Cytokine stimulation of the choriocarcinoma cell line JEG-3 leads to alterations in the HLA-G expression profile
Introduction
The expression of paternally-derived antigens by fetal cells during pregnancy requires a temporary induction of tolerance by the maternal immune cells to allow the human fetus to grow and thrive for nine months in the uterus of the mother. The development of the placenta, the primary contact zone between mother and fetus, is reliant on the invasion of fetal cytotrophoblast cells into the maternal tissue and interactions with the maternal immune cells to induce immunological tolerance and acceptance of the semi-allogenic fetal cells [1], [2]. Insufficient immune tolerance and poor placentation have been associated with pregnancy complications such as recurrent spontaneous abortion, implantation failure, fetal growth restriction and preeclampsia [1], [3]. One of the important factors that allow generation of tolerance is the expression of the non-classical human leukocyte antigen (HLA)-G by the trophoblast cells in the placenta (reviewed by [4]). HLA-G has been shown to be involved in development of immunological tolerance towards the fetus by mediating apoptosis in allo-reactive cytotoxic CD8+ T cells and NK cells, and induce development of tolerant antigen-presenting cells and CD4+ T cells [5], [6], [7]. In contrast to the classical HLA class Ia (HLA-A, -B and -C), HLA-G has a restricted tissue expression pattern and low allelic polymorphism [4], [8], [9], [10]. Due to its broad immune tolerogenic properties, HLA-G is emerging as an important immune checkpoint molecule also relevant in cancer, autoimmune diseases and in transplantation [8].
Although the use of HLA-G is of potential therapeutic interest, its use is complicated as HLA-G is expressed in several different splice variants, supposedly with different functions [11], [12]. The HLA-G gene consists, like other HLA class I genes, of eight exons. At present there are discrepancies regarding the exon/intron structure. The Ensemble database includes an exon (exon 1) up-stream of the primary translational start site found in exon 2, while the International Immunogenetics Database (IMGT/HLA) consider this exon as the first exon, and includes a small un-transcribed exon 7 in the 3′ end. To avoid confusion, the nomenclature from the Ensemble database will be used throughout this study (Fig. 1). Due to alternative splicing of the HLA-G mRNA, four membrane bound (HLA-G1-4) and three soluble (HLA-G5-7) isoforms have been recognized [13], [14]. In addition to the full length, membrane-bound protein HLA-G1, smaller isoforms have been identified that either lack the α2 domain (HLA-G2 and -G6), the α3 domain (HLA-G4) or both (HLA-G3 and -G7) due to skipping of exon 4 and 5, respectively. The retention of intron 3 and 5 introduce a pre-terminal stop codon, resulting in isoforms lacking the transmembrane region, thus coding for the proteins being secreted (sHLA-G5-7; cf. Fig. 1). A 14 base pair (bp) insertion (ins) polymorphism in the 3′ untranslated region results in further splicing of a 92 bp fragment from exon 8 ([12], [15], Fig. 1). Additionally, the identification of novel HLA-G isoforms has recently been reported, which include isoforms with transcription start site in exon 1, 4 or 5, the latter two yielding isoforms lacking the α1 domain; skipping of exon 6 and 7, which would yield novel soluble isoforms; retention of intron 1, and retention of intron 3, either in combination with intron 4 or intron 5 [16]. The HLA-G isoforms have different ability to dimerize and are found associated with or without β2-microglobulin (β2m) [17], [18]. Further complicating the picture, HLA-G has also been found secreted in exosomes released from first trimester placentas [19], and the membrane-bound HLA-G1 can be shed from the cell surface by proteolytic cleavage, thus existing as a soluble form [20].
The dissemination and function of these isoforms in different tissues and in pathological settings are unknown. However, it is suggested that the HLA-G isoforms might play different functions, as they seem to interact differently with the receptors. Whereas the inhibitory receptor Ig-like transcript 2 (ILT-2) (LILRB1/CD85j) is expressed by a diverse set of immune cells including monocytes, B cells, T cell subsets and NK cells and binds HLA-G1 and HLA-G5 in a β2m-dependent fashion, the myeloid-specific ILT-4 (LILRB2/CD85d) receptor, selectively expressed by myelomonocytic cells, has highest affinity for the β2m-free isoforms [21], [22], [23], and has been shown to bind HLA-G1/-G5 and HLA-G2/-G6, but not HLA-G3 [17]. The receptor KIR2DL4 (CD158d), expressed by NK cells and a subset of T cells, recognizes the α1 domain, which is present in HLA-G1-7, and has been shown to harbor both activating and inhibitory functions [24], [25]. In addition, CD160 expressed by NK cells, subsets of T cells and endothelial cells, and the CD8 receptor found among T and NK cell subsets has also been shown to bind HLA-G isoforms, although the exact mechanism and function is not clear [5], [26], [27]. Altogether, HLA-G expression pattern is complex, proposing multiple functions and means of regulation.
Which factors and mechanisms that drive the expression of these different isoforms are not known, nor is the exact function. A successful pregnancy is highly dependent on a balance between Th1, Th2, Th17 and regulatory responses dynamically modulating responses between maternal and fetal cells throughout pregnancy [28], [29]. Interestingly, Svendsen et al. [30] showed, by using the retinal pigment epithelium cell line ARPE-19, that interferon (IFN)-γ induced expression of HLA-G1, G3, G2/4 and low levels of G5, and tumor necrosis factor (TNF)-α induced expression of G2/4 in addition to G1, whereas the unstimulated cells expressed only low levels of HLA-G1.
As it is difficult to study trophoblast cell function in vivo, the HLA-G+ choriocarcinoma cell line JEG-3 is widely used as an in vitro model for the human trophoblast cells. JEG-3 cells express HLA class Ib and have no expression of HLA class Ia, except HLA-C, thus resembling extravillous trophoblasts. HLA-G expression in JEG-3 can be induced by interferons, including IFN-γ, although others have failed to prove this [30], [31], [32]. The regulation of expression of HLA-G isoforms by the microenvironment in this cell line has to our knowledge not previous been studied systematically and in details.
We hypothesized that the local environment, i.e. the cytokine milieu, could regulate both the expression and the isoform profile of HLA-G. Therefore, the aim of the present study was to study the effect of anti- and pro-inflammatory cytokines (IL-2, IL-4, IL-6, IL-10, IL-12, IL-15, IL-17A, TGF-β1, TNF-α and IFN-γ1b) on the regulation of HLA-G expression in JEG-3 as an in vitro model for the fetal trophoblast cells.
Section snippets
Cells and culture conditions
The human choriocarcinoma cell line JEG-3 (ATCC, #HTB-36, Manassas, Virginia, USA) was cultured in growth medium containing Eagle’s Minimum Essential Medium (EMEM; ATTC®30-2003™) supplemented with 10% heat-inactivated Fetal Bovine Serum (FBS; Sigma-Aldrich, St. Louis, Missouri, USA) at 37 °C and 5% CO2. For the cytokine assays, the cells were grown in six-well plates to a 40–50% confluency and stimulated for 24 or 72 h with fresh growth medium containing varying concentrations of IL-2, IL-6,
HLA-G expression is down-regulated in JEG-3 cells in response to both anti- and pro-inflammatory cytokines
The flow cytometry gating strategy is shown in Fig. A in the Appendix. Incubation with IL-12, IL-15, IL-17A and TGF-β1 had no effect on the level of HLA-G surface expression at any of the time points studied (Fig. B, Appendix). As shown in Fig. 2A, 500 U/ml IL-4 led to a small and transient increase in MFI after 24 h of stimulation (P = 0.0282), while HLA-G surface expression level was significantly decreased upon incubation with 2, 20 and 200 U/ml IL-2 for 72 h (P = 0.0014; 0.0055 and 0.006,
Conclusion
The impact of a range of cytokines on the expression of HLA-G mRNA isoforms has to our knowledge not previously been analyzed systematically in JEG-3 cells, which is a widely used cell line. A summary of the findings can be found in Table 2. In the current study we combined mRNA analysis with protein analysis using flow cytometry and ELISA. We show that the level of HLA-G protein surface expression is down-regulated by IL-2, IL-6 and TNF-α, while upregulated by IL-4. Contrasting with findings
Acknowledgements
Thanks to the technical personnel at the Department of Clinical Biochemistry, with a special thanks to Asker Marquard Thomasen and Iben Weisdorf, who assisted in performing fragment analysis and ddPCR, respectively. Also, thanks to Sara Maj Hyldig Matzen and Line Lynge Nilsson, always willing to answer any questions related to flow analysis and ELISA.
Funding
This project was generously supported by The Region Zealand Health Sciences Research Foundation and Zealand University Hospital through The ReproHealth Research Consortium ZUH.
Declaration of interest
The authors declare that they do not have any commercial or financial relationships that could be construed as a potential conflict of interest.
Author’s contribution
GP and TVFH planned and designed the experimental setup. For the JEG-3 analysis GP and JBSB conducted the cell assays and flow cytometry analysis, while GP performed ddPCR, final fragment analysis and ELISA. TGL assisted in cell assays, while CI contributed to fragment analysis and data interpretation of JEG-3. AMB and MSB designed and optimized the ddPCR analysis. Final statistical analysis and data interpretation was done by GP and TVFH. GP wrote and drafted the manuscript with contributions
References (57)
- et al.
The role of decidual immune cells on human pregnancy
J. Reprod. Immunol.
(2017) - et al.
Immune modulation treatments—where is the evidence?
Fertil. Steril.
(2017) - et al.
Identification of HLA-G7 as a new splice variant of the HLA-G mRNA and expression of soluble HLA-G5, -G6, and -G7 transcripts in human transfected cells
Human Immunol.
(2000) - et al.
Immunomodulatory molecules are released from the first trimester and term placenta via exosomes
Placenta
(2012) - et al.
Soluble HLA-G1 inhibits angiogenesis through an apoptotic pathway and by direct binding to CD160 receptor expressed by endothelial cells
Blood
(2006) - et al.
Expression and differential regulation of HLA-G isoforms in the retinal pigment epithelial cell line, ARPE-19
Hum. Immunol.
(2017) - et al.
Cytokine regulation of HLA-G expression in human trophoblast cell lines
J. Reprod. Immunol.
(1995) - et al.
Low plasma HLA-G protein concentrations in early gestation indicate the development of preeclampsia later in pregnancy
Am. J. Obstet. Gynecol.
(2005) - et al.
Differential in vivo and in vitro HLA-G expression in melanoma cells: potential mechanisms
Hum. Immunol.
(2003) - et al.
A novel bridge between oxidative stress and immunity: The interaction between hydrogen peroxide and human leukocyte antigen G in placental trophoblasts during preeclampsia
Am. J. Obstet. Gynecol.
(2012)