Expression Profile of MicroRNAs in Breast Milk of Women With Inflammatory Bowel Disease: Correlation With Disease Activity and Medical Treatments

Abstract

Background

Although most inflammatory bowel disease (IBD) medications are considered safe during pregnancy, their impact on microRNAs (miRNAs) in breast milk is largely unknown. MiRNAs in milk, carried by milk-derived extracellular vesicles (MDEs), are transmitted to the newborn’s gut to regulate genes. Aberrant miRNA expression profiles have been found in IBD within tissue, blood, and feces, but data on mother’s milk are scarce.

Methods

We collected breast milk samples from 32 mothers with Crohn’s disease (CD), 14 mothers with ulcerative colitis (UC), and 44 healthy controls. We analyzed miRNA expression through qualitative real-time polymerase chain reaction and Affymetrix miRNA chips. Target genes of differentially expressed miRNAs were predicted using miRATBase. Statistical analyses were conducted using GraphPad Prism software with Mann–Whitney tests.

Results

Milk-derived extracellular vesicles from mothers with IBD showed altered miRNA profiles compared to controls. Specifically, miR-21 and miR-320 were downregulated, while Let-7a was upregulated in IBD mothers. The expression patterns varied between CD and UC, with significantly lower MiR-21 in UC and higher Let-7a in CD. Additionally, anti-tumor necrosis factor treatment during pregnancy was associated with reduced miR-21 and miR-148a levels in MDEs. Pathway analysis revealed that these miRNAs are involved in immune regulation, particularly interleukin signaling pathways.

Conclusions

This study highlights that miRNAs in breast milk are differentially expressed in mothers with IBD, influenced by the disease and its treatments. These findings emphasize the impact of maternal health on milk composition and potential implications for infant immune development. Understanding these findings may guide personalized treatment strategies for mothers and promote breastfeeding among mothers with IBD.

Introduction

Inflammatory bowel disease (IBD) frequently affects women of childbearing years, and has a strong impact on conception, pregnancy, and lactation. Many IBD medications have been found to be safe in pregnancy, providing more benefit than harm, especially given the risks associated with active disease.¹ Consequently, effective medical treatment of IBD is essential to optimize successful pregnancy outcomes, and anti-tumor necrosis factor (TNF) agents, as well as other biological agents, are used routinely in pregnant women with IBD.² Although several studies have demonstrated detectable biological drug levels in the breast milk of nursing mothers with IBD,^3-5 it is generally accepted that these low drug levels pose minimal risk for the newborn. Nevertheless, there is still much concern related to medication safety during lactation, resulting in low breastfeeding rates among mothers with IBD.⁶

Human milk is one of the richest sources of microRNAs (miRNAs), small noncoding RNAs involved in post-transcriptional gene regulation. Various miRNAs identified in human milk have implications for gene regulation, some of which are critical for infant development.⁷ Milk-derived extracellular vesicles (EVs) (MDEs), like exosomes, are nanovesicles that carry these miRNA molecules, which are transmitted safely to the infant’s gut and then undergo uptake to the circulation to target tissues for gene regulation. The highly expressed miRNAs, especially those packaged in MDEs, include miRNAs related to infant immunological maturation and disease prevention.^8,9 In IBD, aberrant expression of miRNAs has been found to be involved in the pathogenesis of Crohn’s disease (CD) and ulcerative colitis (UC). For example, Cao et al. have shown that the level of plasma exosome miR-149-3p was significantly downregulated in patients with IBD compared to healthy controls. Interestingly, this downregulation was prominent in active UC and CD patients, compared with patients in remission.¹⁰ Recently, the expression of circulating Let-7e and of miR-126 was suggested to predict clinical remission in patients with CD treated with anti-TNFs.¹¹ While the roles of circulating, fecal and inflamed mucosal miRNAs have been studied thoroughly,¹² research on miRNA expression in human breast milk of women with IBD is still scarce. The aim of the current study is to explore changes in the expression of selected beneficial miRNAs in MDEs collected from mothers with CD and UC. Furthermore, we sought to analyze whether changes in miRNA expression are correlated with the type of medical treatment that was given during pregnancy and breastfeeding periods. A more comprehensive understanding of the role of miRNA changes in breast milk may shed more light on the benefit of breastfeeding for women with IBD, as well as on recommendations and the possible impact on newborns.

Methods

Population

Women with IBD were enrolled in the “IBD MOM” unit at the Gastroenterology Institute of Shaare Zedek Medical Center (SZMC), Jerusalem, Israel. The “IBD MOM” unit provides women with CD and UC with comprehensive consultation during the preconception period, throughout pregnancy, and postpartum, which includes conception planning, medication safety, and overall management. The “IBD MOM” unit collects and stores demographic, clinical, and laboratory data. Using these medical records, we obtained maternal demographic variables (age, pregnancy number) and IBD-related variables (diagnosis, disease location, and behavior and medical treatments). Disease activity was quantified by standard indices: Harvey–Bradshaw Index for CD and Partial Mayo index for UC. Women with IBD were compared with healthy women enrolled at SZMC and in Hadassah Medical Center. The study was conducted in accordance with the Declaration of Helsinki and was approved by the SZMC Institutional Review Board (SZMC-21-0165) and the Hadassah Institutional Review Board (study number HM0-0101-13). Informed consent was given by all participants.

Medication Use

Information was collected on medications used during pregnancy and postpartum. Exposure was defined as the use of various medical agents in the year prior to sample collection. Women were assigned to 1 of 4 groups based on drug exposure: unexposed, conventional therapy, anti-TNF, and other biologics. We repeated analyses evaluating only anti-TNFs, excluding other biologics.

Sample Collection

Milk samples were collected during the first 3 months of lactation from women with IBD and from healthy controls. Upon collection, milk samples were immediately placed on ice, divided into 1 mL aliquots, and then stored at −80 °C until isolation of MDEs and further analysis.

Isolation of Milk-Derived Extracellular Vesicles

Milk-derived extracellular vesicles within milk samples were isolated by sequential ultracentrifugation and filtration as described previously.¹³ Briefly, milk samples were fractionated by centrifugation at 3500 × g for 30 min at 4 °C and then the skim milk fraction was separated and centrifuged at 12 000 × g for 60 min at 4 °C to remove debris. The filtered supernatant was then ultra-centrifugated at 135 000 × g for 90 min at 4 °C to pellet the MDEs. The MDE pellet was left overnight in phosphate-buffered saline (PBS) at 4 °C to dissolve the MDEs. The protein content of the exosome’s preparation was measured by a bicinchoninic acid (BCA) protein assay.

Characterization of MDEs

Electron microscopy

Milk-derived extracellular vesicles were analyzed by electron microscopy as previously described¹⁴ and were examined using a Jem-1400 Plus transmission electron microscope (Jeol).

Nanoparticle analysis

Nanoparticle tracking analysis (NTA) was performed using an NS300 nanoparticle analyzer (NanoSight, Malvern), which was used to measure the size distribution of MDEs. Briefly, PBS-suspended MDEs were loaded into the sample chamber of the Nanosight unit, a blue laser source at 488 nm was applied to the diluted MDE suspension, and a video was recorded for 60 s at a frame rate of 24.98 fps. The movement of particles was analyzed using NTA software.

Dynamic light scattering

We performed dynamic light scattering and zeta potential determinations using a Zetasizer nanoseries instrument (λ = 532 nm laser wavelength; Malvern Nano-Zetasizer). The MDE size data was referred to the distribution of scattering intensity (z average).

Immunoblotting

Protein lysates from MDE were separated and transferred onto a polyvinylidene difluoride (PDVF) membrane. The membranes were probed with antibodies and detected using enhanced chemiluminescence detection. Primary antibodies were anti-CD81 (1:1000; Cosmo Bio), anti-HSP70 (1:1000; SBI System Biosciences), and anti-Alix (1:1000; Abcam). The secondary antibody was horseradish peroxidase (HRP)-conjugated goat anti-mouse or anti-rabbit (1:3000; Cell Signaling Technology).

RNA Isolation From MDEs

Following MDE isolation, Trizol reagent was added to the MDE pellet, and extraction of RNA was performed as previously described.¹³

Quantitative Real-Time Polymerase Chain Reaction

Briefly, 100 ng of RNA extracted from MDEs and circulating EVs were used to prepare cDNA.

The quality of extracted RNA was assessed using NanoDrop spectrophotometer. The qScript micro-RNA cDNA Synthesis Kit (Quantabio) was used to reverse-transcribe the RNA to cDNA, according to the manufacturer’s instructions, and the resulting cDNA was used to assess the expression of selected miRNAs and RNU6, which was used as a reference for normalization of miRNA expression levels. The PerfeCTa SYBR Green SuperMix (Quantabio) was used together with Quantabio micro-RNA qPCR primers. The quantitative real-time polymerase chain reaction (qRT-PCR) was run using a 2-step cycling protocol as previously described. Normalization and relative expression level calculation were carried out using the 2^ (−ΔΔ CT) method.¹³

Affymetrix miRNA Chip

Total RNA (300 ng) enriched from each sample was processed and labeled using the FlashTag Biotin HSR RNA Labelling Kit (Applied Biosystems/Affymetrix). Subsequently, samples were loaded on GeneChip miRNA 4.1 Array Plate and processed on the GeneTitan system (Applied Biosystems/Affymetrix) using the kit for miRNA Array plates (Applied Biosystems/Affymetrix) according to the manufacturer’s instructions. Raw data were processed and visualized using Partek Genomics Suite software (Partek).

Statistics

Categorical variables such as disease location, involved site, perianal disease, prior treatments, and patient clinical outcome were described by absolute and relative frequencies. Quantitative variables such as age were reported by way of mean ± standard deviations (Table 1). All statistical analyses were conducted using GraphPad Prism software. MicroRNA expression in milk samples of mothers with IBD and healthy controls was compared using the Mann–Whitney test.

Results

Study Population

To determine whether there are differences in the expression levels of miRNA in the milk of mothers with IBD versus healthy mothers, we collected breast milk samples from 46 mothers with IBD and 44 healthy mothers during the first 3 months of lactation. Clinical and demographic data of all participants are presented in Table 1. Mothers with IBD were younger than healthy controls (28 ± 5.6 for CD and 29 ± 4.4 for UC versus 31 ± 4.5 years, respectively, P < .04); 35% of milk samples were collected from IBD mothers after their first pregnancy, compared to 9% among healthy controls. Nearly 70% of IBD mothers were CD patients. Eighty-one percent of CD patients and 93% of UC patients in the current study were in clinical remission during the year prior to milk collection. During the study, 10 IBD patients (22%) were treated with conventional therapy (aminosalicylates, corticosteroids and immunomodulators); 15 (33%) were treated with anti-TNF therapy (infliximab or adalimumab); 15 (33%) were treated with other biologics (vedolizumab, ustekinumab, and rizankinumab); and 6 (13%) did not receive medical treatment.

Isolation and Characterization of MDEs

Maternal milk contains functional miRNAs encapsulated within MDEs. To determine differences in the miRNA cargo profiles from MDEs of mothers with IBD compared to healthy controls, we first isolated and characterized the MDE samples of these groups.

Milk-derived extracellular vesicles were isolated from individual human milk samples and subjected to NTA demonstrating a mean size of 145.4 ± 1.4 nm (Figure 1A). The expression of the exosome-related protein markers: CD81 and Alix was detectable by western blot analysis in milk samples of IBD and healthy mothers, indicating that a substantial portion of the EVs were exosomes (Figure 1B). To visualize EVs and to evaluate their morphology, transmission electron microscopy (TEM) was used. The TEM demonstrated a typical cup-shaped or spherical appearance with a diameter of around 100 nm (Figure 1C).

miRNAs Are Differentially Expressed in IBD

In our previous study, where we examined miRNA expression profiles in various mammalian milk sources, we identified specific miRNAs that were highly expressed in human milk.⁸ In the current study, we compared the expression of 7 of these miRNAs between MDEs from mothers with IBD and MDEs from healthy controls. The 7 miRNAs were selected based on their high expression in human breast milk and their known roles in immune regulation and inflammation, as reported in previous studies.^8,14Figure 2 shows significant changes in the expression of several miRNAs between MDEs from mothers with IBD and controls. We found that the expression of miR-21a and miR-320 was significantly lower in mothers with IBD compared to healthy mothers (0.74 ± 0.09 vs. 1.53 ± 0.25 and 1.68 ± 0.18 vs. 2.94 ± 0.33, respectively, P < .01). In contrast, expression of Let-7a was significantly higher in MDEs from mothers with IBD compared to healthy controls (1.09 ± 0.13 vs. 0.68 ± 0.11, respectively, P < .05). We did not find significant changes in the expression of miR-375, miR-30, miR-26, and miR-148 between IBD mothers and healthy controls (Figure 2). We did not find any correlation between mild and moderate disease activities and the expression of the various miRNAs.

Target Prediction Analysis of Selected miRNA in MDEs of Mothers With IBD Compared to Controls

Based on the findings described in Figure 2, we selected miR-21, miR-320, and Let-7a to assess their potential functions on target genes. The targets of these differentially expressed miRNAs were predicted using the Target Interaction database: MiRTarBase and (https://miRTarBase.cuhk.edu.cn/¹⁵ for miRNA sequences. We identified 99 genes as potential targets for miR-21, 22 genes for miR-320a, and 42 genes for Let-7a (Figure S1). For this analysis, only genes validated through at least 3 distinct experimental methods were included (highlighted in yellow). Based on the predicted target genes of miR-21, miR-320a, and Let-7a, we performed an analysis based on the Reactome biological pathways database.¹⁶ The most relevant molecular pathways (ranked by P-value) are depicted in Figures S2 and S3. The top-level pathways encompass signal transduction, diseases, gene expression (transcription), cell cycle, programmed cell death, and the immune system pathways. Interestingly, the overrepresented pathways associated with the target genes of miR-21 and miR-320a are related to interleukin (IL) signaling families, such as IL-12 or IL-10 (Table S1). Conversely, the overrepresented pathways associated with the target genes of Let-7a pertain to the Ras pathway (Table S2).

Differential Expression of miRNA in UC and CD

Our IBD cohort included 32 CD and 14 UC patients. We evaluated the expression of miR-21a, miR-320, and Let-7a within each subgroup. Figure 3 shows the expression profiles of these miRNAs for CD and UC patients and for healthy mothers, as well as for miR-148. We found that the expression of these selected miRNAs, miR-21a, miR-320, and Let-7a, is also significantly different in at least 1 IBD type. The expression of Let-7a was significantly higher in mothers with CD compared to healthy women (1.19 ± 0.17 vs. 0.58 ± 0.09, P < .01). In contrast, the expression of miR-21 was significantly lower in MDEs from mothers with UC compared to healthy controls (0.58 ± 0.19 vs. 1.5 ± 0.22, respectively, P < .05), while expression of miR-320 was significantly lower in MDEs from mothers with UC and CD compared to healthy mothers (1.46 ± 0.35 vs. 2.82 ± 0.3 and 1.75 ± 0.21 vs. 2.82 ± 0.3, respectively, P < .05). Interestingly, the expression of miR-148 was significantly higher in MDEs from CD mothers compared to UC mothers (5.21 ± 0.76 vs. 1.53 ± 0.41, respectively, P < .05); however, the differences between UC mothers and healthy controls, and between CD mothers and healthy controls, did not reach statistical significance (Figure 3).

The Expression of Selected miRNAs Is Affected by IBD Medications

Most of our 46 mothers with IBD received medical treatments during their pregnancy as detailed in Table 1. We divided these medications into 3 subgroups: conventional therapies, anti-TNFs, and other biologics. The expression of highly selected miRNAs in MDEs was compared for mothers who received these 3 types of treatment and a fourth group that did not receive any treatment. Figure 4 shows the comparison between different treatment subgroups and untreated mothers. We found that the expression of miR-21a and miR-148a was significantly lower in MDEs from mothers with IBD who received anti-TNFs compared to MDEs from untreated mothers (0.54 ± 0.13 vs. 3.88 ± 2.09, p<0.01 and 3.16 ± 0.62 vs. 14.95 ± 8.73, P < .05, respectively). No significant differences were observed when comparing the expression of miR-21a and miR-148a between mothers treated with conventional therapies or other biological treatments with those from untreated mothers (Figure 4). However, expression of miR-26 was significantly lower in MDEs from mothers who received anti-TNFs compared with those from mothers who have received other biological treatments (0.20 ± 0.05 vs. 0.46 ± 0.09, P < .05). In UC patients, when we analyzed miR-21 expression according to treatment type, including anti-TNFs, no statistically significant differences were observed between those treated with anti-TNFs, those receiving other treatments, and those not receiving any treatment (Figure S4).

Target Prediction Analysis of Selected miRNAs Affected by IBD Medications

Based on the results described above, and to assess the potential functions of miR-21 and miR-148, we analyzed the predicted targets of these differentially expressed miRNAs. Relative to the expression levels in untreated mothers with IBD, miR-21 and miR-148 expression was lower in mothers with IBD that were treated with anti-TNFs. According to the findings from MiRTarBase,¹⁵ 99 genes were potential targets for miR-21, 31 genes for miR-148a, and 52 genes for miR-26 (Figure S1). For the analysis, only genes validated through at least 3 distinct methods were included (highlighted in yellow). Based on the predicted target genes of miR-21 and miR-148, the 25 most relevant pathways (ranked by P-value) are depicted in Table S3. The top-level molecular pathways encompass diseases, signal transduction, gene expression (transcription), and development biology (Figure S5). Interestingly, the overrepresented pathways associated with the target genes of miR-21 and miR-148a are related to ILs and cytokines (Table S3). Based on the predicted target genes of miR-26, the 25 most highly ranked target pathways (ranked by P-value) are depicted in Table S4. The top-level pathways encompass cell cycle, diseases, signal transduction, cellular response to stimuli, and DNA replication (Figure S6).

Gene Enrichment Analysis of miRNAs in MDEs of Mothers With IBD Compared to Controls

To achieve a thorough understanding of miRNAs involved in specific molecular pathways, our goal was to assess the expression levels of all known miRNAs, without restricting ourselves solely to those predominantly expressed in milk. For that purpose, we used the platform of the Affymetrix miRNA array containing all known miRNAs. For the current Affymetrix profiling, we used samples from 9 IBD patients and 5 healthy controls. The volcano plot shows significant up- or downregulation with the fold changes >2.0 in miRNA levels in MDEs from IBD patients versus healthy mothers (Figure 5A). We used the miRNA enrichment analysis tool (https://ccb-compute2.cs.uni-saarland.de/mieaa/)¹⁷ to analyze the enriched pathways in which miRNA expression is altered in IBD relative to healthy MDEs. We found that specific miRNAs were upregulated or downregulated in MDEs isolated from mothers with IBD when compared to those from healthy mothers (Figures 5A and S7). Relative to the expression levels in healthy controls, the miRNAs that were downregulated in IBD mothers were those that were enriched in signaling pathways, such as mTOR, AMPK, and neurotrophin signaling (Figures 5B, 5C and S8). Conversely, no enriched pathways were discerned for the upregulated miRNAs when comparing MDEs from mothers with IBD to those from healthy mothers.

Differential Expression of All Known miRNAs in MDEs Based on IBD Treatment Variations

We further employed the Affymetrix gene array to compare the expression of all known miRNAs in MDEs from mothers with IBD undergoing anti-TNF treatment and those not receiving medical treatment. For the current Affymetrix profiling, we used samples from 4 anti-TNF-treated patients and 3 untreated patients. We found that distinct miRNAs were identified as either upregulated or downregulated in MDEs isolated from mothers that received anti-TNF treatment when compared to those that did not receive any medication. In the heatmap presented in Figure 6A, we show the miRNAs that were significantly upregulated or downregulated with fold changes >2.0 in MiRNAs in MDEs from mothers with IBD who did not receive treatment compared to those who were treated with anti-TNFs (Figures 6A and S7). The miRNA enrichment analysis tool (https://ccb-compute2.cs.uni-saarland.de/mieaa/)¹⁷ shows that remarkably, relative to the expression levels in untreated mothers with IBD, the miRNAs that were downregulated in anti-TNF-treated mothers were those that were associated with T-cell receptor and TNF signaling pathways (Figures 6B, 6C and S8). In contrast, no enriched pathways were identified for downregulated miRNAs when comparing MDEs from anti-TNF-treated mothers with IBD to those who were untreated.

Discussion

Breast milk is a key source of nutrition for newborns, but also contains many other beneficial bioactive components, including miRNAs that play a crucial role in post-transcriptional gene regulation. miRNAs are carried by MDEs and may significantly influence infant growth, development, and maturation of the immune system. Accumulating evidence demonstrates that miRNAs are dysregulated in IBD and are involved in the pathophysiology of the disease. IBD affects the expression levels of miRNAs, as has been demonstrated in tissue, blood, and feces.¹² However, the influence of IBD on the expression of miRNAs in human milk remains unknown. Our approach in the current study was to focus on the expression of milk-derived miRNAs of nursing mothers with IBD.

Using qRT-PCR, we analyzed the expression of miR-21, miR-320, Let-7a, miR-148, miR-375, miR-26, and miR-30 that are among the most abundant miRNAs in MDEs.¹⁴ We also studied the expression of all miRNAs in MDEs using Affymetrix chip analysis. Our study has demonstrated as-yet undocumented alterations in the expression of miRNAs in the breast milk of mothers with IBD. Furthermore, we have shown that some of these miRNAs were differentially expressed in CD and UC and were affected by various IBD treatments. We found significant changes in the expression of 3 out of 7 highly expressed miRNAs between MDEs from mothers with IBD and healthy controls. MiR-21 and miR-320 were downregulated and Let-7a was upregulated in MDEs from IBD mothers. The decreased expression of miR-21 in MDEs was significant in mothers with UC, while miR-320 was significantly reduced in both CD and UC. Additionally, the upregulation of Let-7a was only significant in MDEs from mothers with CD. These results imply that specific miRNA expression patterns in breast milk are associated with UC and CD, in accordance with specific miRNA expression patterns that were recently reviewed by Afaifi et al. in blood, tissue, and feces.¹²

MiR-21, a key mediator of inflammation and one of the most frequently reported miRNAs in IBD has been suggested to regulate impairment of the intestinal barrier function.¹⁸ We found that miR-21 was downregulated in MDEs of mothers with IBD, especially UC, compared to healthy mothers. Furthermore, miR-21 was downregulated in mothers with IBD treated with anti-TNF. This decrease due to medications may hypothetically modulate the effect of breastfeeding on the newborn’s health.¹⁹ However, further studies with larger cohorts are needed. Other studies have demonstrated the role of miR-21 in IBD. In a meta-analysis conducted by Yan et al., the expression of miR-21 in colon tissue was significantly higher in UC and CD patients, but these increased levels were correlated only with active disease, especially UC.²⁰ Furthermore, inhibition of miR-21-5p mediates the IL-6/STAT3 pathway in UC and decreases the levels of inflammation.²¹ Reports on the expression of miR-21 in peripheral blood are contradictory probably due to the heterogeneity of studies

In our study, miR-320 was downregulated in MDEs of mothers with CD and UC compared to healthy mothers. These results are in accordance with those of other studies. Pierdomenico et al. have studied the relationship between NOD2, the first gene associated with CD, and the miR-320 family, demonstrating an inverse correlation between miR-320 levels and NOD2 mRNA. In the inflamed mucosa of pediatric patients with CD and UC, miR-320 expression was decreased and NOD2 expression was increased.²² MiR-320 was also shown to be downregulated in fecal samples of UC patients versus controls.²³ In contrast, miR-320 was found to be increased in the circulating blood of IBD patients.²⁴ We found that Let-7a was upregulated in MDEs of mothers with IBD, especially CD, compared to healthy controls. In accordance, in a small study, Takagi et al. reported elevated levels of several miRNAs, including Let-7a, in colon tissues of active UC patients compared with healthy controls.²⁵ Interestingly, Fujioka et al. found significant elevation of Let-7d and Let-7e in serum samples of CD patients who were treated with anti-TNF.²⁶

Not surprisingly, our miRTarBase analysis of the overrepresented molecular pathways and biological properties related to the target genes of miR-21, miR-320a, and Let-7a suggests a connection to the immune system, a key player in IBD. We identified key processes that are regulated through IL-4, IL-13, and IL-12 pathways. To gain a deeper insight into these pathways, we performed miRNA profiling of all known miRNAs and identified mTOR (mechanistic target of rapamycin), AMPK (AMP-activated protein kinase), and neurotrophins pathways that are potentially regulated by these miRNAs. The involvement of mTOR and AMPK in IBD is noteworthy as the balance between these 2 pathways is crucial for the proper functioning of the immune system and the maintenance of gut homeostasis. Dysregulation of either pathway can contribute to the development and progression of IBD. These pathways were shown to be involved in regulating immune responses and inflammation in UC.^27,28 mTOR promotes Th17 cell differentiation and inflammation,²⁹ while AMPK reduces inflammation.³⁰

One of the main findings of our study is that medical treatment of IBD may influence the expression of milk-derived mRNAs. We have previously demonstrated that oxytocin, which is normally administered during delivery, modulates the expression of major milk-derived miRNAs.¹³ In the current study, we showed that IBD treatments, especially anti-TNFs, that were given during pregnancy and breastfeeding periods, influence the expression of miRNAs in MDEs. TNF-alpha is a pro-inflammatory cytokine, triggering chronic intestinal inflammation and tissue damage in IBD. Blocking TNF-alpha paved the way for targeted therapeutics in IBD,³¹ and today, many biologics are used routinely during pregnancy. In our study, we found that miR-21 and miR-148 were significantly downregulated in MDEs of mothers treated with anti-TNF agents compared to untreated IBD mothers. These findings correlate with the role of miR-21 in regulating intestinal epithelial tight junction permeability. Expression of miR-21 was shown to increase by TNF-alpha in a dose-dependent manner in a Caco-2 cell model and TNF-alpha was suggested to upregulate miR-21 through the NF-kappaB signaling pathway.³² The significant effect of anti-TNF treatment on the expression of miR-21 and miR-148 was further strengthened by Affymetrix analysis that showed associations within T-cell receptor and TNF signaling pathways.³³ Importantly, although most corticosteroids and immunomodulators are present in breast milk³⁴ and may potentially affect milk-derived miRNAs, we did not find any significant effects of these medications on milk-derived miRNAs. It is possible that a larger cohort may reveal statistically significant differences.

There is increasing evidence that breast-milk-derived miRNAs, particularly those encapsulated within MDEs, are stable and can be absorbed through the infant’s digestive system. Milk-derived extracellular vesicles protect miRNA from degradation allowing them to survive the gastrointestinal tract.³⁵ Studies have shown that these miRNAs are taken up by intestinal epithelial cells and accumulate in tissues such as the liver, spleen, and brain.^36,37 This stability and bioavailability suggest their role in immune regulation, metabolic programming, and epigenetic modulation. Notably, miR-21 and miR-148a, which were significantly altered in our study, are key players in immune regulation and epigenetic programming.³⁸ These findings highlight the need for further research linking milk-derived miRNA profiles with infant health outcomes.

Several limitations of our study should be acknowledged including the small sample size of untreated mothers with IBD, which is understandable since it is rare that women with IBD do not receive any medical treatment during pregnancy. Smaller sample sizes were generated for all groups when patients were subgrouped based on their medication use. To achieve the statistical power required to identify significant differences, especially among other biologics-treated groups, expansion of the IBD cohort is needed. Our cohort of IBD mothers was clinically in remission, enabling us to investigate the impact of different medical treatments on milk-derived miRs. Nevertheless, further research is needed to examine the distribution of these miRs in the breast milk of mothers experiencing flares. While our study focused on milk samples collected up to 3 months postpartum, we recognize the potential differences in miRNA expression between colostrum and mature milk, as demonstrated in our previous studies.³⁹ Researching maternal factors that influence the miRNA profile is crucial for gaining a deeper understanding of breast milk’s role in infant development. It has already been suggested that breast milk miRNAs play a role in shaping the developing infant immune system, exhibiting diverse effects. Therefore, it is important to understand these influences to fully grasp how maternal health and treatments can impact infant development through breast milk.³⁹ In summary, our study provides a comprehensive insight into the profile of MDE miRNAs in IBD for the first time. Our findings highlight the differential expression of miRNAs in CD and UC, the influence of the diseases themselves, and their associated medical treatments. While most biologics for IBD treatment are regarded as low risk during pregnancy, with only minimal traces reaching breast milk, our findings hold significant implications for the development of the newborn immune system. More studies are warranted to shed more light on the influences of IBD medications on the miRNA profile in breast milk and the implications on an infant’s health and consequently on breastfeeding recommendation in IBD.

Supplementary Data

Supplementary data is available at Inflammatory Bowel Diseases online.

Funding

This research was supported in part by the Kenneth Rainin Foundation and by the Israel Science Foundation.

Conflicts of Interest

S.R. is the CMO of EXOSOMM and R.G.-G is the CTO of EXOSOMM. The other authors declare no conflicts of interest.

References

Author notes