3.1 Mamba venoms, a source of V2R inhibitory toxins
Mambas (genus Dendroaspis ) are venomous African Elapidae snakes,
whose bites are often deadly without proper medical treatment. They live
throughout sub-Saharan Africa in savannas and in tropical rain forest
(Ainsworth et al., 2017). MQ1 was discovered in the Dendroaspis
angusticeps (Da) venom by function-based screening (Ciolek et al.,
2017). The same strategy was applied to three other mamba venoms (Fig.
2). One gram of each Dendroaspis polylepis (Dp),Dendroaspis jamesoni (Dj) and Dendroaspis viridis (Dv)
venoms were firstly fractionated by strong cation exchange liquid
chromatography (Fig. 2A-C). Seven fractions (DpS, DpL, DjkN, DvC, DvG,
DvK, DvL) inhibited tritiated vasopressin (3H-AVP)
binding on the human V2R stably expressed in a CHO cell line. These
fractions were submitted to reverse phase liquid chromatography
(Supplementary Fig. 1) to isolate DvGB (MQ2), DpSH (MQ3), DvKN (MQ4),
DjkNE (MQ5), DvLM (MQ6) and DvCK (MQ9) containing one toxin each fully
sequenced by mass fragmentation strategies (Supplementary Fig. 2). The
DpLK fraction contains two peptides, one previously described as DTX-B
(Strydom and Joubert, 1981), that we refer to as MQ7, and a novel
homologue with the A27S substitution (MQ8). The list of the active
fractions, their masses and number of disulfide bridges, as well as
their proportions in the venom are summarized in Supplementary Table 1.
Synthetic analogues of MQ2 to MQ6 were produced by solid phase chemical
synthesis (Supplementary Fig. 3). We succeeded in the synthesis of the
linear forms of the three MQ7, MQ8 and MQ9, but failed in their
oxidation steps. Consequently, their pharmacological characterizations
were limited by the availability of natural products found in the
venoms. In addition, MQ7 and MQ8 could not be separated from each other,
thus we characterized the effects of the MQ7/8 mix (DpLK fraction).
Binding competition experiments on hV2R showed that these novel MQs
inhibit 3H-AVP binding with affinities in the 1-100 nM
range (Fig. 2D, Table 1). Moreover, MQs dose-dependently inhibited
AVP-induced cAMP production measured on a CHO cell line stably
expressing hV2R (Fig. 2E, Table 1), highlighting their antagonist
activity (MQs never induced a cAMP production by themselves). Thein vivo diuretic effects of the synthetic MQs (MQ1-MQ6) were
investigated in Sprague-Dawley male rats after a single i.p. injection.
Control rats urinated an average of 43 ml/day/kg BW. MQ1-5 (30 nmol/kg
BW) increased diuresis by 6.5-fold without any loss of electrolytes
(Table 2). MQ6, displaying a lower affinity for V2R and a lower potency
to inhibit cAMP production in vitro was tested at 100 and 300
nmol/kg BW, and a 2-fold increase in diuresis at the highest dose was
observed. Irrespective of the MQ tested or the dose used, urine
osmolality variation is inversely proportional to aquaresis modification
(Table 2), demonstrating a pure aquaretic effect of the MQs.