The ISL LIM‐homeobox 2 transcription factor is negatively regulated by circadian adrenergic signaling to repress the expression of Aanat in pinealocytes of the rat pineal gland

Melatonin is synthesized in the pineal gland during nighttime in response to nocturnal increase in the activity of the enzyme aralkylamine N‐acetyltransferase (AANAT), the transcription of which is modulated by several homeodomain transcription factors. Recent work suggests that the homeodomain transcription factor ISL LIM homeobox 2 (ISL2) is expressed in the pineal gland, but its role is currently unknown. With the purpose of identifying the mechanisms that control pineal expression of Isl2 and the possible function of Isl2 in circadian pineal biology, we report that Isl2 is specifically expressed in the pinealocytes of the rat pineal gland. Its expression exhibits a 24 h rhythm with high transcript and protein levels during the day and a trough in the second half of the night. This rhythm persists in darkness, and lesion studies reveal that it requires intact function of the suprachiasmatic nuclei, suggesting intrinsic circadian regulation. In vivo and in vitro experiments show that pineal Isl2 expression is repressed by adrenergic signaling acting via cyclic AMP; further, Isl2 is negatively regulated by the nocturnal transcription factor cone‐rod homeobox. During development, pineal Isl2 expression is detectable from embryonic day 19, preceding Aanat by several days. In vitro knockdown of Isl2 is accompanied by an increase in Aanat transcript levels suggesting that ISL2 represses its daytime expression. Thus, rhythmic expression of ISL2 in pinealocytes is under the control of the suprachiasmatic nucleus acting via adrenergic signaling in the gland to repress nocturnal expression, while ISL2 itself negatively regulates daytime pineal expression of Aanat and thereby suggestively enhances the circadian rhythm in melatonin synthesis.


| INTRODUCTION
2][3] Melatonin synthesis and release from the pineal gland is rhythmic, being confined to nighttime.The general regulation of melatonin synthesis by the endogenous clock of the suprachiasmatic nucleus (SCN) is well established.A multisynaptic pathway from the SCN governs the nocturnal release of the neurotransmitter norepinephrine (NE) in the gland from sympathetic neurons, the soma of which is located in the superior cervical ganglia. 4,57][8][9] In the pinealocyte, the principal cell type of the pineal gland, 5 melatonin is produced from tryptophan, which is turned into serotonin by the sequential action of tryptophan hydroxylase 1 (TPH1) and aromatic-L-aminoacid decarboxylase. 102][13] Rhythmic melatonin synthesis is controlled by daily oscillations in AANAT activity. 14In the rodent pineal gland, the binding of NE to adrenergic receptors on pinealocytes induces Aanat expression via the second messenger cyclic AMP and protein kinase A-dependent activation of the transcription factor CREB, [15][16][17][18] which in combination with posttranslational modifications 19 leads to increased AANAT activity during the night. 14,20Tph1 expression seems to display a dampened rhythm, while Asmt transcript and protein levels are stable throughout the day. 21,22issue-specific expression of Aanat and further modulation of the daily profile is ensured by several homeobox gene-encoded transcription factors.These homeodomain proteins are primarily known for their role in developmental processes, but as a special feature homeobox gene expression persists into adulthood in the pineal gland. 235][26][27][28] Recently, large-scale screening efforts identified the ISL LIM homeobox 2 (Isl2) gene as being rhythmically expressed in the pineal gland of the adult rat. 29,30Isl2 is known to regulate axon pathfinding, as well as motor neuron and retinal development [31][32][33] and to play a role in tumor progression. 34,35However, nothing is currently known about the possible role of Isl2 in the pineal gland.
Here, we aim to identify the mechanisms that control the pineal expression of Isl2 and the function of Isl2 in circadian biology of the pineal gland.The results of these efforts are presented below.

| Animals
Sprague Dawley rats were obtained from Janvier Labs and housed under a 12 h light:12 h dark schedule (LD) with food and water ad libitum.For circadian analyses, animals were kept in total darkness (constant darkness [dark:dark schedule]-DD) for 24 h before the start of the experiment.For developmental experiments, timed-pregnant mothers or pups were euthanized at the annotated timepoints.Embryonic day (E) 0 corresponds to the plug date and postnatal day (P) 0 corresponds to day of birth.Heads and brains were prepared as previously described. 24or tissue distribution and daily expression experiments, male rats (200-300 g) were killed at the annotated timepoints.The respective tissues were dissected and immediately frozen on dry ice.For cell culture experiments, 15-24 male rats (200-300 g) were killed at Zeitgeber time (ZT) 6-8, and the pineal glands pooled.For pharmacological stimulation in vivo, male rats (200 g) were intraperitoneally injected with isoproterenol (10 mg/kg; Tocris Bioscience, 1747) at ZT5 and euthanized 3 h later.Lesions of the SCN (SCNx) were carried out as previously described and confirmed both histologically and by telemetric registration of running activity and body temperature. 36Euthanasia was carried out by anesthetizing with CO 2 and subsequent decapitation.If animals were killed during the dark period, euthanasia was performed under dim red light.For immunohistochemistry, male rats (250-300 g) were perfusion fixed as previously described. 26ll animal experiments were in accordance with the guidelines of EU directive 2010/63/EU and approved by the Danish Council for Animal Experiments (2022-15-0201-01122) and the Faculty of Health and Medical Sciences, University of Copenhagen (P22-338, P22-341).

| Immunohistochemistry
Cryostat sections of 12-μm thickness were mounted on slides and blocked in 5% normal donkey serum (Abcam).Immunohistochemistry was performed as previously described. 22using the antibodies listed in Table 2.Sections were incubated in primary antibody at 4°C overnight and in secondary antibody at room temperature for 4 h, or 1 h if used for whole section imaging.In case of CRX detection, antigen retrieval was performed by incubating sections for 25 min in Na-citrate buffer at 80°C before antibody staining.The signal was amplified with tyramide (Invitrogen) using an ABC-vectastain kit (Vector).An Axioplan microscope (Zeiss) with a Plan-Neofluar 40× objective (Zeiss) equipped with an Axiocam 702 mono camera (Zeiss) was used for image acquisition of single-cell resolution images.Images of whole sections were taken on an Azure Sapphire Biomolecular Imager (Azure Biosystems).Fluorescence signal intensity after background substraction was quantified with the Azure Spot software.Contrast and brightness were adjusted in Adobe Photoshop 2022.

| Western blot
Samples were prepared as previously described. 25undred micrograms protein per lane was loaded into a NuPage 12% Bis-Tris gel (Thermo Fisher) followed by electrophoresis and blotting onto a nitrocellulose membrane.After blocking in 2% skim milk, membranes were incubated overnight in primary antibodies against ISL2 (2 μg/ml; Biotechne, AF4244) and ACTB (1 μg/mL; Abcam, ab8227) at 4°C, followed by 3 h of incubation in secondary antibodies (2 μg/mL; Invitrogen, A21102 or A11374) and a GAPDH loading control (Invitrogen, MA5-15738-D800).For image acquisition, an Azure Sapphire Biomolecular Imager was used.The density of bands was quantified with the Azure Spot software.Contrast and brightness were adjusted in Adobe Photoshop 2022.

| Preparation of cell cultures
Pineal cell cultures were prepared as previously described. 38To knock down gene expression, each culture was treated with 40 nM of small interfering RNA (siRNA) (Table 3).To mimick nighttime conditions and synchronize cells, they were treated with 3 μM NE (Sigma-Aldrich) or 500 μM dibutyryl-cyclic AMP (DBcAMP) (Sigma-Aldrich) for 6 or 9 h.
2.6 | RNA isolation, complementary DNA (cDNA) synthesis, and real time quantitative polymerase chain reaction (RT-qPCR) RNA was extracted using Trizol (Thermo Fisher) and treated with DNAse (Invitrogen).cDNA was synthesized according to the Superscript III protocol (Invitrogen).RT-qPCRs were run in a Lightcycler 96 (Roche) using FastStart Essential DNA Green Master (Roche) and the primers shown in Table 4. Serial dilutions of plasmids containing the target sequence were used to calculate the standard curves.Cq values above 32 were interpreted as negative results.The geometric mean of Gapdh and Actb levels was used to normalize the transcript levels of target genes.

| Statistical analysis
Statistical analyses were performed in GraphPad Prism 9.5.0.Student's t-test was used to compare data from two groups; if variances of compared groups were significantly different as determined by F-test, Welch's t-test was performed.Data from several groups were compared by use of one-way or two-way analysis of variance (ANOVA) with Bonferronicorrected multiple comparisons.p-Values below .05were considered statistically significant.
T A B L E 1 DNA oligo probes used for in situ hybridization.

| Pineal Isl2 expression begins during late embryonic development and precedes the appearance of Aanat transcripts
To establish the developmental dynamics of Isl2 expression in late prenatal and early postnatal stages, radiochemical in situ hybridization was performed on brain sections (Figure 1).No Isl2 expression was detected outside of the pineal gland in any section.In the pineal gland, Isl2 transcripts were first detected at E19 and increased until P10 (Figure 1A,B).No diurnal variation of Isl2 expression was observed until P10; however, beyond that stage nocturnal transcript levels decreased, so that a significant difference between day and night expression was seen at P30 (p = .0475,Bonferroni post hoc test) (Figure 1A,B).The expression of Aanat was analyzed to compare it to the Isl2 developmental profile.Aanat transcripts were first detectable after birth and reached their maximal values at P10.Diurnal variation in Aanat expression could be detected from P2 (p < .0001,Bonferroni post hoc test).This stage was characterized by substantial daytime expression levels, which subsequently decreased so that no transcripts were detected during daytime in later development, resulting in a highly significant diurnal difference of expression at P10 and P30 (p < .0001,Bonferroni post hoc test) (Figure 1A,C).Thus, pineal Isl2 expression preceded that of Aanat by four developmental days.The gradual decrease of nocturnal Isl2 expression from P2 on correlated with an increase of nocturnal Aanat up to its maximal values.Conversely, Aanat transcripts became undetectable during daytime, when daytime Isl2 expression reached its maximum at P10 (Figure 1A,D).

| Isl2 is expressed primarily in the melatonin-producing cells of the adult pineal gland
To establish the tissue distribution of Isl2 expression in the adult rat, samples from 13 different tissues and brain regions were dissected at ZT6 and ZT18 and transcript levels were quantified using RT-qPCR.Isl2 expression was only detected in the pineal gland and retina (Figure 2A).Pineal expression was more than 100-fold higher than that of the retina and displayed a diurnal variation, with lower expression at nighttime (p < .0001,Bonferroni post hoc test).In contrast, retinal expression was found to be nonrhythmic (p > .9999,Bonferroni post hoc test).In accordance with the pineal-specific transcript expression pattern, immunohistochemistry on brain sections revealed ISL2 protein exclusively in the pineal gland (Figure 2B).ISL2 was found to colocalize with the pinealocyte marker ASMT and thus to be specifically expressed in the melatonin-producing cells of the gland (Figure 2C).

| Pineal Isl2 transcript and protein levels exhibit a more than twofold circadian rhythm with a nadir during nighttime
To establish the daily expression pattern of Isl2 in rats housed in LD, transcript levels were quantified in pineal glands collected in 3 h intervals throughout a 24 h period (Figure 3A).Pineal Isl2 displayed a significant rhythm (p < .0001,one-way ANOVA) with expression being consistently high throughout the day but decreasing 3 h after the onset of the night to a nadir at ZT21, where transcript levels were approximately one-third of average daytime expression.Thus, the daily expression pattern of pineal Isl2 is inverted as compared to the expression of Aanat. 18,27,39To determine whether the expression of Isl2 is truly circadian, pineal glands were collected from animals kept in DD.The resulting 24 h expression profile resembled the one seen in animals kept in a regular lighting schedule, including the characteristic trough at circadian time 21 (Figure 3B).To determine whether the nocturnal downregulation of Isl2 transcription is translated into oscillating protein levels, ISL2 protein was quantified by western blot (Figure 3C) and immunohistochemistry (Figure 3D) on pineal glands from animals killed at ZT10 and ZT22.Differences in the density of the ISL2-containing western blot bands between the two timepoints were evident (p = .0441,Student's t-test) (Figure 3E).Similarly, a clear difference in signal intensity was detected between immunohistochemically reacted day-and nighttime pineal glands (p = .0117,Student's t-test) (Figure 3F) confirming rhythmic expression of ISL2 protein in the pineal gland.

| Expression of
Isl2 is regulated by the SCN, noradrenergic signaling, and the homeodomain transcription factor CRX NE and the second messenger cyclic AMP induce nocturnal pineal melatonin synthesis 5,13,33 and seem to also influence Isl2 expression. 23,24Quantification of Isl2 transcript levels in cultured pinealocytes after 6 h of incubation with either NE or the cyclic AMP analog DBcAMP revealed that Isl2 expression was reduced by F I G U R E 1 (See caption on next page).
more than 50% by both night-time mimicking substances (p = .0062;p = .0042,Bonferroni post hoc tests) (Figure 4A), in accordance with the nocturnal downregulation of Isl2 seen in vivo.To corroborate the influence of NE in vivo, animals were injected with the β-adrenergic agonist isoproterenol during daytime.Following this treatment, pineal Isl2 messenger RNA abundancy was reduced to 50% of the levels seen in phosphate-buffer saline-treated controls (p = .0143,Welch's t-test) (Figure 4B).
To determine whether the circadian master clock of the SCN regulates rhythmic Isl2 expression in the pineal gland, SCN lesion experiments were performed (Figure 4C).Following lesion, day-night variation of pineal Isl2 expression was undetectable (p = .5722,Bonferroni post hoc test) due to an increase of nocturnal transcript levels (p = .0363,Bonferroni post hoc test).
Homeobox gene-encoded transcription factors are known to modulate pineal melatonin synthesis; among these, Crx is upregulated during nighttime to promote the expression of Aanat, Tph1, and Asmt. 19,21Knockdown of Crx in cultured pinealocytes resulted in elevated Isl2 transcript levels (p = .0244,Student's t-test), indicating that CRX is an endogenous repressor of Isl2 expression (Figure 4D).Immunohistochemistry confirmed that CRX and ISL2 are expressed in the same pineal cells (Figure 4E).Thus, Isl2 is regulated by the same neural pathway, signaling molecules, and transcription factors as Aanat although with the opposite response.

| Knockdown of Isl2 in pinealocyte cultures upregulates rhythmic transcripts encoding melatonin-synthesizing enzymes
To test if ISL2 controls the expression of the enzymes catalyzing melatonin synthesis, Isl2 was knocked down in cultured pinealocytes.The expression of target genes was quantified after 9 h of stimulation with NE, since this would mimic the time with the highest Aanat expression in vivo (Figure 5). 9,27Isl2 expression was significantly downregulated by treatment with siRNA to less than 20% of control levels (Figure 5A).At the same time, Aanat was significantly upregulated (p = .0316,Student's t-test) (Figure 5B).Additionally, it was found that Isl2 knockdown had no effect on the expression of Asmt (p = .4648,Student's t-test) (Figure 5C), but significantly increased Tph1 transcript levels (p = .0495,Student's t-test) (Figure 5D).Thus, ISL2 is a downregulator of the rhythmic genes that encode melatonin-synthesizing enzymes.

| DISCUSSION
5][26][27][28] The transcription factor ISL2 investigated here negatively regulates pineal expression of Aanat and Tph1, two genes encoding melatonin-synthesizing enzymes, which is in line with an inverted 24 h Isl2 rhythm in the pineal gland compared to Aanat and Tph1.Notably, neurotransmitters and receptor interactions suppressing adrenergic induction of pineal melatonin synthesis have been previously reported [40][41][42] ; in this regard, our data suggest an alternative transcriptional pathway for repressing expression of melatoninproducing enzymes specifically during daytime.
This report represents the first set of functional data on Isl2 in the pineal gland, but a repressing effect of ISL2 on gene expression in other systems has been described previously.Knockout of Isl2 in human pancreatic tumor cells resulted in an equal number of up-and downregulated genes. 34Additionally, the well-studied effect of ISL2 on axonal pathfinding in development depends on the downregulation of target genes. 32,43This effect could F I G U R E 1 Ontogenetic expression of Isl2 and Aanat in the developing rat pineal gland.(A) Representative images of either Nisslstained sections (left column) to visualize brain morphology or radiographs of in situ hybridization for detection of Isl2 and Aanat transcripts (middle and right columns, respectively).Animals were killed at the annotated daily timepoints and developmental ages.Nissl staining was performed on the same sections used for Isl2 in situ hybridization, while Aanat was detected on distinct sections from the same animals with a distance of 10-140 μm between sections.An arrowhead marks the location of the pineal gland in Nissl-stained sections.Scale bar, 2 mm.(B) Quantification of the radiochemical Isl2 signal intensity in the developing pineal gland.Isl2 is detectable from E19. Day-night differences in expression are only seen at P30 (p = .0475,Bonferroni-corrected multiple comparison following two-way ANOVA).Two-way ANOVA revealed a significant effect of developmental time (p < .0001) on Isl2 expression.(C) Quantification of the radiochemical Aanat signal intensity in the developing pineal gland.A nighttime signal is first seen at P2. Daytime expression is only observed at P2. Significant diurnal variation is detectable at all postnatal developmental stages (p < .0001,Bonferroni-corrected multiple comparison following two-way ANOVA).Two-way ANOVA revealed a significant effect of developmental (p < .0001)and diurnal timepoints (p < .0001).(D) Comparison of the expression levels of Isl2 (black) and Aanat (gray) plotted as mean values connected by a dashed line (ZT6) or a straight line (ZT18).Values are plotted as fraction of the respective highest expression values measured for either gene.Aanat, aralkylamine N-acetyltransferase; ANOVA, analysis of variance; E, embryonic day; Isl2, ISL LIM homeobox 2; P, postnatal day; ZT, Zeitgeber time.be achieved through intermediate factors or by direct binding to regulatory regions.Over 12 000 genomic binding sites of ISL2 have been described, 4884 of which in a promoter region. 34Based on the proteins DNAbinding motif, 44 we have identified several potential ISL2 binding sites on the rat Aanat promoter and in the first intron (Supporting Information: Figure S3). 45ISL1, which shares a highly similar binding motif was previously found to bind to the Aanat promoter. 28,34owever, although both transcription factors have similar functions in motor neuron development, 46,47 we here show that the impact of ISL2 on melatonin synthesis is opposite of that described for ISL1 in the porcine pineal gland. 28Another repressor of Aanat expression is ICER, which competes for DNA-binding with CREB, the driver of Aanat transcription. 45,48In a similar way, ISL1 and ISL2 might compete to either promote or repress Aanat expression.However, no Isl1 expression was detected in the rodent pineal gland in a previous study, 29,30 and interspecies comparisons are impeded by the fact that the relative contribution of transcriptional and posttranslational molecular mechanisms which F I G U R E 2 Localization of Isl2 expression in the adult rat.(A) Transcript levels of Isl2 in day (ZT6) and night (ZT18) samples from 13 tissues quantified by RT-qPCR and normalized against the expression of Actb and Gapdh.Isl2 is only detectable in the retina and primarily the pineal gland.Its transcription is significantly affected by both the timepoint and tissue (time effect: p = .0210,tissue effect: p < .0001,two-way ANOVA).Pineal Isl2 mRNA levels differ between day and night, while the expression in the retina is non-rhythmic (p < .0001;p > .9999,Bonferroni-corrected multiple comparison following two-way ANOVA).(B) Immunohistochemical detection of ISL2 in sagittal (upper) and coronal (lower) brain sections from animals killed at ZT6. ISL2 is only detected in the pineal gland.Scale bars = 5 mm.(C) Immunohistochemical detection of ISL2 and ASMT in sections of the superficial pineal gland from animals killed at ZT6. ISL2 is detected in all cells that express ASMT.The bottom row shows a higher magnification of one group of cells in the upper images.For comparison, the same cell is marked with an arrowhead in all pictures.Scale bar in the upper panel, 50 μm.Scale bar in the lower panel, 10 μm.Actb, actin beta; ANOVA, analysis of variance; Gapdh, glyceraldehyde 3-phosphate dehydrogenase; Isl2, ISL LIM homeobox 2; RT-qPCR; real-time quantitative PCR; ZT, Zeitgeber time.
F I G U R E 3 Diurnal dynamics of Isl2 expression in the pineal gland of the adult rat.(A) Expression of Isl2 in pineal glands collected in 3 h intervals at the annotated timepoints from animals kept in LD.The values at ZT0 correspond to those at ZT24.Transcript levels were quantified by RT-qPCR and normalized against the expression of Actb and Gapdh.Pineal Isl2 expression displays a highly significant diurnal variation (p < .0001,one-way ANOVA).(B) Pineal expression of Isl2 in rats housed in DD and killed in 3 h intervals at the annotated timepoints.Normalized transcript levels display a highly significant diurnal variation (p < .0001,one-way ANOVA).(C) Two representative western blot lanes of samples containing protein extracted from pineal glands were harvested at ZT10 and ZT22 to visualize the difference of ISL2 abundancy in daytime versus nighttime samples.Rat ISL2 protein (NP_065204) has a predicted molecular weight of 39.7 kDa.Housekeeping proteins ACTB and GAPDH were used as loading controls.A neocortical control sample was used to ensure the specificity of the ISL2 antibody.Incubation of membranes in alternative blocking solutions, corresponding to the one used in immunohistochemical experiments, showed that the band at 150 kDa is a result of unspecific binding (Supporting Information: dynamically regulate AANAT activity differ between mammalian species. 49sl2 transcript levels are downregulated by NE and cyclic AMP, a pathway that induces gene expression mediated by activation of the transcription factor CREB. 16,17 Thus, an intermediate factor might be responsible for the repressing effect of NE stimulation described here.Accordingly, Isl2 expression begins to (p = .0062;p = .0042,Bonferroni-corrected multiple comparison following two-way ANOVA).Successful stimulation of pinealocytes was confirmed by quantification of Aanat transcript levels (Supporting Information: Figure S2A).(B) Isl2 expression in rat pineal glands following injection of isoproterenol (Iso) during the day normalized against expression of housekeeping genes.Animals were injected at ZT5 and killed at ZT8. Isl2 mRNA levels are significantly reduced by injection of isoproterenol as compared to PBS (p = .0142,Welch's t-test).Quantification of Aanat transcription confirmed successful stimulation of pineal cells (Supporting Information: Figure S2B).(C) Pineal Isl2 expression at the middle of the day (ZT6) and night (ZT18) following stereotaxic lesion of the SCN (SCNx) or sham lesion normalized against the expression of housekeeping genes.The significant diurnal difference in Isl2 expression in the sham group (p = .0214)is not observed in the SCNx group (p = .5722),due to an increase of nighttime expression (p = .0363),while daytime expression is not significantly altered (p > .9999,Bonferroni-corrected multiple comparison following two-way-ANOVA).Fully functional lesions were previously confirmed by histology and telemetric registration of running activity and body temperature. 36(D) Normalized expression of Isl2 in pinealocyte cultures following knockdown of the transcription factor encoding gene Crx and subsequent 9 h of stimulation with norepinephrine to synchronize the cells.Compared to cells treated with nontargeting siRNA, Isl2 expression was significantly higher following knockdown of Crx (p = .0244,Student's t-test).Crx was successfully knocked down to <15% of control levels (Supporting Information: Figure S2C).(E) Immunohistochemical detection of CRX and ISL2 in sections of the superficial pineal gland from rats killed at ZT6.Both proteins are localized in the same cells.The bottom row shows a cutout of the upper images in higher magnification.For comparison, the same cell is marked with an arrowhead in all images.Scale bar in the upper row, 50 μm.Scale bar in the lower row, 10 μm.Aanat, aralkylamine N-acetyltransferase; Actb, actin beta; ANOVA, analysis of variance; Crx, cone-rod homeobox; DBcAMP, dibutyryl cyclic AMP; Gapdh, glyceraldehyde 3-phosphate dehydrogenase; Isl2, ISL LIM homeobox 2; NE, norepinephrine; PBS, phosphate-buffer saline; SCN, suprachiasmatic nucleus; siRNA, small interfering RNA; ZT, Zeitgeber time.decrease approximately 1.5 h after the increase in pineal NE. 9 CRX is a possible candidate for this mediation since it downregulates Isl2 and its own increase in expression during night precedes the beginning of Isl2 transcript decrease. 27ranscription factors that modulate melatonin synthesis have previously been described as being expressed exclusively in the adult pineal gland and retina, two evolutionary-related tissues that share a substantial overlap of gene expression, specifically homeobox genes. 21,24,50,51Here, we find the same confined expression pattern for Isl2.Retinal expression of Isl2 is relatively low (100-fold lower than in the pineal gland), likely because it is confined to a subset of retinal ganglion cells in postnatal development. 33,52Accordingly, Isl2 expression was previously found to be enriched in the rat pineal gland but not the retina. 21Interestingly, a more diverse distribution of Isl2 expression was reported in adult tissues of other vertebrate species 53,54 ; however, these previous studies did not include pineal and retinal samples for comparison.
Our data on the developing rat pineal gland show that Isl2 transcripts are detectable at E19 and peak during the first two postnatal weeks before declining.Interestingly, the onset and peak of pineal Isl2 expression occur at the same times as those of Crx, 55 contradictory to the CRXmediated downregulation of Isl2 seen in adult pinealocytes, indicating that other factors are required to mediate this interaction.Isl2 onset of expression precedes that of Aanat by 4 days and might possibly suppress the latter during late embryonal development since nighttime Aanat transcripts substantially increase following the decrease in Isl2 expression.Similarly, daytime Aanat expression is fully suppressed at the same time when Isl2 peaks, again supporting the repressing role of ISL2 on Aanat expression.However, clearly ISL2 is not the only factor controlling Aanat expression, since Aanat is rhythmic as early as P2.Interestingly, a previous study carried out in Wistar rats reported Aanat transcription to become rhythmic not earlier than P5, followed by a slow gradual decline of daytime transcripts until adulthood. 56e propose that, at least in our model, pineal Aanat expression is rhythmic earlier than previously thought.In accordance with this, beginning innervation of the pineal gland by sympathetic nerve terminals is first seen at P1 and key factors of the signaling cascade that promotes Aanat expression are already functional at this point. 57,58ISL2 could potentially be involved in this maturation process.
Based on the distinct temporal and spatial expression patterns and the results of knockdown experiments, we propose a repressing role of ISL2 on Aanat expression, impeding the production and release of melatonin during the day, while downregulation of ISL2 may facilitate the peak of melatonin synthesis that occurs during the second half of the night.F I G U R E 5 Effect of Isl2 knockdown on transcripts encoding melatonin-synthesizing enzymes.Cultured pinealocytes were treated with siRNA (40 nM) for 24 h.The siRNA was replaced with 3 μm NE for 9 h to mimic nighttime before harvesting the cells.Target gene expression normalized against housekeeping genes from cells treated with siRNA targeting Isl2 (siIsl2) was compared to the control group treated with a nontargeting siRNA.Isl2 was successfully knocked down to <20% of control levels (p = .0003,Welch's t-test).In the same cells, the expression of Aanat and Tph1 are significantly upregulated (p = .0316;p = .0495,Student's t-tests), while expression of Asmt is not affected by Isl2 knockdown (p = .4648,Student's t-test).
Figure S1).(D) Two representative immunohistochemical stainings in brain sections from brains removed at ZT10 and ZT22 to visualize differences in ISL2 abundancy.Scale bars = 1 mm.(E) Quantification of ISL2 levels in the individual lanes of the western blot shown in (C).The signal intensity of the ISL2containing bands was normalized against the signal intensity of the respective bands containing the housekeeping proteins.ISL2 levels are significantly lower in nighttime samples compared to daytime samples (p = .0441,Student's t-test).(F) Quantification of ISL2 levels in the immunohistochemical stainings shown in (D), as measured by signal intensity inside the pineal gland.ISL2 protein levels are significantly reduced during the night (p = .0117,Student's t-test).Actb, actin beta; ANOVA, analysis of variance; DD, constant darkness (dark:dark schedule); Gapdh, glyceraldehyde 3-phosphate dehydrogenase; Isl2, ISL LIM homeobox 2; LD, 12 h light:12 h dark schedule; RT-qPCR, realtime quantitative PCR; ZT, Zeitgeber time.

F I G U R E 4
Regulation of pineal Isl2 expression.(A) Expression of Isl2 in cultured pinealocytes after 6 h of stimulation with NE or DBcAMP normalized against expression of Gapdh and Actb.Both signaling molecules significantly reduce the transcription of Isl2 AUTHOR CONTRIBUTIONS Martin F. Rath conceived the study.Kuno M.-J.Mattern and Martin F. Rath designed experiments, analyzed data, and prepared the manuscript.Kuno M.-J.Mattern, Aurea S. Blancas-Velázquez, Mikaella T. Ngo, Signe Bille, Henrik Hertz, Tenna Bering, and Martin F. Rath performed experiments.All authors approved the final manuscript.
T A B L E 3 siRNA used for knockdown in pinealocytes cultures.Primer sets used for RT-qPCR.