Effect of Combined Levothyroxine (L-T4) and 3-Iodothyronamine (T1AM) Supplementation on Memory and Adult Hippocampal Neurogenesis in a Mouse Model of Hypothyroidism

Mood alterations, anxiety, and cognitive impairments associated with adult-onset hypothyroidism often persist despite replacement treatment. In rodent models of hypothyroidism, replacement does not bring 3-iodothyronamine (T1AM) brain levels back to normal. T1AM is a thyroid hormone derivative with cognitive effects. Using a pharmacological hypothyroid mouse model, we investigated whether augmenting levothyroxine (L-T4) with T1AM improves behavioural correlates of depression, anxiety, and memory and has an effect on hippocampal neurogenesis. Hypothyroid mice showed impaired performance in the novel object recognition test as compared to euthyroid mice (discrimination index (DI): 0.02 ± 0.09 vs. 0.29 ± 0.06; t = 2.515, p = 0.02). L-T4 and L-T4+T1AM rescued memory (DI: 0.27 ± 0.08 and 0.34 ± 0.08, respectively), while T1AM had no effect (DI: −0.01 ± 0.10). Hypothyroidism reduced the number of neuroprogenitors in hippocampal neurogenic niches by 20%. L-T4 rescued the number of neuroprogenitors (mean diff = 106.9 ± 21.40, t = 4.99, pcorr = 0.003), while L-T4+T1AM produced a 30.61% rebound relative to euthyroid state (mean diff = 141.6 ± 31.91, t = 4.44, pcorr = 0.004). We performed qPCR analysis of 88 genes involved in neurotrophic signalling pathways and found an effect of treatment on the expression of Ngf, Kdr, Kit, L1cam, Ntf3, Mapk3, and Neurog2. Our data confirm that L-T4 is necessary and sufficient for recovering memory and hippocampal neurogenesis deficits associated with hypothyroidism, while we found no evidence to support the role of non-canonical TH signalling.


Introduction
Thyroid hormones (TH) play an essential role in brain development during gestational and perinatal periods and contribute to maintaining brain health in adulthood [1].Adultonset hypothyroidism is associated with depression, anxiety [2], and memory deficits [3,4] both in humans and in animal models [5][6][7][8][9].Poor memory performance is accompanied by altered synaptic plasticity in hippocampal circuits [10][11][12] and impaired neurogenesis in the neurogenic niche of the sub-granular zone (SGZ) of the dentate gyrus [13][14][15][16].Adult neurogenesis is highly sensitive to the hormonal, neurotransmitter, and growth factor milieu within the neurogenic niche, including TH signalling [7,12,17].Levothyroxine (L-T 4 ) monotherapy is the treatment of choice for hypothyroidism [18].However, about 10-15% of patients complain of persisting symptoms of TH deficiency, such as cognitive disturbances, fatigue, mood swings, and anxiety [19], despite thyroid-stimulating hormone (TSH) levels being within the normal range.Serum TSH may not reflect intracellular TH levels [19,20].Additionally, generation of triiodothyronine (T 3 ) from thyroxine (T 4 ) by deiodinases may not be equivalent to thyroidal secretion of T 3 [21].Finally, we hypothesise that persistent symptom of hypothyroidism may be due to decreased brain levels of downstream TH derivatives, in particular, 3-iodothyronamine (T 1 AM).In fact, in mice treated with methimazole-a thyroperoxidase inhibitor-and L-T 4 -replacement, brain T 1 AM levels remained undetectable despite an almost full recovery of free T 4 and T 3 levels (fT 4 and fT 3 ) [22].T 1 AM is an endogenous trace amine involved in the regulation of metabolic homeostasis and central nervous system (CNS) function [23].T 1 AM showed pro-cognitive and anti-amnestic effects [24], and neuroprotective actions in several models, including seizure-related excitotoxic damage, altered autophagy, amyloidosis, and ischemia-reperfusion injury [25][26][27][28].It is unknown if T 1 AM participates in the regulation of adult hippocampal neurogenesis.Here, we investigated whether augmenting L-T 4 treatment with T 1 AM improves neurocognitive and neurobiological alterations associated with adult-onset hypothyroidism and has an effect on hippocampal neurogenesis.In a rodent pharmacological model of hypothyroidism, we compared performance in memory, locomotion, and anxiety-and depression-related tasks in response to different replacement strategies combining L-T 4 and T 1 AM.We evaluated the effect of these replacement strategies on neurogenetic capacity in the SGZ using histological markers and the expression of genes involved in neurogenetic pathways.

Behavioural Correlates of Hypothyroidism and Replacement Treatments
Data relative to behavioural measures are reported in Table 1.In the elevated plus maze (EPM), hypothyroid mice showed a significantly lower number of stretch-attend postures (SAP), indicating reduced risk-assessment behaviours compared to euthyroid mice (t = 3.03, p corr = 0.006, surviving Bonferroni correction for multiple comparison).None of the other EPM measures were significantly different between hypothyroid and euthyroid mice.One-way ANOVA did not detect any significant differences in primary and secondary anxiety-related behaviours or decision-making and risk-assessment measures between treatments (Figure S1B-F).

Discussion
In this study, we confirmed that L-T4 monotherapy is necessary and sufficient to recover deficits in memory and adult hippocampal neurogenesis associated with adultonset hypothyroidism.We observed no further improvement in neurocognitive and neurobiological alterations from augmenting L-T4 with T1AM.
In our model of adult-onset hypothyroidism, we found no alterations in behavioural measures of locomotion, anxiety, or depression, in contrast with empirical clinical observation and previous reports [7,[29][30][31].Our hypothyroidism induction protocol almost completely suppressed serum TT4 levels, while we observed no changes in serum TT3 levels, suggesting that our protocol induced moderate hypothyroidism.Residual

Discussion
In this study, we confirmed that L-T 4 monotherapy is necessary and sufficient to recover deficits in memory and adult hippocampal neurogenesis associated with adult-onset hypothyroidism.We observed no further improvement in neurocognitive and neurobiological alterations from augmenting L-T 4 with T 1 AM.
In our model of adult-onset hypothyroidism, we found no alterations in behavioural measures of locomotion, anxiety, or depression, in contrast with empirical clinical observation and previous reports [7,[29][30][31].Our hypothyroidism induction protocol almost completely suppressed serum TT 4 levels, while we observed no changes in serum TT 3 levels, suggesting that our protocol induced moderate hypothyroidism.Residual thyroperoxidase activity may be able to ensure a sufficient production of T 3 to prevent alteration in measures of depression or anxiety.
On the other hand, our model of moderate hypothyroidism showed alterations in memory, suggesting that cognitive functions are more susceptible to TH deficiency.The effect of TH deficit on cognitive functions and synaptic plasticity has been extensively demonstrated in rodents [7][8][9][10]32] and corresponds to clinical evidence of deficits in spatial, associative and verbal memory in hypothyroid patients [3,4].In our hypothyroid mice, we found a 20% decrease in newborn neuroblasts in the SGZ of the dentate gyrus.A 4-week treatment with L-T 4 was necessary and sufficient to recover both memory and number of neuroblasts, consistent with previous evidence of the rescuing effects of L-T 4 on hypothyroidism-induced behavioural, electrophysiological, morphological, and molecular alterations [7,9,17,32].TH play a crucial role toward the postmitotic survival of adult dentate granule cell progenitors [17].It remains unclear if the decline in survival is due to a delay/block in neuronal differentiation or whether these cells may be targeted for cell death.The susceptibility of cognitive functions and neurogenesis to moderate TH changes may explain the relationship between TH and neurodegenerative diseases, such as Alzheimer's disease, observed in clinical populations [33][34][35], including those with subclinical hypothyroidism [2].
We found no significant differences between L-T 4 monotherapy and L-T 4 +T 1 AM in memory and number of DCX+ cells.T 1 AM did not show any effect per se.We focused on T 1 AM because we hypothesise that some of the neurocognitive symptoms of hypothyroidism traditionally attributed to lack of TH are partly caused by reduced availability of downstream metabolites.It has been demonstrated that T 1 AM remains undetectable in the tissues of animals pharmacologically depleted of TH, even following full recovery of fT 4 and fT 3 [22].Additionally, evidence has been accumulating about the effects of T 1 AM in the CNS.T 1 AM is a modulator of noradrenergic, dopaminergic, and histaminergic transmission [23].It has been observed to modulate sleep [36], exploratory activity, memory acquisition and retention [24].T 1 AM showed neuroprotective actions in models of seizure-related excitotoxic damage, altered autophagy, amyloidosis, ischemia-reperfusion injury, and neuroinflammation [25][26][27][28]37].In these studies, T 1 AM was administered either through perfusion on ex vivo brain acute/organotypic slices or through intracerebral microinjection, while we used a systemic administration.Studies employing radioactively labelled T 1 AM showed that T 1 AM crosses the blood-brain barrier [38].
We found an effect of replacement treatment on expression of several genes, including Ngf, Kdr, Kit, L1cam, Ntf3, Mapk3, and Neurog2.In particular, L-T 4 +T 1 AM-treated mice displayed an upregulation of Mapk3 and Neurog2 as compared to L-T 4 -treated mice.However, we did not observe significant differences between L-T 4 +T 1 AM and T 1 AM in Mapk3 and Neurog2 expression, suggesting that changes in gene expression could be induced by T 1 AM on its own.These genes are involved in neurogenetic mechanisms, such as increased proliferation of DCX+ cells in the hippocampus [39], initiation, and progression through the G1 phase of the cell cycle [40]; neuronal cell faith determination [40]; and suppression of glial faith promoting neurogenesis [41].TH regulate gene transcription through thyroid receptors (TR), widely expressed in different stages of granule cells progenitors [17].T 1 AM does not bind to TR but to several molecular targets, including the trace amine-associated receptor 1 (TAAR1), alpha2 adrenergic receptors, transient receptor potential channels, and ApoB-100 [42].How its distinct target profile links to gene expression will need to be clarified in future studies.However, we hypothesise a role for TAAR1 due to growing evidence of its involvement in neuropsychiatric disorders [43,44].Furthermore, we have previously demonstrated that T1AM neuroprotective actions in models of amyloidosis, ischemia-reperfusion injury, and neuroinflammation are dependent on TAAR1 [27,28,37].TAAR1 was reported to activate distinct signalling pathways depending on its localization in different cellular compartments.When on the plasma membrane, TAAR1 activation leads to increased intracellular cAMP levels and activation of the protein kinase A through coupling to Gαs proteins and adenylyl cyclase activation [45][46][47].Intracellular TAAR1 activation induces its coupling with Gα13 subunits to activate the GTPase RhoA [48].TAAR1 activation is also associated with G protein-coupled inwardly rectifying potassium channels [49,50] and G-protein-independent inhibition of glycogen synthase kinase 3β [51].
The present study has some limitations.First, our findings do not clarify if the effects of replacement treatments on memory are causally mediated by changes in hippocampal neurogenesis.Future studies should investigate if blocking neurogenesis during L-T4 and/or L-T 4 +T 1 AM treatment prevents memory rescue.Second, for some comparisons, the sample size was not sufficient to have enough statistical power.This is probably due to our protocol inducing moderate hypothyroidism (compensated TT 3 levels), which may have caused behavioural changes of small-to-moderate effect size, while our statistical analysis was powered to detect larger effect sizes, based on previous reports in thyroidectomized or transgenic rodents.Finally, to test our hypothesis that some of hypothyroidism-induced cognitive impairments are explained by T 1 AM deficiency, we planned to measure brain T 1 AM in different treatments.However, similarly to others [52], we found no clear endogenous signal of T 1 AM (below detection limit) in most of our pilot samples.Due to these technical reasons, we therefore decided to abort this objective.
Our data show that moderate TH changes are sufficient to induce significant impairments in memory and neurogenesis.L-T 4 was necessary and sufficient to recover both memory and number of neuroprogenitors to the euthyroid state, while we found no evidence to support the role of non-canonical TH signalling.This was despite an effect of T 1 AM on expression of genes involved in neurogenetic pathways, which should be further explored in future studies.

Mouse Model of Pharmacological Hypothyroidism
Six-week-old C57BL/6J male mice (maintained in a 12 h hight/dark cycle with ad libitum access to diet and water) were given methimazole (0.20 mg/g/day) and potassium perchlorate (0.30 mg/g/day) [22] in drinking water for 49 days, while control littermates received water.At day 21, mice were anesthetised with 1.5% isoflurane and implanted with ALZET ® subcutaneous osmotic pumps (Charles River Laboratories, Milan, Italy) delivering replacement treatments for 28 days (reservoir volume of 100 µL, delivery rate of 0.11 µL/h).Mice were divided into 5 groups: euthyroid, hypothyroid, L-T 4 (0.04 µg L-T 4 /g BW/day [53]), L-T 4 +T 1 AM (0.04 µg L-T 4 + 0.004 µg T 1 AM/g BW/day) and T 1 AM (0.004 µg T 1 AM/g BW/day) [27] (Figure 1A).All drugs were purchased from Sigma Aldrich-Merck (Milan, Italy).BW was monitored weekly throughout treatment.All experiments were conducted in accordance with the guidelines for the care and use of laboratory animals of the Italian Ministry of Health (Legislative Decree n. 116/92) and the European Community (European Directive 86/609/EEC).The Italian Ministry of Health approved the use of animals in this protocol (65E5B.10,n.734/2017-PR, 10 October 2017).

HPLC-MS/MS Analysis of Thyroid Hormones
Serum samples were processed to perform a HPLC-MS/MS quantification of TT 3 and TT 4 based on the previously reported methods [54] described in the Supplementary Materials and Table S1.

Behavioural Tests
Before testing, mice were habituated to experimental handling daily for 10 min.Over the last 3 days before the experiment, mice were habituated to the testing room for 1 h, i.e., the experimenter moved them from their usual holding room to the testing room, a soundproof environment with 100 lux illumination level.Afterwards, mice underwent behavioural tests over three consecutive days, in the following order: day 1, EPM; day 2, OF and ORT; day 3, TST.Mouse behaviour was videotaped and scored manually by investigators blind to the experimental condition.For the automatic scoring of the OF, we used the toolbox developed by Patel [55].

Elevated Plus Maze Test
The EPM apparatus consists of 4 arms (25 cm length × 5 cm width) forming a plus sign, elevated 50 cm above the floor.Two opposite arms have walls (16 cm height, closed arms), while the other two do not (open arms) (Figure S1A).Mice were placed in the centre and allowed to freely explore the maze for 5 min.Recorded variables and their interpretation [56][57][58][59] are detailed in Table 2. Mice were allowed to freely explore the OF arena for 10 min (Figure 2A).We scored total travelled distance, and thigmotaxis, i.e., time spent in outer zones versus inner zones.

Novel Object Recognition Test
After a 10 min habituation, mice were allowed to freely explore two identical objects for 3 min (familiarization).Then, one of the two objects was replaced with a novel object.In the test phase, mice were left to freely explore the objects for 3 min (Figure 3A).Exploration time was counted when the mice were within 2 cm from the object with the nose directed towards it, sniffing, or touching the object with the nose but not when the nose was pointing away from the object even if the mice were beside the object, running around it, sitting, or climbing on it.DI were calculated according to the formula: DI = (time at novel − time at familiar)/(time at novel + time at familiar)

Tail Suspension Test
Mice were suspended from a shelf, with climb-stoppers around the tail (Figure S2A).Escape-related behaviours were counted when mice: (a) tried to reach the apparatus walls and suspension bar; (b) shook the body; (c) showed movements akin to running.Escape was not counted when mice showed small movements confined to the front legs without involvement of the hind legs or in cases of oscillations due to the momentum gained during mobility bouts.Lack of escape-related behaviour was considered as immobility and was taken as a measure of depression.

Gene Expression Analysis
RNA was isolated from hippocampi dissected and homogenized in TRIzol™ (Ther-moFisher Scientific, Milan, Italy).RNA yield was determined with the Qubit Fluorometer (ThermoFisher Scientific).After DNase digestion, RNA was retrotranscribed using the iScript Clear cDNA Synthesis kit (Bio-Rad, Coralville, IA, USA).qPCR analysis was performed in PrimePCR TM "Neurogenesis Tier 1 M96" collection panel (Bio-Rad), a predesigned 96-well PCR plate containing primer sets for 88 gene targets involved in neurogenesis pathway for use with SYBR ® Green (Table S2) on a CFX Connect Real Time System (BioRad).Ct values > 35 were considered to be no expression.Fold change calculation by ∆∆Ct method [60] was performed with the CFX Maestro™ Software version 2.2 (Bio-Rad, Hercules, CA, USA) using three reference genes for normalization: TATA-binding protein (Tbp), glyceraldehyde-3-phosphate dehydrogenase (Gadph), and Hypoxanthine-guanine phosphoribosyltransferase (Hprt1).

Statistical Analysis
Outliers were removed with the ROUT method (Q = 1%).Normal distribution was assessed with the D'Agostino and Pearson test.Data were reported as mean ± standard error of the mean (SEM) or as median and interquartile ranges as appropriate.BW were analysed using two-way ANOVA for repeated measures, followed by Sidak's post hoc tests.Differences between groups in hormonal level, behavioural performance, immunofluorescence, and gene expression were assessed with one-way ANOVA, followed by Tukey's post-hoc tests or with a Kruskal-Wallis test as appropriate.For the ORT, one-sample t-test was used to determine whether the average DI for each group was different from chance (hypothetical value = 0).Statistical significance was set at p < 0.05.Statistical analyses were conducted using GraphPad Prism 6 (GraphPad Software Inc., La Jolla, CA, USA).

Figure 1 .
Figure 1.Murine model of pharmacological hypothyroidism.(A) Experimental protocol.Animals were divided into 5 groups: euthyroid (n = 17), hypothyroid (n = 15), hypothyroid treated with L-T 4 (n = 18), hypothyroid treated with L-T 4 +T 1 AM (n = 17), and hypothyroid treated with T 1 AM (n = 15).MMI, methimazole; KClO4, potassium perchlorate.(B) Body weight (BW) changes throughout the induction of hypothyroidism and replacement treatment.BW of hypothyroid mice started to diverge from euthyroid mice at week 2. Hypothyroid mice treated with L-T 4 and L-T 4 +T 1 AM partially recovered BW, while T 1 AM-treated and untreated hypothyroid mice remained significantly lighter than euthyroid mice.Two-way ANOVA testing the effect of treatment, time and their interaction and Tukey's post-hoc comparison were used to test differences between the groups (* p < 0.05, ** p < 0.01, *** p < 0.001).(C) Serum TT 4 levels were almost completely suppressed in the hypothyroid group and restored by L-T 4 and L-T 4 +T 1 AM.(D) We observed no significant changes in serum TT 3 levels between the groups.ANOVA and Tukey's multiple comparison were used to test differences between the groups (** p < 0.01, *** p < 0.001, n.s.not significant).Results are plotted as mean ± SEM using GraphPad Prism 6 software.Baseline n = 28, hypothyroid n = 6, euthyroid n = 7, L-T 4 n = 9, L-T 4 +T 1 AM n = 8, T 1 AM n = 6.The remaining samples were used to analyse serum fT 4 and fT 3 on an AIA 2000LA immunoassay platform.

Figure 2 .
Figure 2. Effect of hypothyroidism and different replacement treatments on locomotion.(A) Open field apparatus and example trajectories in the open field test (10 min duration) for experimental mice in the five experimental groups.We found no between-group differences in total distance travelled (B) or thigmotaxis (C) in the open field test.Hypothyroid n = 10, euthyroid n = 12, L-T4 n = 9, L-T4+T1AM n = 11, T1AM n = 10.Results are plotted as mean ± SEM using GraphPad Prism 6 software.

Figure 2 .
Figure 2. Effect of hypothyroidism and different replacement treatments on locomotion.(A) Open field apparatus and example trajectories in the open field test (10 min duration) for experimental mice in the five experimental groups.We found no between-group differences in total distance travelled (B) or thigmotaxis (C) in the open field test.Hypothyroid n = 10, euthyroid n = 12, L-T 4 n = 9, L-T 4 +T 1 AM n = 11, T 1 AM n = 10.Results are plotted as mean ± SEM using GraphPad Prism 6 software.Int.J. Mol.Sci.2023, 24, x FOR PEER REVIEW 6 of 16

Figure 4 .
Figure 4.The effect of different replacement treatments on hippocampal neurogenesis.(A) Representative confocal microscopy images of sections illustrating immunofluorescence of DCX+ (red) and Ki67+ (green).DAPI was used to counterstain nuclei (blue).The arrows point at proliferating cells, and the arrowheads point at some neuroblast neurites.Scale bars 20 µm.(B) DCX+ cell count.Each point represents the sum of DCX+ cells from four coronal sections taken across the hippocampus of a single mouse.Hypothyroidism caused a 20% reduction in the number of DCX+ cells in hippocampal neurogenic niches.L-T 4 rescued the number of DCX+ cells while L-T 4 +T 1 AM caused a 30.61% rebound relative to the euthyroid state.Pair-wise comparisons between the five groups using a t-test and Bonferroni-corrected for multiple comparisons (** p < 0.01, n.s.not significant).Hypothyroid n = 4, Euthyroid n = 4, L-T 4 n = 4, L-T 4 +T 1 AM n = 4, T 1 AM n = 4.

Author Contributions:
Conception of the work, G.R., L.M., C.K., N.O.and R.Z.; data collection, G.R., N.G., A.B., S.F., V.C., C.I., S.M. and A.S.; data analysis and interpretation, G.R., A.B., N.G., V.C., A.S., N.O.and R.Z.; drafting and critical revision of the article, G.R., A.B., A.S., N.O.and R.Z.All authors have read and agreed to the published version of the manuscript.Funding: This work was supported by the University of Pisa (PRA 2018 to R.Z.) and the NextGenerationEU-MIUR-CNR project "THE-Tuscany Health Ecosystem" (ECS 00000017, CUP B83C22003930001 to N.O.).Institutional Review Board Statement:All experiments were conducted in accordance with the guidelines for the care and use of laboratory animals of the Italian Ministry of Health (Legislative Decree n. 116/92) and the European Community (European Directive 86/609/EEC).The Italian Ministry of Health approved the use of animals in this protocol (65E5B.10,n. 734/2017-PR, 10 October 2017).

Table 1 .
Measures of anxiety-related behaviours, spontaneous locomotor activity, memory, and depression-related behaviours in the mouse model of hypothyroidism and replacement treatments.

Table 2 .
Behavioural interpretation of the variables recorded in the elevated plus maze test.