*2.3. Effect of CRY1 and CRY2 Silencing in HAC15 Cells*

Our observation of the regulation of *CRY* genes by AngII, together with the experimental evidence available from Cry-null mice [8], prompted us to investigate the effect of *CRY* silencing on gene expression in HAC15 cells.

Silencing *CRY* genes by transfection of siRNA resulted in a 62% reduction in *CRY1* mRNA levels, and a 70% reduction in *CRY2* mRNA levels, measured by real-time PCR (Figure 5A,B). Notably, silencing *CRY1* induced a significant upregulation of *CRY2* (1.3 ± 0.2-fold, *p* = 0.005) (Figure 5B), which resulted in less efficient *CRY2* silencing when the double *CRY1* and *CRY2* siRNA assay was performed, for this reason simultaneous silencing of both genes was not allowed.

The expression of mRNA-encoding key enzymes involved in the production of aldosterone was examined. Transfection with *CRY1* siRNA resulted in a significant upregulation of *HSD3B2* expression (1.30 ± 0.23-fold, *p* = 0.009) (Figure 5F), and a trend toward the upregulation of *HSD3B1* (1.20 ± 0.5-fold, *p* = not significant) (Figure 5E), while the transfection with *CRY2* siRNA did not affect the expression of either *HSD3B1* or *HSD3B2*. Similarly, the expression of *CYP11B2*, and its main transcriptional factor *NR4A2* were not significantly modified at the evaluated timepoint (42 h post-transfection) (Figure 5C,D).

**Figure 4.** (**A**) Real-time PCR analysis of *CRY1* gene expression. \* *p*-Value < 0.001, # *p*-value = 0.007, and § *p*-value = 0.017 when compared with basal. (**B**) Real-time PCR analysis of *CRY2* gene expression. \* *p*-value < 0.001 and # *p*-value = 0.003 when compared with basal. (**C**) Real-time PCR analysis of *HSD3B1* gene expression. \* *p*-value = 0.022, # *p*-value = 0.03, § *p*-value < 0.001, and ¥ *p*-value = 0.023 when compared with basal. (**D**) Real-time PCR analysis of *HSD3B2* gene expression. \* *p*-value < 0.001, # *p*-value = 0.001, and § *p*-value = 0.009 when compared with basal. (**E**) Real-time PCR analysis of period (*PER1*) gene expression. \* *p*-value = 0.001 and # *p*-value < 0.001 when compared with basal. (**A**–**E**) Each point expresses the mean fold change over basal expression in at least three independent experiments. (**F**) Real-time PCR analysis of *PER1* gene expression at 6 h, after stimulation with 100 nM AngII ± 1 μM irbesartan. (**G**) Real-time PCR analysis of *CRY1* gene expression at 6 h, after stimulation with 100 nM AngII ± 1 μM irbesartan. (**H**) Real-time PCR analysis of *CRY2* gene expression at 12 h, after stimulation with 100 nM AngII ± 1 μM irbesartan. (**F**–**H**) Each bar represents the mean ± SD of relative fold change of gene expression in three independent experiments.

**Figure 5.** Effect of silencing *CRY1* and *CRY2* on gene expression in HAC15 adrenocortical cells. Real-time PCR analysis of *CRY1* (**A**), *CRY2* (**B**), *HSD3B1* (**C**), *HSD3B2* (**D**), *NR4A2* (**E**), and *CYP11B2* (**F**) gene expression. Each bar expresses the mean ± SD fold change over the expression in cells transfected with a control small interfering RNA (siRNA; Ctrl siRNA) of at least five independent experiments. No significant differences were detected between the cells transfected with Ctrl siRNA, and electroporated cells (Nucleofected).

#### **3. Discussion**

Over the last few years, significant knowledge about the molecular mechanisms that regulate aldosterone overproduction was gained from both next-generation sequencing studies [17], and murine models of primary aldosteronism [18]. The *Cry*-null mice, lacking the core-clock components CRY1 and CRY2 [8], displayed hyperaldosteronism and salt-sensitive hypertension, most likely sustained by the upregulation of the type VI 3β-hydroxyl-steroid dehydrogenase (*Hsd3b6*), corresponding to the human type I 3β-hydroxyl-steroid dehydrogenase (*HSD3B1*) gene.

Immunohistochemistry studies in normal human adrenals showed that HSD3B2 was the predominant isoform, while HSD3B1 localized mainly in the outermost layer zona glomerulosa [8,13,14]. In adrenal pathology, HSD3B1 appeared to be strongly expressed in the hyperplastic zona glomerulosa cells of BAH samples, while its expression was low in a series of eight APAs, composed predominantly of zona fasciculata-like cells [13]. Based on these results, it was hypothesized that HSD3B1 overexpression might represent the molecular mechanism responsible for autonomous aldosterone overproduction in BAH [19].

So far, the role and clinical significance of *CRY1* and *CRY2* genes in the regulation of aldosterone production and APA development, together with their potential interplay with *HSD3B* isoforms, were not explored in humans.

In this study, we demonstrated, for the first time, that *CRY1* is overexpressed, while *CRY2* is downregulated in APA tissue, when compared with the paired adjacent adrenal cortex, which represents the optimal control tissue, given the multiplicity of factors that influence the transcription of the core-clock genes [9]. In agreement with previous reports [15], we observed that *HSD3B2*, being over 300-fold more expressed than *HSD3B1*, is the principal isoform in APAs. A previous study showed that HSD3B1 (evaluated as H-score) was more expressed in APAs carrying somatic mutations in the *KCNJ5* gene [15], while in our cohort we did not detect any significant association between the expression of *HSD3B1* (evaluated by real-time PCR) and the mutational status of the samples. On the contrary, we observed that both *HSD3B1* expression and the relative *HSD3B1*/*HSD3B2* ratio were significantly more elevated in APAs composed mainly of zona glomerulosa-like cells (while APAs carrying a mutation in *KCNJ5* are composed mainly of zona fasciculata-like cells [20]).

Additionally, this study demonstrated, for the first time, that the expression of both *CRY1* and *CRY2* genes is modulated by AngII through activation of the AT1R. Similarly, the negative regulator *PER1* showed an AngII-dependent regulation. It was previously reported that stimulation with AngII for three hours induced the negative regulator of the core-clock protein PER1 in H295R adrenocortical cells [21]. Additionally, overexpressing *PER1* in H295R cells was able to induce CYP11B1 and CYP11B2 promoter activity [21]. A role for the circadian-clock protein PER1 in the regulation of aldosterone production was recently reported in both in vitro and in vivo studies. *Per1* knock-out mice displayed lower aldosterone levels when compared with wild-type animals, and also a lower expression of *Hsd3b6* in adrenal gland tissue [22]. Silencing *PER1* in H295R cells was able to decrease the expression of *HSD3B1* isoform by 58%, supporting the hypothesis that *PER1* is involved in the modulation of serum aldosterone levels [22].

In the presented study, we showed that AngII stimulation triggers the expression of both *HSD3B1* and *HSD3B2* in HAC15 cells; our results differ from those reported by Ota T. et al. [23], showing that AngII can induce the expression of *HSD3B1*, but not *HSD3B2* in H295R cells.

To further investigate the potential role of *CRY1* and *CRY2* in the regulation of *HD3B* isoforms, we transfected HAC15 cells with *CRY1* and *CRY2* siRNA. Contrary to what was expected from the *Cry*-null and the *Per* knock-out murine models, silencing *CRY* genes did not modify the expression of *HSD3B1*; however, we observed a mild upregulation of *HSD3B2* in HAC15 cells transfected with *CRY1* siRNA. However, as previously described [24], silencing *CRY1* resulted in a significant upregulation of *CRY2*, which did not allow us to perform an efficient double silencing, and could, therefore, have affected the results, representing a limitation of the presented study.
