Statistical Optimisation of Streptomyces sp. DZ 06 Keratinase Production by Submerged Fermentation of Chicken Feather Meal

Abstract

This study focused on the isolation of actinobacteria capable of producing extracellular keratinase from keratin-rich residues, which led to the selection of an actinobacterial strain referenced as Streptomyces strain DZ 06 (ES41). The Plackett–Burman screening plan was used for the statistical optimization of the enzymatic production medium, leading to the identification of five key parameters that achieved a maximum activity of 180.1 U/mL. Further refinement using response surface methodology (RSM) with a Box–Behnken design enhanced enzyme production to approximately 458 U/mL. Model validation, based on the statistical predictions, demonstrated that optimal keratinase activity of 489.24 U/mL could be attained with 6.13 g/L of chicken feather meal, a pH of 6.25, incubation at 40.65 °C for 4.11 days, and an inoculum size of 3.98 × 10⁷ spores/mL. The optimized culture conditions yielded a 21.67-fold increase in keratinase compared with the initial non-optimized standard conditions. The results show that this bacterium is an excellent candidate for industrial applications when optimal conditions are used to minimize the overall costs of the enzyme production process.

1. Introduction

The poultry industry is the main driver of global meat production [1]. This sector continues to grow and industrialize in many parts of the world. The increase in population, purchasing power, and urbanization have been powerful drivers of growth for this sector. The latest forecast of the Food and Agriculture Organization of the United Nations (FAO) predicts that world poultry meat production is expected to reach 146 million tonnes in 2024, increasing by 0.8% year-on-year [2]. The Agricultural Outlook 2024–2033 of the Organisation for Economic Cooperation and Development (OECD) and the FAO predicted that worldwide, poultry, pigmeat, beefmeat, and sheepmeat consumption is projected to grow 16%, 8%, 11%, and 16%, respectively, by 2033. By that year, poultry meat is expected to account for 43% of the total protein from meat sources, with pig, beef, and sheepmeat following in consumption [3]. This development is unfortunately inseparable from the increase in accompanying organic products, in the form of viscera, feet, heads, bones, blood, and feathers [4]. Chicken feathers from poultry processing industries are generated in abundance, as iskeratin-rich waste, which poses a serious ecological concern due to the large meat quantities produced and their highly resistant characteristics in terms of degradation [1,4]. The consumption of poultry meat leads to an annual production of around 8.5 million tonnes of feather waste worldwide [5].

In nature, keratin represents the third most abundant biomass. It represents almost 90% of the feather’s total weight, thus constituting an important source of amino acids for various biotechnological applications [6,7]. The recalcitrant nature of these wastes makes conventional disposal methods such as incineration and landfills inappropriate, given the ecological constraints involved, affecting both human health and the whole environment [8]. The development of non-polluting methods guaranteeing the production of interesting products becomes essential. In this regard, the use of microorganisms degrading keratin materials represents a useful alternative biotechnological solution [9]. Keratinolytic microorganisms include fungi such as members of the genera Trichoderma [10], Trichosporum [11], Aspergillus, and Fusarium [12,13], bacteria such as the members of the genera Bacillus [14], Serratia [15], Fervidobacterium [16], Pseudomonas [17], or specifically actinobacteria such as Streptomyces [18], Actinomadura [19], Arthrobacter [20], and Brevibacterium [21]. These microorganisms can grow on media containing keratin substrates as carbon and nitrogen sources, mainly due to their aptitude to producekeratinolytic proteases/keratinases that exhibit high specific hydrolytic activity towards keratin substrates [22].

Keratinases (E.C 3.4.21/24/99.11) consist of a particular class of protein-hydrolyzing enzymes synthesized by microorganisms with the ability to degrade insoluble keratinous substrates into value-added products [6,23,24]. These catalysts generally hydrolyze soluble proteins efficiently compared to insoluble ones, like keratins, and have advantages over conventional proteolytic enzymes regarding stability over a wide class of environmental conditions [24,25]. Most keratinases are classified as serine proteases, while others belong to the metalloproteases class [26]. They are generally inducible, extracellular enzymes secreted by various microorganisms in a keratin-containing medium and exhibit high specificity towards the keratin substrate [5,27,28]. Keratinolytic proteases are promising tools in different biotechnological fields like biotransformation of keratin waste into animal feed and nitrogen fertilizers, as well as application in the cosmetic and detergent industries, pharmaceutical fields, textile manufacturing, and leather tanning [29,30].

Among the numerous microorganisms involved in the production of keratinases, actinobacteria are known as a rich source of bioactive molecules for various industry applications [31]. In nature, this Gram-positive bacterial group represents enormous potential in terms of biodegradation/bioconversion of animal or plant complex biopolymers. They are capable of synthesizing varied types of biocatalysts like proteases, keratinases, lipases, cellulases, and chitinases [32].

On the other hand, the optimal culture conditions for keratinase production vary from one organism to another; the identification and examination of physicochemical factors running the production of these enzymes from a given isolate for a better production of these molecules, is of great importance [33]. The limitations of classic optimization methods, which consist of modifying one factor at a time while maintaining all other factors invariable (OFAT: One-Factor-at-A-Time), have given rise to alternative statistical approaches using experimental designs [33,34]. Thus, studies of factor selection influencing a response given by linear estimation of the variables main effects can be carried out via a two-level factorial design with minimal trials, such as the Plackett–Burman design. The Plackett–Burman’s selected variables can be optimized by response surface methodology (RSM) analysis using Box–Behnken design, which is a statistical and mathematical optimization strategy that estimate the correlation between the variables and predicts the response in an efficient and reproducible experimental design [35].

Here we report the isolation, identification, and description of a new thermophilic actinobacterial strain from the Kabylia region in the north-eastern of Algeria. Isolation and fermentation media, as well as buffered solutions based on keratinious substrates, were prepared from locally collected chicken feathers. The strain showed rapid growth on a medium based on chicken feather meal as the only source of carbon and nitrogen. Keratinolytic activity was detected under fermentation conditions, and statistical approaches were used to select and determine the optimal cultivation conditions required for keratinase production.

2. Materials and Methods

2.1. Substrates and Chemicals

Unless specified otherwise, all substrates, chemicals, and reagents were of analytical grade or the highest available purity and were purchased from Sigma–Aldrich Co. (St. Louis, MO, USA)

2.2. Culture Medium Substrate Preparation

The culture substrate used was chicken feather meal prepared from locally collected poultry feathers. The feathers, which had not undergone any physicochemical treatment, were washed with tap water, then distilled water, air-dried, and then dried in an oven. The dried feathers were ground and sieved [17].

2.3. Sampling and Sample Pretreatment

Samples of chicken compost manure (CCM), soil amended with chicken feathers (CFS), and decomposing feathers (DF) were taken from various sites located in the Bejaia district of Kabylia in north-eastern Algeria. A total of 21 samples were collected from three locations, including nine samples from the DjebiraBoukhelifa site, seven samples from the Fenaia Ilmathen site, and five samples from the Ighil Ali site (Figure 1).

For all samples, the top five centimeters of decomposing manure, soil, and feathers were removed, and sufficient quantities of samples were taken sterilely and put into sterile plastic pouches, then transferred directly to the laboratory and stored at 4 °C until use.

To facilitate the isolation of actinobacteria, all the samples underwent physical and chemical pretreatment to minimize the development of non-mycelial bacteria, which grow faster than actinobacteria, and fungi, which form invasive colonies and inhibit the development of actinobacteria. The samples were heated to 60 °C for 1 h, followed by the addition of 1% (w/w) calcium carbonate (CaCO₃) (1 g of CaCO₃ for every 100 g of sample in sterile vials) [36]. The presence of calcium carbonate alkalized the pH of treated samples, which favored the development of actinobacteria [37]. To promote the development of thermophilic actinobacteria, the containers were incubated at 45 °C for 21 days [19].

2.4. Actinobacterial Isolation and Culture

During the incubation of the samples, 10 g of each sample was taken after 7, 14, and 21 days. Ten-fold dilutions were performed after a 10% (w/v) stock solution was made in sterile distilled water. The stock solutions and their dilutions were spread-plated (200 μL) onto feather basal salt medium (FBM), which contained (g/L): 20.0 chicken feather meal, 1.0 dipotassium hydrogen phosphate (K₂HPO₄), 0.5 magnesium sulfate heptahydrate (MgSO₄·7H₂O), 3.0 calcium carbonate (CaCO₃), 0.01 iron sulfate heptahydrate (FeSO₄·7H₂O), 0.5 sodium chloride (NaCl), and 20.0 agar, with the pH adjusted to 8.2 ± 0.2. After incubating each plate for seven to ten days at 45 °C, the existence of colonies with the morphological characteristics of actinobacteria with the presence or absence of a distinct hydrolysis halo surrounding the colony was evaluated [19]. The colonies of interest were subcultured to obtain pure cultures. A selection of strains was maintained at room temperature (22 ± 3 °C) on solid FBM medium (pH 8.2 ± 0.2) and stored at −20 °C in a 20% (v/v) glycerol suspension.

2.4.1. Solid-State Screening of Keratinase-Producing Strains

Using the same isolation medium containing chicken feather meal (2%w/v) as the sole source of organic carbon and nitrogen, pH 8, an initial screening for keratinase-producing actinobacteria was performed. The media were then inoculated with the isolated strains and incubated at 45 °C for 4 days. The strains presenting pronounced growth with or without clear hydrolysis zones around the colonies were selected for further work [38].

2.4.2. Screening in Liquid Media for Keratinase-Producing Strains

Strains exhibiting pronounced growth with or absence of hydrolysis zones on FBM agar were inoculated into FBM liquid medium (pH 8) and incubated at 45 °C for 2 days. The pre-cultures were used to inoculate 50 mL FBM broth containing chicken feather meal (2%, w/v) as substrate, which was supplemented with 1 mL of trace salts solution (g/L: 0.9 ZnSO₄, 0.7 CaCl₂, 0.2 MnSO₄·7H₂O, 0.3 KCl). The pH was adjusted to 8.2 ± 0.2. The cultures were cultivated for 10 days at 45 °C under agitation at 150 rpm. After 3, 5, 7, and 10 days, 1 mL samples were taken and centrifuged at 10,000× g for 20 min at 4°C. The cell-free supernatant obtained was then used for keratinase activity assay (Section 2.6) [19,24]. All the strains were cultivated in triplicate.

2.5. Bacterial Strain, Growth Conditions, and Preparation of Spore Solutions

A spore suspension was prepared from FBM agar plates inoculated with the strain that exhibited the highest keratinase activity that was previously incubated for 4 days at 45 °C. The spores were gently harvested to prevent mycelial detachment by mixing 10 mL of sterile distilled water with 1% (v/v) Tween 80, then collecting the mixture in sterile flasks to serve as an inoculum for the synthesis of keratinase. The suspension was diluted 1:10 (v/v), and the spores were counted in a counting chamber (Malassez REF 06 106 10 MARIENFELD, Lauda-Königshofen, Germany) [39]. For submerged fermentation experiments, 250-mL flasks containing 50 mL of FBM broth were prepared, adjusted to varying pH levels, autoclaved, inoculated with strain ES41, and then placed in an orbital shaker at given rotational speeds and temperatures for specified times. After the required incubation, the culture broths were harvested and centrifuged at 10,000× g for 20 min at 4 °C. The supernatants were then used for enzyme assays using chicken feather meal as a substrate [40]. For selection and optimization studies of the factors involved in keratinase production, the culture medium composition and conditions were varied according to the experimental data.

2.6. Enzyme Activity Assay

The standard operating conditions for measuring keratinolytic activity were followed as described by Wawrzkiewicz et al. [41]. The enzyme solution (0.2 mL) was added to 1.8 mL of 0.5% (w/v) soluble keratin in 0.05 M Tris-HCl buffer (pH 8.2 ± 0.2) and then incubated in a water bath for 10 min at 50 °C. After placing the reaction mixture in an ice bath and adding two milliliters of 0.4 M trichloroacetic acid (TCA), the reaction was halted. Without adding keratin, the enzyme solution was incubated with 2.0 mL TCA to create the control. After that, the mixture was centrifuged for 20 min at 4 °C and 5000× g. At 280 nm, the supernatant’s absorbance was measured in comparison to the control. Under the previously mentioned experimental conditions, a rise in absorbance at 280 nm with the blank of 0.01 per minute is defined as one unit per milliliter (U/mL) of keratinolytic activity. The activity is calculated by the following Equation (1) [38]:

$$A_{(U/mL)}=V\cdot N\cdot A_{280}/(0.01\cdot T)$$

(1)

where, A is the keratinase activity; N is the dilution factor; V is the final reaction volume (mL); T is the incubation time (min); A₂₈₀ is the absorbance of the reaction mixture measured at 280 nm against the control. Each assay in this study was performed in triplicate.

2.7. Morphological Characterization and Molecular Identification of the Keratinase-Producing Strain, ES41

For better guidance in the identification and classification process of the selected strain, macromorphological characterization (size, pigmentation, shape, and appearance of colonies) and micromorphological characterization (Gram staining, morphology, and arrangement of cells and spores of the isolate studied) were carried out on media recommended for culturing actinobacteria [42].

Molecular identification was carried out using the GF-1 Nucleic Acid Extraction Kit (Vivantis Technologies Sdn. Bhd, Shah Alam, Selangor DE, Malaysia). The polymerase chain reaction (PCR) was performed using universal 16S rRNA primers (27: 5′–AGA GTT TGA TCC TGG CTC AG–3′, and 1478 R: 5′–CCG TCA ATT CCT TTG AGT TT-3′) [43]. A thermal cycler (Bio-Rad bicycler, Hercules, CA, USA) was used to conduct the PCR, whereas the measure of the amplicon concentrations was performed using a nanodrop spectrophotometer (NanoDropTM 2000, Thermo Fisher Scientific, Waltham, MA, USA) [44]. The sequencing was performed using the Sanger method by electrophoretic migration profile analysis of the DNA fragments. The sequences obtained were compared with similar sequences by submission to BLAST (Basic Local Alignment Search Tool, National Center for Biotechnology Information) and then identified using the GenBank database. The phylogeny of the keratinase-producing strain was established using a web application (Phylogeny.fr) via the neighbor-joining method. [45].

2.8. Determination of Influencing Physicochemical Parameters Using the Plackett-BurmanApproach (PBD)

Using the Plackett–Burman experimental design, the critical factors influencing the response, namely the production of keratinases when strain ES41 is cultured in a liquid medium based on chicken feather meal as the only carbon and nitrogen sources, were determined [46]. Ten two-level independent variables represent the initial requirements for keratinase synthesis in strain ES41. These include sodium chloride (NaCl): X₁, inoculum size: X₂, incubation time: X₃, initial pH: X₄, dipotassium hydrogen-phosphate (K₂HPO₄): X₅, orbital agitation: X₆, chicken feather meal: X₇, calcium carbonate (CaCO₃): X₈, incubation temperature: X_9, and magnesium sulphate heptahydrate (MgSO₄·7H₂O): X₁₀ (

Table 1. Selected independent variables and their level variation in Plackett–Burman design (PBD) for the optimization of keratinase production by strain ES41.

Study Type	Screening
Design Type	Plackett-Burman
Design Mode	Principal Effect	First Order Model	No Blocks Runs
Response R	Keratinase Activity		20
Factor	Name	Units	Type	Level (−1)	Level (+1)
X1	NaCl	g/L	Numeric	0	3
X2	Inoculum size	Spores/mL	Numeric	1.5 × 104	1.5 × 1010
X3	Incubation time	Days	Numeric	2	14
X4	initial pH		Numeric	4	12
X5	K2HPO4	g/L	Numeric	0	3
X6	Orbital agitation	rpm	Numeric	0	250
X7	Chicken feather meal	% (w/v)	Numeric	0.2	2
X8	CaCO3	g/L	Numeric	0	4
X9	Incubation temperature	°C	Numeric	30	50
X10	MgSO4·7H2O	g/L	Numeric	0	3

).

The higher (+1) and lower (−1) variation levels of the factors were defined taking into account the experimental limits of development and keratinase production by strain ES41 when using the OFAT (One-Factor-at-A-Time) method and spaced to effectively determine the factors affecting significantly the response (Table 1) [40,47].

The Minitab 19 statistical software (trial version) used to elaborate the Plackett–Burman trials was constructed and organized according to the chosen selection plan. The ten-factor, two-level PBD method comprised 20 randomized experiments (

Table 2. The PBD trial plan layout and associated responses for keratinase activity of strain ES41.

Run	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	Factor 6	Factor 7	Factor 8	Factor 9	Factor 10	Response
	X1	X2	X3	X4	X5	X6	X7	X8	X9	X10	R
	Sodium Chloride	Inoculum Size	Incubation Time	InitialpH	Dipotassium Hydrogeno Phosphate	Orbital Agitation	Chicken Feather Meal	Calcium Carbonate	Incubation Temperature	Magnesium Sulphate Hyptahydrate	Keratinase Activity
	g/L	Spores/mL	Days		g/L	rpm	%(m/v)	g/L	°C	g/L	U/mL
1	0	1.5 × 1010	2	12	3	250	2	0	30	3	180.1 ± 5.08
2	3	1.5 × 104	2	12	3	0	2	4	30	0	142.3 ± 4.89
3	0	1.5 × 104	14	12	0	250	2	0	30	0	150.6 ± 3.91
4	3	1.5 × 1010	14	4	0	250	2	0	50	3	71.3 ± 4.47
5	0	1.5 × 104	2	12	0	250	0.2	4	50	3	77.2 ± 2.50
6	0	1.5 × 1010	14	4	0	0	0.2	4	30	3	59.7 ± 1.63
7	3	1.5 × 1010	2	4	0	0	2	0	50	0	110.8 ± 2.72
8	0	1.5 × 1010	14	4	3	250	0.2	0	30	0	62.3 ± 1.55
9	3	1.5 × 104	2	4	0	250	0.2	4	30	3	69.2 ± 3.36
10	0	1.5 × 104	2	4	0	0	0.2	0	30	0	51.4 ± 3.07
11	3	1.5 × 1010	14	12	0	0	2	4	30	3	162.6 ± 5.90
12	3	1.5 × 104	14	12	3	250	0.2	0	50	3	51.4 ± 1.16
13	0	1.5 × 1010	2	12	0	250	2	4	50	0	143.8 ± 5.30
14	3	1.5 × 104	14	12	3	0	0.2	0	50	0	47.4 ± 1.19
15	3	1.5 × 104	14	4	3	250	2	4	30	0	92.6 ± 3.29
16	3	1.5 × 1010	2	4	3	250	0.2	4	50	0	53.9 ± 5.18
17	0	1.5 × 104	2	4	3	0	2	0	50	3	87.1 ± 4.09
18	0	1.5 × 1010	14	12	3	0	0.2	4	50	0	59.6 ± 3.51
19	3	1.5 × 1010	2	12	3	0	0.2	0	30	3	133.2 ± 5.49
20	0	1.5 × 104	14	4	3	0	2	4	50	3	43.9 ± 4.17

).

The trials were carried out in three replicates. The experimental response average obtained wasevaluated by a first-order polynomial model using the following Formula (2):

$$R_{(U/mL)}={\beta }_0+{\displaystyle \sum _{i=1}^k{\beta }_i{X}_i}$$

(2)

where R is the response; β₀ is the regression coefficient, β_i is the linear coefficient, and X_i is the level of the independent variable.

The set of results was evaluated by analyzing the variance (ANOVA). The factor’s impact on keratinase production was estimated using the probability value ‘p’ for each factor. An effect is statistically significant when the p-value indicates a value less than 0.05 [48].

2.9. Determination of Optimal Physicochemical Parameters Using a Response Surface Model Based on the Box-Behnken Design

Based on the PBD experimental results, keratinase yield in strain ES41 was optimized by applying the response surface methodology (RSM). In the current investigation, the Box–Behnken (BBD) model given by the Design Expert 13^® software (version 13.0.12.0, Statease, Minneapolis, MN, USA—trial version) was chosen to guide the experimental design with five variables selected from those tested by the two-stage PBD experimental design. The selected factors were chicken feather meal: A, initial pH: B, incubation temperature: C, incubation time: D, and inoculum size: E. The parameters were defined at 3 levels: low (−1), medium (0), and high (+1), while keratinase activity was defined as the response (

Table 3. Operating parameters and their level variations in the Box-Behnken design (BBD).

Study Type	Response Surface		Subtype		Randomized
Design Type	Box-Behnken		Runs 46
Design Mode	Quadratic		No Blocks		Levels
Factor	Name	Units	Type	low (−1)	high (+1)	Medium (0)
A	Chicken feather meal	g/L	Numeric	2	8	5
B	Incubation time	Days	Numeric	4	8	6
C	Initial pH		Numeric	5	9	7
D	Incubation temperature	°C	Numeric	35	45	40
E	Inoculum size	Spores/mL	Numeric	1.00 × 106	1.00 × 108	5.05 × 107
Response	Name	Unit	Observations	Analysis Polynomial
R	Keratinase activity	U/mL	46

) [49].

The five-factor, three-level BBD method comprised 46 randomized experiments with triplicates at the central point (Table 3). The required experimental number (N) was defined according to Equation (3):

$$N=2k\cdot (k-1)+C_0$$

(3)

where N represents the number of experiments; k and C₀ are the numbers of factors and central points of the experiments (6), respectively.

The developed and organized 46-trial matrix was executed according to the chosen optimization plan. The tests were carried out in triplicate (

Table 4. The BBD trial plan layout and associated responses for keratinase activity of strain ES41.

	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	Response
Run	A: Chicken Feather Meal g/L	B: Incubation Time Days	C: Initial pH	D: Incubation Temperature °C	E: Inoculum Size Spores/mL	R: Keratinase Activity U/mL
1	2	6	9	40	5.05 × 107	268.12 ± 2.5
2	5	6	7	40	5.05 × 107	451.70 ± 0.4
3	2	4	7	40	5.05 × 107	289.53 ± 1.9
4	5	6	7	40	5.05 × 107	448.69 ± 1.9
5	2	6	7	40	1.00 × 106	287.20 ± 3.2
6	5	8	7	40	1.00 × 106	324.47 ± 0.3
7	5	6	9	40	1.00 × 108	214.17 ± 4.6
8	5	6	7	35	1.00 × 106	267.00 ± 1.7
9	5	4	7	35	5.05 × 107	225.00 ± 0.8
10	5	6	5	35	5.05 × 107	263.92 ± 1.7
11	5	6	7	40	5.05 × 107	448.23 ± 2.7
12	8	6	7	40	1.00 × 108	318.00 ± 6.3
13	8	6	7	45	5.05 × 107	204.00 ± 1.8
14	5	8	7	45	5.05 × 107	179.00 ± 4.6
15	8	6	9	40	5.05 × 107	259.00 ± 2.3
16	5	6	5	40	1.00 × 106	295.78 ± 5.1
17	8	6	5	40	5.05 × 107	363.52 ± 0.4
18	5	8	7	40	1.00 × 108	322.25 ± 0.7
19	8	6	7	40	1.00 × 106	353.00 ± 1.6
20	5	4	9	40	5.05 × 107	251.54 ± 1.4
21	5	6	7	35	1.00 × 108	251.00 ± 2.7
22	5	6	7	40	5.05 × 107	449.88 ± 0.8
23	8	6	7	35	5.05 × 107	326.73 ± 4.9
24	5	6	7	40	5.05 × 107	450.32 ± 3.1
25	2	6	7	40	1.00 × 108	230.00 ± 4.3
26	5	4	7	45	5.05 × 107	376.26 ± 5.6
27	8	4	7	40	5.05 × 107	423.00 ± 1.5
28	5	6	7	40	5.05 × 107	455.00 ± 0.8
29	5	6	9	35	5.05 × 107	174.40 ± 1.7
30	5	6	5	45	5.05 × 107	184.00 ± 2.2
31	5	6	9	45	5.05 × 107	233.12 ± 0.9
32	5	6	7	45	1.00 × 108	177.15 ± 6.2
33	2	6	5	40	5.05 × 107	204.00 ± 0.7
34	8	8	7	40	5.05 × 107	326.23 ± 2.4
35	5	4	7	40	1.00 × 108	327.28 ± 3.4
36	5	8	5	40	5.05 × 107	234.00 ± 0.6
37	5	6	9	40	1.00 × 106	322.98 ± 3.6
38	5	8	7	35	5.05 × 107	342.81 ± 2.1
39	2	6	7	45	5.05 × 107	245.64 ± 3.8
40	5	4	5	40	5.05 × 107	421.00 ± 1.2
41	2	6	7	35	5.05 × 107	147.98 ± 1.8
42	5	8	9	40	5.05 × 107	355.00 ± 4.1
43	2	8	7	40	5.05 × 107	312.00 ± 0.7
44	5	4	7	40	1.00 × 106	458.00 ± 1.3
45	5	6	7	45	1.00 × 106	288.00 ± 2.3
46	5	6	5	40	1.00 × 108	295.43 ± 0.8

).

The Design Expert 13^® software, having provided the Box-Behnken experimental design for optimized keratinase production in strain ES41, was used to analyze the obtained results, validate the statistical model, and predict the optimal operating parameters for enzyme production.

Keratinase production was assessed by multiple regression treatment of the data via examination of the response surfaces and ANOVA. Examination of the response surfaces traced by varying the values of two factors while maintaining those of the other factors constant at level (0) enabled us to detect, the interactions existing between the various factors by highlighting linear, quadratic, and interaction effects, while the ANOVA allowed us to determine the adequacy of the model and thus develop a second-order polynomial equation (quadratic model) according to the general Equation (4) used for predicting the ideal requirements for keratinase synthesis.

$$\begin{array}{c}{\mathrm{R}=\mathsf{\beta }}_0{+\mathsf{\beta }}_{\mathrm{A}}\mathrm{A}+{\mathsf{\beta }}_{\mathrm{B}}{\mathrm{B}+\mathsf{\beta }}_{\mathrm{C}}{\mathrm{C}+\mathsf{\beta }}_{\mathrm{D}}{\mathrm{D}+\mathsf{\beta }}_{\mathrm{E}}{\mathrm{E}+}_{{\mathsf{\beta }}_{\mathrm{AB}}}{\mathrm{AB}+\mathsf{\beta }}_{\mathrm{AC}}\mathrm{AC}+ {\mathsf{\beta }}_{\mathrm{AD}}{\mathrm{AD}+\mathsf{\beta }}_{\mathrm{AE}}{\mathrm{AE}+\mathsf{\beta }}_{\mathrm{BC}}{\mathrm{BC}+\mathsf{\beta }}_{\mathrm{BD}}\mathrm{BD}\\ {+\mathsf{\beta }}_{\mathrm{BE}}{\mathrm{BE}+\mathsf{\beta }}_{\mathrm{CD}}{\mathrm{CD}+\mathsf{\beta }}_{\mathrm{CE}}{\mathrm{CE}+\mathsf{\beta }}_{\mathrm{DE}}{\mathrm{DE}+\mathsf{\beta }}_{\mathrm{AA}}{\mathrm{A}}^2{+\mathsf{\beta }}_{\mathrm{BB}}{\mathrm{B}}^2{+\mathsf{\beta }}_{\mathrm{CC}}{\mathrm{C}}^2{+\mathsf{\beta }}_{\mathrm{DD}}{\mathrm{D}}^2{+\mathsf{\beta }}_{\mathrm{EE}}{\mathrm{E}}^2+\mathsf{\epsilon }\end{array}$$

(4)

Rdesignates response surfaces (keratinase activity); β₀ is the steady term (y-intercept), A, B, C, D, and E represent independent variables; β_A, β_B, β_C,β_D, and β_E represent linearcoefficients; β_AA, β_BB, β_CC, β_DD, and β_EE represent quadratic coefficients; β_AB, β_AC, β_AD, β_AE, β_BC, β_BD, β_BE, β_CD, β_CE, and β_DE represent the interaction coefficients; and ε is a random error component that represents other sources of response variability notaccounted for in the model, including effects such as measurement error on the response, inherent system variation such as instrumental background, and the effects of unstudied variables [50,51]. To assess the statistical validity of the model used, it is necessary to evaluate the coefficient of regression (R²), which indicates the quality of the fit, and the adjusted coefficient of regression (R²adj), representing the proportion of variance described by the model. The closer the value of the coefficient of determination (R²) is to 1, the better the model [51].

2.10. Experimental Model Validation

The relevance of the statistical model was verified by performing predicted experiments generated by the Design Expert 13.0.12.0 software by analysis of the quadratic model under optimal conditions and the determination of the appropriate values of the parameters involved in the production of keratinase by strain ES41 [48]. The predicted experiments were verified by applying two solutions proposed by the model, comprising predicted optimal values of the selected variables and predicted responses under these conditions.

3. Results and Discussion

A total number of 35 strains were obtained on a solid medium composed of chicken feather meal, mainly based on the morphological features of actinobacteria and also the appearance of hydrolysis zones around the colonies. After the selection of pure colonies, a second selection in a submerged medium was carried out by quantitative estimation of keratinase production. Through this selection, five strains were preserved for subsequent studies, particularly for their ability to produce keratinase in a liquid medium.

3.1. Isolation of Keratinase-Producing Actinobacterial Strains

Selective isolation of actinobacterial strains on solid FBM culture medium yielded 35 distinct bacterial strains with morphological characteristics consistent with actinobacteria. Figure 2a,b showsthe proportions of isolates obtained on agar medium by sampling site and type of sample, respectively. Of the strains selected, 15 strains representing 43% were isolated from the Djebira (Boukhelifa) sampling site, compared with 13 and 7 strains comprising 38% and 19% from the Fenaia-Ilmathen and Ighil Ali sampling sites, respectively (Figure 2a). When evaluated by sampling type, chicken feather amended soil (CFS) yielded the highest number of isolates (68%, 58%, and 53%), followed by chicken compost manure (CCM) (19%, 27%, and 33%) and decomposing feathers (DF) (13%, 15%, and 14%) for the three sampling sites, respectively (Figure 2b).

3.1.1. Pre-Screening on FBM Solid Medium

Based on the existence of pronounced growth of colonies with the appearance or absence of clear zones of hydrolysis around them due to the secretion and diffusion of keratinases produced by the isolates identified on the selection medium, the isolation on solid medium made it possible to select five putative keratinase-producing isolates (Figure 3) [52].

Actinobacteria are largely found in nature, especially in soil. They constitute a notable proportion of the telluric microbial flora [53]. The organic-rich soil samples containing chicken feathers as a potential carbon source promote the growth of actinobacterial strains with keratinous material’s ability degradation. A large number of actinbacterial genera/species isolated from different soil sites have been reported as keratinase producers, mainly by the hydrolysis of several keratin materials, citing wool, feathers, and hair. The production ofkeratinolytic enzymes is considered to be inductive; their synthesis takes place mainly in response to the presence of a keratin substrate [30,54].

In this perspective, exploring diverse sources of keratinous waste, including by-products from animal slaughterhouses (such as horns, hooves, and hides), as well as waste and effluents from textile industries and tanneries, is crucial. This investigation aims to isolate microorganisms with keratinolytic capabilities and to develop methods for the treatment and valorization of these wastes.

3.1.2. Screening in FBM Submerged Medium

The results of the submerged screening performed to quantify the keratinase activity of the five most promising isolates are shown in Figure 4. Strain ES41 (Figure 3a) isolated from Djebira (Boukhelifa) chicken compost manure showed the highest enzymatic activity at 22.4 ± 0.5 U/mL. It was followed by strain EP41 (Figure 3b) from the same site and sample type as the preceding strain, with a recorded keratinase activity of around 18.3 ± 0.2 U/mL.Then followed strains ES31 and EP33 (Figure 3c,d) isolated from Fenaia-Ilmathen chicken compost manure and decomposing feathers, respectively, which successively achieved enzyme yields of 17.7 ± 0.2 and 13.8 ± 0.3 U/mL. Finally, strain EP22 (Figure 3e) isolated from Ighil Ali chicken feather amended soil showed a keratinase production of 10.3 ± 0.3 U/mL.

Statisticalanalysis using t-tests indicated significant differences in keratinase activity among the isolates. For instance, ES41 exhibitedsignificantlyhigher keratinase activity comparedto: EP41 (p = 0.03), ES31 (p = 0.01), EP33 (p = 0.01), and EP22 (p < 0.001), whereasstrains EP41 and ES31 show a non-significant difference in enzyme yields (p > 0.05), as shown in Figure 4.

Bacteria and fungi are regularly cited in the literature as producers of keratinase. Most fungal producers like Trichophyton and Microsporum and Gram-negative bacteria, citing Vibrio sp. strain kr2, Citrobacterdiversus, and Pseudomonas aeruginosa 4-3, have limited applications due to a certain degree of pathogenicity [6,55,56,57,58].

Gram-positive bacteria, represented mainly by the genus Bacillus and Actinobacteria, are the most recommended for the degradation of keratin materials. Keratinase producers from the Bacillus genus include Bacillus licheniformis PWD-1, Bacillus subtilis, Bacillus amyloliticus, Bacillus cereus, and Bacillus thuringiensis, while the main actinobacterial keratinase producers belong to the Streptomyces genus, including Streptomyces albicans and Streptomyces fradiae [59,60]. Compared to a few applications mainly from B. licheniformis strains and based on previous research, actinobacterial species have not been widely applied for keratin waste bioconversion. Actinobacteria offer advantages when used in keratinase production, as they present broad physiological tolerances, the capacity and adaptability to grow on different types of biopolymeric substrates, particularly on a wide range of keratin substrates, and the ability to synthesize a wide range of bioactive substances such as keratinases. Moreover, the metabolic variability of actinobacteria enables them to be used in both submerged and solid-state fermentation, depending on their need for water potential. Actinobacteria thus offer potential advantages in terms of their multiple industrial uses [61]. For this reason, this research aimed to explore the ability of actinobacteria to produce keratinases and the possibilities of applying them in several industrial fields.

3.2. Identification and Classification of the Keratinase-Producing Strain ES41

3.2.1. Morphological Characterization of Strain ES41

The macromorphological and micromorphological characteristics of strain ES41 were found to be important for the identification and classification of the microorganism. Colonies of strain ES41 grown on various agar media at 45 °C for 24 to 48 h were circular with a powdery and cottony appearance on Williams and International Streptomyces Project 2 (ISP-2) media (Figure 5a,b) and a powdery and rough appearance on glucose yeast extract agar (GYEA) medium (Figure 5c), with a diameter ranging from 2 to 10 mm.

The isolate exhibited abundant growth and sporulation on Williams medium. Growth was relatively average on GYEA medium but weak on the ISP-2 medium. The aerial mycelium was white, stable, and not fragmented. The substrate mycelium presented various colors depending on the medium used, ranging from yellow on Williams medium to dark brown on GYEA medium and light brown on ISP-2. Diffusible pigments were not produced on any of the test media.

Spore chain arrangements and Gram type determination were observed using a Euromex optical microscope (at 100× magnification). The aerial mycelium takes the form of coiled and entangled thick filaments bearing chains of short spores arranged in primary or secondary whorls. The verticilliums are made up of short sporophores that emerge from a common point and carry the spore chains (Figure 5d). ES41 strain’s Gram staining showed that it is Gram-positive (Figure 5e). All these characteristics are in favor of assigning the isolate to the genus Streptomyces [42].

3.2.2. Molecular Typing of ES41 Strain

To identify and classify the ES41 strain, the 16S rRNA gene was amplified and sequenced using PCR and the Sanger method. Alignment of the obtained sequence with those of the NCBI (National Center for Biotechnology Information) database from the BLASTn (Basic Local Alignment Search Tool) program showed that ES41 belongs to the genus Streptomyces (99.66% identity).

The submission of this sequence (1478 bp) to the GenBank database with the name Streptomyces sp. DZ 06 had been assigned the following accession number: OQ195253.1 [43]. A phylogenetic tree based on partial 16S rRNA sequence is presented in Figure 6, which illustrates the relationship between Streptomyces sp. strain DZ 06 (ES41) and other strains within the same as well as related actinobacteria [45]. The 16S rRNA sequences of Bacillus cabrialesii strain NOK82 (ON 287158.1), Brevibacillusagri strain IHB B 1387 (GU186123.1), and Escherichia coli strain JMC 1649 (LC069032.1) were utilized as out-groups for neighborhood joining.

3.3. Screening of Critical Factors Affecting Keratinase Production via the Plackett-Burman Design

The production of keratinase from bacterial genus/species has often been optimized using a single-step statistical experimental design and response surface methodology. Among these investigations were enzymatic bioconversion of feather waste with keratinases of Bacillus cereus PCM 2849 [1], statistical optimization of keratinase production by Bacillus cereus GJBBR [62], enhanced production, purification, and characterization of alkaline keratinase from Streptomyces minutiscleroticus DNA38 [63], keratinase production by Bacillus pumilus GHD in solid-state fermentation using sugar cane bagasse [64], plus response surface methodology optimization of keratinase production from alkali-treated feather waste and horn meal using Bacillus sp. MG-MASC-BT [65]. However, few studies have been reported using a complete statistical experimental process, including statistical methods for screening the various influencing factors and optimizing their variation levels, as reported in this study.

In the present study, the determination of keratinase production influencing factors by isolating Streptomyces sp. strain DZ 06 (ES41) was carried out by practical testing of the PBD plan generated by the statistical software Minitab 19 (trial version). The results of the obtained PBD experiments are presented in Table 2 and Figure 7 and Figure 8. Table 2 shows the PBD design matrix provided for screening ten two-level independent variables representing the initial production conditions and corresponding responses. The collected response data showed variable keratinase activities ranging from 43.99 U/mL (minimum) to 180.13 U/mL (maximum). The different combinations of high and low levels of the various influencing factors are at the origin of the variations in response [66]. This wide variation reflected the eminence of optimizing production conditions in order to reach high levels in terms of enzyme production. The maximum production of keratinase was obtained in the 1st series (180.13 U/mL), with inoculum size (X₂), initial pH (X₄), K₂HPO₄(X₅), orbital agitation (X₆), chicken feather meal (X₇), and MgSO₄·7H₂O (X₁₀) present at high levels. While NaCl (X₁), incubation time (X₃), CaCO₃ (X₈), and incubation temperature (X₉) were present at low levels (Table 2).

The adequacy of the statistical model was verified by highlighting the effects of the variables examined via Fisher’s F test and analysis of variance. For p-values < 0.05, factors were considered to have a significant effect on response [67]. The ANOVA and the p-value of the model and for each parameter are presented in

Table 5. ANOVA for principal effect model used for the identification of key factors influencing keratinase production by Streptomyces sp. Strain DZ 06 (ES41).

Source	DF	Adjusted Sum of Squares	Adjusted Mean Square	F-Value	p-Value
Regression	8	28,305.9	3538.23	52.45	0.000
NaCl	1	211.5	211.48	3.13	0.137
Inoculum size	1	2037.9	2037.93	30.21	0.003
Incubation time	1	1141.4	1141.42	16.92	0.009
Initial pH	1	4925.3	4925.29	73.01	0.000
K2HPO4	1	218.2	218.24	3.23	0.132
Orbital agitation	1	327.5	327.5	4.85	0.079
Chicken feather meal	1	6903.8	6903.8	102.34	0.000
Incubation temperature	1	3497.5	3497.52	51.84	0.001
Error	5	337.3	67.46
Total	13	28,643.2
R2		98.82%
Adjusted R2		96.94%
Predicted R2		91.12%

DF = the degree of freedom; F-Value = the value of F obtained by carrying out an F test. Prob (P) > F = the value of p, indicating the probability and applying it to significance.

.

Among the variables studied and in decreasing order of influence, the factors identified, chicken feather meal (p = 0.000), initial pH (p = 0.000), incubation temperature (p = 0.001), inoculum size (p = 0.003), and incubation time (p = 0.009), showed highly significant effects on keratinase production with p values < 0.01. The remaining variables showing p values > 0.05 are considered non-influential factors. This suggests that low concentrations of these factors are sufficient for keratinase production within the strain studied [66].

At the same time, the Pareto chart of the effects of the independent variables revealed the factors influencing the response represented by horizontal bluebars exceeding the red line representing the significance level, thus confirming the results obtained from the analysis of the statistical model.(Figure 7) [68].

Figure 8 shows the main plots of the effect of the significant factors on the response and confirms the results reported in Table 5 and Figure 7. Increases in chicken feather meal concentrations, initial pH values, and inoculum size greatly influenced keratinase production, while increases in incubation temperature and incubation time values hurt keratinase activity detected [69].

The regression model “F” value (52.45) was found to be significant. The coefficient of determination R² provides an explanation of the model fit goodness, which is used to assess the explanatory power of regression models and reflects the variation in response in the proposed statistical model. [70].

The regression equation acquired showed an R² value of 0.9882, indicating that 98.82% of the total variation detected for the responses could be interpreted by the model, demonstrating that the design is highly significant in predicting the factors effects on keratinase activity by the studied strain. An R² value > 0.75 indicates fit to the biological models [71]. The correlation of the predicted obtained responses in the present study is explained by the closeness of the R², adjusted R², and predicted R² values: 98.82%, 96.94%, and 91.12%, respectively (Table 5). The design performance was valuated by a first-order model analysis, showing its fitting to the experimental data using the following Equation (5):

$$\mathrm{R}=52.07+29.29{\mathrm{X}}_2-18.91{\mathrm{X}}_3+45.53{\mathrm{X}}_4+50.69{\mathrm{X}}_7-37.85{\mathrm{X}}_9$$

(5)

The PBD results showed that the factors, chicken feather meal (% w/v), initial pH, incubation temperature (°C), incubation time (days), and inoculum size (spores/v), exert significant effects on keratinase production, with chicken feather meal concentration forming a major contributor. This result is identical to those obtained by Abdul Gafar et al. and Laba et al. when PBD was utilized in the screening procedure.It was shown that the concentration of chicken feather meal was the most important influencing factor for keratinase synthesis by Bacillus sp. UPM-AAG1 and the actinobacterium strain Kocuriarhizophila p3-3, respectively [67,72]. In work undertaken by Manivasagan et al. and Demir et al., the same result was obtained where the yield in keratinase production was observed when chicken feather meal was employed as culture substrate during submerged fermentation using Actinoalloteichus sp. MA-32 and Streptomyces sp. 2M21, respectively [73,74]. The use of chicken feather meal as an organic material source shows the ability of the strain Streptomyces sp. DZ 06 (ES41) to grow and to obtain its carbon and nitrogen requirements directly from this substrate. Furthermore, variations in the nutritional needs of each feather-degrading microorganism as well as the type of the keratinolytic proteases generated by the producer microorganisms may be the primary causes of the disparities in the role of the substrate in keratinase production among various feather-degrading bacteria, e.g., which carbon and/or nitrogen sources would act as inducers [75,76].

Similar outcomes to those obtained in the present research showing the significant effect of the initial culture pH during optimization work appeared in the literature. Among these works are those carried out by Fakhfakh–Zouari et al. using Bacillus pumilus A1 and Abd El-Aziz et al. on Streptomyces swerraensis KN23 [75,77]. The pH of the medium culture affects the microbial growth, effectiveness of feather degradation, and keratinase synthesis via influencing the reaction environment, enzymatic process, and movement of nutrients across the cell membrane of bacteria [1,20].

Apart from this, the important role of incubation temperature in the production of keratinases is revealed during this work and in those conducted by Demir et al. on Streptomyces sp. 2M21 and by Matikevičienė et al. using Actinomyces fradiae 119 [74,78]. The importance of incubation temperature is crucial for best enzyme production due to modifications in the structure and characteristics of microbial proteins with changes in temperature. Metabolic activities are reduced at temperatures below or above the optimal temperature, resulting in inhibition of growth and enzyme synthesis [79].

The role of incubation time as a determinant factor for keratinase production obtained during this work was similar to studies on keratinase production by Streptomyces sp. 2M21 and Streptomyces swerraensis KN23 [74,77]. Incubation time is an important parameter for keratinase production. It varies according to the microorganism, nature of the substrate used, and the production medium conditions [80].

Similarly, it was reported that the inoculum size used had a significant effect on keratinase production by Amycolatopsis sp. strain MBRL 40, A. fradiae 119, and B. licheniformis ALW1 [40,78,81]. The inoculum size is the initial bacterial mass required to carry out a fermentation, hence the need to test the selection of this factor and determining its variation levels. Keratinolytic activity is often increased with increasing inoculum size. No discernible increase or even decrease in activity is seen above the necessary initial rate [82].The selected parameters influencing keratinase production in Streptomyces sp. DZ 06 (ES41) are factors usually considered to influence most microorganisms when testing the production of different types of bioactive substances, in particular keratinases; this suggests that the strain studied has no particular requirements and that testing these factors for optimization via response surface methodology using the Box–Behnken design would enable keratinase production to be better controlled in terms of both quantitative and qualitative yields.

3.4. Box-Behnken Examination of Keratinase Synthesis-Based on RSM Design

Plackett–Burman Design, as indicated above, made it possible to select the main parameters influencing keratinase production by Streptomyces sp. strain DZ 06 (ES41). These independent variables, namely chicken feather meal, initial pH, incubation temperature, incubation time, and inoculum size, were evaluated at three levels (−1, 0, +1) to investigate their interaction and their effects on keratinase production applying a Box–Behnken design of experiment. For growth and keratinase production by Streptomyces sp. strain DZ 06 (ES41), the insignificant factors, including mineral salts and orbital agitation, were used at low levels during the enzyme production process [74].

The matrix experiments and results of the different trials performed by the Box–Behnken design are shown in Table 4. The highest keratinase production of 458 U/mL was obtained with 5 g/L of chicken feather meal (A), after 4 days of incubation time (B), at initial pH 7 (C), at incubation temperature of 40 °C (D), and an inoculum size of 1.00 × 10⁶ spores/mL (run 44). In addition, that corresponding to the central points of the factor values tests (run: 2, 4, 11, 22, 24, and 28) showed the closeness and the repeatability of the responses obtained (448.23 ± 1.7 to 450.32 ± 0.4) (U/mL) and consequently the relevance of the statistical model used (Table 4). The performance and adequacy of the quadratic model were verified through variance analysis (ANOVA) using Design Expert 13^® software (

Table 6. Determined regression coefficients for the analysis of variance (ANOVA) and the quadratic polynomial model.

Source	Sum of Squares	Df *	Mean Square	F-Value	p-Value
Model	3.423 × 105	20	17,116.77	10.68	<0.0001	significant
A-Poultry feather meal	22,724.06	1	22,724.06	14.17	0.0009
B-Incubation time	7503.02	1	7503.02	4.68	0.0403
C-Initial pH	2500.00	1	2500.00	1.56	0.2233
D-Incubation temperature	688.28	1	688.28	0.4293	0.5183
E-Inoculum size	8778.75	1	8778.75	5.48	0.0276
AB	4176.39	1	4176.39	2.61	0.1191
AC	6290.08	1	6290.08	3.92	0.0587
AD	12,135.23	1	12,135.23	7.57	0.0109
AE	17.10	1	17.10	0.0107	0.9186
BC	22,126.56	1	22,126.56	13.80	0.0010
BD	24,818.85	1	24,818.85	15.48	0.0006
BE	1973.58	1	1973.58	1.23	0.2778
CD	4802.49	1	4802.49	3.00	0.0958
CE	2914.92	1	2914.92	1.82	0.1896
DE	4088.32	1	4088.32	2.55	0.1229
A²	57,800.81	1	57,800.81	36.05	<0.0001
B²	9623.53	1	9623.53	6.00	0.0216
C²	82,397.69	1	82,397.69	51.40	<0.0001
D²	1.649 × 105	1	1.649 × 105	102.85	<0.0001
E²	43,849.98	1	43,849.98	27.35	<0.0001
Residual	40,079.96	25	1603.20
Lack of Fit	39,980.57	20	1999.03	4.57	0.1781	not significant
Pure Error	99.38	5	19.88
Cor Total	3.824 × 105	45
R2	0.8952
Adjusted R2	0.8113
C.V. %	13.15
Adequate precision	11.2756

* Degree of freedom.

).

In the present study, the larger F-value of 10.68 and the low p-value < 0.001 indicate greater significance of model terms [83]. For p-values less than 0.05, the significant terms of the model are as follows: A (chicken feather meal), B (incubation time), E (inoculum size), AD (chicken feather meal vs. incubation temperature), BC (incubation time vs. initial pH), BD (incubation time vs. incubation temperature), A² (chicken feather meal²), B² (incubation time²), C² (pH²), D² (incubation temperature²), and E² (inoculum size²) (Table 6). On the other hand, the p-values were inferior to 5% for the variables C (initial pH), D (incubation temperature), and the interactions AB (chicken feather meal vs. incubation time), AC (chicken feather meal vs. initial pH), AE (chicken feather meal vs. inoculum size), BE (incubation time vs. inoculum size), CD (initial pH vs. incubation temperature), CE (initial pH vs. inoculum size), and DE (incubation temperature vs. inoculum size), showing that these factors and interactions are not significant and therefore may be excluded from the regression model. The lack of fit of a regression model of 4.57 compared to the random pure error with a value of 99.38 shows that the regression model lack-of-fit is notably lower than a random pure error (p-value > 0.05) (Table 6), indicating that the regression model is pertinent [67]. According to Bezerra et al., in this particular context, a model that exhibits both a substantial regression and a non-significant lack of fit would be well matched to the experimental data [84].

In the present study, the coefficient of determination R² value of 0.8952 reported in Table 6, reveals that the model is acceptable for predicting response values from experimental data. The R² value greater than 0.75 is sufficient for the model to explain the majority of variations in responses [85]. The model accuracy can be estimated by the signal-to-noise ratio (adequate precision), which should be superior to 4. The value of the adequate precision calculated is 11.2756 (Table 6), indicating the highest precision and an adequate response ratio. This adequate precision value and that of the coefficient of variation (CV %; 13.15%) implies that the statistical model is valid and reproducible and can be used to provide navigation in the conception data space. [86,87].

Figure 9 clearly shows that the predicted and actual values for keratinase production are evenly distributed close to the straight line, indicating an ideal match between these values. The model is well adjusted and quite realistic; even a small difference between certain real and predicted values should be noted. Consequently, this indicates that the response variables for the experimental data can be fully expected to be sufficiently predicted by the quadratic model chosen. [83]. It may be concluded from its statistical properties that this model is suitable for determining the primary, quadratic, and interaction effects of implied factors in keratinase production by Streptomyces sp. strain DZ 06 (ES41).

Based on the coded factors, the following is the second-order polynomial quadratic regression Equation (6) generated for keratinase production:

$$\begin{array}{ll}\mathrm{Keratinase} \mathrm{Activity}& =+450.55+37.69\mathrm{A}-21.65\mathrm{B}-12.50\mathrm{C}-6.56\mathrm{D}-23.42\mathrm{E}-32.31\mathrm{AB}-39.66\mathrm{AC}\\ & – 55.08\mathrm{AD}+2.07\mathrm{AE}+74.38\mathrm{BC} - 78.77\mathrm{BD}+22.21\mathrm{BE}+34.65\mathrm{CD}-26.99\mathrm{CE}\\ & -31.97\mathrm{DE}-81.38{\mathrm{A}}^2-33.21{\mathrm{B}}^2-97.17{\mathrm{C}}^2-137.46{\mathrm{D}}^2-{70.88\mathrm{E}}^2\end{array}$$

(6)

where A: chicken feather meal (g/L), B: incubation time (days), C: initial pH, D: incubation temperature (°C), and E: inoculum size (spores/mL).

The positive sign preceding the equation terms indicates a synergistic effect, whereas the negative sign indicates an antagonistic effect. The primary, quadratic, and interaction impacts of factors on the production of keratinase enzyme are indicated by positive or negative values of linear, quadratic, and interaction coefficients, respectively [83]. The model based on the coded variable development is useful for identifying the most important factors having an impact on the response [88].

3.4.1. Interaction Analysis between Critical Parameters

Based on the regression analysis of the BBD, interaction plots and three-dimensional response surface plots (Figure 10a–f) were used to investigate the interplay of relevant factors and their impacts on response. The analysis of different interaction types obtained during this study was carried out by varying the values of each variable involved while maintaining the other factors at their optimal level.

The fluctuation in response (keratinase activity) and the significant p-value (0.0109) for the interaction between AD (chicken feather meal and incubation temperature) demonstrated the presence of positive interactions between these two variables (Figure 10a). Figure 10b depicts the effect of AD on keratinase production, which increased significantly (p < 0.05) when increasing the chicken feather meal quantity and the incubation temperature up to 5.6 g/L and 41 °C, respectively, then followed by a slight decrease with increasing levels of both parameters. On the other hand, keratinase production depended mainly on chicken feather meal quantity and incubation temperature as their quadratic effects, which were highly significant (p < 0.0001) and on the linear effect (p = 0.0009) of the first factor, confirming the single-factor experimental results (Table 6) and the influence in the metabolic rate of the interaction between these two parameters.

Previous investigations conducted by Siddharthan et al. and Ahmadpour [89,90], showed that the interaction of chicken feather meal and incubation temperature was significant in the optimization of keratinase production by Geobacillusthermodenitrificans PS41 and Bacillus cereus during submerged fermentation. However, Abdul Gafar et al. and Demir et al. [67,74] found that the effect of the interaction between chicken meal and incubation temperature was insignificant on keratinase activity when Bacillus sp. UPM-AAG1 and Streptomyces sp. 2M21 were cultivated in a chicken feather meal medium.

The positive interaction between chicken feather meal and incubation temperature (AD) was found to be evident in this study. In this context, Singh et al. [34] revealed that carbon is a critical component of the medium, as it provides energy to the microorganisms and is necessary for their development and the synthesis of primary and secondary metabolites. On the other hand, Revankar et al. [91] reported that incubation temperature is a critical feature in the production process, impacting in turn not only microbial development but also enzyme synthesis.

Figure 10c,d showsthe second interaction influencing the keratinase production yield. This interaction occurred between incubation time and initial pH (BC). Figure 10c clearly shows the interaction effects of the two factors on keratinase activity, as also shown in Table 6 (p = 0.0010). This combination maximizes keratinase activity detected with increasing incubation time and initial pH values of up to 7 days and 6, respectively. Beyond these values, a significant decrease in terms of enzymatic yield is recorded (Figure 10d). The positive influence of this interaction is consistent with the results presented in Table 6, showing a highly significant quadratic effect of initial pH (p < 0.0001) and significant quadratic and linear impacts of incubation time (p < 0.05) on the enzyme yield, confirming the single-factor trial results.

Khalil et al. andDhiva et al. [92,93] both reported that the interaction between incubation time and initial pH was significant for keratinase production by Pichia kudriavzevii and Bacillus sp. CBNRBT2. The first study was carried out on a medium consisting only of chicken feather meal, while in the second study, glucose was added. However, in work conducted by Abd El-Aziz et al. [77], a non-significant result for the incubation time-initial pH interaction was obtained when using the feather-degrading keratinase-producing Streptomyces swerraensis KN23 strain through submerged fermentation in cultivation medium with sucrose as an added carbon source.

In this study, the interaction between incubation time and initial pH (BC) was critical for keratinase production. Bhari et al. indicated that incubation time is an important parameter for metabolite production. Its effects may be explained by the significant changes in the rheological properties of the cultivation broth during the fermentation period. Keratinase, as a primary metabolite, is generated by actively developing cells using chicken plumage as cultural substrate. A continued decrease in keratinase production is then recorded. This may be assigned to nutrient depletion, accumulation of inhibitory byproducts in the medium, and catabolic repression of enzyme production. All of these adjustments have the potential to alter the medium pH, which is crucial for determining the best possible physiological function for bacterial cells as well as the passage of different nutrients through the cell membrane for the highest possible enzymatic production [83].

The third and last interaction affecting keratinase activity levels was between incubation time and incubation temperature (BD), as seen in Figure 10e,f. This combination presented a significant p-value of 0.0006 (Table 6), proving the interaction influence of the two factors as shown by Figure 10e. Keratinase production was improved for low values of incubation time and high values of incubation temperatures. Whereas for inverse values, the keratinase activity fell sharply (Figure 10f). The combination effects of the incubation time-incubation temperature pair on enzymatic production yield wereobviously observed via the high quadratic impact of temperature (p < 0.0001), as well as the quadratic and linear effects of incubation time with p < 0.05 (Table 6).

During work conducted by Demir et al., Dhiva et al., and Ahmadpour [74,90,94], it was shown that the interaction between incubation time and incubation temperature significantly affects keratinase activity observed for Streptomyces sp. 2M21, B. cereus, and Pseudomonas aeruginosa SU-1 on chicken feather meal medium using statistical optimization methods. However, during a study reported by Dhiva et al. [93], using Bacillus sp. CBNRBT2, it was found that the combination of incubation timeand incubation temperature was insignificant towards keratinase production.

The combination of incubation time and incubation temperature presented a significant effect on keratinase activity in the present work. As shown by Bhari et al., the essential role of incubation time depends on the changes that take place in the production medium during the fermentation period [83]. Moreover, as mentioned above, production temperature is a crucial factor that affects both microbial growth and enzyme synthesis, according to Revankar et al. [91].

Microbial degradation of chicken feathers is a complex process. In the present work, highlighting the different main, quadratic, and interaction effects of the various factors influencing keratinase production and the hydrolytic capacities against an insoluble keratin substrate, in this case chicken feather meal in Streptomyces sp. DZ 06, make it a potential candidate for use in various industrial fields, in particular the bioconversion of poultry industry waste.

3.4.2. Approval of Proposed Model

The response surface equations were approved by numerical optimization using Design Expert 13 software, using the desirability function on the one hand, and analysis of the disparity between the observed values and the predicted values resulting from the response regression on the other hand. This approach involves a desirability function scale between d = 0, suggesting that the response is totally unacceptable, and d = 1, suggesting that the response corresponds exactly to the desired value. The value progresses from 0 to 1 as the desirability of the corresponding response increases [83].

The RSM as an analysis model and the regression equation were approved by testing two solutions of the most suitable operating conditions proposed by the model. In order to anticipate and confirm the correctness of the mathematical model, a new position of the experiment was computed using the ideal conditions that had been established and the response predicted.

Table 7. The performed assays for the response surface model validation.

	Factors					Keratinase Activity
Solutions Number	Chicken Feather Meal (g/L)	Incubation Time (Days)	Initial pH	Incubation Temperature °C	Inoculum Size (Spores/mL)	Predicted Value (U/mL)	Experimental Value (U/mL)	Desirability
1	6.13	4.11	6.25	40.65	3.98 × 107	485.44	489.24	0.998
2	6.12	4.37	6.29	40.51	4.06 × 107	482.86	484.37	0.980

displays the system’s most appropriate alternative optimization outcomes for each parameter based on their chosen optimal values and the desirability value.

Two tests were conducted under identical circumstances to verify the predicted ideal parameters experimentally. Among the two solutions suggested by the model, the optimal conditions turned out to be a quantity of chicken feather meal of 6.13 g/L, an incubation time of 4.11 days, an initial pH of 6.25, an incubation temperature of 40.65 °C, and an inoculum size of 3.98 × 10⁷ spores/mL for an optimal value of 485.44 U/mL of keratinase activity as illustrated by the desirability ramp for numerical optimization of the five independent variables (Figure 11). The results showed a keratinase production of 489.24 U/mL with a desirability rate of 0.998. The great dependability and validation of the keratinase production model were strongly supported by the good agreement between the anticipated values and the repeated results.

Analysis of the quadratic model under optimal conditions and determination of the appropriate values for the parameters involved in keratinase production by Streptomyces sp. DZ 06 attests to the reproducibility, relevance, and validity of the statistical model used.

4. Conclusions

Keratinous waste, including poultry feathers and by-products from industrial slaughterhouses, can act as a carbon and nitrogen source to enhance and optimize the production of keratinases in fermentation media by bacterial isolates under various experimental conditions. This makes it possible to valorize a resource often considered a potential environmental pollutant into avalue-added product. In this study, the use of keratin-rich residues led to the selection of an actinobacterial strain referenced as Streptomyces strain DZ 06 (ES41).

The utilization of the Plackett–Burman and Box–Behnken designs for the selection and optimization of the physiological and nutritional factors of the culture medium for keratinase production showed the relevance of optimizing culture conditions using statistical tools. The Plackett–Burman screening plan resulted in the selection of five influencing parameters for a maximum activity of 180.1 U/mL. Optimization of the selected factors using response surface methodology (RSM) via Box–Behnken design resulted in an enzyme production of around 458 U/mL.Validation of the model by applying the solutions proposed by the statistical model resulted in optimum enzyme production of around 489.24 U/mL, close to the predicted value (485.44 U/mL), using chicken feather meal as substrate (6.13 g/L), at an incubation temperature of 40.65 °C, an initial pH of 6.25, and an inoculum size of 3.98 × 10⁷ spores/mL during 4.11 days of fermentation. The Plackett–Burman and Box–Behnken designs exploited in the present study allowed us to identify and investigate the cultural conditions that would ultimately result in a 21.67-fold increase in keratinase production compared to standard, non-optimized initial conditions.

The optimized parameters and characteristics of Streptomyces sp. strain DZ 06 (ES41), including optimal growth using an inexpensive substrate under conditions of near-neutral pH and mesophilic temperature, with a standard inoculum size for an ideal incubation time, can support sustainable growth and attractive keratinase yield without additional external requirements. These qualities make this bacterium an excellent candidate for industrial application using the optimal conditions predicted and validated in this study and reducing the overall costs of the enzyme production process.