Discussion
Main Findings In this study, we found that higher Tfh1/Tfh2 ratio before pregnancy was seen in females with early AAM, and this increased Tfh1/Tfh2 ratio was significant with higher TSH level at early pregnancy. Moreover, the aberrant Tfh1/Tfh2 ratio was also related to a higher risk and a shorter time to spontaneous miscarriage. Lastly, The contribution of early AAM on hallmarks of Tfh cells was further validated by a two sample MR test.
Interpretation Tfh cells are differentiated from Naïve CD4 cells, and is required for B cell development(29). They interact with antigen-specific B cells in germinal center (GC), and promote somatic hypermutation, class switch recombination and affinity maturation of B cells(30). In addition, Tfh cells also facilitate the differentiation of GC B cells, transforming them into memory B cells and long-lived plasma cells(31). The hallmarks of the Tfh cells will lead to the uncontrolled activation of B cells, thus produce a large amount of antibodies(32). There were three subsets of Tfh cells in peripheral blood, Tfh1(CXCR3+CCR6-), Tfh2(CXCR3-CCR6-) and Tfh17 (CXCR3-CCR6+) cells28,29. Among of them, Tfh1 cells are major regulators of proinflammatory stress, and can strengthen the immune reactions(33); While Tfh2 cells induce the differentiation of naïve B cells into plasmablasts(34). Therefore, increase in Tfh1/Tfh2 ratio usually implicates an inflammation.
To the best of our knowledge, the present study revealed the association of early AAM with Tfh cells in women for the first time. The present study showed that genetic factors of AAM contributes to the variances of PD1- naïve Tfh cells. Compared to PD1+ Tfh cells, PD1- naïve Tfh cells are less proliferative and produce less cytokines(35). It implied that early AAM will impair Tfh progenitor cells, thus aberrant Tfh1/Tfh2 ratio and unregulated humoral immunity. In shared genetic factors of AAM and Tfh cells, LIN28B encodes a key repressor of let-7 microRNA biogenesis and cell pluripotency(36). Transgenic Lin28b mice demonstrate altered pubertal growth(37). Moreover, LIN28B/let-7 pathway influences Naïve CD4 cells survival(38) and B cell development(39). let-7 is also aberrantly expressed in the embryonic chorion tissue of spontaneous abortion(40). Therefore, it is biological plausible thatLIN28B involves the etiology of early AAM, aberrant Tfh cell level, and miscarriages. For chromosome region 9p12, it includes three AAM-related genes LINC01505 (26), TAL2 (27) andTMEM38B (28). Among of them, TAL2 is crucial for thymocyte development(41), and chromosome translocation involving this gene occurs in T cell acute lymphoblastic leukemia(42). Therefore, chromosome region 9p12 is linked to both AAM and immunity. Taken together above reports, our findings on the association of AAM and Tfh cells are supported by genetic evidences. Additionally, the longer hormone exposure and obesity, which are produced by early AAM(13), both are found to play pivotal roles in the function of Tfh cells in recent years(43-45). Thus, our findings on the relationship between early AAM and aberrant Tfh cells function are credible.
Another important finding of this study is the association between pre-pregnancy Tfh1/Tfh2 ratio and spontaneous miscarriage. Because humoral immunity plays a pivotal role in miscarriages(46), imbalance of pro-inflammatory and anti-inflammatory cytokines may lead to impairment of pregnancy(47). The relationship between Tfh cells and spontaneous miscarriage is biological plausible. Different types of Tfh cells can secrete different antibodies: Tfh1 cells secrete IFN-γ, which can trigger to produce IgG2α, then cause abortion; Tfh2 secrete IL-4, which can trigger to produce IgG 1 and IgE, and was down-regulated in abortion women(48); Tfh17 secrete IL-17, which triggers IgA(49, 50). It can be concluded that the balance of Tfh1/Tfh2 is one of the important immune factors for abortion. When Tfh1 is dominant, the proportion of IFN-γ increases and IL-4 decreases, which will lead to abortion. Therefore, It is credible that Tfh cells, especially Tfh1 cells, mediates the process of early AAM related miscarriages, which was found in this study.
In this study, we also found that the pre-pregnancy ratio of Tfh1/Tfh2 was associated with TSH at early pregnancy. Thyroid diseases in women of childbearing age are mainly caused by autoimmune diseases(51). Autoimmune thyroid disease (AITD), including autoimmune hypothyroidism (Hashimoto’s thyroiditis, HT) and autoimmune hyperthyroidism (Graves’ disease, GD). It has been found that the percentage of Tfh cells is higher in HT patients and Tfh cells exist in the thyroid, suggesting that Tfh cells drive autoimmunity and participate in the pathogenesis(52). Abnormal immune microenvironment results in increased expression of CXCR5 in thyroid tissue, which leads to immune response of Tfh cells and the development of AITD(53). The disorder of thyroid function in pregnancy can cause hyperthyroidism or incomplete thyroid function. Women with hyperthyroidism during pregnancy are more likely to have obstetric complications such as miscarriage, premature birth, abruptio placentae, thyroid store, and heart failure or adverse neonatal outcomes such as prematurity, LBW, intrauterine growth restriction (IUGR) and still birth than normal women(17). Some studies have pointed out that even in healthy women without thyroid dysfunction, the risk of miscarriage, fetal death or neonatal death increases as TSH levels rise(54). Thus, we infer that the changes of the pre-pregnancy ratio of Tfh1/Tfh2immune function before pregnancy will affect the thyroid function during pregnancy, thus affecting the outcome of pregnancy.
Because early AAM also relates to many diseases, such as cardiovascular diseases(5), autoimmune diseases(7) and cancers(55), our study may hold potential to initiate new avenues of research.
Strengths and Limitations The most strength of the present study is using multiple analyses to show the mediation of Tfh cells in the process of early AAM-induced subsequent pregnancy events. First, we performed Spearman correlation analyses to study the relationship between AAM and Tfh cells at the pre-pregnancy examinations. Second, we used Manny-Whitney U tests, Spearman correlation analyses, Kaplan-Meier estimate and Cox regression analyses to investigate the association between AAM-related pregnancy indexes by a follow-up study. The pregnancy indexes included pregnancy status, thyroid function at early pregnancy, and metabolic indexes at mid pregnancy, gestational weeks and birth weight of newborns. Last but not the least, we validated the association between early AAM and aberrant function of Tfh cell by a two-sample MR test. Therefore, this preliminary work may provide a proof-of-concept for evaluating the role of Tfh cells in early AAM-related reproductive events.
The present study has some limitations: First, the pilot study was conducted in a small Chinese population, while the genetic data that support its findings were collected from two consortiums that were of European origin. The generalizability of our findings needs to be confirmed. Second, other productive events such as ectopic pregnancies cannot be analyzed because of the limited sample size. Third, many women have improved their awareness of pregnancy examination, such as blood glucose and blood pressure during pregnancy, which have been well controlled. Forth, there were only 497 females in TwinsUK. To avoid overestimate of locus-specific effects from GWAS data(56), the genetic contribution of Tfh cells to spontaneous miscarriages was not analyzed in this study.
Conclusion The present study found that early AAM may produce hallmarks of Tfh cells, then induce subsequent reproductive events, such as miscarriages and higher TSH levels. Our results may provide a new concept on the etiology of the AAM-related reproductive events. Larger population studies and functional studies are warranted.