الخلاصة:
This work aims to isolate Actinobacteria strains with a growth promotion ability and the biocontrol potential
of Fusarium culmorum, the wheat root rot-causing fungi. The exploration of terrestrial and aquatic Algerian
ecosystems, i.e. the wheat rhizosphere in the Tiffeche region (Souk-Ahras) and the aquatic sediments of Lake
Oubeira (El Taref), respectively, resulted in 102 native isolates. Therefore, 37 have morphological and cultural
characteristics similar to the genus Streptomyces. These isolates were screened for their plant growth-promoting
traits. These activities consist of the production of hydrogen cyanide (HCN), ammonia (NH3), and indole-3-
acetic acid (IAA), as well as in vitro antagonism against F. culmorum. Among the auxin family, IAA constitutes
a crucial phytohormone regulating specific tropic responses of plants and functions as a chemical signal between
host plants and their symbionts. IAA derived from Actinobacteria grown on agricultural waste represents a more
economical alternative than its synthetic homologous. Rhizospheric isolate AW 22 was positive for HCN and
NH3 production, growth inhibition of F. culmorum with an index of 67.320±8.99% and high IAA content of
23.999±1.126 μg. mL-1 in standard growth conditions on yeast-tryptone broth (YTB) amended with 0.2% (w/v)
L-Tryptophan. Thus, the AW22 isolate was selected for a polyphasic chemotaxonomic characterization and the
optimization of the production process of this phytohormone. Molecular and phylogenetic analysis identified
isolate AW22 as Streptomyces rubrogriseus, and its sequence was deposited in Genbank under accession ID
OP176004. Analysis of the putative IAA produced by S. rubrogriseus AW22 on YTB using thin-layer
chromatography (TLC) and (HPLC) revealed Rf values equal to 0.69 and a retention time of 3.711 min,
equivalent to the authentic IAA. Artificial intelligence-based approaches (i.e. Behnken design from response
surface methodology (BBD-RSM) with artificial neural networks (ANNs) coupled with the genetic algorithm
(GA)) were employed to bioengineer in vitro and silico a suitable medium for maximum IAA bioproduction.
According to the Box Behnken Design matrix, data were based on empirical studies involving the inoculation
of AW22 in various cultural conditions and low-cost feedstocks notably, the spent coffee grounds (SCGs). Four
input variables comprising L-Trp (X1), incubation T° (X2), initial pH (X3) and SCG concentration (X4) were
screened via Plackett-Burman design (PBD) and served as BBD and ANN-GA inputs. The IAA yield constituted
the output variable (Y in µg. mL-1). Upon training the model, the optimal conditions suggested by the ANN-GA
model were X1= 0.6%, X2= 25.8°C, X3 = 9, X4=30%). An R2 of 99.98%, adding to an MSE of 1.86x10-5 at
129 epochs, postulated higher reliability of the ANN-GA approach in predicting responses, compared with
BBD-RSM modeling exhibiting an R2 of 76,28%. Using the process parameters generated by ANN-GA AW 22
achieved a maximum IAA yield of 188.290±0.38 µg. mL-1. This optimization resulted in a 4.55-fold and 4.46-
fold increase in IAA secretion after eight days of incubation, corresponding to ANN-GA and BBD-RSM models,
respectively. These results confirm the validity of both models in maximizing IAA yield from the
multifunctional S. rubrogriseus AW22 isolated for the first time in Algeria.
