Abstract
Devastating diseases of the central nervous system (CNS), such as neurodevelopmental disorders and neurodegenerative diseases, carry severe neurological symptoms, including cognitive impairment –a decline characterized by learning and memory problems. Currently, no efficacious treatment is available for preventing or treating these diseases; therefore, there is a great unmet need to find novel, effective therapeutics. Previous studies have shown that insulin-like growth factor 2 (IGF2) and mannose 6-phosphate (M6P), two major ligands of the cation-independent M6P/IGF2 receptor (CIM6P/IGF2R), significantly enhance memory and prevent forgetting in healthy animals. Furthermore, they reverse most core symptoms in mouse models of neurodevelopmental disorders, such as Angelman syndrome (AS mice) and autism spectrum disorder, and of neurodegenerative diseases. Starting from the M6P chemical structure, we designed the novel prodrug PMP1. We tested its effects on learning and memory in healthy mice as well as on behavioral deficits in AS mice. We report that both subcutaneous and oral administrations of PMP1 significantly enhance memory strength and persistence in healthy mice. Furthermore, they reverse memory impairments and motor deficits in AS mice. The effect of PMP1 requires the expression of CIM6P/IGF2R in the hippocampus, and the treatment does not elicit any detectable adverse effects. The highest effects were achieved with the same PMP1 maximum effective dose via either route of administration; however, oral administration produced a more prolonged effect. Thus, the novel small-molecule prodrug PMP1, through CIM6P/IGF2R, is a promising therapeutic candidate for Angelman syndrome and an effective memory enhancer.
Introduction
Memory and cognitive impairments are found in several neuropsychiatric disorders, including neurodegenerative diseases [1, 2] and neurodevelopmental disorders [3]. Currently, there are no effective treatments for these devastating and widespread diseases; therefore, there is an urgent need to discover new therapies based on novel compounds that target new mechanisms. Studies in healthy mice and rats have shown that the administration of IGF2 (or IGFII), a small protein of 67 amino acids belonging to the insulin-IGF system [4, 5], promotes stronger and more prolonged memories [6,7,8,9,10,11,12]. Furthermore, IGF2 reverses memory impairment in aged rats [11], as well as memory impairments and several other behavioral deficits in a variety of rodent models of neurodevelopmental disorders and neurodegenerative diseases [13,14,15,16,17]. Where tested, the beneficial effects of IGF2 have been shown to be mediated by the high-affinity receptor for IGF2 – the CIM6P/IGF2R, and not by the IGF1 receptor [6, 8, 18, 19]. Similar to IGF2, another main ligand of CIM6P/IGF2R, M6P, also significantly enhances memory in healthy mice and rats and reverses major symptoms, including memory and motor impairments as well as repetitive behavior, in a mouse model of the neurodevelopmental disorder Angelman syndrome [12, 19]. These results support the conclusion that the CIM6P/IGF2R triggers memory enhancement as well as recovery from cognitive and other behavioral impairments in several models of neurodevelopmental disorders and neurodegenerative diseases.
The CIM6P/IGF2R is an approximately 300 kDa transmembrane receptor that, in addition to binding IGF2 and transporting it to lysosomes for degradation, binds with high affinity to M6P moieties conjugated to newly synthesized acid hydrolases in the trans-Golgi network (TGN) for their subsequent transfer to lysosomes. CIM6P/IGF2R, via endosomes, can either recycle back to the TGN or be trafficked to the plasma membrane [20]. CIM6P/IGF2R can bind to a variety of M6P-containing ligands, including lysosomal enzymes, transforming growth factor-β (TGF-β), granzyme B, glycosylated leukemia inhibitory factor (LIF), proliferin, and thyroglobulin (see [21] for review). Other ligands that may bind to the receptor through a non-M6P-based interaction are retinoic acid, urokinase-type plasminogen activator receptor (uPAR), and plasminogen itself (see [21] for review).
An emerging body of literature indicates that CIM6P/IGF2R plays a crucial role in the CNS. Compared to other brain cell types, CIM6P/IGF2R is expressed at higher levels in neurons, where it is required for the formation of hippocampus-dependent long-term memories [19]. These are the memories of life episodes, contexts, space, conspecifics, and time, which decay in cognitive impairments [22]. CIM6P/IGF2R also plays an important role in layer 2/3 of the auditory cortex for the precision of recent memories, the stability of remote memories, and the development of fear generalization [23]. A key molecular mechanism of neuronal CIM6P/IGF2R is the control of the increase in de novo protein synthesis evoked by learning, a process that is coupled to autophagy and is a fundamental requirement for the formation of long-term memories [6, 14, 19, 24]. Impaired protein metabolism, and particularly defective protein degradation via autophagy and lysosomes, is a shared problem of several CNS diseases, including neurodevelopmental disorders like autism spectrum disorder and Angelman syndrome [25,26,27] and neurodegenerative diseases such as Alzheimer’s disease, Huntington’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis [28,29,30,31,32]. We hypothesized that ligands of CIM6P/IGF2R may represent new, potential treatments for cognitive impairments and other symptoms of these diseases.
CIM6P/IGF2R binds IGF2 at the extracellular domain 11 [33, 34] and M6P at domains 3, 5, 9 and 15 [21]. The structure of the M6P binding site has been used to generate small-molecule binders, such as phosphonate and carboxylate analogs of M6P [35,36,37]. Previous work highlighted the potentially positive role of CIM6P/IGF2R binding to M6P-modified proteins in diseases such as the lysosomal storage disorder Pompe disease [38] and in cancer [39,40,41]. These findings inspired the development of M6P analogs, often with higher affinity for CIM6P/IGF2R than M6P, to deliver therapeutic compounds to lysosomes [42]. Moreover, ligands targeting CIM6P/IGF2R have also been used for controlling protein degradation. In fact, lysosome targeting chimeras (LYTACs) have been developed as tools that engage CIM6P/IGF2R for targeted degradation of extracellular and membrane-bound proteins [43, 44].
Here, we designed a novel prodrug phosphonate analog of M6P, called PMP1, that, by exposing the phosphonate group, targets the M6P binding site of the CIM6P/IGF2R. Our molecule features a “disoproxil” phosphorester, which has been shown to be a highly effective phosphonate masking group in the antiviral drug tenofovir disoproxil [45, 46]. This design aimed to create a compound that would selectively bind CIM6P/IGF2R after conversion into an active principle (AP) and be efficacious when administered orally, thereby offering opportunities for translational drug development.
We tested the efficacy of PMP1 on memory strength and persistence in healthy (wild-type, WT) mice as well as on memory and motor deficits in a widely studied mouse model of Angelman syndrome [47] (referred to as AS mice). We show that PMP1, administered either via subcutaneous (s.c.) injection or oral gavage, significantly enhances memory in healthy mice and reverses memory impairment and motor deficits in AS mice. CIM6P/IGF2R expressed in the hippocampus is required for the PMP1 effect evoked by systemic treatment. Moreover, no toxic or adverse effects were detected after PMP1 treatment. These data suggest that the novel compound PMP1 could be developed as a potential therapeutic for Angelman syndrome and several other CNS diseases.
Materials and methods
PMP1 synthesis
Compounds 2-8 were prepared according to literature procedures [48,49,50,51,52]. Compounds 8, 10, and 11 (PMP1) were synthesized as follows:
Methyl 2,3,4-Tri-O-benzyl-6-deoxy-6-dihydroxyphosphinylmethylene-α-d-mannopyranoside (8)
Phosphonic acid 8 was prepared according to a published procedure [53]. To a solution of 7 (0.146 g, 0.245 mmol, 1 eq) in anhydrous CH3CN (5.6 mL) under nitrogen was added pyridine (31 μL, 0.392 mmol, 1.6 eq) and trimethylsilyl bromide (0.32 mL, 2.45 mmol, 10 eq) with stirring at ambient temperature. After 2 h, the reaction mixture was cooled to 0 °C and was added pyridine (51 μL, 0.634 mmol, 2.6 eq) and H2O (185 μL, 10.3 mmol, 42 eq) then warmed to ambient temperature and stirred. After 2 h, the reaction mixture was diluted with CH2Cl2 and 2 M HCl (4 mL) and water (4 mL). The organic layer was extracted with CH2Cl2, dried over anhydrous sodium sulfate, and concentrated under reduced pressure to yield 8 as a brown oil. The crude residue was used directly in the following procedure.
Methyl 2,3,4-Tri-O-benzyl-6-deoxy-6-diisopropyloxycarbonyloxy-methyl-phosphinyl- methylene-α-d-mannopyranoside (10)
A reported procedure was modified and used to prepare phosphonate 10 [54]. Crude 8 from the previous procedure in anhydrous CH3CN under nitrogen was treated with DIPEA (0.480 mL, 2.76 mmol, 9.9 eq), TBAB (93.1 mg, 0.289 mmol, 1.0 eq), and chloromethyl isopropyl carbonate (0.30 mL, 2.24 mmol, 8.1 eq) then was heated to 60 °C. After stirring for 16 h, the reaction mixture was concentrated under reduced pressure. The crude residue was purified by flash column chromatography (30 → 100% ethyl acetate/ hexanes) to yield 10 as a colorless oil (116 mg, 0.150 mmol, 54%).
Rf = 0.49 (EtOH/EtOAc/hexanes 1.5:1.5:7).
1H NMR (CDCl3, 400 MHz) δ 7.40 - 7.29 (m, 15H), 7.10 (ddd, J = 24.5, 17.2, 3.8 Hz, 1H), 6.40 - 6.17 (m, 1H), 5.80 - 5.65 (m, 6H), 4.81 - 4.59 (m, 7H), 4.22 - 4.14 (m, 1H), 3.91 (dd, J = 9.3, 3.1 Hz, 1H), 3.83 - 3.78 (m, 1H), 3.74 (t, J = 9.5 Hz, 1H), 3.30 (s, 3H), 1.32 - 1.29 (m, 12H) ppm.
13C NMR (CDCl3, 101 MHz) δ 153.5, 138.7, 138.5, 138.3, 128.8, 128.7, 128.6, 128.2, 128.1, 127.9, 99.7, 84.5 (d, J = 5.7 Hz), 84.4 (d, J = 6.8 Hz), 80.5, 78.3 (d, J = 2.1 Hz), 75.8, 75.0, 73.5 (d, J = 3.5 Hz), 73.3, 72.7, 71.3 (d, J = 22.3 Hz), 55.3 ppm.
31P NMR (162 MHz, CDCl3) δ 26.3 ppm.
Methyl 6-Deoxy-6-bis(isopropyloxycarbonyloxymethyl)-phosphinylmethyl-α-d-mannopyranoside (PMP1) (11)
A reported hydrogenation procedure was adapted and used to prepare PMP1 [55] (11). In an oven-dried vial 10 (36.0 mg, 0.047 mmol, 1 eq) was dried and degassed under high vacuum. To this was added 10% Pd/C (36.6 mg, 0.344 mmol, 7.4 eq) and rinsed down with CH2Cl2 (2 mL) and EtOH (2 mL). The reaction mixture was subsurface sparged with N2 for 1 min. The reaction mixture was then degassed under reduced pressure and the atmosphere was replaced by H2 (5×). The reaction mixture was stirred vigorously under H2 for 4 h, after which time the reaction mixture was degassed under reduced pressure and refilled with N2 (5×). The reaction mixture was diluted with CH2Cl2 (2 mL) and filtered over a plug of wet celite. The filtered organic layer was concentrated under reduced pressure and the crude residue was purified by HPLC (40 → 85% [H2O + 0.1% FA]:[CH3CN + 0.1% FA]) to afford PMP1 (11) (10.1 mg, 0.020 mmol, 43%) as a white solid (Fig. S1). All 13C-31P coupling constants match reported values [56].
tR = 7.00 min (HPLC (40 → 85% [H2O + 0.1% FA]:[CH3CN + 0.1% FA]))
1H NMR (400 MHz, CDCl3) δ 5.68 (2 H, dd, J = 20.5 Hz, J = 5.3 Hz, H8), 5.65 (2 H, dd, J = 18.3 Hz, J = 5.4 Hz, H8′), 4.93 (2 H, hept, J = 6.3 Hz, H10), 4.68 (1 H, s, H1), 3.95 - 3.86 (1 H, br, H5), 3.74 (1 H, m, H2), 3.58 (2 H, m, H3, H4), 3.35 (3 H, s, OCH3), 3.22 - 3.07 (1 H, m, OH), 2.95 (2 H, m, 2 × OH), 2.27 - 2.07 (2 H, m), 2.06 - 1.86 (2 H, m, H6, H6′, H7, H7′), 1.32 (12 H, d, J = 6.2 Hz, H11) ppm.
13C NMR (101 MHz, CDCl3) δ 153.6 (d, J = 3.7 Hz, C9), 101.2 (s, C1), 84.5 (d, J = 6.3 Hz, C8), 84.3 (d, J = 6.3 Hz, C8′), 73.7 (d, J = 3.2 Hz, C10), 72.0 (s, C2), 70.9 (d, J = 16.1 Hz, C5), 70.6 (s), 70.5 (s, C3, C4), 55.3 (s, OCH3), 23.8 (d, J = 4.5 Hz, C6), 22.4 (s, C11), 21.7 (d, J = 142.3 Hz, C7) ppm.
31P NMR (162 MHz, CDCl3) δ 34.4 ppm.
FT-IR (neat, cm-1): ν(O–H) = 3409 (br), ν(C–H) = 2923 (m), ν(C = O) = 1760 (s), ν(P = O) = 1269 (s).
MS (ESI–) calcd for [M + HCOO]−: 549.2; found: 549.2.
Two batches of PMP1 were synthesized at NYU, one batch was generously donated by Ritrova Therapeutics, which conducted the synthesis at EVOTEC-Apuit (Verona, Italy).
Ethics approval and consent to participate
All methods were performed in accordance with the relevant guidelines and regulations.
Animal housing and experiments were conducted in the animal facility provided by the Office of Veterinary Resources (OVR), located at the Washington Square Campus of New York University and were in strict accordance with the institutional guidelines for care and use of laboratory animals outlined by OVR in conjunction with the University Animal Welfare Committee (UAWC). This study and included experimental procedures comply with the NIH Guide for the Care and Use of Laboratory Animal and were approved by the Institutional Animal Care and Use Committee (IACUC) (Protocol #19-1511), which oversees the use of animals for research conducted at or by NYU.
Mice
The study was performed on young adult wild type (WT) C57BL/6J (Jackson Laboratory, Maine, USA), maternal Ube3a-deficient [25] mice and Igf2rfl/fl mice [57] at approximately 8 weeks of age at the start of the experiments. To generate maternal Ube3a-deficient mice (Ube3am−/p+), male mice carrying a paternally imprinted Ube3a knockout mutation on a C57BL/6J background (B6.129S7-Ube3atm1Alb/J; The Jackson Laboratory; Stock No. 016590) were paired with C57BL/6J female mice (F0 generation). The obtained female heterozygotes [25] (Ube3am+/p−) were bred with male WT mice (F1 generation) to obtain Ube3am−/p+ (AS) and WT littermates (F2 generation). AS mice from F2 generation were used for all the experiments of this study. WT male and female littermates were used as controls. Igf2r floxed mice were generously provided by Dr. R.L. Jirtle (Duke University) [57]. Tail biopsies from 4 weeks old mice were used for genotyping that was carried out by TransnetYX (Cordova, TN, USA). Male and female littermates were included at equivalent or near equivalent ratios in all the experiments, and mice were randomly assigned to the experimental groups. For all experiments, data from both males and females were combined and analyzed as a single group because statistical analyses of separate sex groups did not show statistically significant differences (Two-way ANOVA followed by Tukey’s post-hoc test, significance was considered starting at p < 0.05; the separate sex groups had a n/group of 3–8). Mice were group-housed on a 12:12 light/dark cycle with ad libitum access to food and water. All experiments were carried out during the light cycle. Each behavioral paradigm employed dedicated groups of WT and AS mice, except for hindlimb clasping. As hindlimb clasping is a very quick test consisting of a 30-s suspension by the tail (see also the dedicated hindlimb clasping section for details), this behavior was assessed immediately before each novel object recognition (nOR) test. All animal procedures were approved by the Institutional Animal Care and Use Committee of the New York University and were performed in accordance with guidelines of the U.S. National Institutes of Health.
Pharmacological treatments
Each group of mice was injected 20 min (min) before each behavioral exposure (nOR, limb clasping, Y-maze, open field, contextual fear conditioning, or observational battery) with vehicle or one of three different compounds: recombinant mouse IGF2 (R&D systems, 792-MG-050, IGF2), M6P (Sigma M3655) or PMP1. The administration time of 20 min before behavioral exposure, such as learning, was chosen based on several prior studies showing that other ligands of IGF2R, specifically IGF2 and M6P, administered 20 min before behaviors effectively enhance memory and reverse impairments in disease models [11, 12, 18]. Each s.c. injection was given in a volume of 0.01 mL/g of body weight. IGF2 was dissolved in 0.1% bovine serum albumin-phosphate-buffered saline (0.1% BSA-PBS, pH 7.4) and injected s.c. at 30 μg/kg [12, 18]. M6P was dissolved in PBS (pH 7.4) on the day of the experiment and injected s.c. or orally at 850 μg/kg [12]. PMP1 was dissolved in PBS, pH 7.4/10% of dimethyl sulfoxide (DMSO, Sigma-Aldrich, 276855) and diluted to the final concentrations in PBS, pH 7.4/1% DMSO on the day of the experiment. A dose-response curve of PMP1 injected either s.c. or orally was done to identify the maximum effective dose of the compound, as detailed in the results.
In all experiments, mice were handled for 2–3 min per day for 5 days before behavioral procedures. Mice were randomly assigned to receive treatment with either IGF2, M6P, PMP1, or vehicle. In the behavioral experiments where multiple tests were carried out over time (longitudinal studies), the mice were repeatedly tested at the times indicated in each experiment, as detailed in the Results and Figure legends. Validation of data was ensured by repetitions of experiments conducted on different days and on different groups of mice (n = 4–5/group per experiment) and performed by different experimenters. All experiments and analyses were performed blind to treatment.
Hindlimb clasping
Hindlimb clasping behavior was used to assess motor function on AS mice and was performed as previously described [12]. Each mouse was suspended by the tail 10 cm above their home cage for 30 s. The session was video recorded, and time was scored offline by independent experimenters blind to genotype and treatment conditions. The score of limb clasping was defined as follows: 0- no limb clasping, if the limbs were consistently splayed outward, normal escape extension; 1- incomplete splay of either hind limbs or forelimbs and loss of mobility; 2- either hind limbs or forelimbs clasping together and loss of mobility; 3- three limbs clasping together and loss of mobility; 4- four limbs clasping together and loss of mobility. Hindlimb clasping was scored immediately before nOR testing, therefore in the same groups of mice.
Novel Object Recognition (nOR)
NOR was performed as previously described [12, 18]. On day 1 (habituation), each mouse was placed in a clean square novel arena for 5 min to allow habituation. On day 2, each mouse was trained in the same arena, now containing two identical objects (Mega Bloks 120) and allowed explore freely for 3 min (training). Memory tests were performed at 4, 24 h, 1, 2, 3, 4, or 5 weeks after training: each mouse was returned to the same arena, but this time one of the objects was replaced by a novel object, and the time spent exploring both objects was scored. Each session was video-recorded and the videos were automatically analyzed by EthoVision XT software (Noldus). Time spent interacting/sniffing each object over 5 min was recorded in seconds, and memory retention was expressed as the percentage index exploration preference for the novel object 100 × (novel object/novel object + old object in seconds) over 5 min.
Contextual Fear Conditioning (CFC)
CFC was carried out as previously described [25]; the conditioning chamber consisted of a rectangular Perspex box (30.5 × 24.1 × 21.0 cm3) with a metal grid floor (Model ENV-008-VP, Med Associates, Fairfax, VT) through which foot shocks were delivered via a constant current scrambler circuit. During training, each mouse was left to explore the chamber for 2 min and then received an unsignaled 0.7 mA footshock for 2 s. After 1 additional min in the chamber, the mouse was returned to its home cage. Memory testing was performed at 24 h, 1, 2, 3, 4, and 5 weeks after training: each mouse was placed back in the conditioning chamber and allowed to freely explore it for 3 min in the absence of a foot shock. Each session was videorecorded for offline analysis. Freezing, defined as lack of movement except for heartbeat and respiration, was recorded every 10th s by independent experimenters blind to treatment conditions. Percent freezing was calculated as total number of freezing events divided by the total number of bouts assessed for each mouse × 100, as previously reported [25].
Open field
To test locomotor activity and anxiety, the open field test was performed as previously described [18]. Briefly, the light intensity of the box during open field testing was maintained at 195 lux. Mice were allowed to freely explore an open-field arena (43.2 × 43.2 × 30.5 cm3; ENV-515, Med Associates, Fairfax, VT), designated into 16 identical squares, for 5 min. The activity was analyzed with Ethovision-XT (Noldus). Locomotor activity was measured by tracking the total distance traveled by the mouse in the arena. Anxiety-like behaviors were measured as the total entries as well as the time spent in the four central squares out of 16 total squares.
Spontaneous alternation in Y-maze
The spontaneous alternation test was performed as previously described [9, 18] on a three-arm Y-maze consisting of three black polycarbonate arms (7.6 × 12.7 × 38.1 cm3). Mice were allowed to freely explore the arms from the center of the maze for 8 min while being video recorded and analyzed using Ethovision-XT (Noldus). One spontaneous (correct) alternation was defined as successive entries into each of the three labeled arms on overlapping triplet sets (e.g., ABC, BCA, CAB, etc.). Percentage of correct alternations was determined as the fraction of spontaneous alternations to possible alternations (total arm entries−2) × 100.
Observational battery
The observational battery was carried out as previously described [12] with minor modifications. The battery was assessed once prior to s.c. injection of PMP1 or vehicle and a second time at 30 min after injection. The list of each observational screen sub-component was completed in the following order: General Behaviors, Sensorimotor Reflexes, and General Characteristics. Each mouse was observed and video recorded in an empty cage for 1 min, and general behavioral observations were documented. Sensorimotor reflexes and simple motor responses were then tested in the order outlined in Table 1. Body temperature was taken from the mouse lower abdomen with a digital external infrared thermometer (Exergen, TAT-2000C) [58,59,60] and physical characteristics were recorded. Observers blind to experimental procedures scored all experiments.
Cloning of Cre recombinase into plasmid and AAV production
Coding sequences for NLS-T2A with or without Cre recombinase were amplified from LFNG:GFPnls-T2A-ER-Cre-ER (AddGene, #200107, AddGene) by PCR using Q5 HighFidelity 2x Master Mix (M0492S, NEB, Ipswich, MA). The two amplified fragments were cloned into the BamHI site of pAAV-Syn-GFP (AddGene #58867) using HiFi Assembly 2xMaster Mix (NEB, E2621S) to generate pAAV-Syn-Cre-NLS-T2A-eGFP. Primers used are listed in the table below.
Primer table.
Name | Sequence |
|---|---|
CRE-F | CCTGAGAGCGCAGTCGAGAGGCCACCATGAATTTACTGACCGTACACC |
CRE-R | TCTTTTTTGGGTATCTCTGACCGGAGTC |
NLS-T2A-F | TCAGAGATACCCAAAAAAGAAGAGAAAGGTAGATCCAAAAAAGAAGAGAAAGGTAGATC |
NLS-T2A-R | CCTTGCTCACCATGGTGGCGGCCGGCCATGACAGGGCC |
The obtained AAV plasmids were propagated in chemically competent E. coli OneShot Stbl3 (C737303, Thermo Fisher, Waltham, MA) to minimize recombination events. All constructs were verified by sequencing.
Both GFP-encoding plasmids pAAV-Syn-GFP and pAAV-Syn-Cre-T2A-GFP were encapsidated for targeted central neurotropism with pUCmini-iCAP-PHP-N (AddGene #127851) by the Gene Vector and Virus Core (Wu Tsai Neurosciences Institute, Stanford, CA).
Hippocampal AAV injections to knockout CIM6P/IGF2R in the dorsal hippocampus
Adult male and female Igf2rfl/fl mice were anesthetized using isoflurane mixed with oxygen. The skull was exposed, and holes were drilled in the skull bilaterally above the dorsal hippocampus (dHC). A Hamilton (Reno, NV) syringe with a 33-gauge needle mounted onto a NanoPump (KD Scientific, Holliston, MA) was stereotactically inserted into the dHC (−2 mm AP, 1.5 mm lateral, 2.3 mm DV from bregma) and 0.5 μL virus (4.5 × 1012 vg/mL) was injected bilaterally into the dHC at a rate of 0.2 μL/min. The injection needle was left in place for 3 min following injection to allow diffusion of the solution in the brain tissue and then the wound was stapled closed. Meloxicam was used as an analgesic treatment after surgeries. After recovery from the surgery, mice were returned to their home cage for 14 days prior to further experiments. The viruses used were AAV.PHP.N-hSyn1-Cre-T2A-eGFP-WPRE and AAV.PHP.N-hSyn1-eGFP-WPRE. At the end of the behavioral experiments, mice were anesthetized with an intraperitoneal (i.p.) injection of chloral hydrate (750 mg/kg) and transcardially perfused with 4% paraformaldehyde in 1x PBS (pH 7.4). Mouse brains were post-fixed in 4% PFA/PBS for 24 h at 4 °C and transferred sequentially into PBS containing 15% sucrose then 30% sucrose for 24 h each. Coronal sections (20 μm) were collected by cryosection for immunofluorescence analysis.
Immunofluorescence
Twenty micron coronal sections from the dHC (−1.8 mm from bregma) free-floating in 1× PBS were incubated with a blocking solution [0.25% Triton X-100, 5% normal donkey serum (ab7475, Abcam, Waltham, MA), and 1% bovine serum albumin (BSA, 10735078001, Millipore-Sigma, Burlington, MA) in 1× PBS, 7.4 pH] on a rocking platform for 2 h at room temperature. Rabbit anti-CIM6P/IGF2R antibody (1/16,000, ab124767, Abcam) was diluted in the incubation solution (0.1% Triton X-100, 5% normal donkey serum, 1% BSA in 1× PBS) and applied to the brain sections for 24 h, on a rocking platform at 4 °C. Sections were then washed three times in washing buffer (0.1% Triton X-100 in PBS) and incubated with incubation solution containing the secondary antibody goat anti-rabbit Alexa Fluor-594 (1/1000, A-11012, ThermoFisher) and DAPI (300 nM, D3571, Thermo Fisher) on a rocking platform for 2 h at room temperature. Following three washes with washing buffer, sections from the dHC were mounted on slides using ProLong Diamond Antifade Mountant (P36961, Thermo Fisher). Images were collected by an Olympus VS120 virtual slide microscope (Olympus, Tokyo, Japan) at 10× magnification under non-saturating conditions. Quantification was performed using the Fiji software [61] by experimenters who were blind to the mouse genotype, AAV, and treatment conditions. For each image, the mean fluorescence of the background in each channel was subtracted prior to quantification. Regions of interest (ROIs) were manually drawn around the hippocampal formation and used as masks to overlay and quantify the integrated density of fluorescence from eGFP and CIM6P/IGF2R channels.
Statistical analyses
Statistical analyses were performed using GraphPad Prism software 10 (GraphPad Software Inc.). All experiments were carried out at least on 2–3 independent cohorts of mice to test for data replicability. Sample sizes were calculated using the online software G*Power (v3.1.9), and effect sizes were estimated based on our previous work [12, 18, 19]. Data acquisition and analyses were performed blind to experimental group conditions. One-way ANOVA followed by Tukey’s post hoc tests were used when one factor was compared. Two-way ANOVA followed by Tukey’s post hoc tests were used when two factors were compared. P < 0.05 was considered significant. See Table S1 for a detailed description of the statistical analyses used in each experiment.
Results
Design and synthesis of PMP1
We designed a prodrug-modified phosphonate analog of M6P, termed PMP1, with the aim of producing a ligand of the CIM6P/IGF2R with higher oral bioavailability compared to M6P (Fig. 1A, B). The synthesis started from commercially available methyl α-d-mannopyranoside (1). Tritylation followed by exhaustive benzyl protection gave 3 [48,49,50] (Fig. 1C). Selective deprotection then afforded primary alcohol 4 [51], which was subjected to a Parikh–Doering oxidation to yield aldehyde 5 [52]. Aldehyde 5 was used directly in a subsequent Horner–Wadsworth–Emmons olefination to afford phosphonate 7 (Fig. 1C). Aldehyde 5 is a versatile intermediate that can be elaborated with a range of phosphonates, carboxylates, or sulfonates compatible with HWE reaction conditions. Previous studies found that a linker length of two carbons between the phosphorus atom and the mannose scaffold maximizes CIM6P/IGF2R binding affinity [35].
A Structural model of CIM6P/IGF2R with the three N-terminal domains enlarged (in the gray circle); from PDB 6UM2 [91] and 1SYO [92]. M6P bound to domain 3 is shown as a space-filling model. Bound IGF2 is shown in orange. B Design of a non-hydrolyzable phosphonate that mimics M6P and its prodrug PMP1 (C) Synthesis of PMP1. TEMDP tetraethyl methylenediphosphonate.
Phosphonate 7 was dealkylated using a procedure first report by McKenna et al. [53], in which phosphonic acid dialkyl esters are converted to bis(trimethylsilyl) esters under mild conditions (Fig. 1C). The trimethylsilyl ester of 7 was hydrolyzed to the corresponding phosphonic acid 8 [53]. Phosphonic acid 8 was then alkylated with chloromethyl isopropyl carbonate (9), affording the corresponding phosphonate 10. We chose isopropyl methyl carbonate as a prodrug protecting group based on its high stability, solubility, and improved oral bioavailability compared to other prodrug alternatives, as extensively studied by Arimilli et al. [45]. Finally, the benzyl-protected phosphonate 10 was deprotected and hydrogenated to afford PMP1 (11).
PMP1 enhances novel object recognition memory in WT mice and reverses memory and motor deficits in AS mice without causing adverse effects
Previous studies showed that a bilateral hippocampal injection or a systemic (s.c.) injection of IGF2 or M6P enhances memory retention and persistence in both rats and mice [6, 9, 12, 19]. Furthermore, a s.c. injection of either 30 μg/kg of IGF2 or 850 μg/kg of M6P 20 min before behavioral exposure rescued several behavioral deficits in multiple mouse models of neurodevelopmental disorders, including autism spectrum disorder and Angelman syndrome [9, 18]. Here we tested whether a single s.c. injection of PMP1 affects novel object recognition (nOR) memory in WT mice and also examined the impact of PMP1 on cognitive and motor deficits in AS mice. Toward this end, we performed dose-response curves to determine the maximum effective dose of PMP1 in WT and AS mice using nOR (Fig. 2A, B). We also tested the dose-response curve on the motor task hindlimb clasping in AS mice (Fig. 2C). We selected the doses of 283, 850 and 2550 µg/kg based on previous results [12] indicating that a s.c. injection of the M6P at 850 μg/kg in WT mice produced the most prolonged effect on memory enhancement. NOR is a non-aversive, hippocampus-dependent learning paradigm that is based on rodents’ innate preference for novelty [62], and it is a valuable task for testing context and recognition memory. WT mice were first habituated to an arena, and 24 h later, they received a single s.c. injection of a different dose of PMP1, i.e., 283, 850, or 2550 μg/kg, or vehicle. Twenty minutes later they underwent nOR training. Memory retention, measured as the ability to retain the memory of a previously encountered object, was assessed through longitudinal testing at 4, 24 h, 1, 2, 3 and 4 weeks after training and expressed as exploration preference for the novel object (Fig. 2A, Table S1). In WT mice, at 4 h after training, only the 850 μg/kg dose significantly increased memory retention compared to vehicle (Fig. 2A), whereas the other two doses did not show any significant effect (Fig. 2A, Table S1). In agreement with previous data [9, 18, 63], at 24 h after training, vehicle-injected WT mice lacked nOR memory. WT mice injected with any of the three doses of PMP1 significantly enhanced nOR memory retention (Fig. 2A, Table S1). The memory of WT mice injected with 850 μg/kg remained significantly elevated at 1, 2, and 3 weeks after training, and returned to control levels (i.e., no memory) at 4 weeks after training (Fig. 2A, Table S1). The lowest dose (283 μg/kg) significantly increased memory retention, although to a lesser degree compared to 850 μg/kg, at 24 h and 1 week, but not at 2 weeks after training, revealing a relatively shorter and smaller effect on memory retention compared to that evoked by 850 μg/kg (Fig. 2A, Table S1). The highest dose (2550 μg/kg) significantly increased memory retention at 24 h, 1 week, and 2 weeks after training but had no effect at 3 weeks after training (Fig. 2A, Table S1); hence, this dose also had a shorter effect compared to 850 μg/kg. The effect on nOR was not due to changes in motivation to explore the objects, as both vehicle and PMP1-injected WT mice spent similar amounts of time exploring the objects during training (Fig. 2A, Table S1). We concluded that a single s.c. injection of PMP1 enhances memory in WT mice, and its effect is long-lasting (up to 3 weeks), with a maximum effective dose of 850 μg/kg.
Dose-response curves of a single s.c. injection of PMP1 on (A) nOR in WT mice, B nOR on AS mice, and (C) hindlimb clasping in AS mice. Experimental timeline is shown above graphs. A, B: after habituation (hab) to the arena, mice received a s.c. injection of a dose of PMP1 (283 μg/kg, 850 μg/kg, 2550 μg/kg) or vehicle (veh) 20 min before nOR training and were longitudinally tested for memory retention 4, 24 h, 1, 2, 3, and 4 weeks after training. From left to right, graphs show: total time spent exploring both objects during training expressed in seconds (s), percent of time spent exploring the new object (% preference) during the test session at 4h, 24 h, 1, 2, 3, and 4 weeks after training. N = 6–9 per group for WT mice and n = 8–9 per group for AS mice; each, 2 independent experiments). Dots represent the percent of preference of each mouse (blue dots are for males, and pink dots are for females). Data are expressed as mean ± s.e.m; one-way ANOVA followed by Tukey’s post-hoc test was used. *p < 0.05, **p < 0.01, ***p < 0.001. C Mice received a s.c. injection of a dose of PMP1 (283, 850, 2550 μg/kg) or veh 20 min before testing their hindlimb clasping behavior. The hindlimb clasping behavior was tested longitudinally at 20 min, 4h, 24 h, 1, 2, 3, and 4 weeks after the injection. N = 8–9 per group; 2 independent experiments. Dots represent the clasping score for each mouse (blue dots are for males, and pink dots are for females). Data are expressed as mean ± s.e.m. relative to clasping score and one-way ANOVA followed by Tukey’s post-hoc test; *p < 0.05, ***p < 0.001. The red bars show significant inter-group comparisons between non-vehicle groups.
We then investigated a dose-response curve of PMP1 on AS mice to determine the maximum effective dose that could help with their core deficits. AS mice were injected s.c. with either 283, 850, 2550 μg/kg, or vehicle 20 min before nOR training, and their memory retention was measured 4, 24 h, 1, 2, 3, and 4 weeks after training (Fig. 2B). Vehicle-injected WT mice were used as the reference group (Fig. 2B). In agreement with previous findings [12], vehicle-injected AS mice had no retention of the previously encountered object 4 h after training, demonstrating memory impairment compared to the vehicle-injected WT control mice (Fig. 2B, Table S1). Conversely, AS mice that received any of the three doses of PMP1 showed significant nOR memory retention at 4 h after training, reaching a level comparable to that of WT mice (Fig. 2B, Table S1). At 24 h after training, 283 μg/kg had no effect on AS mice’s memory retention, which, in fact, returned to chance level, similar to that of vehicle-injected AS mice (Fig. 2B, Table S1). Injections of 850 or 2550 μg/kg significantly enhanced memory retention in AS mice, and the enhancement lasted up to 2 weeks after training (Fig. 2B, Table S1). At 3 weeks after training, while the 850 μg/kg dose continued to significantly enhance memory in AS mice, the 2550 μg/kg dose returned to chance levels (i.e., no memory, Fig. 2B, Table S1). At 4 weeks after training, the effect of all three PMP1 doses on AS memory ceased, as comparative exploration time between the two objects returned to chance level (Fig. 2B, Table S1). AS memory impairment and the pharmacological effect of PMP1 were not related to changes in the time spent exploring the objects (total exploration time) as both vehicle and PMP1-injected AS mice spent similar amounts of time exploring the objects during training (Fig. 2B, Table S1).
We then tested the dose-response effect of PMP1 on motor deficits in AS mice. Toward that end, we assessed the effect of PMP1 on hindlimb clasping, shown previously to be abnormally high in AS mice [12, 64]. Hindlimb clasping is a behavioral procedure used to evaluate motor deficits in rodent models, particularly in disease models that exhibit ataxia [64,65,66]. AS mice received a single s.c. injection of one of three doses of PMP1 (283, 850, 2550 μg/kg) or vehicle, and their limb clasping was measured through longitudinal repeated testing at 20 min, 4, 24 h, 1, 2, 3 and 4 weeks after the injection (Fig. 2C, Table S1). Vehicle-injected WT mice were used as a control reference, providing the clasping score of normal mice (Fig. 2C). In agreement with previous studies [12, 64], 20 min after s.c. injection, vehicle-injected AS mice showed a significant increase in the clasping reflex with a prevalence of 3–4 limbs clasping together, compared to vehicle-injected WT mice. In vehicle-injected groups, AS mice’s clasping behavior remained significantly higher than that of WT mice at all testing time points (Fig. 2C, Table S1). In contrast, AS mice injected with any dose of PMP1 showed a significant reduction of clasping at 20 min after the injection. The effect lasted up to 1 week after the injection (Fig. 2C, Table S1). Two weeks after the injection, the 283 and the 850 μg/kg dose of PMP1 still significantly reduced clasping compared to vehicle injection, whereas 2550 μg/kg no longer had any effect on the AS clasping, which returned to the level of vehicle-injected AS mice (Fig. 2C, Table S1). At three and four weeks after the injections, none of the PMP1 doses had any effect on limb clasping relative to vehicle, as all clasping scores returned to the disease levels (Fig. 2C, Table S1).
These data, along with those on nOR, led to the conclusion that 850 μg/kg is the optimal dose for a s.c. injection to enhance memory in WT mice and reverse memory impairment and motor deficits in AS mice.
Finally, at the optimal dose of 850 μg/kg, we determined whether the PMP1 s.c treatment produced adverse effects. Toward this end, we injected PMP1 in WT and AS mice and exposed them to a standard observational battery, which evaluated physical characteristics, general behavior, and sensorimotor reflexes in mice [67]. As shown in Table 1, no differences were observed before and after PMP1 treatment in either WT or AS mice. We concluded that a s.c. injection of PMP1, at the optimal dose, does not cause major acute adverse effects.
In summary, our results indicate that PMP1 is a significant memory enhancer in healthy mice and has therapeutic effects in an AS mouse model. The maximum effective dose of PMP1 is 850 μg/kg for both memory enhancement in WT mice and recovery from memory and motor deficits in AS mice, without causing adverse effects.
PMP1 shows greater efficacy compared to IGF2 and M6P
We compared the efficacy of the maximum effective dose of PMP1 with the maximum effective doses of IGF2 (30 μg/kg) and M6P (850 μg/kg), both of which had been identified and shown to significantly enhance memory via s.c. injection in both WT and AS mice [9, 12, 18]. WT and AS mice received a single s.c. injection of either IGF2 (30 μg/kg), M6P (850 μg/kg), PMP1 (850 μg/kg), or vehicle 20 min before nOR training. The mice underwent longitudinal repeated testing at 4, 24 h, 1, 2, 3, and 4 weeks after training. WT and AS mice with all treatments spent a similar amount of time exploring the objects during the training session without showing any preference for either object (Fig. 3A, Table S1). In WT mice, at 4 h after training, PMP1, IGF2, and M6P increased nOR memory retention compared to vehicle, although only PMP1 showed a statistically significant effect (Fig. 3A, Table S1). At 24 h and 1 week after training, IGF2, M6P, and PMP1 significantly enhanced memory retention to a similar extent. The effect of PMP1 as a memory enhancer in WT mice persisted for 3 weeks and ceased at 4 weeks after training (Fig. 3A, Table S1). The memory enhancement evoked by IGF2 lasted for 1 week but ceased by 2 weeks after training, while M6P-evoked memory enhancement lasted for 2 weeks and ceased by 3 weeks after training (Fig. 3A, Table S1).
Comparison of the efficacy of maximum effective s.c. dose of IGF2, M6P, PMP1 relative to vehicle (veh) on memory performance (A, B) and motor deficits (C) in WT and AS mice. Experimental timeline is shown above the graphs. A nOR. Mice were habituated (hab) to the arena the day before training. One day after hab, the mice received a s.c. injection of the maximum effective dose of IGF2 (30 μg/kg), M6P (850 μg/kg), PMP1 (850 μg/kg) or veh 20 min before nOR training and were longitudinally tested for memory retention 4h, 24 h, 1, 2, 3, and 4 weeks after training. From left to right, graphs show total time spent exploring both objects during training expressed in seconds (s), percent of time spent exploring the new object (% preference) during the test session at 4h, 24 h, 1, 2, 3, and 4 weeks after training. N = 7–10 per group, 2 independent experiments. Dots represent the percent of preference of each mouse. B CFC. Mice received a s.c. injection of the effective dose of IGF2 (30 μg/kg), PMP1 (850 μg/kg), or veh 20 min before CFC training and were longitudinally tested for memory retention at 4h, 24 h, 1, 2, 3, 4, and 5 weeks after training. Memory retention is expressed as percent of time spent freezing (% freezing). N = 7–12 per group, 3 independent experiments. Dots represent the % of freezing value for each mouse. C Hindlimb clasping. Mice received a s.c. injection of the maximum effective dose of IGF2 (30 μg/kg), M6P (850 μg/kg), PMP1 (850 μg/kg) or veh 20 min before testing limb clasping behavior, which was longitudinally assessed at 20 min, 4h, 24 h, 1, 2, 3, and 4 weeks after the injection. N = 6–10 per group; 2 independent experiments. Dots represent clasping score for each mouse. For all graphs (A–C), blue dots are for males, and pink dots are for females. Data are expressed as mean ±s.e.m.; two-way ANOVA followed by Tukey’s post-hoc test was used. *p < 0.05, **p < 0.01, ***p < 0.001. The red bars show significant inter-group comparisons between non-vehicle groups.
In AS mice, IGF2, M6P, and PMP1 reversed the deficits of nOR memory at 4 h after training (Fig. 3A, Table S1). At 24 h and 1 week after training IGF2, M6P, and PMP1 significantly increased memory retention compared to AS mice that received vehicle injection (Fig. 3A, Table S1). At 2 weeks after training, IGF2 no longer had any effect, and memory level returned to chance, while the effects of M6P and PMP1 persisted (Fig. 3A, Table S1). PMP1, but not M6P, continued to significantly enhance memory at 3 weeks after training; its effect ceased at 4 weeks after training (Fig. 3A, Table S1). We concluded that PMP1 is the most effective compound for enhancing memory in healthy animals and represents the best potential therapeutic compound for alleviating symptoms of Angelman syndrome.
Next, we extended the investigation of the PMP1 effects in WT and AS mice to other hippocampus-dependent types of memories. We assessed the effects of PMP1 on episodic, aversive contextual memory using contextual fear conditioning (CFC) [68]. Previous studies showed that a s.c. injection of IGF2 or M6P enhances retention and persistence of CFC memory up to 1 week after training in WT mice [9, 12, 19] and rescues CFC memory impairment in AS mice [12]. Here we compared the efficacy of the optimal doses of PMP1 (850 μg/kg) and IGF2 (30 μg/kg) on WT and AS mice in CFC. S.c. injections of either IGF2, PMP1, or vehicle were given 20 min before CFC training, and memory retention was assessed longitudinally at 24 h, 1, 2, 3, 4, and 5 weeks after training (Fig. 3B, Table S1). As depicted in Fig. 3B, at 24 h after training, PMP1 and IGF2 significantly enhanced CFC memory in WT mice, with similar efficacy (Fig. 3B, Table S1). Both treatments produced significant memory enhancement at 1 week after training; however, PMP1 maintained a significant enhancing effect for up to 4 weeks post-training, while IGF2 lost its effect by 2 weeks after training (Fig. 3B, Table S1). In AS mice, as expected, the vehicle-injected group showed impaired CFC compared to WT, which was detected at all testing time points (Fig. 3B, Table S1). The injection of IGF2 or PMP1 restored the memory function at 24 h and 1 week after training, as shown by the memory retention of AS mice, which reached the level of vehicle-injected WT mice (Fig. 3B, Table S1). At 2 weeks after training, both treatments still significantly enhanced memory, but PMP1 produced the strongest effect, which was significantly higher relative to that of IGF2 (Fig. 3B, Table S1). At 3 weeks after training, the effect of IGF2 on CFC memory had ceased, as memory retention of IGF2-injected AS mice was similar to that of vehicle-injected AS mice (Fig. 3B, Table S1). In contrast, PMP1 still produced a significant memory-enhancing effect, which lasted up to 4 weeks after training (Fig. 3B, Table S1). Finally, neither IGF2 nor PMP1 had any effect, compared to vehicle, at 5 weeks after training in both WT and AS mice (Fig. 3B, Table S1).
We finally compared the efficacy of a s.c. optimal dose injection of IGF2 (30 μg/kg), M6P (850 μg/kg), or PMP1 (850 μg/kg) in reversing motor deficits in AS mice measured by hindlimb clasping (Fig. 3C). We longitudinally tested the effects of treatments at 20 min, 4, 24 h, and 1, 2, 3, 4 weeks after treatment. In agreement with previous studies [12, 69], vehicle-injected AS mice had a significantly higher limb clasping score compared to vehicle-injected WT mice throughout all time points tested (Fig. 3C, Table S1). At 20 min, 4, and 24 h after treatment, IGF2, M6P, and PMP1 significantly reduced the clasping score, decreasing it to levels close to those of WT mice (Fig. 3C, Table S1). At 1 week after treatment, all three compounds significantly lowered the clasping score; however, IGF2 showed a smaller effect relative to M6P and PMP1 (Fig. 3C, Table S1). At 2 weeks after treatment, IGF2 no longer had any effect on reducing limb clasping of AS mice as shown by the return of the clasping score to disease level, whereas M6P and PMP1 still significantly decreased the clasping score of AS mice (Fig. 3C, Table S1). At 3 weeks after treatment, all three compounds lost their efficacy, and the clasping behavior returned to the level of vehicle-injected AS mice (Fig. 3C, Table S1). No effect of treatment was observed in WT-injected groups at all the time points tested (Fig. 3C, Table S1).
These results showed that, compared to IGF2 and M6P, PMP1 has the strongest efficacy as a memory enhancer in WT mice on both aversive and non-aversive hippocampus-dependent memory tasks. They also demonstrated that PMP1 is more efficacious compared to IGF2 and M6P in reversing memory impairments and motor deficits in AS mice.
The effect of PMP1 systemic treatment is rapid, targets memory consolidation as well as working memory, and requires CIM6P/IGF2R in the brain
Given that the effects of treatments were assessed through repeated testing, and although the multiple tests conducted over longitudinal studies were spaced in time and did not involve additional compound administration, here we investigated whether multiple testing could affect the treatment outcomes. Toward this end, we chose to investigate CFC, as this memory could potentially undergo extinction with repeated testing. We injected WT and AS mice with PMP1 or vehicle 20 min before CFC training and tested memory retention at 3 weeks after training (Fig. 4A). We found that contextual fear memory in both WT and AS mice treated with PMP1 was similar and significantly enhanced relative to corresponding WT or AS mice treated with vehicle (Fig. 4A, Table S1). The memory enhancements detected in PMP1-treated WT and AS groups tested only once at 3 weeks after training were comparable to those found in the same groups of mice at 3 weeks after training that underwent multiple testing (Fig. 3B). We concluded that repeated weekly testing was not a confound of memory performance and treatment.
Efficacy of a s.c. dose of PMP1 relative to vehicle (veh) in WT and AS mice. (A–E) on long-term memory (A), short-term memory (B), memory consolidation (C), locomotor/anxiety response (D), spatial working memory (E), and the role of CIM6P/IGF2R in memory formation and enhancement evoked by PMP1 in WT mice (F). Experimental timeline is shown above the graphs. A CFC. Mice received a s.c. injection of the optimal effective dose (850 μg/kg) of PMP1 or veh 20 min before CFC training and were tested 3 weeks after training. Memory retention is expressed as percent of time spent freezing (% freezing). N = 7–10 per group, 2 independent experiments. Dots represent the % of freezing value for each mouse. B CFC. Mice received a s.c. injection of the optimal effective dose (850 μg/kg) of PMP1 or veh 20 min before CFC training and were tested for memory retention 1 h after training. Memory retention is expressed as percent of time spent freezing (% freezing). N = 8–9 per group, 2 independent experiments. Dots represent the % of freezing value for each mouse. C CFC. Mice received a s.c. injection of the optimal effective dose (850 μg/kg) of PMP1 or veh immediately after CFC training and were tested 24 h after training. Memory retention is expressed as percent of time spent freezing (% freezing). N = 10–13 per group, 3 independent experiments. Dots represent the % of freezing value for each mouse. D OFT. Mice received a s.c. injection of the maximum effective dose of PMP1 (850 μg/kg) or veh 20 min before testing open field behaviors. From left to right, distance travelled (in cm), time spent in the center of the open field (in seconds), and total entries into the center of the arena expressed relative to WT mice injected with vehicle. N = 7–8 per group, 2 independent experiments. Dots represent the associated behavior score for each mouse. E Y-maze. Mice received a s.c. injection of the maximum effective dose of PMP1 (850 μg/kg) or veh 20 min before testing in the Y-maze. Working memory performance is expressed as the percentage of correct alternations in WT and AS mice. N = 8–9 per group, 2 independent experiments. Dots represent the % of correct alternations relative to total alternations for each mouse. F CFC and CIM6P/IGF2R knockout in the dorsal hippocampus. Mice expressing floxed CIM6P/IGF2R (Igf2rfl/fl) received hippocampal injections of AAV.PHP-hSyn1-Cre-T2A-GFP to knock-out neuronal CIM6P/IGF2R, or injections of the control virus AAV.PHP-hSyn1-GFP. Three weeks later, the mice received a s.c. injection of the effective dose of PMP1 (850 μg/kg) or veh 20 min before CFC training and were tested 24 h after training. The mice were then evaluated for dorsal hippocampal expression of CIM6P/IGF2R. From left, the bar graphs report: memory retention is expressed at training and 24 h testing as percent of time spent freezing (% freezing). N = 11–14 per group, 3 independent experiments. Dots represent the % of freezing value for each mouse. CIM6P/IGF2R relative quantifications in the dorsal hippocampus based on immunohistochemistry (IHC) and representative images of IHC from AAV-Cre-injected mice relative to CIM6P/IGF2R from AAV-GFP-injected mice for each corresponding treatment (AAV-GFP + PMP1 or veh = 100%). Dots represent dorsal hippocampal CIM6P/IGF2R expression for each mouse. N = 8–11 per group, 3 independent experiments. Tissue was immunostained with anti-CIM6P/IGF2R and nuclei were stained with DAPI. For all graphs (A-F), blue dots are for males and pink dots are for females. Data are expressed as mean ±s.e.m.; two-way ANOVA followed by Tukey’s post-hoc test was used. *p < 0.05, **p < 0.01, ***p < 0.001. The red bars show significant inter-group comparisons between non-vehicle groups.
Next, to determine whether the treatment effect acts on encoding, short-term memory, and/or memory consolidation, we carried out the following experiments using CFC. First, we measured the effects of a PMP1 s.c. injection given 20 min before training on memory retention measured at 1 h after training (short-term memory, Fig. 4B). We found that PMP1 treatment significantly enhanced WT mice’s memory at 1 h after training and reversed the impairment of AS mice (Fig. 4B, Table S1), indicating that PMP1 effects are rapid and detected at short times following learning.
Second, we determined whether PMP1 treatment, similar to IGF2 [6, 9], acts on memory consolidation – the process that stabilizes long-term memory after encoding [6]. WT and AS mice were trained in CFC and were injected s.c. with PMP1 or vehicle immediately after training and then tested at 24 h after training (Fig. 4C). At testing, AS mice showed the typical significant memory impairment compared to WT controls and PMP1 treatment significantly reversed this impairment. WT mice showed a small but significant memory enhancement with PMP1 treatment (Fig. 4C, Table S1).
Collectively, these results revealed that PMP1 enhances short-term CFC memory in WT mice, reverses memory impairment in AS mice, and acts on memory consolidation.
In order to exclude that changes in memory performances were due to other behavioral responses such as locomotion and/or anxiety, we carried out an open-field test (Fig. 4D). WT and AS mice were injected s.c. with PMP1 or vehicle and tested 20 min later in the open field. As depicted in Fig. 4D and consistent with previous reports [12, 70], we found that the total distance traveled by AS mice was significantly less than that of WT mice. No difference in time spent in the center of the arena between AS and WT mice was found (Fig. 4D, Table S1). Furthermore, there was no difference with treatment: the AS mice treated with PMP1 or vehicle showed similar distances traveled, time spent in the center, or total entries to the center (Fig. 4D, Table S1). These data excluded that differences in motor or memory testing observed in AS mice treated with PMP1 were due to changes in locomotion or anxiety-like responses.
In order to expand the characterization of the effects of PMP1 treatment on other types of memories, we measured spatial working and reference memory using spontaneous alternation in Y-maze [71]. WT and AS mice received a s.c. injection of PMP1 or vehicle and 20 min later were tested in Y-maze (Fig. 4E). This three-arm maze exploits the natural preference of mice for novel environments and thus their tendency to explore the least recently visited arm [9, 72]. Execution of this task requires the engagement of multiple brain regions, including the hippocampus, medial septum, prefrontal cortex, and basal forebrain [73]. In agreement with previous studies [12], we found that AS mice that received a vehicle injection had a significant impairment in performing correct alternations relative to vehicle-injected WT mice. PMP1-treated AS mice significantly increased the rate of correct alternations relative to the AS-vehicle group, reaching a level similar to those of WT injected with vehicle or PMP1 (Fig. 4E, Table S1). The proportion of correct alternations of WT mice treated with vehicle or PMP1 was similar (Fig. 4E, Table S1). The outcome of these experiments, along with those of the open field test, is comparable to the data obtained with IGF2 or M6P treatment [12], suggesting that PMP1 functions similarly.
We then tested whether the systemic administration of PMP1 engages the CIM6P/IGF2R in the brain. Toward this end, we silenced neuronal CIM6P/IGF2R expression in the mouse hippocampus and determined the effect of PMP1 on CFC performance (Fig. 4F). Adult Igf2rfl/fl mice, which allow for tissue-specific knock-out of Igf2r using the Cre/loxP system [57], were injected bilaterally in the dHC, with an AAV-Syn-Cre-T2A-eGFP or AAV-Syn-eGFP, which expressed either Cre-recombinase fused with eGFP or eGFP only (control), respectively, under the transcriptional control of the neuronal promoter of synapsin-1 (for selective neuronal targeting). Three weeks after viral injection, the mice were injected s.c. with PMP1 or vehicle and trained in CFC 20 min later. The mice were tested 24 h after training. As shown in Fig. 4F, PMP1 treatment in mice expressing eGFP (control) significantly enhanced memory relative to vehicle (Fig. 4F, Table S1). In contrast, vehicle-injected Cre-expressing mice had a significantly impaired memory, in agreement with previous data showing that CIM6P/IGF2R is required for long-term memory [19]. PMP1 treatment failed to change this memory impairment (Fig. 4F, Table S1), suggesting that the effect of PMP1 in the hippocampus requires neuronal CIM6P/IGF2R. Immunohistochemical quantification of CIM6P/IGF2R immunofluorescence in the dHC of the behaviorally tested mice, expressing either Cre recombinase-eGFP or eGFP, validated the CIM6P/IGF2R knockout. In fact, the dHC of Cre recombinase-expressing mice decreased the expression of CIM6P/IGF2R by 84% in the vehicle groups and 92% in the PMP1 groups relative to the level of CIM6P/IGF2R of the dHC of mice expressing only eGFP (Fig. 4F).
Collectively, these results revealed that PMP1: (i) acts on short- and long-term memory consolidation, (ii) acts on associative and spatial working memories, (iii) does not affect locomotion or anxiety responses, and (iv) requires hippocampal CIM6P/IGF2R to enhance memory in WT mice.
Oral administration of PMP1, but not of M6P, significantly enhances memory
PMP1 was designed as a prodrug that could be effective through oral administration. To test whether PMP1 influences memory when given orally, we conducted a dose-response curve of PMP1 administered via oral gavage in WT mice. The mice received a single oral dose of either the vehicle or PMP1 at 283, 850, or 2550 µg/kg, 20 min before nOR training. Their memory retention was evaluated longitudinally at 4, 24 h, 1, 2, 3, 4, and 5 weeks after training. 850 μg/kg, but not the other doses, significantly increased memory retention at 4 h after training compared to the WT vehicle-injected group (Fig. 5A, Table S1). At 24 h, all three doses, 283, 850, and 2550 μg/kg, significantly enhanced memory. However, 850 μg/kg had the strongest effect, which was statistically significant relative to the lowest dose (283 μg/kg, Fig. 5A, Table S1). At 1 and 2 weeks after training, the lowest dose, 283 μg/kg, lost its efficacy relative to vehicle, while 850 and 2550 μg/kg still significantly enhanced memory with similar efficacy (Fig. 5A, Table S1). At 3 weeks after training, both 850 and 2550 μg/kg of PMP1 significantly increased memory retention, but 850 μg/kg had a more robust and significant effect (Fig. 5A, Table S1). At 4 weeks after training, only 850 μg/kg significantly enhanced memory (Fig. 5A, Table S1), and the effect ceased at 5 weeks after training (Fig. 5A, Table S1).
A Dose response curve of a single oral gavage (o.g.) administration of PMP1 in WT mice. Experimental timeline is shown above graphs; mice were habituated (hab) to the arena the day before training. One day after hab, the mice received via o.g. a dose of PMP1 (283, 850, 2550 μg/kg) or vehicle (veh) 20 min before nOR training and were longitudinally tested for memory retention at 4h, 24 h, 1, 2, 3, 4, and 5 weeks after training. From left to right, graphs show total time spent exploring both objects during training expressed in seconds (s), percent of time spent exploring the new object (% preference) during the test sessions. N = 5–9 per group, 2 independent experiments. B Comparison of the efficacy of the maximum effective dose of PMP1 and M6P orally administered. Experimental timeline is shown above graphs: mice were habituated (hab) to the arena the day before training. One day after hab, the mice received via o.g. either M6P (850 μg/kg), PMP1 (850 μg/kg) or veh 20 min before nOR training and were longitudinally tested for memory retention at 4h, 24 h, and 1 week after training. Graphs show total time spent exploring both objects during training expressed in seconds (s), percent of time spent exploring the new object (% preference) during the test sessions. N = 7–8 per group; 2 independent experiments. For (A) and (B), dots represent the percent of preference of each mouse; blue dots show data in males and pink dots data in females. Data are expressed as mean ± s.e.m; one-way ANOVA followed by Tukey’s post-hoc test was used. *p < 0.05, **p < 0.01, ***p < 0.001. The red bars show significant inter-group comparisons between non-vehicle groups.
Finally, we determined whether M6P had any effect through oral administration and to compare its efficacy to that of PMP1. We used their respective maximum effective dose (850 μg/kg) and administered WT mice via oral gavage 20 min before nOR training and compared them to vehicle. Memory retention was longitudinally tested at 4, 24 h, and 1 week after training (Fig. 5B, Table S1). M6P showed no effect relative to vehicle at any of the testing time points, whereas, as found in the dose-response curve, PMP1 significantly enhanced memory retention compared to vehicle at 4, 24 h, and 1 week after training (Fig. 5B, Table S1). Thus, PMP1, but not M6P, enhances memory retention via oral administration, with a maximum effective dose at 850 μg/kg – the same as that found with s.c. injection.
Oral administration of PMP1 reverses memory and motor impairments in AS mice
We tested whether orally administered PMP1 affected memory and motor impairments in AS mice. WT and AS mice received an oral administration of PMP1 at 850 μg/kg 20 min before nOR training, and memory retention was tested longitudinally at 4, 24 h, 1, 2, 3, 4, and 5 weeks after training (Fig. 6A, Table S1). At training, WT and AS mice had no preference for either object (Fig. 6A, Table S1). As expected, at 4 h after training, vehicle-treated AS mice showed memory impairment relative to the vehicle-treated WT group (Fig. 6A, Table S1). At 4 h after training, PMP1 significantly increased memory retention in WT mice and rescued memory impairment in AS mice (Fig. 6A, Table S1). PMP1 significantly increased nOR memory retention in both genotypes for up to 4 weeks after training; the effect ceased at 5 weeks after training (Fig. 6A, Table S1). Moreover, orally administered PMP1 at 850 μg/kg significantly reversed hindlimb clasping in AS mice starting 20 min after treatment, and the effect continued for up to 4 weeks after treatment (Fig. 6B, Table S1).
A Effect of a single oral gavage (o.g.) administration of PMP1 in WT and AS mice tested on nOR. Experimental timeline is shown above graphs: mice were habituated (hab) to the arena the day before training. One day after hab, the mice received an o.g. of the maximum effective dose of PMP1 (850 μg/kg) or veh 20 min before nOR training and were longitudinally tested for memory retention at 4h, 24 h, 1, 2, 3, 4, and 5 weeks after training. Graphs show the total time spent exploring both objects during training expressed in seconds (s), and the percent of time spent exploring the new object (% preference) during the test sessions. N = 12–14 per group, 2 independent experiments. Dots represent the percent of preference for each mouse (blue dots are for males, and pink dots are for females). Data are expressed as mean ± s.e.m. relative to % of preference for familiar or novel object and two-way ANOVA followed by Tukey’s post-hoc test was used. ***p < 0.001. B Effect of a single o.g. administration of PMP1 in WT and AS mice tested on limb clasping. Experimental timeline is shown above graphs: mice received an o.g. administration of the maximum effective dose of PMP1 (850 μg/kg) or vehicle (veh) 20 min before testing limb clasping behavior, which was longitudinally tested at 20 min, 4h, 24 h, 1, 2, 3, 4 and 5 weeks after the oral administration. N = 8–14 per group; 2 independent experiments. Dots represent clasping score for each mouse (blue dots show data in males and pink dots in females). Data are expressed as mean ± s.e.m. relative to clasping score and two-way ANOVA followed by Tukey’s post-hoc test was used. ***p < 0.001.
Collectively, these results show that orally administered PMP1 significantly ameliorates both memory impairment and motor deficits in AS mice, and the effect is more persistent than that of a s.c. injection.
Discussion
Several CNS diseases, such as neurodevelopmental disorders, including Angelman syndrome [25], and neurodegenerative diseases, including Alzheimer’s disease, Huntington’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis [28,29,30,31,32], which affect millions of people worldwide, are awaiting effective therapies. Notably, these diseases present a common problem: the disruption of protein metabolism homeostasis with the accumulation of certain proteins in the nervous system, leading to impaired neural functions and, in some cases, neuronal degeneration [74]. Therefore, a potentially transformative approach for developing new therapies to treat these diseases could be the identification and targeting of novel mechanisms that effectively rebalance neuronal protein metabolism, hence correcting the overaccumulation of proteins in the nervous system.
Here we reported the design of PMP1, a novel small-molecule prodrug conceived to target the CIM6P/IGF2R, a receptor that is highly expressed in neurons [19], where it regulates protein synthesis, trafficking, and degradation [14, 19, 24]. We showed that both s.c. and oral administration of PMP1 are highly effective in promoting memory enhancement and persistence in healthy mice and reversing cognitive impairments and motor deficits in a mouse model of Angelman syndrome. This model has been shown to be associated with protein accumulation in the hippocampus [25].
Compared to M6P and IGF2, the administration of PMP1 produced the most potent and prolonged effects, both in memory enhancement in healthy animals and in reversing deficits in AS mice, suggesting that this small synthetic molecule is more effective than natural ligands of the CIM6P/IGF2R. Additionally, a key and unique advantage of PMP1 relative to the natural ligands is that it has the highest efficacy when taken orally.
We found that the effects of PMP1 are rapid and detected starting at 20 min after treatment. Furthermore, similar to IGF2 [9], the effect of PMP1 treatment acted on the memory consolidation process, providing support for the sustained effects of treatments with CIM6PR/IGF2R ligands. The effect of a single treatment was, in fact, found to last for weeks, aligning with the time frame of memory consolidation itself, which is to produce strengthening and persistence of memory over weeks. The effect on memory persistence was not due to multiple longitudinal testing, as repeated testing (at the time intervals we used) did not significantly change the memory’s strength. In addition to long-term memory, PMP1 also reversed short-term memory and working memory deficits in AS mice, indicating that it impacts different types of memory processes. PMP1, like IGF2 and M6P [12], did not change the open field test response, suggesting that its effects on memory were not due to changes in anxiety or motor responses.
Tested on the s.c.-evoked memory enhancement, the PMP1 effect required the CIM6P/IGF2R expression in hippocampal neurons, indicating that systemic treatment engages CIM6P/IGF2R in the CNS. The efficacy of PMP1 via both s.c and oral administration had a similar maximum effective dose of 850 μg/kg, and the treatment did not show any adverse or toxic effects. Hence, PMP1 has potential therapeutic effects for Angelman syndrome and acts as a strong memory booster.
Our data on PMP1 significantly extend the findings that the CIM6P/IGF2R natural ligands IGF2 and M6P, when administered either intracerebrally or s.c. enhance memory in healthy rodents and reverse core symptoms in several neurodevelopmental disorders and neurodegenerative diseases [11,12,13,14, 75, 76]. These unique features of PMP1 make this novel small molecule attractive as a potential therapeutic compound.
The design of our compound was guided by a desire to make it orally active as a prodrug. Phosphonate ester and amide masking groups have been employed in the development of effective HIV reverse transcriptase inhibitors, such as tenofovir [46]. Other phosphate or phosphonate prodrugs, such as alafenamides, could be implemented [77].
Our results, along with previous drug development studies, indicate that targeting CIM6P/IGF2R has great translational potential for several conditions, mostly of which are linked to protein accumulation. Previous investigations that targeted CIM6P/IGF2R focused on the receptor’s role in either delivering lysosomal enzymes to lysosomes or binding bioactive proteins such as TGF-β1, renin, and proliferin, for which the receptor may function to promote their activation and/or degradation [78]. For example, CIM6P/IGF2R has been targeted for developing treatments for Pompe disease [79] and as a promising transporter for drug delivery treating lysosomal storage disease, liver fibrosis, and cancer [42, 80]. Targeting fibrosis was designed by either increasing, through the CIM6P/IGF2R binding, the delivery of anti-fibrotic agents to intracellular compartments or by employing CIM6P/IGF2R inhibitors to antagonize the activation of TGF-β1.
An important advantage of targeting the CIM6P/IGF2R for CNS diseases is that the CIM6P/IGF2R itself, or similar proteins acting as transporters are present in the blood brain barrier (BBB) [81,82,83,84], allowing for efficient brain delivery of the ligands. Affinity cross-linking studies suggested that the BBB expresses an IGF receptor, distinct from the 300 kDa CIM6P/IGF2R found in peripheral tissues and the brain stroma [12, 81], and is instead a 141 kDa transporter for both IGF1 and IGF2 [81]. These results are consistent with the hypothesis that the human BBB IGF receptor is a transport system for the circulating IGFs, including IGF2, and other ligands like M6P or derivatives. More studies are needed to characterize the expression of the CIM6P/IGF2R or similar receptor/transporters in the BBB. The CIM6P/IGF2R is widely expressed throughout the brain, but it appears to be more enriched in the hippocampus, choroid plexus, meninges, and neocortex [85, 86]; however, further studies are needed to better characterize the cellular and subcellular distribution of CIM6P/IGF2R. Given that CIM6P/IGF2R is highly expressed in neurons [19], regulates activity-dependent neuronal protein metabolism [19, 87, 88], and plays a key role in lysosomal targeting for protein degradation [20, 89, 90], it is reasonable to hypothesize that PMP1 counteracts the neuronal dysregulation of protein metabolism by accelerating protein degradation. Future studies should be able to test this hypothesis.
The next important questions that remain to be addressed is the full characterization of the mechanisms of action of PMP1 administration via s.c. or per os, their targets in the brain, and throughout the body, and the cellular mechanisms of action. Future metabolic profiling studies are needed to identify the PMP1-derived active principle as well as determine the pharmacokinetic-pharmacodynamic (PKPD) properties after s.c. and oral administration. These investigations are essential for determining the concentration of PMP1, active principle, and their metabolites in body compartments, especially in the brain. Given that the CIM6P/IGF2R is a membrane receptor that traffics intracellularly via endosomes to target lysosomes, the Golgi network, and protein recycling, we speculate that the mechanisms of action in the recovery of functions of the various diseases will converge on protein degradation and protein quality control [14]. Despite commonly targeting protein metabolism rebalancing, we cannot, however, exclude that the specific intracellular target mechanisms of each disease or in healthy animals may differ [14]. Although no toxic or other adverse effects were detected after a single s.c. administration of PMP1, future studies are needed to test off-target effects and toxicity with the appropriate protocols required for translational applications. Future studies should also investigate the effects of multiple dosing, including toxicity, in a variety of disease models that carry brain protein overaccumulation in order to frame potential translational protocols for treating devastating diseases.
In conclusion, PMP1 may represent a novel small molecule that could be readily developed for the treatment of Angelman syndrome and other neurodevelopmental disorders, as well as several neurodegenerative diseases.
Data availability
Data supporting the findings of this study are available from the corresponding author upon reasonable request.”
References
Dugger BN, Dickson DW. Pathology of neurodegenerative diseases. Cold Spring Harb Perspect Biol. 2017;9:a028035.
Erkkinen MG, Kim MO, Geschwind MD. Clinical neurology and epidemiology of the major neurodegenerative diseases. Cold Spring Harb Perspect Biol. 2018;10:a033118.
Parenti I, Rabaneda LG, Schoen H, Novarino G. Neurodevelopmental disorders: from genetics to functional pathways. Trends Neurosci. 2020;43:608–21.
Bergman D, Halje M, Nordin M, Engström W. Insulin-like growth factor 2 in development and disease: a mini-review. Gerontology. 2013;59:240–9.
O’Dell SD, Day IN. Insulin-like growth factor II (IGF-II). Int J Biochem Cell Biol. 1998;30:767–71.
Chen DY, Stern SA, Garcia-Osta A, Saunier-Rebori B, Pollonini G, Bambah-Mukku D, et al. A critical role for IGF-II in memory consolidation and enhancement. Nature. 2011;469:491–7.
Agis-Balboa RC, Arcos-Diaz D, Wittnam J, Govindarajan N, Blom K, Burkhardt S, et al. A hippocampal insulin-growth factor 2 pathway regulates the extinction of fear memories. EMBO J. 2011;30:4071–83.
Ouchi Y, Banno Y, Shimizu Y, Ando S, Hasegawa H, Adachi K, et al. Reduced adult hippocampal neurogenesis and working memory deficits in the Dgcr8-deficient mouse model of 22q11.2 deletion-associated schizophrenia can be rescued by IGF2. J Neurosci. 2013;33:9408–19.
Stern SA, Kohtz AS, Pollonini G, Alberini CM. Enhancement of memories by systemic administration of insulin-like growth factor II. Neuropsychopharmacology. 2014;39:2179–90.
Shahmoradi A, Radyushkin K, Rossner MJ. Enhanced memory consolidation in mice lacking the circadian modulators Sharp1 and -2 caused by elevated Igf2 signaling in the cortex. Proc Natl Acad Sci USA. 2015;112:E3582–9.
Steinmetz AB, Johnson SA, Iannitelli DE, Pollonini G, Alberini CM. Insulin-like growth factor 2 rescues aging-related memory loss in rats. Neurobiol Aging. 2016;44:9–21.
Cruz E, Descalzi G, Steinmetz A, Scharfman HE, Katzman A, Alberini CM. CIM6P/IGF-2 receptor ligands reverse deficits in Angelman syndrome model mice. Autism Res. 2021;14:29–45.
Pardo M, Cheng Y, Sitbon YH, Lowell JA, Grieco SF, Worthen RJ, et al. Insulin growth factor 2 (IGF2) as an emergent target in psychiatric and neurological disorders. Review Neurosci Res. 2019;149:1–13.
Alberini CM. IGF2 in memory, neurodevelopmental disorders, and neurodegenerative diseases. Trends Neurosci. 2023;46:488–502.
Xia L, Zhu X, Zhao Y, Yang G, Zuo X, Xie P, et al. Genome-wide RNA sequencing analysis reveals that IGF-2 attenuates memory decline, oxidative stress and amyloid plaques in an Alzheimer’s disease mouse model (AD) by activating the PI3K/AKT/CREB signaling pathway. Int Psychogeriatr. 2019;31:947–59.
Pascual-Lucas M, Viana da Silva S, Di Scala M, Garcia-Barroso C, González-Aseguinolaza G, Mulle C, et al. Insulin-like growth factor 2 reverses memory and synaptic deficits in APP transgenic mice. EMBO Mol Med. 2014;6:1246–62.
Mellott TJ, Pender SM, Burke RM, Langley EA, Blusztajn JK. IGF2 ameliorates amyloidosis, increases cholinergic marker expression and raises BMP9 and neurotrophin levels in the hippocampus of the APPswePS1dE9 Alzheimer’s disease model mice. PLoS One. 2014;9:e94287.
Steinmetz AB, Stern SA, Kohtz AS, Descalzi G, Alberini CM. Insulin-like growth factor II targets the mTOR pathway to reverse autism-like phenotypes in mice. J Neurosci. 2018;38:1015–29.
Yu XW, Pandey K, Katzman AC, Alberini CM. A role for CIM6P/IGF2 receptor in memory consolidation and enhancement. Elife. 2020;9:e54781.
Hawkes C, Kar S. The insulin-like growth factor-II/mannose-6-phosphate receptor: structure, distribution and function in the central nervous system. Brain Res Brain Res Rev. 2004;44:117–40.
Wang Y, MacDonald RG, Thinakaran G, Kar S. Insulin-like growth factor-II/Cation-independent mannose 6-Phosphate receptor in neurodegenerative diseases. Mol Neurobiol. 2017;54:2636–58.
El Haj M, Roche J, Gallouj K, Gandolphe MC. Autobiographical memory compromise in Alzheimer’s disease: a cognitive and clinical overview. Geriatr Psychol Neuropsychiatr Vieil. 2017;15:443–51.
Concina G, Renna A, Milano L, Manassero E, Stabile F, Sacchetti B. Expression of IGF-2 receptor in the auditory cortex improves the precision of recent fear memories and maintains detailed remote fear memories over time. Cereb Cortex. 2021;31:5381–95.
Pandey K, Yu XW, Steinmetz A, Alberini CM. Autophagy coupled to translation is required for long-term memory. Autophagy. 2021;17:1614–35.
Aria F, Pandey K, Alberini CM. Excessive protein accumulation and impaired autophagy in the hippocampus of Angelman syndrome modeled in mice. Biol Psychiatry. 2023;94:68–83.
Deng Z, Zhou X, Lu JH, Yue Z. Autophagy deficiency in neurodevelopmental disorders. Cell Biosci. 2021;11:214.
Ham A, Chang AY, Li H, Bain JM, Goldman JE, Sulzer D, et al. Impaired macroautophagy confers substantial risk for intellectual disability in children with autism spectrum disorders. Mol Psychiatry. 2025;30:810–24.
Blokhuis AM, Groen EJ, Koppers M, van den Berg LH, Pasterkamp RJ. Protein aggregation in amyotrophic lateral sclerosis. Acta Neuropathol. 2013;125:777–94.
Nixon RA. The role of autophagy in neurodegenerative disease. Nat Med. 2013;19:983–97.
Valenzuela V, Nassif M, Hetz C. Unraveling the role of motoneuron autophagy in ALS. Autophagy. 2018;14:733–7.
Klionsky DJ, Petroni G, Amaravadi RK, Baehrecke EH, Ballabio A, Boya P, et al. Autophagy in major human diseases. EMBO J. 2021;40:e108863.
Im E, Jiang Y, Stavrides PH, Darji S, Erdjument-Bromage H, Neubert TA, et al. Lysosomal dysfunction in Down syndrome and Alzheimer mouse models is caused by v-ATPase inhibition by Tyr(682)-phosphorylated APP betaCTF. Sci Adv. 2023;9:eadg1925.
Schmidt B, Kiecke-Siemsen C, Waheed A, Braulke T, von Figura K. Localization of the insulin-like growth factor II binding site to amino acids 1508-1566 in repeat 11 of the mannose 6-phosphate/insulin-like growth factor II receptor. J Biol Chem. 1995;270:14975–82.
Brown J, Esnouf RM, Jones MA, Linnell J, Harlos K, Hassan AB, et al. Structure of a functional IGF2R fragment determined from the anomalous scattering of sulfur. EMBO J. 2002;21:1054–62.
Vidil C, Morère A, Garcia M, Barragan V, Hamdaoui B, Rochefort H, et al. Synthesis and biological activity of phosphonate analogs of mannose 6-phosphate (M6P). Eur J Org Chem. 1999;1999:447–50.
Vidal S, Garcia M, Montero JL, Morère A. Synthesis and biological evaluation of two mannose 6-phosphate analogs. Eur J Org Chem. 2000;2000:3433–7.
Jeanjean A, Garcia M, Leydet A, Montero JL, Morère A. Synthesis and receptor binding affinity of carboxylate analogues of the mannose 6-phosphate recognition marker. Bioorg Med Chem. 2006;14:3575–82.
Lim JA, Li L, Raben N. Pompe disease: from pathophysiology to therapy and back again. Front Aging Neurosci. 2014;6:177.
Hankins GR, De Souza AT, Bentley RC, Patel MR, Marks JR, Iglehart JD, et al. M6P/IGF2 receptor: a candidate breast tumor suppressor gene. Oncogene. 1996;12:2003–9.
DaCosta SA, Schumaker LM, Ellis MJ. Mannose 6-phosphate/insulin-like growth factor 2 receptor, a bona fide tumor suppressor gene or just a promising candidate? J Mammary Gland Biol Neoplasia. 2000;5:85–94.
Osipo C, Dorman S, Frankfater A. Loss of insulin-like growth factor II receptor expression promotes growth in cancer by increasing intracellular signaling from both IGF-I and insulin receptors. Exp Cell Res. 2001;264:388–96.
Gary-Bobo M, Nirdé P, Jeanjean A, Morère A, Garcia M. Mannose 6-phosphate receptor targeting and its applications in human diseases. Curr Med Chem. 2007;14:2945–53.
Banik SM, Pedram K, Wisnovsky S, Ahn G, Riley NM, Bertozzi CR. Lysosome-targeting chimaeras for degradation of extracellular proteins. Nature. 2020;584:291–7.
Stevens CM, Zhou Y, Teng P, Rault LN, Liao Y, Tang W. Development of oligomeric Mannose-6-phosphonate conjugates for targeted protein degradation. ACS Med Chem Lett. 2023;14:719–26.
Arimilli MN, Kim CU, Dougherty J, Mulato A, Oliyai R, Shaw JP, et al. Synthesis, in vitro biological evaluation and oral bioavailability of 9-[2-(phosphonomethoxy)propyl]adenine (PMPA) prodrugs. Antivir Chem Chemother. 1997;8:557–64.
De Clercq E, Holy A. Acyclic nucleoside phosphonates: a key class of antiviral drugs. Nat Rev Drug Discov. 2005;4:928–40.
Jiang YH, Armstrong D, Albrecht U, Atkins CM, Noebels JL, Eichele G, et al. Mutation of the Angelman ubiquitin ligase in mice causes increased cytoplasmic p53 and deficits of contextual learning and long-term potentiation. Neuron. 1998;21:799–811.
Tennant-Eyles RJ, Davis BG, Fairbanks AJ. Peptide templated glycosylation reactions. Tetrahedron-Asymmetry. 2000;11:231–43.
Traboni S, Bedini E, Iadonisi A. Solvent-free one-pot diversified protection of saccharide polyols via regioselective tritylations. Chemistryselect. 2017;2:4906–11.
Hofmann A, Sommer R, Hauck D, Stifel J, Göttker-Schnetmann I, Titz A. Synthesis of mannoheptose derivatives and their evaluation as inhibitors of the lectin LecB from the opportunistic pathogen Pseudomonas aeruginosa. Carbohydr Res. 2015;412:34–42.
Jaramillo C, Chiara JL, Martinlomas M. An effective strategy for the synthesis of 6-O-(2-Amino-2-Deoxy-Alpha-D-Glucopyranosyl)-D-Chiro and D-Myo-Inositol 1-Phosphate related to putative insulin mimetics. J Org Chem. 1994;59:3135–41.
Tojo G, Fernandez M. Oxidation of alcohols to aldehydes and ketones: a guide to current common practice. Basic Reactions in Organic Synthesis. Springer: New York, NY; 2006. pp. 1–375.
McKenna CE, Higa MT, Cheung NH, McKenna MC. The facile dealkylation of phosphonic acid dialkyl esters by bromotrimethylsilane. Tetrahedron Lett. 1977;18:155–8.
Graham TH, Yu T, Waddell ST, McCauley JA, Stamford A, Liu W, et al. Spirocyclic Pyridotriazine Derivatives useful as HIV integrase inhibitors. U.S. Patent US10624912B2, 2020.
Jeanjean A, Gary-Bobo M, Nirdé P, Leiris S, Garcia M, Morère A. Synthesis of new sulfonate and phosphonate derivatives for cation-independent mannose 6-phosphate receptor targeting. Bioorg Med Chem Lett. 2008;18:6240–3.
Buchanan GW, Benezra C. Stereochemical dependence of C-13 shieldings and C-13-P-31 couplings in phosphonates of known geometry - is there a karplus-type relationship for P-C-C-C coupling. Can J Chem. 1976;54:231–7.
Wylie AA, Pulford DJ, McVie-Wylie AJ, Waterland RA, Evans HK, Chen YT, et al. Tissue-specific inactivation of murine M6P/IGF2R. Am J Pathol. 2003;162:321–8.
Kawakami Y, Sielski R, & Kawakami T Mouse body temperature measurement using infrared thermometer during passive systemic anaphylaxis and food allergy evaluation. JOVE. https://doi.org/10.3791/5839, 2018.
Mei J, Riedel N, Grittner U, Endres M, Banneke S, Emmrich JV. Body temperature measurement in mice during acute illness: implantable temperature transponder versus surface infrared thermometry. Sci Rep. 2018;8:3526.
Newsom DM, Bolgos GL, Colby L, Nemzek JA. Comparison of body surface temperature measurement and conventional methods for measuring temperature in the mouse. Contemp Top Lab Anim Sci. 2004;43:13–8.
Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, et al. Fiji: an open-source platform for biological-image analysis. Nat Methods. 2012;9:676–82.
Antunes M, Biala G. The novel object recognition memory: neurobiology, test procedure, and its modifications. Cogn Process. 2012;13:93–110.
Dix SL, Aggleton JP. Extending the spontaneous preference test of recognition: evidence of object-location and object-context recognition. Behav Brain Res. 1999;99:191–200.
Ciarlone SL, Wang X, Rogawski MA, Weeber EJ. Effects of the synthetic neurosteroid ganaxolone on seizure activity and behavioral deficits in an Angelman syndrome mouse model. Neuropharmacology. 2017;116:142–50.
Lalonde R, Strazielle C. Brain regions and genes affecting limb-clasping responses. Brain Res Rev. 2011;67:252–9.
Guyenet SJ, Furrer SA, Damian VM, Baughan TD, La Spada AR, & Garden GA A simple composite phenotype scoring system for evaluating mouse models of cerebellar ataxia. J Vis Exp. 2010:1787. https://doi.org/10.3791/1787.
Brown J, Jones EY, Forbes BE. Keeping IGF-II under control: lessons from the IGF-II-IGF2R crystal structure. Trends Biochem Sci. 2009;34:612–9.
Fanselow MS. From contextual fear to a dynamic view of memory systems. Trends Cogn Sci. 2010;14:7–15.
Mandel-Brehm C, Salogiannis J, Dhamne SC, Rotenberg A, Greenberg ME. Seizure-like activity in a juvenile Angelman syndrome mouse model is attenuated by reducing Arc expression. Proc Natl Acad Sci USA. 2015;112:5129–34.
Sonzogni M, Wallaard I, Santos SS, Kingma J, du Mee D, van Woerden GM, et al. A behavioral test battery for mouse models of Angelman syndrome: a powerful tool for testing drugs and novel Ube3a mutants. Mol Autism. 2018;9:47.
Vincent SB. The white rat and the maze problem: the introduction of a tactual control. J Anim Behav. 1915;5:175.
Kraeuter AK, Guest PC, Sarnyai Z. The Y-Maze for assessment of spatial working and reference memory in mice. Methods Mol Biol. 2019;1916:105–11.
Lalonde R. The neurobiological basis of spontaneous alternation. Neurosci Biobehav Rev. 2002;26:91–104.
Wilson MR, Satapathy S, Vendruscolo M. Extracellular protein homeostasis in neurodegenerative diseases. Nat Rev Neurol. 2023;19:235–45.
Lee Y, Lee YW, Gao Q, Lee Y, Lee HE, Ryu JH. Exogenous insulin-like growth factor 2 administration enhances memory consolidation and persistence in a time-dependent manner. Brain Res. 2015;1622:466–73.
Fitzgerald GS, Chuchta TG, McNay EC. Insulin-like growth factor-2 is a promising candidate for the treatment and prevention of Alzheimer’s disease. CNS Neurosci Ther. 2023;29:1449–69.
Mehellou Y, Rattan HS, Balzarini J. The ProTide prodrug technology: from the concept to the clinic. J Med Chem. 2018;61:2211–26.
El-Shewy HM, Luttrell LM. Insulin-like growth factor-2/mannose-6 phosphate receptors. Vitam Horm. 2009;80:667–97.
Liang Q, Catalano F, Vlaar EC, Pijnenburg JM, Stok M, van Helsdingen Y, et al. IGF2-tagging of GAA promotes full correction of murine Pompe disease at a clinically relevant dosage of lentiviral gene therapy. Mol Ther Methods Clin Dev. 2022;27:109–30.
Dalle Vedove E, Costabile G, Merkel OM. Mannose and Mannose-6-Phosphate receptor-targeted drug delivery systems and their application in cancer therapy. Adv Healthc Mater. 2018;7:e1701398.
Duffy KR, Pardridge WM, Rosenfeld RG. Human blood-brain barrier insulin-like growth factor receptor. Metabolism. 1988;37:136–40.
Pardridge WM. Transport of insulin-related peptides and glucose across the blood-brain barrier. Ann N Y Acad Sci. 1993;692:126–37.
Reinhardt RR, Bondy CA. Insulin-like growth-factors cross the blood-brain-barrier. Endocrinology. 1994;135:1753–61.
Lewitt MS, Boyd GW. The role of insulin-like growth factors and insulin-like growth factor-binding proteins in the nervous system. Biochem Insights. 2019;12:1178626419842176.
Kar S, Chabot JG, Quirion R. Quantitative autoradiographic localization of [I-125] insulin-like growth factor-I, [I-125] insulin-like growth factor-Ii, and [I-125] insulin-receptor binding-sites in developing and adult-rat brain. J Comp Neurol. 1993;333:375–97.
Fernández-Monreal M, Brown TC, Royo M, Esteban JA. The balance between receptor recycling and trafficking toward lysosomes determines synaptic strength during long-term depression. J Neurosci. 2012;32:13200–5.
Pandey K, Bessières B, Sheng SL, Taranda J, Osten P, Sandovici I, et al. Neuronal activity drives IGF2 expression from pericytes to form long-term memory. Neuron. 2023;111:3819–36e8.
Schmeisser MJ, Baumann B, Johannsen S, Vindedal GF, Jensen V, Hvalby ØC, et al. IkappaB kinase/nuclear factor kappaB-dependent insulin-like growth factor 2 (Igf2) expression regulates synapse formation and spine maturation via Igf2 receptor signaling. J Neurosci. 2012;32:5688–703.
Wang ZQ, Fung MR, Barlow DP, Wagner EF. Regulation of embryonic growth and lysosomal targeting by the imprinted Igf2/Mpr gene. Nature. 1994;372:464–7.
Ghosh P, Dahms NM, Kornfeld S. Mannose 6-phosphate receptors: new twists in the tale. Nat Rev Mol Cell Biol. 2003;4:202–12.
Wang R, Qi X, Schmiege P, Coutavas E, Li X. Marked structural rearrangement of mannose 6-phosphate/IGF2 receptor at different pH environments. Sci Adv. 2020;6:eaaz1466.
Olson LJ, Dahms NM, Kim JJ. The N-terminal carbohydrate recognition site of the cation-independent mannose 6-phosphate receptor. J Biol Chem. 2004;279:34000–9.
Acknowledgements
This work was supported by the Foundation for Angelman Syndrome Therapeutics–Italia and National Institutes of Health (Grant No. MH065635 to CMA). CJA and DT thank NYU for a MacCracken Fellowship.
Author information
Authors and Affiliations
Contributions
CMA, DT, FA, CJA and EP contributed to conceptualization; FA, KP, SP, CJA, EP, and SG-G contributed to the methodology and data acquisition; DT and CJA designed the prodrug; FA, EP, SG-G, KP, LM, SP, and CJA contributed to formal analysis and validation; CMA, DT provided supervision; CMA, DT, FA, EP, KP, and CJA assisted with data visualization; CMA and DT provided funding; CMA, DT, FA, CJA, and EP wrote the manuscript; all authors reviewed and edited the manuscript and approved the final draft.
Corresponding authors
Ethics declarations
Competing interests
CMA, CJA, and DT are inventors of patents filed by NYU on the use of IGF2 receptor agonist ligands for the treatment of Angelman syndrome and autism and of memory and cognitive impairments. The other authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Aria, F., Arp, C.J., Prikas, E. et al. A prodrug targeting CIM6P/IGF2R enhances memory in healthy mice and reverses deficits in an Angelman syndrome mouse model. Transl Psychiatry 15, 438 (2025). https://doi.org/10.1038/s41398-025-03610-1
Received:
Revised:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41398-025-03610-1
