CD138 Expression Promotes Accumulation and Activation of T Cells in Autoimmune MRL/lpr Mice

CD138+ T cells that accumulated in Fas-deciency lupus mice, had been identied as autoreactive T cells in SLE which signicantly promoted autoantibody secretion. In present study, we found CD138 expression in T cells played a key role in the progression of SLE in MRL/lpr mice. Our results indicated CD138+ T cells apoptosis was in Fas dependent way. However, CD138 expression of T cells in MRL/lpr mice could signicantly prevent T cells apoptosis, contribute to accumulation of T cells and DN T cells and simultaneously promote T cells activation. Importantly, CD138 expression in DN T cells signicantly increased FasL expression of DN T cells enhancing the cytotoxity of DN T cells. Phorbol 12-myristate 13-acetate and Ionomycin (PI) stimulation could signicantly prevent CD138+ T cells accumulation by strikingly inducing their specic apoptosis. Moreover, PI stimulation signicantly activated CD138+ T cells with increased CD69 expression in them. Importantly, our results showed CD69 expression in CD138+ T cells could signicantly increase the apoptosis level of them. That indicated PI stimulation could induce specic apoptosis of CD138+ T cells via increasing CD69 expression in CD138+ T cells. In addition, our results showed CD138- T cells in MRL/lpr mice had signicant defects in activation. However, to activate T cells could signicantly prevent CD138 expression in CD3+ T cells of MRL/lpr mice. Our results suggested CD138 expression in CD3+ T cells of MRL/lpr mice was probably caused by the failure of activation in autoreactive T cells before self-antigens exposure to immune system.


Introduction
Systemic lupus erythematosus (SLE) is the autoimmune disease characterized by the production of multiple autoantibodies including anti-nuclear antibody (ANA) and anti-double-stranded DNA (dsDNA) antibody 1,2 . Productions of these autoantibodies make a detrimental effect on multiple tissues and organs [3][4][5] . However, SLE is the autoimmune disease with a so complex mechanism that both T cells and B cells participate in the progression 6-8 .
CD138 is a marker of plasma cells which have been believed to be originated from activated B cells 9,10 .
While recent years, CD138+ T cells, which express both CD3 and CD138 on their surface, were reported to be plasmablastic B-cell neoplasms in clinical cases 11 , and also have been identi ed in SLE murine model [12][13][14] . CD138+ T cells have been shown to signi cantly accumulate in Fas-de ciency lupus mice [12][13][14] .
Previous research has demonstrated double negative (DN) T cells play an important role in the progression of lupus and signi cantly contribute to the tissue injury of SLE 6, 15,16 . Plasma cells accumulation were also the cardinal feature of SLE 17,18 . While interestingly, our present studies showed that the majority of CD138+ cells in SLE murine model were CD138+ T cells. Morever, most of CD138+ T cells were also DN T cells that were CD4 and CD8 double negative. Importantly, recent research has demonstrated that CD138+ T cells signi cantly could signi cantly promote autoantibody production both in vivo and in vitro with a CD4 receptor dependent way indicating that CD138+ T cells are those autoreactive T cells that accumulated in MRL/lpr mice 12 . So, CD138+ T cells may be key to uncover the underlying mechanism in the progression of SLE. However, the mechanism of CD138+ T cells accumulation in splenocytes of MRL/lpr mice was still undeciphered. In present study, we tried to investigate the mechanism of CD138+ T cells accumulation in MRL/lpr mice.
Phorbol 12-myristate 13-acetate (PMA) and Ionomycin are commonly used to promote cellular activation in vitro culture 19,20 . Surprisingly, we found PMA and Ionomycin (PI) stimulation could signi cantly prevent CD138+ T cells accumulation in splenocytes of MRL/lpr mice. PI stimulation progressively decrease the CD138+ T cells frequency as the time of PI stimulation increased. Based on these evidences, we further sought the underlying mechanisms of CD138 expression in CD3+ T cells and of PI stimulation preventing CD138-expressing T cells accumulation.

Animals
Female MRL/lpr mice and MRL/MPJ mice were purchased from the Slac Laboratory (Shanghai, China).
Mice were housed at 22 ± 1°C with a relative humidity of 50-60% with a 12-hour light/dark cycle. All

Cell Culture in vitro
Mice were euthanized at the 17-18th week age and then we obtained the spleen in mice. After harvesting the spleen, single-cell suspensions of splenocytes were obtained by ltering through a 70um cell strainer (BD Bioscience, USA). Splenocytes were cultured in medium RPMI-1640 (Hyclone, USA) with 10% FSC (Gibco, USA) in vitro at 37˚C with 5% CO 2 , or with treatement of PMA (Thermo Fisher Scienti c, USA) and Ionomycin (Thermo Fisher Scienti c, USA). 50ng/mL PMA and 1ug/mL Ionomycin were performed to stimulate splenocytes for 2 hours or 4 hours as PI stimulation.
Measurement of cytokine levels in serum using the Luminex platform Serum levels of multiple cytokines were measured using the Luminex assay kits (Thermo Fisher, USA).
Measurements were performed according to the manufacturer's instructions. Diluted serum samples were added onto 96-well plates coated with magnetic beads and incubated for 120min after vortexing. The beads were then washed, and the detection antibody mixture was added and incubated for 30min at room temperature. After incubation and plate washing, the samples were analyzed on the Luminex™ platform.

Immuno uorescence
Frozen tissue sections were used for immuno uorescence assays. Frozen sections were xed, blocked, and then stained with primary antibodies, i.e., anti-CD3 antibody (ab33429, Abcam) and anti-CD138 antibody (AF3190, R&D Systems). Sections were visualized using Alexa Fluor 488 and Alexa Fluor 594 secondary antibodies (Abcam) both at a dilution of 1:200. Immuno uorescence images were obtained using ZEN Blue lite software (ZEISS, Germany). Representative imaging data were obtained at identical settings of the ZEN Blue lite software and all assays included negative controls, where the primary antibody was omitted (Supplementary Figure A).

Statistical analysis
Data from all experiments were expressed as mean ± SD and were analyzed using the SPSS software (SPSS, Inc., Chicago, IL, USA). Comparisons between the groups were performed for statistical signi cance using one-way analysis of variance. Differences with P values less than 0.05 were considered statistically signi cant.

CD138 expression in T cells leads to defective apoptosis of T cells of MRL/lpr mice.
We observed CD138+ cells frequencies in CD3+ T cells were negligible in splenocytes of MRL/MPJ mice ( Figure 1A). But these abnormal T cells accumulated in splenocytes of Fas-de ciency MRL/lpr mice ( Figure 1A). In addition, CD138+ T cells also evidently in ltrated in kidney of MRl/lpr mice, but not in kidney of MRL/MPJ mice ( Figure 1B). We also observed CD138+ T cells in MRL/MPJ mice had a very high level of apoptosis, but a low level in MRL/lpr mice ( Figure 1C). That indicated CD138+ T cells apoptosis in mice was in a Fas-dependent way and Fas de ciency in MRL/lpr mice led to the accumulation of CD138+ T cells. However, we observed CD138+ T cell in splenocytes of MRL/lpr mice had a signi cant decrease in number of apoptotic cells and an increase in number of live cells, compared with CD3+CD138-T cells ( Figure 1D). According to these results, we conclude CD138 expression in CD3+ T cells leads to the defective apoptosis of CD3+ T cells in Fas-de ciency MRL/lpr mice. Then, we observed CD138+ T cells in fresh splenocytes of MRL/lpr mice had a signi cantly increased CD69+ cells frequency and simultaneously had a decreased CD25+ cells frequency compared with CD138-T cells in fresh splenocytes of MRL/lpr mice ( Figure 1E). CD138 expression signi cantly promoted FasL expression in CD3+ T cells of MRL/lpr mice ( Figure 1E). Importantly, CD138 expression in DN T cells also strikingly increased FasL expression in DN T cells of MRL/lpr mice ( Figure 1F).

CD4+ T cells and Double negative T cells expressed CD138 in MRL/lpr mice
We observed the majority of CD138+ T cells in MRL/lpr mice were CD4 and CD8 double negative, only a proportion of CD138+ T cells expressed CD4 (Supplementary Figure B). CD8+ T cells had a limited frequency of CD138+ cells ( Figure 1G). Frequency of CD138+ cells in DN T cells were more signi cant, compared with CD4+ T cells ( Figure 1G). Previous research also has demonstrated DN T cells frequency among CD3+ T cells are elevated signi cantly and progressively as the increase of ages of MRL/lpr mice 6 . According to these results, it indicated that the majority of CD138+ T cells in MRL/lpr mice would derive from CD138-DN T cells in MRL/lpr mice as increaed DN T cells accumulation of MRL/lpr mice. Additionally, we observed CD4+ T cells in MRL/lpr mice comprised two cells subsets including CD4 hi T cells and CD4 int T cells ( Figure 1H). CD4 int T cells which have a down regulated expression of CD4 also expressed B220 ( Figure 1H). We found CD4+CD138+ T cells which expressed B220 too ( Figure 1H) had a signi cant down regulated CD4 expression and belonged to the B220+CD4int T cells subsets ( Figure 1H).
CD138 expression could improve the defective activation of CD138-T cell in MRL/lpr mice.
Further, we observed multiple cytokines levels in serum of MRL/lpr mice including IFN-γ, TNF, IL-6, IL-10, IL-17, and IL-2 were signi cantly increased compared those in MRL/MPJ mice. That indicates MRL/lpr mice was in in ammation in vivo ( Figure 2A). CD138-T cells in fresh splenocytes of MRL/lpr mice without in vitro stimulation had an increased frequency of CD69+ cells compared with those in fresh splenocytes of MRL/MPJ mice ( Figure 2B). But frequency of CD69+ cells in CD138-T cells of MRL/MPJ mice strikingly elevated compared with that in CD138-T cells of MRL/lpr mice after 5 h in vitro stimulation of fresh splenocytes with PMA and Ionomycin ( Figure 2C). In addition, both CD69 expressions and FasL expressions in CD138-T cells were not signi cantly increased after in vitro stimulation of splenocytes with 4 h PI stimulation ( Figure 2D). These results indicates CD138-T cell in MRL/lpr mice has a signi cantly defective activation. Importantly, we found CD138+ T cells in fresh splenocytes of MRL/lpr mice had signi cantly increased CD69+ cells frequencies ( Figure 1D and 2B) and FasL expressions ( Figure 1D) compared with CD138-T cells in those of MRL/lpr mice. In addition, CD138+ T cells in fresh splenocytes of MRL/lpr mice also had a signi cantly increased CD69+ cells frequency and FasL expression compared with CD138-T cells in those of MRL/lpr mice even after in vitro 5 h stimulation of splenocytes ( Figure 2C and 2E). Our results also indicate CD138 expression could improve the defective activation of CD138-T cells in MRL/lpr mice.

PI stimulation could signi cantly prevent CD138+ T cells accumulation
We tried to use PI to stimulate and activate splenocytes of MRL/lpr mice, and then observed the change of CD138+ T cells frequency in splenocytes after PI stimulation. 4 h in vitro culture of splenocyes in MRL/lpr mice without any stimulation caused signi cant increases in CD138+ T cell frequency in splenocytes of MRL/lpr mice with increased CD138 expression ( Figure 3A and 3D). However CD138+ T cells frequencies were progressively reduced as the time of PI stimulation increased ( Figure 3A). Besides, PI stimulation also signi cantly reduced CD138+ cells frequencies both in DN T cells and CD4+ T cells after PI stimulation ( Figure 3B and 3C). CD138 expression in CD138+ T cells was also signi cantly and progressively down regulated as the time of PI stimulation increased ( Figure 3D). These results indicated PI stimulation could prevent CD138+ T cells accumulation in splenocytes of MRL/lpr mice. In addition to reducing CD138+ T cells frequency in splenocytes, 4 h PI stimulation also signi cantly reduced both CD3+ T cells and DN T cells frequency in splenocytes of MRL/lpr mice ( Figure 4C and 4D). Moreover, we cultured splenocytes of MRL/lpr mice without any stimulation for 24 h. CD138+ T cells frequency in splenocytes evidently and rapidly increased after 24 h culture of splenocytes in vitro ( Figure  4E). The 24 h in vitro culture simultaneously also enhanced CD3+ T cells frequency and DN T cells accumulation in splenocytes ( Figure 4E). These results demonstrated CD138 expression could signi cantly contribute to the increased CD3+ T cells frequency and DN T cells accumulation in splenocytes of MRL/lpr mice.

CD138 expression in T cells could contribute to DN T cells accumulation in MRL/lpr mice T cells and DN T cells accumulations
PI stimulation could induce speci c apoptosis of CD138+ T cells Next, we sought the underlying mechanisms of PI stimulation preventing the accumulation of CD138+ T cells in MRL/lpr mice. Splenocytes activated with 4 h PI stimulation had a signi cantly increased apoptotic cells number and a decreased live cells number in CD138+ T cells ( Figure 5A). But PI stimulation did not evidently affect the numbers of apoptotic and live cells in CD3+CD138-T cells ( Figure  5A). Interestingly, PI stimulation even decreased apoptotic cells number and simultaneously increased live cells number in CD3-CD138+ plasma cells ( Figure 5B). That indicated PI stimulation speci cally induced cellular apoptosis in CD3+CD138+ T cells, but not in CD3+CD138-T cells and CD3-CD138+ plasma cells. So, our results demonstrated PI stimulation signi cantly prevented CD138+ T cells accumulation by inducing the speci c apoptosis of CD138+ T cells. Moreover, the increased levels of apoptosis in CD138+ T cells induced by PI stimulation were not caused by its cytotoxic effect 21,22 . PI stimulation signi cantly promotes the activation of CD138+ T cells CD69 expression in CD138+ T cells results in signi cantly increased level of apoptosis in CD138+ T cells At last, we tried to demonstrate our hypothesis that the activation of CD138+ T cells induced by PI stimulation caused their apoptosis. Consistent with our speculation, we observed CD69+ cells in CD138+ T cells of MRL/lpr mice had a signi cantly increased apoptosis level compared with CD69-cells in CD138+ T cells ( Figure 6E). But CD69 expression failed to increase the apoptosis level of CD138-T cells in MRL/lpr mice ( Figure 6F). Our results indicate CD69 expression in CD138+ T cell could promote its speci c apoptosis and demonstrate PI stimulation results in increase of speci c apoptosis in CD138+ T cells by activating them via increasing their expressions of CD69.

Discussion
Our study has showed that CD138+ T cells apoptosis in MRL/MPJ mice was in Fas dependent way. But

CD138 expression in CD3+ T cells of MRL/lpr mice signi cantly prevented the apoptosis of CD3+ T cells, contributed to CD3+ T cells and DN T cells subsets accumulation and simultaneoulsy promoted CD3+ T cells activation. Importantly CD138 expression could also increase FasL expression in DN T cells promoting the cytotoxicity of DN T cells. Moreover, decreased apoptotic and increased live numbers of CD138+ T cells led to the accumulation of CD138+ T cells in spleen of MRL/lpr mice. Our results demonstrated PI stimulation prevented CD138+ T cells accumulation and decreased CD138+ cells frequencies in DN and CD4+ T cells by inducing the speci c apoptosis of CD138+ T cells. Our results demonstrated PI stimulation caused speci c apoptosis in CD138+ T cells via increasing CD69 expression in them.
Our results also suggested CD138 expression in CD3+ T cells was probably caused by the failure of activation in autoreactive T cells before self-antigens exposure to immune system. Fas (CD95) is the member of the tumor necrosis factor receptor family and interacts with Fas ligand (FasL) after T cell receptor (TCR) activation to induce apoptosis 16 . Fas de ciency leads to DN T cells accumulation in MRL/lpr lupus mice resulting in lymphadenectasis and splenomegaly 23,24 . In addition to DN T cells, our results also showed apoptosis of CD138+ T cells was in a Fas-dependent way and indicated Fas de ciency also led to CD138+ T cells accumulation in MRL/lpr mice as the age increased and the lupus developed 12 . Our results showed CD138+ T cells had a high level of CD69 expression but a low level of CD25 expression. Importantly, the majority of CD138+ T cells were CD4 and CD8 double negative. CD138 expression signi cantly increased FasL expression in CD3+ T cells and their subsets including DN T cells. However, DN T cells in MRL/lpr mice are strongly cytotoxic, overexpressing FasL, which results in autoimmune injuries of multiple tissues that express small amounts of Fas receptor 16,25 . That indicated CD138 expression increased DN T cells cytotoxicity which could promote the lupus development and tissue injuries in MRL/lpr mice.
Our results showed CD3+ T cells accumulated but had a defective proliferation in Fas-de ciency MRL/lpr mice. Moreover the accumulated CD138+ T cells had a signi cant decrease in apoptotic number of cells and simultaneously had a higher number of live cells compared with CD138-T cells demonstrating CD138+ T cells have a defective apoptosis in MRL/lpr mice. Importantly, previous study has shown that CD138+ T cells had a lower level of proliferation compared with CD138-T cells subsets 12,14 . According to these results, we conclude that CD138+ T cells accumulation was caused by Fas de ciency leading to their defective apoptosis but not hyper proliferation. In addition, our results also demonstrated CD138

expression greatly contributes to the accumulation of T cells and DN T cells in MRL/lpr mice by signi cantly decreasing the apoptotic number of T cells and DN T cells.
Production of autoantibodies has a detrimental effect on multiple organs and plays a key role in the progress of SLE 26 . We know immature T cells will experience positive selection and negative selection to be mature single positive T cells that cannot recognize self-antigens 27,28 . Autoreactive T cells will be deleted by Fas-mediated apoptosis during negative selection in the thymus 29 . Fas de ciency give the chance of autoreactive T cells to pass through the negative selection 30,31 . Autoreactive B cells may also avoid the apoptosis in negative selection induced by Fas-dependent apoptosis 32 . Our results has shown Fas de ciency results in the accumulation of CD138+ T cells in MRL/lpr mice. In addition, CD138+ T cells including CD4+CD138+ T cells commonly express B220 which has been demonstrated to express on those autoreactive T cells such as nonselected CD8+ T cells and DN T cells 29,33 . Importantly, CD138+ T cells have been demonstrated to be key in the progression of anti-dsDNA antibody secretion with a CD4 receptor dependent way 12 . CD138+ T cells have been also reported to promote the tissue injuries in condition that self-antigens are exposed to the immune system 7,12 . These results demonstrate accumulated CD138+ T cells had large numbers of the auto-reactive T cells that had avoided Fasdependent apoptosis in negative selection 12,14 . CD4+CD138+ T cells in MRL/lpr mice had a signi cant decrease in CD4 expression compared with CD4+CD138-T cells. It had been reported that CD4+ T cells could convert into DN T cells 34,35 . CD4+CD138+ T cells in MRL/lpr mice may be the precursor of DN T cells which were originated from CD4+ T cells that have the potential of conversion into DN T cells.
CD138 expression in T cells plays a key role in the progression of lupus in MRL/lpr mice. Our results showed CD138 expression could contribute to the accumulation of CD3+ T cells by signi cantly preventing their apoptosis. Morever, CD138+ T cells showed to be activated more easily compared with CD138-T cells indicating CD138 expression could promote the activation of T cells in MRL/lpr mice. That suggested CD138 expression could contribute to autoreactive T cells accumulation and simultaneously promote autoreactive T cells to be activated by autoreactive B cells, by which abnormal plasma cells formation was increased 12 . However, the mechanisms of CD138 expession in these abnormal T cells were still undeciphered. Our results showed CD138-T cells in MRL/MPJ mice were more easily activated compared with CD138-T cells in MRL/lpr mice. PI stimulation also failed to signi cantly activate CD3+CD138-T cells in splenocytes of MRL/lpr mice. These results indicated CD138-T cell had an evidently defective activation in MRL/lpr mice. Our results simultaneously showed PI stimulation signi cantly decreased CD138 expression in CD138+ T cells. PI stimulation was able to signi cantly reduce frequency of CD138+ cell in CD3+ T cell and its cell subsets with increased activation level of CD3+ T cell and CD138+ T cells. Previous research also has showed TCR activation negatively regulated CD138 expression frequency in DN T cells 14 . These results demonstrated CD138 expression in CD3+ T cells could be prevented by the stimulation to activate CD3+ T cells.
Isolated splenocytes from MRL/lpr mice were cultured in vitro in the medium without cellular stimulation. We observed CD138+ T cells frequency increased in the 48 h accompanied with decreased activation level of CD3+ T cells and CD138+ T cells, and then decreased gradually after 48 hours accompanied with increased activation level of CD3+ T cells and CD138+ T cells. That indicated CD138+ T cells frequency was inversely correlated with the activation level of CD3+ T cells and CD138 + T cells. Previous research has demonstrated CD138+ T cells had large numbers of autoreactive T cells that promote autoantibody production in the presence of dsDNA, and enhanced disease progression in SLE by rapidly activating autoreactive B cells when self-antigens are exposed to the immune system 12 . These evidences suggest that CD138 expression in CD3+ T cells was probably due to failure of activation in these abnormal T cells in absence of self-antigens exposure.
In our present study, PI stimulation reduced CD138+ T cells frequency in splenocytes of MRL/lpr mice. Moreover, our results showed PI stimulation led to speci c apoptosis of CD138+ T cells. PI stimulation simultaneously promoted the activation of CD138+ T cells with increased CD69 cells frequency and FasL expression in them. What is important, CD69+ cells in CD138+ T cells had a signi cant increase in apoptosis level compared with CD69-cells in CD138+ T cells. According to our results, we con rmed that the increased apoptosis of CD138+ T cells and subsequent decreased accumulation of CD138+ T cells were induced by PI stimulation via increasing CD69 expression in CD138+ T cells. Our research uncovered the role of CD138 expression in CD3+ T cells, and why to activate CD3+ T cells signi cantly increased the apoptosis of CD138+ T cells and prevented the accumulation of CD138+ T cells in MRL/lpr mice. We also provided the novel insight into the mechansims preventing the autoreactive T cells accumulation in SLE.   Flow cytometry analyses and bar chart denote CD69+ cells frequencies in CD138+ T cells in fresh