Abstract
Background
There are few data on oxidative stresses during and after pregnancy, although aggravation of autoimmune disease is implicated in oxidative stress and occurs frequently in the postpartum period. Thioredoxin (TRX) is a stress-inducible protein, and is used as a good biomarker for oxidative stress. To clarify the changes in the levels of oxidative stress during and after pregnancy, we examined serum TRX levels and the numbers of lymphocyte subsets.
Methods
We measured serum TRX levels by enzyme-linked immunosorbent assay (ELISA), and neutrophils, lymphocytes, and CD4 and CD8 lymphocytes by flow cytometry in peripheral blood from 88 healthy pregnant women, 26 just after delivery women, 77 healthy postpartum women and 19 healthy non-pregnant women.
Results
The serum levels of TRX did not change during pregnancy, but increased in four, seven and 10 months postpartum. Serum TRX levels were correlated with the percentages of neutrophils in normal non-pregnant women and women one month postpartum, and with those of CD8 lymphocytes in early pregnant women and women one and four months postpartum.
Conclusions
Oxidative stress increased in the postpartum period, and the levels at one and four months postpartum were related to CD8 lymphocytes.
Introduction
Pregnancy affects the pathologic condition of various diseases. 1,2 It is well known that autoimmune diseases, such as Hashimoto's disease, 3 rheumatoid arthritis, 4 and multiple sclerosis, 5 often remit during pregnancy and are aggravated again at one to four months postpartum. Aggravation of autoimmune disease has been implicated in oxidative stress. 6 Therefore, it is thought that the activity of oxidative stress is lower during pregnancy and higher in the postpartum period. Thioredoxin (TRX) is a stress-inducible protein, which is released from cells in response to oxidative stress, 7–10 and plays a protective role against oxidant injury. 11,12 Recently, it has been shown that the serum concentration of TRX is a good biomarker of oxidative stress induced by various conditions, such as autoimmune disease, viral infection, postischaemic reperfusion, heart failure, X-ray, ultraviolet irradiation and so on. 6,13,14 In this study, we examined serum concentration of TRX in healthy pregnant and postpartum women in order to examine the level of oxidative stress during and after pregnancy. Furthermore, we analysed the relations between TRX and inflammatory cells such as neutrophils, lymphocytes, and CD4 and CD8 lymphocyte subsets, which change markedly during and after normal pregnancy. 15
Materials and methods
Subjects
The study involved nine groups of healthy women. Group 1 comprised 33 women in the first trimester of pregnancy (7–13 weeks; age, 21–37 y; mean age, 28.2 y), group 2 comprised 26 women in the second trimester (19–25 weeks; age, 22–39 y; mean age, 29.1 y) and group 3 comprised 28 women in the third trimester (30–38 weeks; age, 23–37 y; mean age, 29.3 y). Group 4 comprised 26 women at the time of delivery (age, 25–38 y; mean age, 30.3 y). Groups 5, 6, 7 and 8 were made up of postpartum women: group 5 comprised 27 women one month postpartum (age, 19–39 y; mean, 30.7 y), group 6 comprised 22 women four months postpartum (age, 20–38 y; mean age, 27.5 y), group 7 comprised 14 women seven months postpartum (age, 20–35 y; mean age, 26.6 y) and group 8 comprised 14 women 10 months postpartum (age, 21–38 y; mean age, 28.9 y). Group 9 comprised 19 healthy non-pregnant women as control subjects (age, 21–45 y; mean age, 30.3 y). The mean age did not differ significantly between the groups. None of the subjects were receiving any medication at the time of the study, and none had any associated complications or autoimmune disease, as determined by clinical and laboratory examinations, including serologic tests for antinuclear antibody, rheumatoid factor and anti-thyroid microsomal antibody. Blood samples were collected, and serum samples were separated and stored at −70°C until use. The research protocol was in accordance with the ethical requirements of the Helsinki Declaration and all subjects enrolled gave informed consent.
Enzyme-linked immunosorbent assay (ELISA)
TRX levels were measured in serum samples with a sandwich ELISA kit (Redox Bioscience Inc, Kyoto, Japan). 16 In brief, 96-well microplates were coated with anti-human TRX monoclonal antibody, samples to be measured or standards at final dilutions of 1:10 were applied to the plate and the plates were incubated at 25°C for 2 h. After five washes, the plates were incubated with peroxidase-conjugated anti-human TRX monoclonal antibody. After a washing to remove unbound antibody, the peroxidase substrate was mixed with the chromogen and allowed to incubate for 30 min. An acid solution was then added to each well to terminate the enzyme reaction and to stabilize the developed colour. The optimal density of each well was then measured at 450 nm. Results (expressed as ng/mL) were determined from standard curves with six standards (ranging from 0 to 150 ng/mL).
Neutrophils, lymphocytes and lymphocyte subsets
The proportions of neutrophils and lymphocytes in leukocytes were determined with an automated differential leukocyte counter (H6000; Technicon Instrument Co, Tarrytown, NY, USA). 17
The fluorescein isothiocyanate (FITC)-conjugated anti-CD4 and anti-CD8 monoclonal antibodies (Ortho-McNeil Pharmaceutical, Raritan, NJ, USA) were used. CD8 lymphocytes were analysed with a Spectrum III automated flow cytometer (Ortho Diagnostic Systems Inc, Westwood, MA, USA) with whole-blood samples. 18
Statistical analysis
To assess the statistical differences among the data, we first applied one-way analysis of variance for groups designated as non-pregnancy through third trimester of pregnancy (among the four groups). Then we applied Dunnett's multiple comparison between non-pregnant control and group of each trimester. The same procedure was also applied to groups between delivery through 10 months postpartum. To bridge these two sets of testing, we tested for differences between non-pregnant control and group of women at delivery using student's t test.
To analyse the correlations between the data, linear regression analysis and the Pearson's correlation coefficient were used.
Probability values of less than 0.05 were considered significant.
Results
Serum TRX concentrations
Serum TRX concentrations did not change significantly between the first, second and third trimesters of pregnancy (57.4 ± 25.6, 66.3 ± 25.0 and 65.5 ± 23.2 ng/mL, respectively) compared with the concentration in non-pregnant control subjects (47.9 ± 26.3 ng/mL; Figure 1). Serum TRX concentrations at four, seven and 10 months postpartum (88.0 ± 20.3 ng/mL, P < 0.0001; 86.8 ± 39.3 ng/mL, P < 0.0002 and 80.7 ± 20.6 ng/mL, P < 0.01, respectively) were significantly greater than the concentration at non-pregnant control subjects and delivery (49.9 ± 23.4 ng/mL).
Serum concentrations of TRX in normal pregnant and postpartum women (*P < 0.01, **P < 0.0002, ***P < 0.0001). TRX, thioredoxin
Serum TRX concentrations were not significantly different between the non-pregnant control subjects and women at delivery.
Neutrophils, lymphocytes and lymphocyte subsets
The percentages of neutrophils were significantly increased during pregnancy (first trimester, n = 33, 74.0 ± 5.6%, P < 0.001; second trimester, n = 26, 75.7 ± 3.5%, P < 0.001; third trimester, n = 27, 73.1 ± 3.9%, P < 0.01), but decreased at four and seven months postpartum (4 months, n = 22, 57.9 ± 7.1%, P < 0.05; 7 months, n = 14, 57.0 ± 4.6%, P < 0.05) compared with the percentage in non-pregnant control subjects (n = 19, 65.3 ± 5.4%). The percentages of lymphocytes were significantly decreased during pregnancy (first trimester, n = 33, 20.2 ± 5.2%, P < 0.001; second trimester, n = 26, 18.6 ± 3.4%, P < 0.001; third trimester, n = 27, 20.1 ± 3.7%, P < 0.001), but increased at four and seven months postpartum (4 months, n = 22, 36.3 ± 6.7%, P < 0.05; 7 months, n = 14, 38.2 ± 4.5%, P < 0.01) compared with the percentages in non-pregnant control subjects (n = 19, 29.5 ± 4.6%).
Percentages of CD4 lymphocytes did not change during or after pregnancy, with the exception of a decrease in the percentage of CD4+ T cells in the first trimester of pregnancy (n = 33, 37.0 ± 5.8%, P < 0.05) compared with that in non-pregnant control subjects (n = 19, 40.8 ± 4.2%). Percentages of CD8 lymphocytes did not change during or after pregnancy compared with that in non-pregnant control subjects (25.0 ± 3.2%).
Correlation between TRX and leukocyte subsets
Serum TRX concentrations were positively correlated with the percentages of neutrophils in non-pregnant women (n = 19, r = 0.49, y = 0.10x + 60.4, P < 0.05) and women one month postpartum (n = 27, r = 0.52, y = 0.12x + 54.5, P < 0.01). On the other hand, serum TRX concentrations were positively correlated with the percentages of CD8+ T cells in the first trimester of pregnancy (n = 33, r = 0.46, y = 0.11x + 21.3, P < 0.01) and at one and four months postpartum (n = 27, r = 0.40, y = 0.09x + 20.8, P < 0.05; n = 22, r = 0.55, y = 0.13x + 13.5, P < 0.01, respectively) (Figure 2).
Correlations between serum TRX levels and the percentages of neutrophils and CD8+ lymphocytes in healthy non-pregnant women (a), and in healthy women in the first trimester of pregnancy (b), in women one month postpartum (c), and in four months postpartum (d). TRX, thioredoxin
Discussion
We estimated that oxidative stress and TRX levels may decrease during pregnancy, since autoimmune diseases, such as Hashimoto's disease, rheumatoid arthritis and multiple sclerosis, often remit during pregnancy. 1,3–5 However, TRX mRNA was reported to be increased in the cervix of women during late pregnant and in the postpartum period. 19 In addition, umbilical TRX concentrations were reported to be six to seven times higher than the serum concentrations of adults. 20 These data indicate that oxidative stress and TRX may increase locally in uterus during pregnancy. However, serum TRX levels did not change during pregnancy. Thus, the distinct changes in immune system and uterus may result in overall unchanged serum TRX levels during pregnancy.
We did observe increase in serum TRX concentrations at four to 10 months postpartum. This increase of TRX is considered to be induced by the physiologic increase of oxidative stress at one to four month postpartum, since both helper T lymphocytes and cytotoxic T lymphocytes are increased, 15 and autoimmune diseases, such as Hashimoto's disease, rheumatoid arthritis, and multiple sclerosis, become aggravated in this postpartum period. 3–5 Furthermore, this postpartum increase of TRX continued up to 10 months postpartum. Similar postpartum long-term changes were also observed in peripheral lymphocyte subsets 15,18 and serum cytokine levels, 21 and it took about one year for these levels to recover to the non-pregnant women levels completely. Therefore, serum TRX may take one year and more to recover to the non-pregnant women levels, since TRX increases in response to the oxidative stress such as inflammation. These physiological changes are considered to cause various postpartum changes in various diseases. 22,23
The changes of neutrophils, lymphocytes, and CD4 and CD8 lymphocytes were similar to previous reports. 15,18,24
There have been no reports on the correlations between serum TRX levels and the percentages of inflammatory cells, such as neutrophils, lymphocytes, and CD4 and CD8 lymphocyte subsets. In order to clarify the inflammatory cells related to the level of oxidative stress in peripheral blood, we examine them and found that serum TRX levels were correlated with the percentages of neutrophils in healthy non-pregnant women. This finding means that the level of oxidative stress may be related to neutrophils which always defend the body against foreign materials, such as pathogenic bacteria, in healthy non-pregnant women. Also we found that serum TRX levels were correlated with the percentages of CD8 lymphocytes in the first trimester of pregnancy and at four months postpartum. This finding also means that the level of oxidative stress may be related to CD8 lymphocytes, which react against newly-appearing foreign materials, such as embryo, tumour cells and the cells denatured by viruses and so on, in early pregnancy and at four months postpartum. In this regard, the activity and the number of natural killer cells, which are weakly positive for CD8 molecules, increase physiologically in the first trimester of pregnancy. 15,25 This increase of NK activity is important for the maintenance of pregnancy, 26,27 but the excessive killer activity induces spontaneous abortion. 28 Also, the number of CD8+CD11b− cytotoxic T cells increase physiologically at one to four months postpartum. 15 This increase is considered to be caused by the rebound from the immune suppression in pregnancy and to induce postpartum aggravation of autoimmune diseases in the patients. Furthermore, serum TRX levels were correlated with both the percentages of neutrophils and CD8 lymphocytes at one month postpartum, suggesting both contributions to the level of oxidative stress at one month postpartum.
In conclusion, serum TRX levels did not change during pregnancy, but increased at four to 10 months postpartum, suggesting postpartum long-term increase of oxidative stress. The TRX levels may be related mainly to neutrophils in non-pregnant women, and to CD8 lymphocytes in early pregnant women and in women one to four months postpartum.
DECLARATIONS