Tegaserod mimics the neurostimulatory glycan polysialic acid and promotes nervous system repair

Abstract

Glycans attached to the cell surface via proteins or lipids or exposed in the extracellular matrix affect many cellular processes, including neuritogenesis, cell survival and migration, as well as synaptic activity and plasticity. These functions make glycans attractive molecules for stimulating repair of the injured nervous system. Yet, glycans are often difficult to synthesize or isolate and have the disadvantage to be unstable in a complex tissue environment. To circumvent these issues, we have screened a library of small organic compounds to search for structural and functional mimetics of the neurostimulatory glycan polysialic acid (PSA) and identified the 5-HT4 receptor agonist tegaserod as a PSA mimetic. The PSA mimicking activity of tegaserod was shown in cultures of central and peripheral nervous system cells of the mouse and found to be independent of its described function as a serotonin (5-HT4) receptor agonist. In an in vivo model for peripheral nerve regeneration, mice receiving tegaserod at the site of injury showed enhanced recovery compared to control mice receiving vehicle control as evidenced by functional measurements and histology. These data indicate that tegaserod could be repurposed for treatment of nervous system injuries and underscores the potential of using small molecules as mimetics of neurostimulatory glycans.

Keywords: Drug repurposing; Glycan; Mimetic; Peripheral nerve; Polysialic acid; Regeneration; Tegaserod.

Fig. 1

Tegaserod competes with PSA peptide mimetic for binding to the PSA-specific antibody 735 (mAb 735). (A) Results from a competition ELISA of the dose-dependent ability of tegaserod to interfere with binding of the antibody 735 with the PSA peptide mimetic immobilized at the bottom of the wells, *p < 0.05; ***p < 0.0005 (data were compared by one-way analysis of variance (ANOVA)). (B) Surface plasmon resonance (SPR) profile of the ability of 100 nM of tegaserod to interfere with binding of antibody 735 with the PSA peptide mimetic covalently immobilized to a CM5 sensor chip. The SPR signal is displayed as resonance versus time where 1000 RU (resonance units) represent a shift in resonance angle of 0.1° corresponding to a change in surface antibody concentration of ~1 ng/mm². (C) Surface plasmon resonance (SPR) profile of the ability of 0.1, 1, 10, 20 and 30 µM tegaserod or nitrendipine to interfere with binding of antibody 735 with the PSA peptide mimetic. Only tegaserod interferes with binding of antibody 735 with the PSA peptide mimetic in a concentration dependent manner.

Fig. 2

Structure models of PSA and tegaserod in complex with the surrogate PSA binding pocket of antibody 735 (heavy chain, green; light chain, cyan) derived from molecular modeling and docking. (A) PSA, depicted in yellow, binds to a broad groove in the antibody 735 CDR region forming multiple hydrogen bonds. (B) Tegaserod, depicted in magenta, binds within a groove in one region of the putative PSA-binding site, where it is anchored primarily by a salt-bridge formed between the guanidinium group of the ligand with aspartic acid 105 on the heavy chain located deep within the groove. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

In vitro activity of tegaserod on extension of neurites/processes from and survival of motoneurons, cerebellar granule neurons (cGNs), dorsal root ganglion (DRG) neurons and Schwann cells. (A) Neurite/process extension following varying doses of tegaserod after 24 h. (B) Comparison of the neurite-extending capabilities of tegaserod and related compounds on murine motoneurons from wild type and NCAM deficient (NCAM^−/−) mice. Tegaserod elicits neurite extension comparable to the PSA analog colominic acid and the PSA peptide mimetic, which is not observed in motoneurons from NCAM^−/− mice. 5-HT₄ agonists serotonin and cisapride as well as the 5-HT₄ antagonist GR113808 do not increase neurite extension from motoneurons, nor did GR113808 attenuate the effects of tegaserod. (C–E) Effects of tegaserod and related compounds on cGNs, DRG neurons and Schwann cells. (F) Relative survival of motoneurons and cGNs treated with tegaserod and related molecules. Concentrations of compounds, glycans and peptides: tegaserod, serotonin, cisapride and GR113808 (100 nM), colominic acid (3 µM), PSA peptide mimetic (30 µM). All treatments were performed in duplicates and at least 100 cells were counted for each treatment. Results are from two or more experiments. Mean values ± SEM are shown. (C–E) *p < 0.05, **p < 0.005, ***p < 0.0005 via Student’s t-test.

Fig. 4

Metrics of functional recovery of limb function following femoral nerve injury. (A–B) The foot base angle (FBA) is the angle between the walking surface and the hind foot when the contralateral foot is lifted. (A) In an uninjured mouse, the FBA is 60–70°. (B) Following injury, the FBA of the injured leg increases to 100–110°. As the femoral nerve reinnervates the quadriceps muscle and muscle function is restored, the FBA decreases towards pre-injury levels. (C–D) The protraction limb ratio (PLR) measures the voluntary movement of the hind limbs to grasp an object when hanging upside down. The PLR measures the distance between the base of the tail to the tip of the extended limbs, as measured by dividing the length of the uninjured limb by that of the injured limb. (C) In uninjured animals, both hind limbs extend equal distances resulting in a PLR close to 1. (D) Following femoral nerve injury the limb with the injured nerve fails to extend to the same extent as the uninjured hind limb, giving a PLR of >1. A decline in the PLR toward 1 over time indicates recovery.

Fig. 5

Evaluation of functional recovery of mice via measurement of the foot base angle (FBA) following femoral nerve injury and insertion of a conduit containing vehicle alone, 250 nM tegaserod, or 2,500 nM tegaserod. Pre-injury FBAs are between 60 and 70°, which increase following injury, and a return toward pre-injury angles indicates recovery. (A) FBA of mice that received vehicle, 250 nM tegaserod, or 2,500 nM tegaserod over a 15-week recovery period. ***p < 0.0005 via one-way ANOVA with Tukey’s post-hoc test. (B) Recovery index (RI) of the FBA of individual animals within each treatment group is shown by normalizing the recovery of individual animals to their pre-injury FBA, **p < 0.005, *p < 0.05 via one-way ANOVA with Tukey’s post-hoc test.

Fig. 6

Evaluation of functional recovery via the protraction limb ratio (PLR) following femoral nerve injury. Pre-injury PLR of 1 is increased following injury as the injured limb cannot extend to the same extent as the uninjured limb. A return to 1 indicates recovery of function. (A) PLR of mice that received vehicle, 250 nM tegaserod, or 2,500 nM tegaserod measured over a 15-week recovery period. *p < 0.05 via one-way ANOVA with Tukey’s post-hoc test. (B) Recovery index (RI) of PLR of individual animals within each treatment group after 15 weeks.

Fig. 7

Mass of quadriceps muscle of mice after 15-week recovery period. (A) Images of representative quadriceps muscles from the healthy and injured hind limbs of mice following femoral nerve injury and insertion of a conduit containing vehicle or two doses of tegaserod. (B) Quantification of average quadriceps muscle masses of mice that received vehicle, 250 nM tegaserod, or 2,500 nM tegaserod, and the average muscle mass from the uninjured hind limbs. *p < 0.05 via Student’s t-test.

Fig. 8

Histology of femoral nerves and quantification of regenerated axons after a 15-week recovery period. (A–C) 10× and 100× magnification images of regenerated nerves from animals treated with (A) vehicle control, (B) 250 nM tegaserod, and (C) 2,500 nM tegaserod. (D) Dot plot shows the numbers of axons in the center of the regenerated nerves and averages within groups. *p < 0.05 via Student’s t-test.