking or folate supplementation, but not by maternal age.

Discussion

This comprehensive translational study integrates an epigenetic analysis with clinical data from patients with NSCL/P and their mothers. Our primary objective was to investigate any aberrance in Alu methylation, oxidative stress (AGE, RAGE) and senescence (p16) using blood, lip, and palatal tissue samples. All positive findings are shown in Fig. 3. Our study found no significant differences in total blood (WBC) Alu methylation among cleft subtypes. We hypothesized that CLO would have less severity than CPO, with CLP being the most severe type. Interestingly, we observed lower Alu hypermethylation in CLP in comparison to CLO, a finding consistent with a previous report by Alvizi et al., which showed hypomethylation contributed to CLP29. In addition, a greater decrease in hypermethylated sites was found to be associated with increased severity of cleft penetrance4,29,30,31. These findings support our hypothesis that CLP, the most severe cleft type, exhibits lower Alu hypermethylation, resulting in a decrease in DNA stability, an increased risk of DNA damage, and greater cleft severity. Our study also revealed that CLP and CPO patients had higher levels of partial Alu methylation (mCuC) compared to CLO which is in alignment with previous research linking increased partial Alu methylation (mCuC) loci to poorer prognosis of breast cancer32, and worse prognosis of head and neck cancer33. We also investigated Alu methylation in lip and palatal tissues. There were no associations between cleft subtype and tissue specific Alu methylation, or significant changes in Alu methylation relative to cleft severity. A possible explanation is that no significance was shown when comparing between all impaired tissue such as orbicularis oris in cleft lip and levator veli palatinii in cleft palate. A limitation to this study is that we did not include normal lip and palatal tissue. Thus, we could not compare the alterations of methylation between healthy subjects and patients with cleft. Previous studies in cancer reported that aberrant methylation does differ between healthy individuals and cancer patients, even though minor changes can occur due to age and environmental factors. However, the Alu methylation cannot be used to distinguish between different cancer types or tissues34,35.

Fig. 3
figure 3

Schematic diagram for the positive results. CLO = cleft lip only; CPO = cleft palate only; CLP = cleft lip with cleft palate.

Full size image

The recruitment of mothers onto the study was justified by the results from intrauterine embryologic development in NSCL/P. Interestingly, our study found that mothers of CLP had higher blood total Alu methylation, which corresponded to lower hypomethylation (uCuC) in comparison to mothers of CPO. Our study showed that even though maternal Alu methylation was higher, it did not completely pass to the fetus. Regarding our findings, we hypothesized that pregnant mothers increase the level of folic utilization for themselves resulting in less folic storage for methylation in the children, resulting in hypomethylation and lower Alu hypermethylation alleles in the children. If our hypothesis is true, it is reasonable to assume that taking of a folic acid supplement could prevent CLP. However, we conducted maternal blood examinations at the first visit of the children, not during the time of their pregnancy, which might not accurately represent the maternal methylated status during intrauterine cleft lip and palate development. Alu methylation varies not only between different loci in the same person, but also changes dynamically with the time at different ages, as well as being influenced by various environmental factors35,36.

Alu methylation plays a crucial role in maintaining the stability of the genome by influencing gene recombination and chromosome translocation6. Previous reports have linked Alu hypomethylation to the aging process and chronic diseases such as hypertension and diabetes37,38. Our study revealed that CLO had lower RAGE levels than CPO and CPL, while those with CLP had higher levels of p16. These results align with our hypothesis that CLO is a less severe phenotype than the other two groups, showing less DNA damage and lower exposure to oxidative stress. This is consistent with Kobayashi et al. on NSCL/P dental pulp stem cells exposed to highly reactive oxygen, which caused DNA damage, DSB, and phenotype abnormalities related to NSCL/P39. The use of stem cells in this context is relevant to intrauterine embryonic development39. Additionally, Washausen et al. reported that increased apoptosis and senescence can lead to abnormal branchial arch development40. All these findings support the pathogenetic mechanism that Alu hypomethylation can potentially decrease the youth-DNA-gap, resulting in DNA damage and DDR. DDR causes senescence and an increase in senescence markers.

Alu methylation has recently been used as an adjunctive tool in diagnosing cancer. Early loss of methylation or unmethylation is clinically relevant to tumorigenesis, as seen in BRCA1 and BRCA2 which show extensive Alu hypomethylation32,41,42. We hope that we can adapt the positive results from this study by using Alu methylation and RAGE examination as an early tool for screening for cleft lip and palate incidence in high-risk groups with a family history of cleft lip and palate or even in women who plan for pregnancy.

There were several limitations in this study. Firstly, the small sample size in our study, particularly when further divided by cleft type, posed significant challenges for analysis and correlation calculations. Secondly, we examined CLO, CPO, and CLP groups, all consisting of cleft patients; we did not include non-cleft, healthy children in the study as controls. Ideally, we would have had a control cohort, but ethical concerns prevented us from taking large blood samples from healthy children at such a young age and with low body weight. Thirdly, total Alu methylation is calculated using formula reported as a percentage (%mC). Each value is dynamic, which can affect the summarized calculation and lead to nonsignificant findings of total Alu methylation.

Despite these limitations, this study has notable strengths. We conducted the analysis of Alu methylation using COBRA, a method that allows for assessment of both the quantity and pattern of methylation. Why pyrosequencing may not elucidate methylation patterns, COBRA has been proven to have the capability to detect multiple CpG sites. This choice of method assisted in analyzing inharmonious changes in methylation. This was especially the case when DNA methylation patterns exhibited greater sensitivity in uncovering changes rather than relying solely on methylation levels. In this case, the Alu methylation level was not significant but the Alu methylation pattern was. Thus, COBRA-Alu may be more sensitive and provide more information than pyrosequencing. Hence, we opted for COBRA over pyrosequencing in this study. We also investigated various sample types from both patients and their mothers, providing more comprehensive insights into Alu methylation, oxidative stress, and aging in NSCL/P patients.

The results of this study present a novel discovery regarding the association between senescence markers and NSCL/P occurrence. We hope to increase the participant pool in the future to yield more significant findings.

Materials and methods

Participants in this study

The cross-sectional study protocol is summarized in Fig. 4. All methods were performed in accordance with the relevant guidelines and regulations with the STROBE guidelines for reporting data from a cross-sectional study. Ethical approval was obtained by the Committee on Human Rights Related to Research Involving Human Subjects, the Faculty of Medicine, Chiang Mai University, Thailand (Approval number: ID 08646). Thirty-nine NSCL/P patients and their mothers who had received treatment at the Plastic and Reconstructive Surgery Unit, Department of Surgery, Faculty of Medicine, Chiang Mai University, between January 2022 and January 2023 were recruited. Of the 39 patients, six mothers were unavailable, resulting in a total of 33. Inclusion criteria comprised NSCL/P patients with a parent or guardian able to provide legally informed consent. Informed consent was obtained from all the patients, parent and/or legal guardian. Additionally, patients needed to be free of other associated diseases and in a state of overall health, without infections or inflammation that could potentially influence the study outcomes. Exclusion criteria excluded syndromic cases. Patients were categorized into 3 groups: CLO, CPO, and CLP. Data collection was conducted in three phases based on the type of defect: first visit (age 0–3 months) for all patients, cheiloplasty (age 3–6 months) for cleft lip only and cleft lip with cleft palate, and palatoplasty (age 9–18 months) for cleft palate only and cleft lip with cleft palate. Blood samples were collected during each visit to enable analysis of Alu methylation and senescence markers (AGE, RAGE, and p16 expression). Discarded muscle, orbicularis oris from cheiloplasty and levator veli palatini from palatoplasty, was subjected to tissue-specific Alu methylation analysis. Maternal blood samples were also collected during the first visit and examined for Alu methylation and senescence markers.

Fig. 4
figure 4

Schematic diagram for the assessment schedule of the study protocol. CLO = cleft lip only; CPO = cleft palate only; CLP = cleft lip with cleft palate.

Full size image

DNA extraction protocol

Blood was collected using an EDTA-coated tube then centrifuged at 3000 rpm for 10 min, and the buffy coat layer was collected. Red blood cells (RBC) were lysed with RBC lysing solution (Qiagen, Germany), and the white blood cells were collected. For DNA, we used the DNeasy blood and tissue kit (Qiagen, Germany). DNA yield was evaluated using the Take3 microvolume plate (BioTek Instruments, Inc., USA).

Bisulfate treatment

Bisulfate treatment is used to convert unmethylated cytosines to uracil, while leaving methylated cytosines unchanged. This allows for detection of DNA methylation patterns in genomic DNA. In this modified protocol, denatured genomic DNA was incubated in 0.22 M NaOH at 37 °C for 10 min, followed by adding 30 µl of 10 mM hydroquinone and 520 µl of 3 M sodium-bisulfite for 16–20 h at 50 °C. Finally, the DNA was purified and incubated in 0.33 M NaOH at 25 °C for 3 min, ethanol precipitated, washed with 70% ethanol, and resuspended in 20 µl of H2O14. The resulting sodium bisulfite-treated DNA was then PCR amplified using 1X PCR buffer (Qiagen, Germany), 0.2 mM deoxynucleotide triphosphate (dNTP) (Promega, USA), 1 mM magnesium chloride (Qiagen, Germany), 25 U of HotStarTaq DNA Polymerase (Qiagen, Germany), and 0.3 µM the primer pair: ALU-Forward and ALU-Reverse, where R = A and G and Y = C and T. For Alu amplification, the program was set as follows: initial denaturation at 95 ̊C for 15 min followed by 40 cycles of denaturation at 95 ̊C 45 s, annealing at 57 ̊C 45 s, extension at 72 ̊C 45 s, and ending with the final extension at 72 ̊C 7 min.

Alu-combined bisulfite restriction analysis (COBRA)

Alu PCR products were subjected to COBRA using 2 U of TaqI (Thermo Scientific, USA), 2 U of TasI (Thermo Scientific, USA) 5X NEB3 buffer (New England Biolabs, USA) and 1 µg/µl bovine serum albumin (BSA) (New England Biolabs, USA). Each digestion reaction was incubated at 65 ̊C overnight and later separated on 8% acrylamide and SYBR (Lonza, USA) gel staining. The band intensity of Alu methylation was observed and measured on the acrylamide gel using a Typhoon FLA 7000 and ImageQuant software (GE Health care, UK)[37.

Alu methylation analysis

According to COBRA, Alu methylation can be categorized into four patterns based on 2 CpG dinucleotide locus: hypermethylation (mCmC=both CpG locus were methylated); hypomethylation (uCuC=both CpG locus were unmethylated); partial methylation (mCuC=methylated followed by unmethylated cytosine and uCmC=unmethylated followed by methylated cytosine and)36. The level of each type of Alu methylation was determined by measuring the intensity of COBRA-digested Alu products using a Typhoon FLA 7000 and ImageQuant software (GE Health care, UK). Alu methylation analysis involved the use of five digested Alu products with sizes of 133, 90, 75, 58, and 43 bp. The Alu methylation analysis was performed using the following formula A = 133 bp/133, B = 58 bp/58, C = 75 bp/75, D = 90 bp/90, E = 43 bp/43, F = [(E + B) − (C + D)]/2. Alu methylation levels were calculated using the following formulas: Alu methylation level percentage (%mC) = 100 x (E + B)/(2 A + E + B + C + D); percentage of mCmC loci (%mCmC) = 100 x F/(A + C + D + F); percentage of uCmC loci (%uCmC) = 100 x C/(A + C + D + F); percentage of mCuC loci (%mCuC) = 100 x D/(A + C + D + F); and percentage of uCuC loci (%uCuC) = 100 x A/(A + C + D + F). The same positive control was used to adjust inter-assay variation37.

P16, AGE and RAGE determination

To determine the levels of p16, AGEs, and RAGEs, we utilized commercially available ELISA kits. The plasma p16 was determined using the Human p16 ELISA kit (Cat# MBS015625, MyBioSource, USA), the plasma AGEs level was determined using the Human AGEs ELISA kit (Cat# MBS267540, Cusabio, USA), and the plasma RAGEs level was detected using the Human RAGE Quantikine ELISA kit (Cat# DRG00, R&D Systems, USA).

Statistical analyses

Continuous variables were presented as mean and standard deviation (SD) or median and interquartile range (IQR). Categorical variables were expressed as frequency and percentage. Continuous variables were compared using a one-way ANOVA and categorical variables were compared using chi-square or Fisher exact tests as appropriate.

The percentage of maternal Alu methylation, compared between groups, was presented by a box and whisker plot and analysed using one-way ANOVA followed by Tukey’s multiple comparisons test. Univariable and multivariable regression were performed to compare Alu methylation and senescence markers. The β-coefficient and 95% confidence interval (95%CI) were reported. A p-value of less than 0.05 was considered statistically significant. Statistical analysis was performed using STATA version 16.1 software (StataCorp LLC, College Station, USA).