Smooth Muscle-Specific BCL6+/− Knockout Abrogates Sex Bias in Chronic Hypoxia-Induced Pulmonary Arterial Hypertension in Mice

The “estrogen paradox” in pulmonary arterial hypertension (PAH) refers to observations that while there is a higher incidence of idiopathic PAH in women, rodent models of PAH show male dominance and estrogens are protective. To explain these differences, we previously proposed the neuroendocrine-STAT5-BCL6 hypothesis anchored in the sex-biased and species-specific patterns of growth hormone (GH) secretion by the pituitary, the targeting of the hypothalamus by estrogens to feminize GH secretion patterns, and the role of the transcription factors STAT5a/b and BCL6 as downstream mediators of this patterned GH-driven sex bias. As a test of this hypothesis, we previously reported that vascular smooth muscle cell- (SMC-) specific deletion of the STAT5a/b locus abrogated the male-dominant sex bias in the chronic hypoxia model of PAH in mice. In the present study, we confirmed reduced BCL6 expression in pulmonary arterial (PA) segments in both male and female SMC:STAT5a/b−/− mice. In order to test the proposed contribution of BCL6 to sex bias in PAH, we developed mice with SMC-specific deletion of BCL6+/− by crossing SM22α-Cre mice with BCL6-floxed mice and investigated sex bias in these mutant mice in the chronic hypoxia model of PAH. We observed that the male-bias observed in wild-type- (wt-) SM22α-Cre-positive mice was abrogated in the SMC:BCL6+/− knockouts—both males and females showed equivalent enhancement of indices of PAH. The new data confirm BCL6 as a contributor to the sex-bias phenotype observed in hypoxic PAH in mice and support the neuroendocrine-STAT5-BCL6 hypothesis of sex bias in this experimental model of vascular disease.

There is extensive neuroendocrinology literature showing that these patterns are derived from the patterned secretion of GH by the anterior pituitary in response to hypothalamic input via neuronal GHRH secretion which is in turn affected by estrogens (e.g., estradiol-17β (E2)) at the level of the arcuate nucleus (and other sites) in the ventromedial hypothalamus [30][31][32][33][34][35][36][37][38][39][40][41][42]. In terms of downstream mechanisms, there is also extensive literature mechanistically linking these GH pulsatility patterns to patterned activation of STAT5a/b (frequency and level of activation by Tyr phosphorylation) with resulting patterned activation of the transcriptional repressor BCL6 and hundreds of downstream genes ( Figure 1) [43][44][45][46][47][48][49][50][51]. Remarkably, it has been established beginning in the late 1970s that the feminizing effect of exogenously administered estrogens on liver gene expression obligatorily required the pituitary in that hypophysectomy blocked this sex-bias effect of estrogens ( Figure 1) [30][31][32][33][34][35][36][37][38][39][40][41][42]. Thus, although there is considerable research on the interactions between the estrogens, the estrogen receptors, and STAT5 and other transcription factors at the level of distal target tissues [47], there is also clear evidence that a major mechanism of the sex-bias effects of exogenously administered estrogens begins with the interaction of sex hormones with the hypothalamus and changes in the subsequent cascade of patterned activation of the GH-STAT5 axis in distal tissues ( Figure 1) . Thus, especially in light of the "estrogen paradox" in PAH [6,9], we proposed the neuroendocrine-STAT5 hypothesis of sex bias in this disease to account for the difference between humans and rodents ( Figure 1) [19][20][21].
Experimentally, after observing that SMCs in obstructive pulmonary arterial lesions in IPAH in both female and male patients showed reduced STAT5a/b and reduced PY-STAT5 [19], we tested the STAT5 hypothesis of sex bias in PAH in Ac tiv ati on R ep re ss io n STAT5a/b BCL6 Sex-bias in growth hormone (GH) patterns M: "pulsatile" versus F: "more continuous" Sex-biased gene expression in distal tissues

Hypothalamus E2
In PAH: ↓BCL6 ↑PAH in females ↓STAT5a/b Figure 1: Schematic of the pathway involving the contributions of neuroendocrine-STAT5-BCL6 mechanisms of sex bias in PAH. Patterns of circulating GH in males typically consist of a low baseline level with few (2-4) but high peak levels (a pattern customarily referred to as "pulsatile") [22][23][24][25][26][27][28][29]. Females have a higher baseline level of GH with small frequent peaks (≥7 per day) (a pattern referred to as "more continuous") [22][23][24][25][26][27][28][29]. E2, estradiol-17β; M, male; F, female; downward short arrows indicate reduction; upward short arrow indicates increase. the chronic hypoxia model in mice. In this model, wild-type (wt) mice show a male-dominant phenotype. We developed tissue-specific knockout mice with heterozygous or homozygous deletion of STAT5a/b in vascular smooth muscle cells (SMCs) and investigated the influence of these deletions on the male dominance in this model of PAH [19]. We observed that SMC-specific heterozygous or homozygous deletion of STAT5a/b abrogated the male-dominant sex bias in that both female and male mice developed PAH (increased right ventricular systolic pressure (RVSP), hypertrophy of pulmonary arterial vessel wall, and right ventricular hypertrophy) to a similar extent [19][20][21].
We then focused on experimentally testing a downstream element in the GH-STAT5 hypothesis ( Figure 1). It is well established that the transcriptional repressor BCL6 is one of the major transcription factors downstream of STAT5 in the sex-biased regulation of expression of hundreds of liver genes [50,51]. We have previously reported that BCL6 was also reduced in SMCs in obliterative pulmonary arterial lesions in both women and men with IPAH concomitant with the reduction in STAT5a/b and PY-STAT5 levels [21]. It is also already known that whole-body homozygous deletion of BCL6−/− in the mouse led to extensive pulmonary edema and pulmonary parenchymal and vascular inflammation [52]. Thus, from the point of view of downstream mechanisms in the STAT5 hypothesis, we investigated (a) whether SMC-specific deletion of STAT5a/b in the mouse had an effect on expression of BCL6 in SMCs in pulmonary arterial tunica media and (b) whether SMC-specific deletion of BCL6 in the mouse had an effect on chronic hypoxiainduced PAH and its male-dominant sex-bias phenotype ( Figure 1). The new data reported here represent a successful test of the BCL6 element in the GH-STAT5-BCL6 hypothesis of sex bias in PAH [19][20][21].

Derivation of Mice with Smooth Muscle Cell-(SMC-)
Specific Heterozygous and Homozygous Deletions of STAT5a/b or BCL6. Mouse breeding, care, and use of animals for these studies was approved by the Institutional Animal Care and Use Committee (IACUC) of the New York Medical College. A mouse line with SMC-specific homozygous deletion of the STAT5a/b locus had been previously produced by mating STAT5a/b fl/fl mice carrying loxP inserts bracketing the entire STAT5a/b locus in C57BL/6:129J mice provided by Dr. Lothar Hennighausen (National Institutes of Health, Bethesda, MD) with SM22-Cre +/+ mice (expressing Cre recombinase under the control of the mouse transgelin (smooth muscle protein 22α) promoter) purchased from The Jackson Laboratories, Bar Harbor, ME (stock number: 004746 Tg (Tagln-cre)1Her/J in C57BL/6:129SJL background) [19]. PCR analyses of the tail DNA using appropriate primers were used to confirm the genotype of the Cre-positive mice with respect to the presence of STAT5a/b wt or fl or Cre alleles [19]. All mice in the present studies designated "Wt" were Cre-positive littermates not carrying any STAT5a/b-floxed alleles [19]. The selective reduction in STAT5 expression in the tunica media of pulmonary arterial segments in the Cre-positive heterozygous STAT5a/b wt/fl (designated "+/−") or homozygous fl/fl (designated "−/−") mice was verified in immunofluorescence and immunohistochemistry assays for STAT5 [19]. Smooth muscle α-actin-(SMA-) positive aortic smooth muscle cells (SMCs) derived from the homozygous knockouts showed a marked loss of STAT5a/b by Western blotting [19]. Lung tissues from Wt and SMC:STAT5a/b−/− mice used in the immunofluorescence analyses for BCL6 were from experiments previously reported in Figure 4 in [19].
Mouse lines (in C57BL/6 background) with SMC-specific heterozygous and homozygous deletions of BCL6 were generated in the same manner described above by cross-mating the BCL6 fl/fl (B6.129S (FVB)-Bcl6 tm1.1Dent /J; stock number: 023727 Bcl6 fl ) and SM22α-Cre +/+ mice (stock number: 004746) purchased from The Jackson Laboratories, Bar Harbor, ME, and confirmed by PCR analyses of tail DNA using appropriate primers designated by the supplier. All mice in the BCL6 studies designated "Wt" were Cre-positive littermates not carrying any BCL6-floxed alleles. Reduced expression of BCL6 in the tunica media and SMCs of pulmonary artery segments of the mutant mice was confirmed by immunofluorescence analyses (Figure 2). These mutant mice, including the SMC-specific STAT5a/b−/− and SMC-specific BCL6+/− and −/− knockouts have no fertility or growth issues. In the present study, we used the more abundantly available male and female heterozygous SMC-specific BCL6 +/− for the chronic hypoxia treatment.
2.2. Exposure to Hypoxia. Groups of Wt, or mutant male and female mice at approximately 8 weeks of age were used in the respective chronic hypoxia experiments (n = 4 -7 mice per group per variable; average group size n = 5 mice) [19]. Mice were exposed to hypobaric hypoxia (0.5 atmospheres in room air equivalent to 10% v/v oxygen) for approximately 4 weeks; normoxia controls were left in ordinary room air [19].

Assessment of PAH and Lung Vascular Remodeling after
Chronic Hypoxia. This was carried out as previously described in [19]. Briefly, mice were anaesthetized using the ketamine/xylazine cocktail (100 mg/kg ketamine, 10 mg/kg xylazine, intraperitoneally). A Millar catheter was used for the in vivo measurement of right ventricular systolic pressure (RVSP) via external jugular vein catheterization (PVR-1030; Millar Instruments Inc., Houston, TX). The animals were then euthanized and the heart, aorta, and left lung removed and formalin fixed (the right lung was snap-frozen in liquid N 2 ). The heart was dissected to obtain the ratio of right ventricle weight to that of the left ventricle plus septum (RV/LV + S) as a measure of right ventricular hypertrophy (RVH). The left lung was used for paraffin block embedding and histology (hematoxylin & eosin, Van Gieson's elastin and Masson's Trichrome stains, and immunofluorescence for smooth muscle actin (SMA)). The extent of vascular remodeling in pulmonary arteries with an outer diameter < 80 μm was quantitated in terms of medial wall thickness expressed as 2x wall thickness in % of average vessel outer diameter measured in two axes perpendicular to each other [19]. Muscularization of pulmonary vessels in lung sections was evaluated as indicated in Section 2.4. All histological quantitation was carried out in a blinded fashion.

Immunofluorescence Analyses of Lung Sections.
Serial sections (5 μm) of formalin-fixed, paraffin-embedded lung tissue were evaluated for BCL6 and smooth muscle α-actin (SMA) immunofluorescence as described earlier [19] using rabbit pAb for BCL6 or mouse mAb for SMA, and respective donkey AlexaFluor 594-tagged pAb (in red) as the secondary antibody. Imaging was carried out using a Zeiss AxioImager M2 motorized microscopy system equipped with a highresolution RGB HRc AxioCam camera and AxioVision 4.8.1 software in a 1388 × 1040 pixel high-speed color capture mode. All data within each experiment were collected at identical imaging settings. Fluorescence intensity was quantitated using the NIH Image J software [19]. Muscularization of pulmonary vessels < 80 μm in diameter was evaluated by immunofluorescence of SMA in lung sections and expressed as the mean number of SMA-positive vessels per 10x objective field. 2.6. Statistical Evaluations. These were performed using the two-tailed Student t-test (as in Figure 2) and single-factor ANOVA methods (NCSS version 8, 2012). Multiple group comparisons were carried out using ANOVA (Tukey-Kramer, Newman-Keuls, and Tukey-Kramer tests).

Results
3.1. SMC-Specific STAT5a−/− Deletion Leads to Reduced BCL6 Expression in Pulmonary Arterial Tunica Media. We used quantitative immunofluorescence methods to investigate whether BCL6 might be downstream of STAT5 in SMCs in the pulmonary arterial vasculature. Groups of male and female Wt and STAT5−/− mice were exposed to hypoxia (10% hypobaric oxygen) for 4 weeks or kept under normoxic conditions, and the development of PAH was first evaluated as previously reported (RVSP, ratio of RV/LV + S weights, increased muscularization of vessels, and increased arterial wall thickness) [19]. From the data previously reported, we observed in this experiment that STAT5−/− deletion abrogated the male-dominant phenotype of PAH in this model-the knockout female and male mice developed equivalently high levels of PAH [19]. There was little change in heart rate or systemic blood pressure in these mice [19]. Subsequently, representative formalin-fixed paraffinblock sections of lungs were quantitatively evaluated for expression of BCL6 in the pulmonary arterial (PA) walls as a ratio to that in bronchial epithelium in different groups ( Figure 3). The immunofluorescence data in Figure 3(a) (arrows) and the quantitation summarized in Figure 3 3.2. Development of SMC-Specific BCL6+/− and BCL6−/− Knockout Mice. We generated lines of C57Bl6 mice with heterozygous or homozygous deletions of the BCL6 locus in vascular SMCs in crosses between BCL6 fl/fl and transgelin (SM22α)-Cre +/+ parents using methods previously described for the generation of SMC-specific STAT5+/− and −/− mice [19]. In addition to molecular characterization (using PCR methods summarized in [19]; data not shown), quantitative immunofluorescence methods were used to verify the reduced expression of BCL6 in the pulmonary arterial tunica media and SMCs in mice with SMC:BCL6+/− deletion (Figures 2(a) and 2(b)) and SMC:BCL6−/− deletion (Figures 2(c) and 2(d)). SMC-specific heterozygous and homozygous deletion of BCL6 reduced the levels of expression of BCL6 in the tunica media/SMCs in PA segments of the respective knockout mice but not in the endothelium (Figures 2(a) and 2(c)).

Abrogation of the Male-Dominant Sex Bias in Hypoxic
PAH in BCL6+/− Mice. We were able to generate sufficient SMC:BCL6+/− mice and Cre-positive "wt" littermates of both sexes to provide groups of 4-7 mice for evaluation indices of PAH per variable as indicated in Figure 4. (Only a limited number of SMC:BCL6−/− mice were generated in this project; these provided confirmatory preliminary data; see below.) Male and female Wt and SMC:BCL6+/− mice (n = 4-7 per group) were exposed to 10% hypobaric oxygen or normoxia for 4 weeks and evaluated for development of PAH. The data in Figure 4(a) show the mean RVSP in these groups in normoxia and in response to chronic hypoxia. There is a clear male-dominant phenotype in the increase in mean RVSP in Wt mice in response to hypoxia. In contrast, this sex bias was absent in BCL6+/− mice; both males and females showed increased RVSP in response to hypoxia. Figure 4(b) shows the RVSP data for individual mice. The data show that even under normoxia, one male and three female BCL6+/− mice had RVSP levels > 25 mm Hg. Moreover, all male and female BCL6+/− hypoxic mice had markedly elevated RVSPs (Figure 4(b)). Overall, the data in Figure 4 provide evidence showing that (a) even an heterozygous deletion of BCL6 in vascular SMCs led to an enhancement of the chronic hypoxia-induced increase in RVSP irrespective of sex, and (b) that there was a detectable increase in RVSP even in male and female knockout mice under normoxic conditions (Figure 4(a)). In preliminary studies, male and female homozygous SMC:BCL6−/− mice also showed an equivalent enhancement of RVSP after chronic hypoxia (data not shown). Figures 5(a) and 5(b) summarize pulmonary arterial wall thickness data from the experiment in Figure 4. Again, the male-dominant sex bias in response to chronic hypoxia in terms of increased wall thickness in Wt mice was abrogated in BCL6+/− mice (Figures 5(a) and 5(b)). Moreover, there appeared to be an increase in wall thickness in both male and female knockout mice even under normoxic conditions ( Figure 5(b)). Figure 6(a) summarizes data pertaining to increased muscularization of pulmonary vessels in lung sections from hypoxic mice, while Figure 6(b) summarizes development of right ventricular hypertrophy measured in terms of the Fulton index (ratio of RV/LV + S weights). As expected, a male-dominant phenotype was observed in Wt hypoxic mice, but there was little sex bias in hypoxic SMC:BCL6+/− mice in terms of these indices of PAH (Figures 6(a) and 6(b)).
Taken together the data in Figures 4-6 provide evidence for a contribution of BCL6 in the sex-bias phenotype of PAH observed in chronically hypoxic mice. Even heterozygous deletion of BCL6 in vascular SMCs rendered female mice equally susceptible to the development of hypoxic PAH as male mice. This observation was similar to our  Figure 4 were processed for histological sections and stained for elastin using Van Giesson's reagent. Wall thickness of PA segments with a diameter < 80 μm was quantitated in terms of the ratio of 2x wall thickness/outer diameter as reported earlier [14]. previous finding that heterozygous SMC-specific deletion of STAT5a/b was sufficient to abrogate the male-specific sex bias in the hypoxic mouse model [19].

Discussion
This study represents a specific test of the downstream BCL6 element in the GH-STAT5-BCL6 hypothesis of sex bias in PAH ( Figure 1) [19][20][21]. In a previous study, we provided experimental support for this hypothesis by showing that vascular SMC-specific deletion of the STAT5a/b locus in mice abrogated the male-dominant sex bias observed in the chronic hypoxic PAH model [19]. In the present study, we have tested the next downstream element-BCL6-in this hypothesis ( Figure 1). Consistent with its placement downstream of STAT5 [49][50][51], we observed reduced BCL6 levels in the tunica media/SMCs in pulmonary arterial segments in male and female mice with homozygous SMC-specific deletion of STAT5a/b (Figure 3). We then investigated whether SMC-specific deletion of BCL6 (+/−) affected the observed sex bias in the chronic hypoxia model. We observed the abrogation of the male-dominance in the development of PAH in the knockout mice. Both male and female SMC:BCL6+/− knockout mice responded equally to chronic hypoxia in terms of PAH development (Figures 4-6). Thus, these data place BCL6 in the STAT5-anchored sex-bias pathway in this experimental model and represent a successful critical test of the GH-STAT5-BCL6 hypothesis of sex bias in a vascular disease.
As elucidated in detail by Waxman et al. and other investigators in the last two-three decades, this neuroendocrine mechanism of sex bias, operating through an axis consisting of the arcuate nucleus, growth hormone releasing hormone (GHRH), growth hormone (GH), and signal transducer and activator of transcription 5 (STAT5), accounts for sexbiased expression of~500-1000 genes in the liver, and also of body growth and body weight (Figure 1) [43][44][45][46][47][48][49][50][51].  Figure 4 were immunostained for SMA and the average number of muscularized vessels per 10x field were enumerated for each (mean ± SE). (b) Hearts derived from the experiment in Figure 4 were fixed in 4% formalin, rinsed in PBS and dissected into the right ventricle (RV) and the left ventricle plus septum (LV + S) portions. The ratio of the wet weights of RV/LV + S (mean ± SE) illustrates the extent of right ventricular hypertrophy irrespective of the total weight of the individual heart. * P < 0 05 by ANOVA; NS, not significant.
We studied STAT5a/b in vascular cells in terms of its contribution to the sex-bias phenotype in PAH induced in mice following chronic hypoxia [19][20][21]. We generated mice with heterozygous or homozygous conditional deletions of the STAT5a/b locus in vascular smooth muscle cells (SMCs). Although, Wt males subjected to hypoxic stress showed significant pulmonary arterial and cardiac remodeling, with Wt females showing minimal changes (a maledominant phenotype), the SMC-specific STAT5 −/− mice hypoxic females showed the severest manifestations of PA remodeling (a female-dominant phenotype) [19]. The reversal of sex bias in this model of PAH did not require the complete loss of STAT5a/b genes. Even heterozygous conditional STAT5a/b+/− mice (thus, with a 50% loss of STAT5a/b) showed a loss of the male-dominant PAH phenotype [19]. The observation in the present study that these SMC:STAT5a/b−/− mice also showed reduced expression of BCL6 in the tunica media/SMCs in pulmonary arterial segments in both normoxic and hypoxic mice irrespective of sex ( Figure 3) was consistent with the known role of STAT5 as a transcription factor activating BCL6 gene expression ( Figure 1) [47][48][49][50][51]. In as much as it is known that BCL6 is a broad-spectrum transcriptional repressor [49][50][51][52], the reduced expression of BCL6 in SMC:STAT5a/b−/− mice prompted us to test the involvement of BCL6 per se in the sex bias observed in this model of PAH by genetically deleting BCL6 in SMCs in mice. Since it has already been shown that mice with whole-body knockout of the transcriptional repressor BCL6 showed increased production of proinflammatory cytokines (IL-6, IFN-gamma, and IL-4), developed generalized vasculitis including pulmonary vasculitis, grew poorly, and lived only 6-8 weeks [52], in the present study we specifically investigated the effects of tissue-specific BCL6 deletion only in vascular SMCs.
We discovered that SMC-specific deletion of BCL6 led to an abrogation of the male-dominant sex bias seen in hypoxic Wt mice. The data provide evidence for the involvement of the BCL6 axis in the sex bias observed in the chronic hypoxia-induced model of PAH in mice. This represents a successful test of the neuroendocrine-STAT5 hypothesis of sex bias in hypoxic PAH in mice in terms of the involvement of a predicted downstream player-the involvement of BCL6.
In preliminary experiments, we have initiated tests of the involvement of the predicted upstream participant-the pituitary-in sex bias in hypoxic PAH in mice ( Figure 1). Preliminary data suggest that hypophysectomy also abrogates the sex bias observed in Wt mice (data not shown).
We have previously reported that a marked decrease in STAT5a/b and BCL6 was observed in SMCs in obliterative lesions in human IPAH-in women and men [19,21]. These observations suggested a role of aberrations in the STAT5-BCL6 axis in the pathogenesis of the human disease. It is long established that 30-50% of patients with acromegaly develop systemic hypertension [58][59][60]. Although respiratory complications in acromegaly are known to include sleep-breathing disorders and "respiratory insufficiency" [60], there is little information about the pulmonary arterial circulation per se in such patients. Whether patients with IPAH have altered GH secretion kinetics could be evaluated using patterns of exercise-induced GH secretion [61][62][63].

Data Availability
The data used to support the findings of this study are included within the article.