Age-specic different immune responses due to capabilities of secreting primal immunoglobulins responding to unexperienced antigens

Among numerous studies on COVID-19, we noted that the infection and mortality rates of SARS-CoV-2 increased with age and that fetuses known to be particularly susceptible to infection were better protected despite various mutations. Hence, we established the hypothesis that a new immune system exists that forms before birth and decreases with aging. To prove this, we analyzed the components from early pregnancy fetal stem cells cultivated in various ex-vivo culture conditions simulating the environment during pregnancy. Resultingly, we conrmed that IgM, a natural antibody produced only in early B-1 cells, immunoglobulins including IgG3, which has a wide range of antigen-binding capacity and anity, complement proteins, and antiviral proteins are induced. Our results suggest that fetal stem cells can form an independent immune system responding to unlearned antigens as a self-defense mechanism before establishing mature immune systems. Moreover, we propose the possibility of new solutions to cope with various infectious diseases based on the factors therein. expression FSCsControl and FSCsEx-vivo. of protein microarray analysis comparing the relative expression levels in supernatants of FSCsControl and to obtained through a typical 2D plate culture, from the newly established distributions in both EVs according to the culture conditions. (D) The results of ow cytometry analysis to compare the expression patterns of exosome markers (CD9, CD63, CD81) in FSCControl-EVs and NAbs-FSC-EVs. The expression aspects that CD63 and CD81 are expressed at higher levels than CD9 are similar in both EVs. (E) The results of co-expression analysis of exosome markers (CD9, CD63, CD81, and syntenin) in FSCControl-EVs and NAbs-FSC-EVs using ExoView equipment. They show expression patterns very similar to the ow cytometry results in [Fig.4D], and these common expression patterns appear to be a unique characteristic of FSCs-derived EVs. (F) The results of ow cytometry analysis comparing the presence of IgG and NAbs (IgG3 and IgM) in FSCControl-EVs and NAbs-FSC-EVs. (G) The results of exosomal cargo analysis to compare the content of NAbs in FSCControl-EVs and NAbs-FSC-EVs using ExoView®. Similar to the ow cytometry results in [Fig.4F], they show that NAbs-FSC-EVs contain higher levels of NAbs than FSCControl-EVs. (H) The results of analyzing the patterns of NAbs content in both EVs captured by each exosomal marker using Exoview®. (I) The results of protein antibody microarray analysis comparing relative expression levels of B cell-specic proteins related to the promotion of Ig transcription induced in FSCControl-EVs and NAbs-FSC-EVs. They show that B cell-specic proteins related to the promotion of Ig transcription are induced in NAbs-FSC-EVs at higher levels than FSCControl-EVs. Error bars represent standard deviation. *, P <0.05; **, P<0.01; ***, P<0.001; Student’s t-test. culture conditions (A) Various FSCs established in different culture conditions to characterize the self-defense mechanism formed by NAbs-FSCs. (B) Cytokine analysis results of EVs from FSCs obtained under each culture condition using human cytokine array. The secretion levels of pro-inammatory cytokines (IL-1β, IL-6, TNF-α) in FSCControl-EVs, itFSC-EVs, and NAbs-FSC-EVs were similar, but the secretion levels of Th1 cytokines (IFN-γ and IL-2), showing increased expression in asymptomatic infected person to COVID-19, were the highest in NAbs-FSC-EVs. (C) The results of ow cytometry analysis to compare the expression aspects of membrane-bound HLA-G1 and intracellular soluble HLA-G5/G6 in each FSC using 87G (detecting β2m of HLA-G1 and HLA-G5 isoforms) and 2A12 (detecting intron 4 of soluble HLA-G5 and HLA-G6 isoforms) antibodies. Similar to the previous study results (28), all HLA-G expressions were increased in NAbs-FSCs and itFSCs compared to FSCsControl. In addition, the expression of HLA-G in NAbs-FSCs was also higher than in itFSCs due to the effect of increased IFN-γ stimulation. (D) sHLA-G (shedding HLA-G1 and HLA-G5) and HLA-G5/6 concentration analysis results of EVs from FSCs obtained under each culture condition through ELISA analysis using MEM-G/9 (detecting β2m of HLA-G1 and HLA-G5 isoforms) and 5A6G7 (detecting intron 4 of soluble HLA-G5 and HLA-G6 isoforms) antibodies, respectively. Similar to the previous study results (28), all HLA-G concentrations were higher in itFSC-EVs and


Introduction
, which has generated worldwide pandemics, has caused previously unexperienced phenomena, such as the rapid expansion of viruses before activating the innate immunity (1), increased infection and mortality rates due to continuous mutations (2,3), severe side-effects across the whole body including the respiratory tract and nervous system (4), induction of incomplete and delayed adaptive immune response (5,6), re-infection after cure (7,8), and post-vaccination breakthrough infection (9). Particularly, we paid attention to the patterns of the asymptomatic infection rate in the young and the relatively high fatality rate of COVID-19 in the elderly. The case fatality rate (CFR) of seasonal in uenza was less than 0.1%, whereas the CFR of COVID-19 was 1.38%, but was reported to be much higher at 13.4% in those over 80 years of age (10). In contrast, the CFRs of 0-9 years old children and 10-19 years old adolescents were reported to be very low compared to adults, 0.0026% and 0.0148%, respectively (10). In addition, many studies (11)(12)(13)(14) reported that children's relatively less experienced immune system could eliminate SARS-CoV-2 much faster than adults'. In particular, fetuses and newborns considered to be susceptible to viral infection also show a relatively low infection rate (15), but the exact cause is not yet known.
However, from the studies that children infected with seasonal coronavirus already have antibodies that cross-react to unexperienced SARS-CoV-2 (16), that live vaccines vaccinated in newborns activate nonspeci c immunity that has a protective effect against a variety of antigens within 2-3 days, much faster than target pathogen-speci c immunity, which is typically activated over several weeks in adults who need to get the u vaccine every year (17), and that SARS-CoV-2-speci c IgM with a macromolecular structure that cannot be transmitted from the mother infected with SARS-CoV-2 exists in infant blood (18), we have established a hypothesis that another immune system may exist even before birth capable of protecting the fetus before the existing innate and adaptive immune systems can respond to new antigens after birth.
Con rming the possibility that FSCs in early pregnancy can produce immunoglobulins There have been reports that FSCs expressed the characteristics of hematopoietic stem cells (HSCs) (25)(26)(27). However, no study has reported that FSCs could secrete Igs produced only from B cells of HSCs lineage. To con rm the possibility of Igs production in the fetus in the early pregnancy, we obtained FSCs from amniotic uid samples donated from a healthy pregnant woman at around 10 weeks of gestation for amniocentesis. Through ow cytometry analysis on the obtained FSCs, we con rmed the expressions of various stem cell markers and Igs. Then, we performed Protein Antibody Microarray and ELISA analysis on the conditioned media obtained by culturing FSCs on a typical 2D plate. As a result, we con rmed that IgM, IgG, IgA, and IgE were expressed at very low levels in microarray analysis. Still, no Ig, especially NAbs acting on the innate immunity, was detected in ELISA analysis. From these results, we con rmed the possibility of Ig expression in FSCs. However, it was con rmed that establishing a new culture condition simulating in vivo environment capable of inducing NAbs secretion from FSCs is necessary to verify the possibility of secretion of NAbs.
Establishment of an ex-vivo culture condition simulating in vivo environment during pregnancy to induce the secretion of NAbs in FSCs In a previous study (28), we established the maternal-fetal interface-like ex-vivo culture condition to investigate the immune tolerance mechanism that completely protects the fetus from the maternal immune system during pregnancy. In addition, we demonstrated for the rst time that extravillous trophoblasts (EVTs) cultured in this condition promoted the continuous expression and secretion of HLA-G, which induces an immune tolerance environment through the activation of self-secretion of pregnancyrelated hormones, including human chorionic gonadotropin (hCG) and progesterone. In particular, we con rmed that the secretion of hCG and progesterone from trophoblast cells could be promoted cultured in maternal-fetal interface-like ex-vivo culture condition by applying temperature pro le based on the woman's body temperature change, pH, and vibration conditions to replace the internal nervous system that regulates the secretion of hormones promoting the expression and secretion of HLA-G protein that protects the fetus from the mother's immune cells. As a result, we could establish immune-tolerized trophoblasts (itTBCs) that express and secrete HLA-G consistently. Based on this, we tried to establish new ex-vivo culture conditions to induce the secretion of NAbs from FSCs in early pregnancy.
FSCs in the rst trimester of pregnancy are protected by layers of trophoblasts in direct contact with the maternal blood and decidua. Moreover, they coexist with immature immune cells of the fetus ( Fig. 2A). To simulate this in vivo environment of FSCs in the early stages of pregnancy, we devised three in vitro culture environment modules as shown in [Fig. 2B]. The rst is the maternal-fetal interface-like in vitro culture condition established in a previous study (28), in which we cultured trophoblasts on hyaluronic acid (HA)-based matrix for in vitro culture containing extracellular vesicles (EVs) obtained through coculture of amniotic uid and amnion-derived stem cells for pH and physical conditions similar to in vivo environment during pregnancy. The second is an in vitro culture condition for inducing the secretion of pregnancy-related hormones. From the possibility of Ig production of FSCs that we have identi ed, we hypothesized that pregnancy-related hormones secreted from trophoblast cells could induce HLA-G secretion protecting the fetus from the maternal immune system, and at the same time, affect the secretion of Igs, especially NAbs, protecting the fetus from foreign antigens. To prove this, we obtained itTBC-derived EVs (itTBC-EVs) containing various pregnancy-related hormones by applying the temperature pro le and vibration conditions similar to women's internal environment to maternal-fetal interface-like culture conditions, as shown in [Fig. 2C, 2D]. Lastly, to stimulate the secretion of NAbs from FSCs, we obtained the EVs (HSC/UCB-MSC CO -EVs) from serum-free co-cultivation of HSCs and umbilical cord blood mesenchymal stem cells (UCB-MSCs) under hypoxic conditions by replacing undeveloped fetal immune cells. And by applying these EVs together with itTBC-EVs, we prepared a new matrix for in vitro culture. Moreover, by applying the above temperature change, pH, and vibration conditions, we nally established the new ex-vivo culture condition simulating the in vivo environment to con rm the induction of NAbs secretion in FSCs.

Establishment and characterization of FSCs secreting NAbs
The expression and secretion of NAbs in FSCs (FSCs Ex−vivo ) cultured in new ex-vivo culture conditions established for inducing NAbs secretion were compared with those of FSCs (FSCs Control ) cultured in general 2D-plate and veri ed. [ Fig. 3D] shows the analysis of the Ig secretion patterns of FSCs Control and FSCs Ex−vivo through Protein Antibody Microarray. As a result of analyzing the expression pattern of Igs in the culture supernatant obtained through the serum-free culture of each FSC, we con rmed that the expression of all Igs such as IgG, IgM, IgE, and IgA included in the microarray were higher in FSCs Ex−vivo than in FSCs Control . These results indicate that our ex-vivo culture conditions induced the expression and secretion of Igs in FSCs.
Finally, we veri ed the secretion of NAbs in FSCs Ex−vivo through IgM and IgG3 ELISA analysis. To prove whether the expression and secretion characteristics of Igs in stem cells, not immune cells, are inherent to FSCs and whether new ex-vivo culture conditions induce the secretion of NAbs, IgM and IgG3 concentrations were analyzed by ELISA for each conditioned medium obtained by serum-free cultivation of various stem cell lines under normal 2D-plate culture conditions (2D) and new ex-vivo culture conditions (Ex-vivo). As shown in [Table.1], IgG3 and IgM were not detected in the conditioned medium of FSCs cultured under normal culture conditions (2D), but only in the conditioned medium of FSCs cultured under newly established in vitro culture conditions (Ex-vivo). In addition, IgG3 and IgM were not detected in the conditioned medium of cord blood, bone marrow, adipose, and umbilical cord-derived MSCs and HSCs regardless of the culture conditions. Moreover, we con rmed that IgG3 and IgM were not detected in all EVs used in new ex-vivo culture conditions, including itTBC-EVs, AF/AM-MSC CO -EVs, and HSC/UCB-MSC CO -EVs. These results veri ed that only FSCs in the early stages of pregnancy have the characteristic of secreting NAbs to protect themselves from external infections, and FSCs cultured in established exvivo culture conditions are NAb-secreting FSCs (NAb-FSCs).
Meanwhile, trophoblast cells secrete estrogen, another critical sex hormone, together with progesterone and hCG during pregnancy (29). We found that hCG, progesterone, and estrogen were induced in itTBC-EVs used in new ex-vivo culture conditions. A study (30) showed that estrogen promotes Ig production in human PBMCs, but no study has been reported in other types of cells. Therefore, our results that Igs were induced in FSCs Ex−vivo cultured in ex-vivo culture conditions applying itTBC-EVs containing estrogen at a higher level than in FSCs Control derived from cultivation without itTBC-EVs verify that estrogen also contributes to the promotion of Ig transcription in FSCs before B cell differentiation. Besides, our results demonstrate our hypothesis that hormones secreted during pregnancy can induce the secretion of HLA-G and Igs, thereby simultaneously protecting the fetus from the maternal immune system and foreign antigens.

Characterization of NAb-secreting FSCs-derived EVs (NAbs-FSC-EVs)
To verify whether EVs (NAbs-FSC-EVs) secreted from FSCs (NAbs-FSCs) cultured under new ex-vivo culture conditions contained NAbs, we performed various characterization and compared the results with transmission microscope (TEM, right) images of FSC Control -EVs and NAbs-FSC-EVs, respectively, and we con rmed that they show very similar size distributions and shapes.
[ Fig. 4D] represents the results of ow cytometry analysis of the expression patterns of exosome (small EVs) markers, such as CD9, CD63, and CD81, in FSC Control -EVs and NAbs-FSC-EVs. Compared to CD63 and CD81, which show similar expression levels in both EVs, we con rmed that the patterns showing a somewhat low expression level of CD9 were similar in both EVs. This result seems to be due to the unique characteristics of FSCs. In addition, [ Fig. 4E] shows the results of re-verifying the co-expression patterns of exosome markers such as CD9, CD63, CD81, and syntenin in each EV using ExoView®, which is very similar to the ow cytometry results shown in [Fig. 4D]. These results show no difference in the expression characteristics of the exosome markers of EVs secreted from each FSCs regardless of the culture conditions and also show that EVs used in the analysis are all normal exosomes. Meanwhile, to study the mechanism by which NAbs are induced at higher levels in NAbs-FSC-EVs, we performed protein antibody array analysis on the components included in FSC Control -EVs and NAbs-FSC-EVs and compared the expression patterns of B cells speci c proteins related to the transcription of Igs in both EVs (Fig. 4I).
B cell-speci c Ig gene expression and secretion are due to tissue-speci c expression of octamer transcription factors (OCT), nuclear proteins that bind to the octamer sequence ATGCAAAT motif in the promoter and enhancer of the Ig heavy chain gene (31), and another transcription factor, NF-kB, is known as an enhancer of the Ig light chain gene (32). In addition, genes specially expressed in early B cells, immunoglobulin lambda-like polypeptide 1 (Igll1, also λ5) and paired box protein 5 (Pax-5), are also known to be expressed during B cell receptor (BCR) development which is converted into a membranebound form of IgM by antigen stimulation (33,34). Furthermore, the cleavage stimulating factor (Cstf), which is known to be involved in the growth and differentiation of B cells, has been reported to convert the membrane-bound form of IgM mRNA to secreted form (35).
Our results shown in [ Fig. 4I] con rmed that various B cell-speci c and Ig transcription-related proteins were expressed in FSCs at the stage before B cell differentiation and showed that these proteins are induced at higher levels in NAbs-FSC-EVs than FSC Control -EVs. In particular, our results that NAbs-FSC-EVs contained higher levels of soluble IgM than FSC Control -EVs while the expression of membrane-bound IgM was decreased in FSCs Ex−vivo (NAbs-FSCs) than in FSCs Control were due to the effect of Cstf induced at higher levels in NAbs-FSC-EVs. Our results represented that NAbs-FSCs established by environmental factors simulated to induce secretion of NAbs may have the ability to secrete higher levels of IgG3 and IgM.

Self-defense mechanism by NAbs and complement proteins induced in NAbs-FSC-EVs
Innate immunity is a primary defense system that removes the foreign antigens by non-speci cally immediately reacting to them and activates antigen-speci c adaptive immunity. Various immune cells, including macrophages, dendritic cells (DCs), and natural killer (NK) cells, are involved in innate immunity, but their cytotoxicities are mainly induced by NAbs and complement proteins together that can bind to various antigens. Therefore, we investigated whether NAbs-FSCs secrete complement proteins capable of eliminating foreign antigens by binding with NAbs.
The complement system contributes to local in ammation, pathogen elimination and death, and the formation of the subsequent adaptive immune response through three main functions, including opsonization, chemotaxis, and lysis (36). Most of the complement proteins in the serum are produced and secreted by liver cells, and a small amount of them is secreted from endothelial cells, epithelial cells, and immune cells such as monocytes, macrophages, and DCs in a local area where the serum is restricted (37). However, there has been no report on whether stem cells secrete complement proteins. From the results of protein antibody microarray and protein absolute quanti cation analysis on the components in FSC Control -EVs and NAbs-FSC-EVs, we con rmed that various complement proteins related to three activation pathways were induced in NAbs-FSC-EVs at a higher level than FSC Control -EVs (Fig. 5A, B, C). Our results show that FSCs in early pregnancy, before the innate immune system is fully established, can form the self-defense system to protect themselves from unexperienced antigens by secreting NAbs, including IgG3 and IgM, and various complement proteins.
Identi cation of the characteristics as an immune system of the self-defense mechanism formed by NAbs-FSCs To identify the characteristics as an immune system of the self-defense mechanism formed by NAbs-FSCs and to investigate the effect of HSC/UCB-MSC CO -EVs applied in ex-vivo culture conditions as the replacement of immature fetal immune cells to establish NAbs-FSCs, we performed cytokine array analysis on FSC Control -EVs, immune-tolerized FSCs-derived EVs (itFSC-EVs), and NAbs-FSC-EVs secreted from three FSCs established under different culture conditions as shown in [Fig. 6A]. As a result, we con rmed that the secretion patterns of pro-in ammatory cytokines such as interleukin (IL)-1β, IL-6, and tumor necrosis factor (TNF)-α were similar in EVs from three FSCs, but Th1 cytokine, IL-2 and interferon (IFN)-γ, were secreted in NAbs-FSC-EVs at a higher level than itFSC-EVs and FSC Control -EVs (Fig. 6B). In these results, we noted that IFN-γ was secreted at a higher level in NAbs-FSC-EVs, obtained from FSCs cultured in new ex-vivo culture conditions including HSC/UCB-MSC CO -EVs, than in itFSC-EVs, because IFN-γ promotes the transcription of the HLA-G gene together with progesterone (38), and at the same time, acts as a transcriptional factor for various anti-viral proteins that affect mainly on the cell membrane and inside the cell, unlike Igs, NAbs, and complement proteins which function outside the cell.
To verify the effect of IFN-γ on HLA-G expression and secretion in FSCs, we compared the expression aspects of HLA-G proteins in each FSC through ow cytometry analysis. As a result, we con rmed that the expressions of membrane-bound HLA-G1 and intracellular soluble HLA-G5/G6 were more increased in NAbs-FSCs than itFSCs (Fig. 6C). Moreover, to compare the secretion aspects of HLA-G proteins in each FSC, we performed ELISA analysis for the content of HLA-G proteins on each EVs. The results showed that sHLA-G (shedding HLA-G1 and soluble HLA-G5) were secreted more in in itFSC-EVs than in FSC Control -EVs, and in NAbs-FSC-EVs than in itFSC-EVs, but HLA-G5/G6 were secreted at similarly higher levels in NAbs-FSC-EVs and itFSC-EVs than FSC Control -EVs (Fig. 6D), which can be interpreted as a difference due to the membrane-bound HLA-G1. Since itTBC-EVs, including progesterone which promotes HLA-G gene transcription, were equally applied in both ex-vivo culture conditions to establish itFSCs and NAbs-FSCs, we hypothesized that the difference in shedding HLA-G1 in NAbs-FSC-EVs and itFSC-EVs was due to the effect of different amounts of IFN-γ secreted from each FSC.
To verify this, we performed protein antibody microarray and protein absolute quanti cation analysis and compared the contents of interferon-inducing proteins in NAbs-EVs and itFSC-EVs. As a result, we con rmed that the contents of interferon-inducing proteins, including interferon-inducible transmembrane protein 3 (IFITM3) and lymphocyte antigen 6 family member E (LY6E), known to play an anti-viral role in the cellular and endosomal membranes (39,40), were induced at a higher level in NAbs-FSC-EVs than itFSC-EVs (Fig. 6E). In addition, we demonstrated that transient receptor potential mucolipin subfamily member 2 (TRPML2), which is expressed only in recycled endosomes similarly to interferon-inducing proteins and is known to inhibit virus replication through activation of anti-viral autophagy (41), was induced at a higher level in NAbs-FSC-EVs than itFSC-EVs (Fig. 6F).
Our results suggest that a more complex and sophisticated self-defense mechanism in the cell membrane and inside the cell by various anti-viral proteins induced by increased IFN-γ stimulation in addition to the extracellular defense mechanism by secretion of NAbs and complement proteins from NAbs-FSCs exists as an immune system. In addition, our ndings that NAbs, complement proteins, and various anti-viral proteins are all contained in immune-tolerized EVs expressing various HLA-G isoforms suggest that they may propose a safe response strategy not causing unnecessary immune rejection to patients in the prevention and treatment of various infectious diseases, as identi ed through previous studies (28).

Discussion
Although many studies have been made through long-term COVID-19 pandemic, con icting reports have been published on newborns and prenatal fetuses born from mothers infected with SARS-CoV-2 due to many limitations, such as di culties in designing large-scale studies and ethical issues. However, data accumulated to date on vertical infections (42,43) were obtained in the last trimester of pregnancy and after birth. Little study has been done on early pregnancy before the fetal immune system was fully established, especially on Igs that protect the fetus against external pathogens.
As a result of culturing FSCs in the rst trimester of pregnancy in ex-vivo culture conditions simulating in vivo environment of FSCs co-existing with the trophoblast layer as a protective barrier against external pathogens from mother and undeveloped fetal immune cells, we con rmed that various Igs, including IgG3, IgG, IgM, and IgA, known to be produced only in mature B cells, are induced from FSCs. We also demonstrated that the B cell-speci c proteins such as OCT-1/2, NF-kb, CstF, and Igll1, which promote the transcription and secretion of Igs, complement proteins playing an essential role in innate immunity along with NAbs, including IgM and IgG3, and various anti-viral proteins are induced. From these ndings, we veri ed that FSCs, before the immune system was thoroughly developed, can form a perfect primal immune system to protect themselves from foreign pathogens.
As below, we propose the following three mechanisms for the new immune system, which we demonstrated, never mentioned in the existing immunity.
First, we suggest the new perspective of fetal protection mechanism by hormones secreted for maintaining normal pregnancy. In other words, just as the expression and secretion of HLA-G proteins that protect the fetus from the maternal immune system are induced by hCG and progesterone secreted from trophoblasts (28), we demonstrated the transcription and secretion of Igs that protect the fetus from foreign antigens could be induced by estrogen, another pregnancy-related hormone. In addition, our results proved for the rst time that the mechanism that estrogen increases the secretion of IL-10, which promotes Igs production in B cells (44,45), was already applied to FSCs before B cell differentiation (Supplemental Fig. 1).
Second, we propose that the self-defense mechanism of FSCs is acting as an "immune system". FSCs in early pregnancy before developing the adaptive immune system by T and B cells should immediately remove the foreign antigens when the infection occurs while minimizing the damage caused by NK cells, which account for most fetal immune cells to maintain normal development. In this regard, we suggest that the following protection mechanisms by various anti-viral factors contained in NAbs-FSC-EVs exist as a sophisticated self-defense system that works organically as below (Supplemental Fig. 2 (47) Although we couldn't obtain the analysis results for more interferon-inducing proteins in NAbs-FSC-EVs due to the limitation of the protein library in the protein antibody array we used, we expect a much more elaborate defense system will protect FSCs from foreign infectious agents than our results. Therefore, we propose to de ne the defense system of FSCs as a new term, "Primal Immune System", to distinguish from the existing innate and adaptive immune systems.
Third, the most notable result is that IgG3, one of NAbs, was detected in NAbs-FSC-EVs only obtained under newly established ex-vivo culture conditions. NAbs are preimmune antibodies generated without antigen stimulation and act as a primary line of defense against infection because they can immediately nonspeci cally cross-react with various exogenous antigens (48). IgM is known as a representative NAbs, but our study shows that early pregnancy FSCs can secrete IgM, IgA, IgG, and especially IgG3, which is recently attracting attention as an ideal alternative to existing Ab-based immunotherapies (18). IgG3 has many differences that distinguish it from other IgG subclasses.
1. IgG3 has a long hinge region consisting of 11 disul de bonds to provide extended exibilities and reaches to bind various antigens of wide ranges. 2. IgG3 has various functions, including signi cant improvement of Fc effector activity, broader neutralizing e cacy, and intensively inducing adaptive immune responses, by high binding a nities to Fcγ receptors of immune cells and complement protein C1q, which induce antibody-dependent cell cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC) (49-53). 3. IgG3 has a more remarkable neutralization ability despite structural modi cation of endocytic virus into cells due to the property to form aggregates more easily than other IgG subclasses at low pH and maintain binding to receptors (54-56).
However, a more critical point than these structural features of IgG3 is that FSCs in early pregnancy possess the most naive, inexperienced Ig repertoire that can immediately recognize and respond to all foreign antigens. The primary Ig repertoire, which is formed in the early stage of B cells development at about 26 weeks of pregnancy (57,58), gradually has antigen-speci c diversity and becomes increasingly di cult to expand the repertoire for unexperienced new antigens because the number of naive B cells contributing to the formation of the primary Ig repertoire decreases, while antigen-experienced B cells accumulate due to various acquired factors with aging (59). These are supported by many studies showing that children's immune systems respond better to COVID-19 than the old and adults (11)(12)(13)(14) and that Ig repertoires responding to COVID-19 exist even in some adults not infected with COVID-19 (60).
We have demonstrated that NAbs-FSCs, established in this study by culturing FSCs obtained at around 10 weeks of gestation, that is, the stem cells in the early stage of HSCs that can differentiate into B cells, could secrete a variety of NAbs, including IgG3. Since these NAbs are not almost exposed to any antigens, they may have the "Primal Ig Repertoire" that can form a full range of primary Ig repertoire that expands later in the B cell development stage against all antigens. Therefore, to differentiate IgG3, a NAb secreted from FSCs in the early stages of pregnancy and having a primal repertoire, from the existing IgG3, which is known to be secreted from B-1 cells, we suggest that it should be rede ned with a new classi cation and terminology as "Primal Immunoglobulin (IgP)". In addition, we would like to de ne IgPsecreting FSCs established under new ex-vivo culture conditions in this study as "immune-tolerized primal immunoglobulin secreting FSCs (itPG-FSCs)". Also, considering the characteristics of the primal immune system by the newly de ned IgP, we propose a human immune response system against external infections, including SARS-CoV-2, as shown in [Supplemental Fig. 3].
Lastly, through additional studies and clinical trials of "itPG-FSC-EVs" not causing immune rejection, as demonstrated in the previous study (28), and containing all of the various anti-viral factors constituting the newly proposed "Primal Immune System", we expect that our results become new immunological strategies that can be applied to fundamentally preventive protection and treatment against foreign antigens of various infection pathways, including SARS-CoV-2.  In vivo-like temperature change, pH, and circulation conditions An in vivo-like culture condition for inducing immune-tolerized cell lines was applied using temperature change, pH, and circulation conditions established in the previous study (28). In summary, temperature change between 36.0 ℃ and 37.0 ℃ with a ve or six-day cycle and an acidic pH of 6.2 to 6.8 by HAbased matrix were applied to induce the secretion of various pregnancy-related hormones as shown in Fig.2C. Simultaneously, a circulation condition with a 24-hr cycle, as shown in Fig.2D, was applied to facilitate the signal transduction between the culture matrix and immune-tolerized cells to promote autocrine of various soluble factors.
Establishment of co-culture system of HSCs and UCB-MSCs The concentration and size of EVs were analyzed by NTA using a NanoSight NS300 with a Blue 488 nm laser (Malvern Panalytical, Malvern, UK). All samples were diluted in PBS to reach concentrations inside the precision range of the NTA machine (2 × 10 8 to 10 × 10 8 particles/ml). EVs were measured at camera level 14 (camera shutter speed: 21.48 shutters / ms, slider gain: 366). After capture, the videos have been analyzed using the in-build NanoSight Software NTA 3.4 Build 3.4.003 with a detection threshold 5.
Flow Cytometry The The concentrations of IgG3 and IgM were analyzed using Human IgG3 ELISA Kit (MyBioSource) and Human IgM ELISA Kit (Abnova, Taipei, Taiwan) according to the manufacturer's instructions, respectively.

Protein Antibody Microarray
The immunoglobulin and protein pro les in EVs secreted from FSCs were analyzed using SET100

Statistical analysis
Statistical analyses were performed using the Student's t-test for the comparison of means in more than two groups. Data presented are mean±S.D. and P<0.05 was considered statistically signi cant. In addition, maternal IgG can be transferred to the fetus through the umbilical cord after 15 weeks of gestation. However, it is not known whether the fetus produces Igs before the innate and adaptive immune system are formed. Modi ed from Fig. 1 in Ref. 23. (B) To investigate the mechanism of selfprotection by FSCs before establishing the innate and adaptive immune system by immune cells in early pregnancy, we established hypotheses and prepared a ow chart to verify them. (1) We con rmed whether Igs, particularly NAbs, including IgG3, which act on the innate immune system, were secreted from FSCs under a typical 2D culture condition. (2) We established an ex-vivo culture condition that simulates in vivo environment of FSCs to induce the secretion of NAbs from them. We con rmed the secretions of Igs and NAbs from FSCs cultured in this culture condition and compared them with FSCs cultured in a typical 2D culture condition.   Analysis results of various complement proteins contained in NAbs-FSC-EVs The complement system is activated by the classical pathway (CP), the alternative pathway (AP), and the lectin pathway (LP). C1

Figures
subcomponents such as C1q, C1r, and C1s form the C1 complex in the CP, and the C5, C6, C7, C8, and C9 complement proteins form the membrane attack complex. C3a, C4a, and C5a proteins, generated in this process, recruit various immune cells to spread the in ammatory response (36). Comparisons of relative expression levels of complement proteins contained in NAbs-FSC-EVs and FSCControl-EVs using SET100 protein antibody microarray (A) and human L493 array (B). (C) Comparisons of absolute quanti cation of complement proteins contained in NAbs-FSC-EVs and FSCControl-EVs using ESI-Q-TOF MS/MS. We con rm that NAbs-FSC-EVs contain higher levels of various complement proteins compared to FSCControl-EVs in all analysis results. Error bars represent standard deviation. *, P <0.05; **, P <0.01; ***, P<0.001; ****, P<0.0001; Student's t-test. Figure 6