Hypothalamic deiodinase type-3 establishes the period of circannual interval timing in mammals

Targeted neuroanatomical injections, using highly precise Guide RNA (gRNA) against Dio3 and Crispr–Cas9 vectors, were essential to ensure the long-term inhibition of Dio3 function in the MBH in hamsters; an approach necessary to reliably examine the functional significance of Dio3 for the duration of circannual interval timing. Custom-made Crispr–Cas9 constructs were packaged in a lentiviral vector by Merck Life Science UK. Lentiviral vectors are established to effectively transfect in adult Syrian hamsters (Gao et al., 2014) and in Siberian hamsters (Munley et al., 2022). The lentiviral vector used was U6-gRNA:ef1a-puro-2A-Cas9-2A-tGFP (Sigma All-in-One vector). gRNA for Dio3 was designed using CHOPCHOP (Labun et al., 2019). Three gRNAs were designed to target the Mus musculus Dio3 gene (gRNA1: CGACAACCGTCTGTGCACCCTGG; gRNA2: GTTCCCGCGCTTCCTAGGCACGG, and gRNA3: GACCCAGCCGTCGGATGGGTGGG). Only gRNA 1 and 2 were identified to align with 5′ end of the Siberian hamster Dio3 gene and were predicted to align with chromosome 12 at locations 110279543 and 110279375, respectively (Figure 3—figure supplement 1). To check the specificity of gRNA sequences, we conducted BLAST using the NCBI dataset. Both gRNA1 and gRNA2 were found to align 100% with Dio3 and did not align 23/23 with any other region.

Adult male hamsters (n = 10) were selected to assess the impact of Dio3 genome modification on circannual interval timing. Hamsters were kept on LP conditions to maintain the summer phenotype. Hamsters were then divided into two treatment groups. The control group included hamsters (n = 6) that received ICV injections of blank Crispr–Cas9 constructs that were packaged into a lentivirus Dio3wt; U6-gRNA:ef1a-puro-2A-Cas9-2A-tGFP (Sigma All-in-One vector). The treatment group consisted of hamsters (n = 4) that received an ICV injection of Crispr–Cas9 constructs that harbored both gRNA1 and gRNA2 (Dio3cc).

ICV injections provide the ability to induce neuroanatomically localized manipulations in Dio3 gene function in the adult brain. Crispr–Cas9 constructs were delivered via stereotaxic ICV injection into the third ventricle to target the MBH tanycytes. Under general anesthesia (5% induction, 2% maintenance of isoflurane mixed with oxygen), a SGE series II syringe was positioned along the midline at Bregma –1.5 mm, to a depth of –5.7 mm. The anatomical coordinates were refined based on estimates taken from adult hamster brains (Steward et al., 2003) and tested using fresh hamster cadavers. Analgesia was administered subcutaneously (5 mg/kg Rimadyl and 0.1 mg/kg buprenorphine). A small ~1 cm incision was made by scalpel to expose the skull and a small hole was drilled using a Bone Micro Drill for Brain surgery (Harvard Apparatus). Lentivirus was administered using a Pump 11 Elite Programmable Syringe injection system (Harvard Apparatus) and 1 µl of viral vector (2–2.5 E + 13 vg/ml) was delivered over 10 min at a rate of 0.1 µl per minute. To allow viral diffusion and prevent backflow, the syringe was kept in place after injection for 1 min and then the syringe was raised by 1 mm and kept in place for another 1 min. Animals were then moved to heated home cages and given mashed food to aid recovery. Hamsters were provided with 2 weeks recovery and were monitored to ensure stable body mass, locomotor activity, and continued food and water intake.

To assess the impact of Dio3cc on circannual interval timing, hamsters were moved to SP for 32 weeks and body mass and pelage were measured biweekly. Pelage score was determined using visual assessment and defined scale in which a score of 1 is the summer dark agouti coat; a score of 2 is a transition between agouti and white; and a score of 3 is a full winter white pelage (Marshall et al., 2024). Body mass was monitored to the nearest 0.1 g using an Ohaus portable scale (TA301). After 32 weeks, hamsters were killed 4–5 hr after lights on by cervical dislocation followed by exsanguination. Brains were dissected and stored immediately at –70°C. Brains were cut into 200 µM sections using a Leica CM1520 cryostat. Anatomical structures (optic tract to the infundibular stem; Bregma –2.12 to –3.80 mm) were used to isolate the MBH. Bilateral tissue punches were performed using an Integra Miltex 1 mm disposable biopsy punch. Tissue punches were stored at –70°C. DNA was extracted from dissected tissues using the DNeasy Blood & Tissue Kit (QIAGEN) as per the manufacturer’s instructions. The Dio3 gene was amplified using OneTaq Quick-Load 2X mastermix, as per the manufacturer’s instructions (New England Biolabs). Amplification was achieved using a SimpliAmp thermal cycler (Applied Biosystems) the following thermal cycling conditions (94°C for 30 s (94°C for 15 s, 62°C for 30 s, 68°C for 3 min) for 45 cycles, 68°C for 5 min, 4°C until further analysis). The resultant amplicon was purified using the Qiaquick PCR purification kit (QIAGEN) as per the manufacturer’s instructions. Isolated Dio3 PCR amplicons were sequenced using the Eurofins LightRun Sanger sequencing service. The forward primer provided to Eurofins was CATGCTCCGCTCCCTGCTGCTTCA (5′–3′). CRISPR-modified transcripts were aligned to a reference wild-type transcript (animal #821) using the Tracy Sanger basecaller and aligner (Rausch et al., 2020). Alignments were visualized using the Ugene integrated bioinformatics tool (Okonechnikov et al., 2012). In case mutation was not obvious, an additional analysis was carried out using the associated online resource SAGE (https://www.gear-genomics.com/), to investigate the decomposition error of the modified transcripts with the wild-type control (821). Four Dio3wt hamsters had Dio3 sequences that fully matched the reference genome (Figure 3—figure supplement 1). Two hamster Dio3 sequences did not automatically align, and the raw reads were visually inspected. No mutations or sequence mutations were detected in the raw sequence trace. Three Dio3cc-treated hamsters had Dio3 sequences with evidence of genomic mutation evidenced by low decomposition error (Figure 3—figure supplement 1). One Dio3cc-treated hamster did not show any genomic mutation (905) and was removed from the final analysis. Hamster 905 had a body mass circannual interval period of 32 weeks and we propose the animal represents a false positive control.