The Targeted Secretion Inhibitor (TSI), SXN101959, activates the rat GHRH receptor, internalises into GH3 cells stably expressing this receptor and cleaves vesicle associated membrane protein 2 (VAMP2)

Verity Cadd, Mark Elliott, Julien Browne, Mary Charteris, Agnieszka Lewandowska, Matthew Beard, Elaine Harper. Syntaxin Ltd, Units 4-10, The Quadrant, Barton Lane, Abingdon, Oxfordshire, UK.

Botulinum neurotoxins (BoNTs) are potent inhibitors of neuronal vesicular secretion (VS) particularly acetylcholine (ACh) at the neuromuscular junction (NMJ) and as such they have been used clinically in the treatment of muscular dystonias. However, although VS plays a significant role in a wide range of diseases, the use of BoNT for treating these conditions is precluded due to its inherent neuronal targeting. TSIs are proteins rationally engineered from BoNT such that they retain the ability to inhibit VS but are retargeted to non-neuronal cells with aberrant secretion. The mechanism of action (MOA) of BoNT is well known and requires activity of all three protein domains (Montal, 2010). The targeting domain directs the toxin to neuronal binding sites to facilitate internalisation into endosomes, the translocation domain (HN) inserts into the endosomal membrane to allow the light chain (LC) into the cytosol and the LC acts as a metalloprotease to cleave SNARE(Soluble N-ethylmaleimide Sensitive Factor Attachment Protein) proteins that are critical to the process of VS. The seven serotypes of BoNT (A-G) have different specificity for SNARE proteins with serotype D specifically cleaving the vesicle associated membrane (VAMP1, 2 and 3) proteins. The TSI, SXN101959, comprises the LC and HN domains of BoNT serotype D and has a targeting domain comprising a modification of growth hormone releasing-hormone (GHRH). It is in development by Syntaxin Ltd. for the treatment of Acromegaly; a disease caused by excess pituiutary growth hormone (GH) secretion. It is hypothesised that the MOA of SXN101959 is essentially the same as BoNT and involves binding and internalisation into target cells, SNARE cleavage and inhibition of secretion. This study was conducted to confirm this hypothesised MOA. To facilitate this, a test cell system was constructed where the rat GHRH receptor (rGHRH-R) was stably expressed in GH3 cells, a rat pituitary cell line which secretes GH and expresses VAMP2 but not this receptor (GH3-GHRH-R; these cells were used for all studies).

To assess the ability to bind to rGHRH-R, cells were incubated (90min) with increasing concentrations of SXN proteins (SXN01959 and SXN101742) and cAMP accumulation measured using a TR-FRET immunoassay. SXN protein internalisation was assessed using high content screening (HCS). Confluent cells were incubated with SXN (5 and 60min) and cellular localization established using a labelled antibody specific to the BoNT D serotype LC. VAMP2 cleavage/ depletion was detected by Western blotting. Cells were incubated (72h) with increasing concentrations of SXN and lysed. Lysate proteins were separated by SDS-PAGE, transferred to nitrocellulose membranes and VAMP2 detected using a specific antibody and densitometry. Data are the mean ± s.e.mean. Both SXN101742 and SXN101959 produced a concentration-dependent increase in cAMP accumulation (pEC50 = 7.53 ± 0.14, n = 3; and 7.90 ± 0.06, n = 5, respectively) and there was no significant difference in the maximal response they induced (Emax nM cAMP; SXN101742 = 26.7 ± 3.9; SXN101959 = 29.5 ± 4.6, t-test p>0.05). SXN101742 was rapidly internalised with punctate staining indicating a predominant endosomal localisation. After 60min incubation, internalised SXN101742 was significantly increased (3µM; 5min RFU = 36.8 ± 3.4x103; 60min = 146.3 ± 7.1x103; n = 3; paired t-test p<0.05) and internalisation was shown to be concentration-dependent (pEC50 = 6.28 ± 0.01, n = 3). SXN101742 produced a concentration-dependent and complete depletion of VAMP2 (pEC50 = 7.63 ± 0.08; depletion = 98.7 ± 0.3%, n = 3). The effect of SXN proteins on GH secretion from GH3-rGHRH cells could not be established because in contrast to GH3 cells, the GH secreted from these cells was below the limit of detection of a commercially available GH ELISA.

These studies, in part, confirm the hypothesised MOA of SXN101959; binding to and activation of the GHRH-R, internalisation into cells and cleavage of the SNARE protein VAMP2.

Montal M, (2010). Annu Rev Biochem., 79, 591-617.