Treatment with a tissue-selective oestrogen complex does not affect disease pathology but reduces pre-BI cells in lupus-prone mice

Objective Systemic lupus erythematosus (SLE or lupus) is an autoimmune disease characterized by B-cell dysfunction, production of autoantibodies, and immune complex formation. Lupus is overrepresented in females, indicating that sex hormones play a role in the pathophysiology. Treatment with a tissue-selective oestrogen complex (TSEC) containing conjugated oestrogens and the selective oestrogen receptor modulator bazedoxifene (BZA) protects against postmenopausal vasomotor symptoms and osteoporosis, but its impact on organ damage in lupus is not fully understood. Method We used ovariectomized MRL/lpr mice, treated with two different physiological doses of 17β-oestradiol-3-benzoate (E2), BZA, or TSEC (E2 plus BZA), to assess early and late B-cell development and to determine histological disease manifestations in the kidneys and salivary glands. Results TSEC treatment reduced the frequency of the pre-BI population in bone marrow to levels equivalent to treatment with physiological doses of E2 alone but did not affect any of the other examined B-cell populations. Our earlier studies indicated that TSEC treatment did not aggravate disease development in ovariectomized MRL/lpr mice, while protecting against trabecular bone loss. Here, we follow up on our previous study and show that neither ovariectomy alone nor TSEC treatment of ovariectomized MRL/lpr mice influenced perivascular lymphocyte infiltration to the kidneys or salivary glands. Conclusion TSEC does not aggravate a mouse model of lupus, when given in doses that protect against postmenopausal lupus-associated bone loss. This indicates that further investigations into TSEC as a treatment for osteoporosis or vasomotor symptoms in postmenopausal women with SLE are warranted.

Systemic lupus erythematosus (SLE or lupus) is a disease characterized by inflammation and multiorgan damage.This is associated with dysfunctional B cells, immune complex formation, and perivascular leucocyte infiltration (1).The incidence of SLE is higher in women (female to male ratio 9:1) and peaks during the reproductive years (2,3), indicating an influence of sex hormones.
Oestrogen has major effects on the immune system, e.g. on B-cell development and function, the complement system, cytokine production, and type 1 interferon (IFN)-regulated genes (4)(5)(6)(7)(8), which all are affected in lupus.However, studies exploring the role of exogenous oestrogens on disease activity in SLE are conflicting.
Two studies reported that oral contraceptives did not increase the risk of flares (9,10), while a slight increase was shown after hormone replacement therapy (HRT) in postmenopausal patients (11).In mouse models of lupus, treatment with supraphysiological doses of oestradiol (E2) resulted in worsening of the disease characteristics and decreased survival (12,13).
Osteoporosis is a common secondary complication in SLE (14).Although oestrogen protects against both postmenopausal bone loss and vasomotor symptoms (VMS), the use of HRT is associated with an increased risk of severe side effects (15,16).The selective oestrogen receptor modulator (SERM) bazedoxifene (BZA) has beneficial effects on bone with little influence on endometrium or breast tissue (17); however, it does not alleviate VMS (18).A tissue-selective oestrogen complex (TSEC) containing the combination of conjugated oestrogens and BZA maintains the established benefits of oestrogen on bone and VMS, while minimizing the adverse effects (19).This combination (Duavive®/Duavee®) is approved for the treatment of VMS and prevention of postmenopausal osteoporosis.However, the efficacy and safety of TSEC for treatment of bone loss or VMS in postmenopausal women with SLE are unknown.
MRL/MpJ-Fas lpr /J (MRL/lpr) mice spontaneously develop a disease phenotype resembling human lupus, and we previously reported that ovariectomy (OVX) of MRL/lpr mice is a suitable model for studies of postmenopausal bone loss in lupus (20).Furthermore, treatment with the TSEC (E2 + BZA) protected the skeleton without influencing disease markers measured in urine and serum in this model (21).However, histopathological examinations of lupus target organs would further clarify whether TSEC is a safe alternative for the treatment of VMS and osteoporosis in postmenopausal women with SLE.
B-cell alterations are a hallmark of lupus disease, and SLE patients exhibit significant changes in the development, function, and distribution of B-cell populations (22,23).We reported previously that TSEC treatment induces the same effect on B lymphopoiesis as E2 alone, using healthy C57BL/6 mice (24).However, the effect of TSEC on B-cell populations in experimental lupus remains to be examined.
In this study, we show that neither the removal of sex hormones by OVX nor treatment of ovariectomized MRL/lpr mice (OVX MRL/lpr mice) with TSEC influences the perivascular infiltration of immune cells to the kidneys or salivary glands, while it reduces the frequency of pre-BI cells in the bone marrow.Thus, TSEC does not aggravate disease development in lupusprone mice, when given in doses that protect against lupus-associated osteoporosis.

Mice
All work was approved by the animal ethical review board in Gothenburg (Dnr: 95-2015) and the study is reported in accordance with ARRIVE guidelines.Female MRL/MpJ-Fas lpr /J (MRL/lpr) mice and control mice MRL/MpJ (MRL/++) mice were purchased from the Jackson Laboratory (Bar Harbor, ME, USA).They were kept in groups of three to 10 mice per cage under controlled environmental conditions and fed soya-free chow and tap water ad libitum.Mice were weighed and controlled for proteinuria and haematuria using dipsticks (Multistix® 7; Siemens Healthcare Diagnostic, Tarrytown, NY, USA) once or twice every week.Proteinuria and haematuria were scored and quantified according to the manufacturer's instructions.

Ovariectomy (OVX)
Mice were ovariectomized by removing the ovaries through a midline incision of the skin followed by flank incisions of the peritoneum, and the skin was closed with metallic clips.The same procedure was performed for sham surgery, except for removal of the ovaries.The mice were sedated with isoflurane (Baxter Healthcare Corporation, Deerfield, IL, USA), and subcutaneous injections of Carprofen (Rimadyl; Orion Pharma Animal Health, Sollentuna, Sweden) were used as postoperative analgesia.
Experimental model 1: MRL/++ and MRL/lpr mice MRL/++ and MRL/lpr mice were subjected to OVX or sham surgery (nine or 10 per group) at 8 weeks of age.The experiment was terminated at the age of 21 weeks.Owing to a weight loss of more than 10%, two MRL/lpr mice were terminated before the endpoint.
High and low doses of E2, BZA, and TSEC were used:

Histological analysis
Formalin-fixed kidneys and mandibular salivary glands were embedded in paraffin, cut at a thickness of 4 µm, and stained with haematoxylin and eosin (H&E).
Images were taken with a Nikon Eclipse 80i microscope and were subsequently fused for estimation of the whole organ area.Whole kidneys and mandibular salivary glands were scored blindly using a 0-4 scoring system (0: no cellular infiltration; 1: minimal; 2: mild; 3: moderate; 4: severe).

Flow cytometry analysis
Bone marrow cells and splenocytes were obtained and analysed by flow cytometry, as described previously (24,25).In brief, spleen was mashed through 70 μm cell strainers to obtain single-cell suspensions.Bone marrow cells were flushed from the femur using a syringe and phosphate-buffered saline (PBS ) were stained for IgM-PE (Southern Biotechnology), B220-FITC (Biolegend), CD21-APC-Cy7 (Biolegend), CD23-PE-Cy7 (eBioscience, Vienna, Austria), and CD93-APC (eBioscience).Forward and side scatter gates were used to discriminate doublets and debris, and fluorescence minus one (FMO) was used as a control.The cells were investigated using FACSVerse (BD) and the data were analysed using FlowJo software (Three Star, Ashland, OR, USA).

Statistical analysis
Statistical analyses were performed using GraphPad Prism (Graph Pad Software, La Jolla, CA, USA).Groups were compared using one-way analysis of variance (ANOVA) followed by Sidak's post-hoc analysis, unless stated otherwise.Histological score data were analysed using Kruskal-Wallis followed by Dunn's multiple comparison test.Kaplan-Meier curves for proteinuria and haematuria versus time were analysed using the log-rank test.In Experimental model 1 (MRL/lpr and MRL/++ mice, sham and OVX surgery), post-hoc tests ware used for comparisons between groups.In Experimental model 2 (treatment with TSEC), sham-operated mice were only compared with vehicle-treated OVX mice, while all OVX groups, independent of treatment, were compared against each other.A p-value ≤ 0.05 was considered statistically significant.

OVX does not affect perivascular infiltration in the kidneys and salivary glands of MRL/lpr mice
To investigate how the absence of sex steroids affects disease symptoms in experimental lupus, MRL/lpr mice and their congenic control strain (MRL/++) were subjected to OVX or sham surgery at the age of 8 weeks (Figure 1A).Successful OVX was confirmed by weighing the uterus at the end of the experiment (Figure 1B).The development of proteinuria and haematuria did not differ between Sham and OVX MRL/lpr mice, although a tendency towards earlier onset and higher frequency of both proteinuria and haematuria was shown for OVX MRL/lpr mice.Sham and OVX MRL/++ mice displayed no signs of proteinuria or haematuria (Online Supplementary Figure 1A, B).Kidney pathology was investigated by assessing the perivascular infiltration of inflammatory cells.MRL/lpr mice showed an increased score of infiltration in the kidney compared to the control MRL/++ strain in both sham-operated and OVX mice, but no differences in the level of infiltration between the Sham and OVX MRL/ lpr groups were detected (Figure 1C, D).
Since the MRL/lpr mouse strain exhibit a phenotype similar to Sjögren's syndrome, we examined perivascular infiltration in the mandibular salivary gland.MRL/ lpr mice displayed an increased score of infiltration to the salivary glands compared to controls, but no differences in the level of infiltration between the Sham and OVX MRL/lpr groups were detected (Figure 1E, F).
Cells from bone marrow were isolated and analysed for T-and B-cell populations in Sham and OVX MRL/+ + and MRL/lpr mice.MRL/lpr mice displayed an increased frequency of CD3 + positive T cells compared to the MRL/++ controls, whereas no difference was shown between Sham and OVX MRL/lpr mice (Figure 2A).No differences in the proportions of CD19 + B cells were found between the groups (Figure 2B).

TSEC treatment does not influence perivascular infiltration in the kidneys and salivary glands of OVX MRL/lpr mice
Given the lack of difference in perivascular cell infiltration in the kidneys and salivary glands between Sham and OVX MRL/lpr mice, in combination with our recent publication showing that TSEC (E2 combined with BZA) can be used as a treatment for osteoporosis in a mouse model of postmenopausal lupus (21), we evaluated the influence of TSEC treatment on kidney and salivary gland pathophysiology in the MRL/lpr lupus model.
Low dose of E2, BZA, or TSEC.MRL/lpr mice were subjected to either sham surgery or OVX.Shamoperated mice were treated with Veh and OVX mice were treated with Veh, low dose E2 (0.15 μg/mouse/ day), BZA (36 μg/mouse/day), or TSEC (E2 + BZA).A schematic overview of the experimental set-up is shown in Figure 3A.Measurements of the uterus weights confirmed successful OVX surgery and showed that treatment of OVX mice with E2 restored the uterus weight to the same levels as in sham-operated mice, while OVX mice treated with BZA or TSEC had uterus weights comparable with the OVX Veh group [data shown in Nordqvist et al (21)].
To determine whether TSEC treatment affected the perivascular infiltration of leucocytes in kidneys and salivary glands, these organs were sectioned and scored.The cellular infiltration in kidneys of MRL/lpr mice did not differ between Sham and OVX mice treated with Veh, or in response to any of the treatments in OVX mice (Figure 3B).In addition, histological scoring of perivascular infiltration in the salivary gland revealed no differences between the groups (Figure 3C).High dose of E2, BZA, or TSEC.The experimental setup was repeated with higher doses of E2 (0.5 μg/mouse/ day), BZA (48 μg/mouse/day), or TSEC (E2 + BZA) (Figure 3D).Even though an increased dose of E2 was used, the uterus weight of the E2-treated mice did not exceed the uterus weight of the sham-operated mice.However, a small but significant increase in uterus weight was shown for the TSEC treatment group compared with the OVX Veh group, suggesting that BZA was not able to fully antagonize the high dose of E2 on the uterus [data shown in Nordqvist et al (21)].
Perivascular kidney infiltration did not differ between any of the treatment groups after high-dose treatment of E2, BZA, or TSEC in MRL/lpr mice (Figure 3E).Moreover, scoring of cellular infiltration in mandibular salivary gland revealed no differences between any of the groups (Figure 3F).
TSEC treatment reduces the frequency of pre-BI cells in the bone marrow of OVX MRL/lpr mice E2, SERM, and TSEC have previously been shown to affect B-cell development in C57BL/6 mice (24,25).Therefore, we investigated B-cell subpopulations in the bone marrow and spleen from Sham and OVX MRL/lpr mice.Sham-operated mice were treated with Veh, and OVX mice with Veh and low or high doses of E2, BZA, or TSEC.A schematic overview of B-cell development in the bone marrow and B-cell differentiation in spleen is shown in Figure 4.
Low dose of E2, BZA, or TSEC.The two earliest populations of B cells during development in the bone marrow are the pro-B cells and pre-BI cells.Low-dose treatment with E2, BZA, or TSEC resulted in similar frequencies of pro-B cells between the groups (Figure 5A).Decreased frequencies of pre-BI cells were observed after treatment of OVX mice with E2, BZA, and TSEC compared to Veh treatment (Figure 5B), while no differences in the frequency of pre-BII or immature B cells were detected between the groups (Figure 5C, D).
High dose of E2, BZA, or TSEC.B-cell development in bone marrow was also assessed after high-dose treatment of OVX mice with E2, BZA, and TSEC.As for the low-dose treatments, no differences in the frequency of pro-B cells were detected between any of the groups (Figure 5E), while treatment of OVX mice with E2, BZA, and TSEC resulted in reduced frequencies of pre-BI cells compared to Veh-treated mice (Figure 5F).No differences in the frequency of pre-BII or Immature B cells were detected between the groups (Figure 5G, H).E2 treatment of OVX mice decreased T1 B cells to the same level as Sham.No differences after treatment of OVX mice with BZA or TSEC were detected (Figure 6B).Frequencies of transitional type 2 (T2) B cells as well as follicular (FO) B cells and marginal zone (MZ) B cells were similar between the groups (Figure 6C-E).

V e h V e h E 2 B Z A T S E C V e h V e h E 2 B Z A T S E C V e h V e h E 2 B Z A T S E C V e h V e h E 2 B Z A T S E C
High dose of E2, BZA, or TSEC.B-cell differentiation in spleen was also investigated after treatment of OVX mice with high-dose E2, BZA, and TSEC.Total spleen weights did not differ between the treatment groups (Figure 6F).In contrast to the low-dose experiment, similar levels of T1 B cells were observed between Sham and OVX mice treated with Veh or E2 (Figure 6G).Furthermore, as shown in Figure 6H, the frequencies of T2 B cells did not differ between the groups.In contrast to the low-dose experiment, FO B cells were increased after OVX compared with Sham, but no differences were found in response to treatments (Figure 6I).Furthermore, treatment with high-dose E2, BZA, or TSEC did not change the frequency of MZ B cells (Figure 6J).

Discussion
In this study we follow up on our previous work which indicated that treatment with TSEC (E2 + BZA) does not worsen lupus disease characteristics in OVX MRL/ lpr mice (21), by further investigating cellular and histological mechanisms.We present encouraging data  showing that treatment of OVX MRL/lpr mice with TSEC does not influence the infiltration of immune cells in the kidneys and salivary glands, although the frequency of pre-BI cells is reduced compared to vehicle-treated mice.
There is a clear female predominance in SLE.X-chromosome-linked genes have been shown to contribute to disease pathophysiology (26), but the fact that the female prevalence peaks during child-bearing years also points towards a hormonal influence on the development of the disease.However, the role of oestrogen in SLE is complex, as the risk of flares is increased during pregnancy (27), while the use of third generation oral contraceptives is well tolerated in patients with stable disease (9,10).
A previous study from our group revealed that depletion of sex hormones by OVX did not alter the disease development in MRL/lpr mice, assessed by proteinuria, haematuria, and disease-specific and inflammatory serum markers (urea, interleukin-6, immunoglobulins G and M, and anti-double-stranded DNA) compared to sham-operated mice (21).Here, we confirm this finding and show no differences in perivascular cell infiltrations in the kidneys and salivary glands between Sham and OVX MRL/lpr mice.This is also in line with research conducted in lupus-prone NZM2410 or NZB/NZW F1 mice, where OVX did not affect the survival of the mice (28,29).
We further show that treatment of OVX MRL/lpr mice with two different doses of TSEC (E2 + BZA) did not affect the perivascular infiltration of inflammatory cells in the kidneys and salivary glands compared to vehicle controls.This is in line with the results from our previous study showing no differences in proteinuria, haematuria, or serum markers in response to TSEC treatment (21).The effect of treatment with E2 or BZA alone did not differ from treatment with TSEC on typical lupus characteristics in the kidneys and salivary glands.However, postmenopausal treatment with E2 is not recommended owing to the increased risk of breast and endometrial cancer (15), and BZA lacks effects on VMS (18).As SLE is associated with both early menopause and an increased risk of developing osteoporosis (30), treatment with TSEC could be beneficial for the prevention of bone loss and alleviation of VMS.Our results encourage further investigations into TSEC as a treatment option in postmenopausal lupus patients.
Alterations in B cells are associated with lupus disease development.SLE patients exhibit significant changes in the development and distribution of B-cell populations due to defects in immunological checkpoints during B-cell development (22,23).Oestrogen is known to influence B lymphopoiesis by inhibiting the differentiation from pro-B cells to pre-B cells in the bone marrow (25,31).Oestrogen also influences later stages of B-cell development in the spleen, where E2 treatment was shown to reduce the T1 B-cell population and increase the number of MZ B cells (24,25).We showed previously that TSEC induces the same effect on B-cell development as treatment with E2 alone in healthy C57BL/6 mice ( 24), but B-cell development after TSEC treatment has not previously been investigated in experimental postmenopausal lupus.This study reveals reduced levels of pre-BI cells in OVX MRL/lpr mice after TSEC treatment.Although both the early and late phases of B-cell development in the bone marrow and spleen were investigated, no other populations were affected.In addition, we showed in our previous publication that immunoglobulin production is unaffected by TSEC treatment (21).In contrast, TSEC treatment of healthy C57BL/6 mice also resulted in reductions in pre-BII, immature, and transitional B-cell populations, as well as increased MZ B cells in response to TSEC (24).The differences between the studies are probably due to the dissimilar settings between healthy C57BL/6 and lupus-prone MRL/lpr mice.The bone marrow of mice with lupus displays an increased number of apoptotic cells and IFN-producing inflammatory cells, such as T cells, which can influence B-cell development (32,33).This is in line with the result from this study showing that the MRL/lpr mice have an increase in T cells in bone marrow compared to MRL/++ controls.There is a large spread between the individual MRL/lpr mice in most of the studied B-cell populations that may conceal possible effects of hormone treatments.Future studies in which B-cell development is investigated at different stages of lupus disease progression would further clarify the picture.
The dose of E2 used is crucial for the outcome of experimental studies in mouse models of lupus, as treatment with E2 in NZB/NZW F1 mice worsened the lupus disease (12,29).Notably, in those studies, silastic tubes containing 6-7 mg of E2 were used, resulting in considerably higher hormone serum levels than the doses given to mice in this study (0.15 and 0.5 µg/ day).Administration of such high doses escalates the risk of urinary retention (34) and subsequently accelerates the development of renal pathology.In this study, none of the E2 doses used led to uterus weights exceeding the uterus weights of sham-operated mice.Thus, the MRL/lpr mice were given physiological doses of E2 alone or in combination with BZA.
The study has some limitations.The treatment period lasted for 3 weeks, and to evaluate the effects of longterm use of TSEC in postmenopausal lupus, longer treatments are needed.However, although the MRL/ lpr strain is suitable for studying lupus disease because of the resemblance to the human disease, severe symptoms such as pronounced lymphadenopathy prevented us from letting the mice age too far.A strain with a slower disease development may be more suitable for these studies.Furthermore, a broader and more specific panel of SLE disease activity markers could have been used to provide additional information about the effects of treatments on lupus disease activity.
In summary, treatment with TSEC in doses that were previously shown to protect against bone loss in this animal model of postmenopausal lupus resulted in a reduced frequency of pre-BI cells in bone marrow.Furthermore, the data from our previous study, which indicated that TSEC treatment does not worsen lupus disease characteristics, were strengthened by the fact that perivascular infiltrations to the kidneys or salivary glands were not affected by TSEC.

Conclusion
The results from this study indicate that TSEC treatment may be well tolerated in lupus, and further investigations into TSEC as a treatment option for osteoporosis and VMS in postmenopausal women with SLE are warranted.

Figure 1 .Figure 2 .
Figure 1.Ovariectomy (OVX) does not affect perivascular leucocyte infiltration in the kidneys or salivary glands of MRL/lpr mice.MRL/lpr mice and their congenic control strain (MRL/+ +) were subjected to sham surgery or OVX at 8 weeks of age and perivascular infiltration was analysed in whole sections of kidney and salivary gland.The severity was scored according to a 0-4 grade scale.(A) Schematic overview of the experimental set-up; (B) uterus weight; (C) perivascular leucocyte infiltration score in the kidney and (D) representative images of kidney infiltrations; (E) perivascular leucocyte infiltration in the salivary gland and (F) representative images of salivary gland infiltrations.Each data point represents an individual animal.The bars indicate the median (B) or median (C and E).Statistical analysis: one-way ANOVA followed by Sidak's post-hoc analysis (B) or Kruskal-Wallis followed by Dunn's multiple comparison test (C and E).*p < 0.05, **p < 0.01, ****p < 0.0001.Scale bar = 200 µm.

Figure 3 .Figure 5 .Figure 4 .
Figure 3. Tissue-selective oestrogen complex (TSEC) does not aggravate perivascular infiltration in the kidneys or salivary glands of OVX MRL/ lpr mice.MRL/lpr mice were subjected to sham surgery or ovariectomy (OVX) at 6-7 weeks of age.In the low-dose experiment, sham-operated mice were treated with vehicle (Veh) and OVX mice were treated 5 days per week for 3 weeks with a low dose of 17β-oestradiol (E2; 0.15 μg/ mouse/day), bazedoxifene (BZA; 36 μg/mouse/day), E2 + BZA (TSEC), or Veh.(A) Experimental set-up; perivascular cellular infiltration in (B) kidney and (C) salivary gland.In the high-dose experiment, sham-operated mice were treated with Veh and OVX mice were treated 5 days per week for 3 weeks with a high dose of E2 (0.5 μg/mouse/day), BZA (48 μg/mouse/day), TSEC, or Veh.(D) Experimental set-up; perivascular cellular infiltration in (E) kidney and (F) salivary gland.Each data point represents an individual animal.The bars indicate the median.Statistical analysis: Kruskal-Wallis followed by Dunn's multiple comparison test.

Figure 6 .
Figure 6.Treatment with low-dose E2 reduces the frequency of transitional type 1 B cells in OVX MRL/lpr mice.MRL/lpr mice were subjected to sham surgery or ovariectomy (OVX) at 6-7 weeks of age.In the low-dose experiment, sham-operated mice were treated with Veh and OVX mice were treated 5 days per week for 3 weeks with a low dose of 17β-oestradiol (E2; 0.15 μg/mouse/day), bazedoxifene (BZA; 36 μg/mouse/day), E2 + BZA (TSEC), or Veh.B-cell populations in spleen were analysed by flow cytometry.(A) Total spleen weight and frequencies of (B) transitional type 1 (T1) B cells, (C) transitional type 2 (T2) B cells, (D) follicular (FO) B cells, and (E) marginal zone (MZ) B cells.In the high-dose experiment, sham-operated mice were treated with Veh and OVX mice were treated 5 days per week for 3 weeks with a high dose of E2 (0.5 μg/ mouse/day), BZA (48 μg/mouse/day), TSEC, or Veh.B-cell populations in spleen were analysed by flow cytometry.(F) Total weight of spleen and frequencies of (G) T1 B cells, (H) T2 B cells, (I) FO B cells, and (J) MZ B cells.Each data point represents an individual animal.The bars indicate the mean.Statistical analysis: one-way ANOVA and Šídák's multiple comparison test.*p < 0.05.