Enhanced keratinase production by Bacillus subtilis amr using experimental optimization tools to obtain feather protein lysate for industrial applications

Abstract

The poultry industry produces millions of tons of feathers waste that can be transformed into valuable products through bioprocess. The study describes the enhanced keratinase and feather hydrolysate production by Bacillus subtilis AMR. The metabolism of each microorganism is unique, so optimization tools are essential to determine the best fermentation parameters to obtain the best process performance. The evaluation of different propagation media indicated the constitutive production of two keratinases of approximately 80 kDa. The combination of Mn²⁺, Ca²⁺, and Mg²⁺ at 0.5 mM improved the keratinolytic activity and feather degradation 1.5-fold, while Cu²⁺ inhibited the enzymatic activity completely. Replace yeast extract for sucrose increased the feather hydrolysate production three times. The best feather concentration for hydrolysate production was 1.5% with an inoculum of 10⁸ CFU/mL and incubation at 30 °C. None of the inorganic additional nitrogen sources tested increased hydrolysate production, although (NH₄)₂SO₄ and KNO₃ improved enzymatic activity. The optimization process improved keratinolytic activity from 205.4 to 418.7 U/mL, the protein concentration reached 10.1 mg/mL from an initial concentration of 3.9 mg/mL, and the feather degradation improved from 70 to 96%. This study characterized keratinase and feather hydrolysate production conditions offering valuable information for exploring and utilizing AMR keratinolytic strain for feather valorization.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13205-022-03153-y.

Introduction

Annually, the poultry industry produces approximately 115 million tons of chicken meat (FAOSTAT 2019). The chicken meat processing industry is developing rapidly worldwide (Tesfaye et al. 2017). This enormous consumption of chicken results in the generation of vast amounts of chicken feathers each year worldwide. Considering that 5–7% of chicken weight correspond to feather (Gessesse et al. 2003; Forgács et al. 2013; Goda et al. 2021), the average of feathers generated as a by-product is approximately 7 million tons annually, creating a severe solid-waste problem. Most of these feathers are discarded or burned as waste. At the same time, a small proportion is converted into feathers meal to feed livestock or used as fertilizer (Mazotto et al. 2017; Peng et al. 2019; Sobucki et al. 2019). Thus, sustainable disposal of feathers is a significant constraint for slaughterhouses and poultry farms (Chauhan and Devi 2020).

Nowadays, there is increasing interest worldwide in developing bioproducts using renewable and sustainable sources, primarily agricultural by-products and co-products (Tesfaye et al. 2017). Keratin-rich animal by-products such as feathers constitute the third most abundant renewable biomass present in nature after cellulose and chitin (Falco et al. 2019a, b). Since chicken feathers contain more than 85% of crude protein (Peng et al. 2019), they can be a raw material for developing several new bioproducts. Among the bioproducts based on feather hydrolysates it can be cited biofertilizer (Gurav and Jadhav 2013a; Nurdiawati et al. 2019; Gurav et al. 2020), feed supplement (Mazotto et al. 2017; Kumar 2021), bioplastic (Dou et al. 2016; Ramakrishnan et al. 2018), and cosmetics (Villa et al. 2013; Mokrejš et al. 2017).

The major protein that compounds feather is keratin, a protein characterized by the resistance to chemical and enzymatic hydrolysis related to the presence of a high degree of cross-linking by disulfide bridges, salt bridges, hydrogen bonds, and hydrophobic interactions (Uttangi and Aruna 2018). Keratin can be classified in α-keratin or β-keratin: α-keratin consists of α-helical-coil coils which are self-assembled into intermediate filaments, while β-keratin is rich in β-pleated (Ghaffar et al. 2018; Sharma and Devi 2018). Almost all keratinous material contains both types of keratins. Feathers have 41–67% of α-keratin, 33–38% of β-keratin, and a small fraction of amorphous keratin (Ghaffar et al. 2018).

Keratinous waste can be transformed into valuable products; however, the previous hydrolysis of rigid feather structure is necessary. Microbial keratinases offer an economical and eco-friendly alternative for degrading and recycling keratin materials (Goda et al. 2021; Moussa et al. 2021). Keratinase [E.C.3.4.21/24/99.11] are peptidases belonging predominantly to the classes of serine proteases or metalloproteases, able to hydrolysate keratin (Ghaffar et al. 2018; Sharma and Devi 2018). These enzymes have numerous applications in the industries of feed addictive, detergents, leather, pharmaceutical, and cosmetics (Ghaffar et al. 2018; Sypka et al. 2021).

Due to the highly rigid, strongly cross-linked structure of keratins, there are relatively few microorganisms keratinolytic described with the capability of using keratin as the only source of C, N, S, and energy (Uttangi and Aruna 2018). Nevertheless, keratinolytic microorganisms are ubiquitous and have been isolated from different environmental niches, including soil and animal gut and intestine (Daroit et al. 2009; Corrêa et al. 2010; Barman et al. 2017; Liu et al. 2017; Wang et al. 2019), but the most frequent isolation site is poultry waste (Lin et al. 1992; Mazotto et al. 2011; Babalola et al. 2020). Keratinase-producing microorganism includes bacteria from Bacillus genera as B. licheniformis PWD-1 (Lin et al. 1992), B. subtilis KD-N2 (Cai and Zheng 2009), B. cereus (Arokiyaraj et al. 2019), and others as Pseudomonas aeruginosa KS-1 (Sharma and Rani 2010), Arthrobacter sp. NFH5 (Barman et al. 2017), and Stenotrophomonas maltophilia BBE11-1 (Peng et al. 2019). Several pathogenic and saprophytic fungi have been described as keratinolytic, for instance, Trichophyton rubrum, Trichophyton mentagrophytes, Microsporum gypseum, Candida albicans, Aspergillus oryzae, Paecilomyces marquandii, Chrysosporium indicum (Sharma and Devi 2018).

The optimization of culture conditions and medium composition is essential in successfully producing an enzyme. Several variables as pH, temperature, carbon and nitrogen sources, media additives, and inoculum size should be evaluated to improve enzymatic production. The most conventional approach to screen the variables is the one-variable-at-a-time method (OVAT). This method is simpler to analyze the generated data but does not establish any combined interactions among various variables; beyond it is a very time-consuming and laborious process (Arokiyaraj et al. 2019; Chauhan and Devi 2020). Optimization of culture medium by statistical approaches, such as fractional experimental design and response surface methodology, is an effective method widely utilized in microbial enzyme production to significantly improve the yield and lower production costs (Silva et al. 2014; Cai and Zheng 2009; Chauhan and Devi 2020). Each microorganism is unique in its requirement for maximum growth and production of enzymes (Arokiyaraj et al. 2019). The lack of a common medium to produce a specific enzyme leads to the necessity of optimizing the culture medium parameters and composition for each microorganism.

The feather hydrolysate produced by Bacillus subtilis AMR previously demonstrated potential as a protein supplement in animal feed (Mazotto et al. 2017). Improving the productivity of this hydrolysate is attractive in industrial terms. Thus, the present study objective was to enhance the keratinase production of B. subtilis AMR using OVAT and factorial approaches to obtain feather keratin hydrolysates.

Material and methods

Materials and chemicals

White chicken feathers, obtained from local slaughterhouse waste, were washed extensively with tap water and neutral detergent, dried at 60 ºC overnight, delipidated with chloroform: methanol (1:1, v/v) for 1 h, dried again at 60 ºC overnight, and stored in plastic bags until use. Gelatin was obtained from Merck (Darmstadt, Germany). Reagents used in electrophoresis and molecular mass standards were acquired from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany) and BioRad (California, USA). All other reagents were of analytical grade.

Microorganism and cultivation

The Bacillus subtilis AMR strain used in this study was isolated as described in Mazotto et al. (2010). Medium YE agar (g/L: yeast extract 5, peptone 5, KCl 20, sucrose, 20, and agar 20) was used to maintain B. subtilis AMR. The medium for keratinase production and feather fermentation contained (g/L): MgSO₄·7H₂O 0.2, NaCl 0.5, CaCl₂·2H₂O 0.1, whole feather 10, sucrose 0.5, in 0.06 M Na₂HPO₄·7H₂O, and 0.04 M KH₂PO₄, pH 8. Cultivation was performed in 250 mL Erlenmeyer flasks containing 100 mL medium for up to 8 days at room temperature (28 ± 2 ºC) with constant shaking (120 rpm). As inoculum, 20% bacteria grown in YE broth for 72 h under the same conditions were used. The cells were washed twice (3000 rpm/ 20 min) with sterilized saline solution (NaCl 8.5 g/L) before inoculum in feather media. At regular time intervals, according to the experiment, the culture was interrupted for harvesting the residual feather and supernatant (3000 rpm/20 min) to determine the feather degradation, protein lysate concentration, and enzymatic activities.

Effects of medium composition for inoculum production

To evaluate if the medium used in inoculum preparation affect the fermentation productivity, the following media were tested: M1 (g/L: yeast extract 5, peptone 5, KCl 20, sucrose, 20); M2 (g/L: MgSO₄·7H₂O 0.2, NaCl 0.5, CaCl₂·2H₂O 0.1, whole feather 10, yeast extract 0.1, in 0.06 M Na₂HPO₄·7H₂O and 0.04 M KH₂PO₄, pH 8); and M3 (g/L: MgSO₄·7H₂O 0.2, NaCl 0.5, CaCl₂·2H₂O 0.1, whole feather 10, yeast extract 1, in 0.06 M Na₂HPO₄·7H₂O and 0.04 M KH₂PO₄, pH 8). B. subtilis AMR was inoculated in the three media and incubated for 72 h at 28 ± 2 ºC under constant agitation (120 rpm/min). At the end of the incubation period, the media were centrifugated (3000 rpm/20 min), and the supernatant was reserved to determine the enzymatic profile in different media. The cells were washed (twice at 3000 rpm/20 min), counted by the plate counting method (count medium g/L: glucose 20, peptone 10, yeast extract 5, agar 20), and a volume corresponding to an initial count of 10⁸ CFU/mL was inoculated into the fermentation medium. The fermentation process occurred for 8 days (120 rpm, 28 ± 2 ºC), and aliquots were harvested every 24 h intervals to determine protein concentrations and enzymatic activities.

Effect of metallic salts and additional carbon source on enzymatic activity and feather degradation

To evaluate the effect of ions in enzymatic activity, feather degradation, and hydrolysate production in submerged cultures, 2³ fractional factorial design (FFD) was used, as shown in Table 1. Two independent experiments were performed; in the first one was evaluated the effect of FeSO₄·7H₂O, CuSO₄·5H₂O e ZnSO₄·7H₂O at 0 (− 1), 0.5 (0) e 1 mM (+ 1) each, in the second one, the effect of MgSO₄.·7H₂O, CaCl₂·2H₂O e MnCl₂·4H₂O, at the same concentrations, was evaluated. After growth of strain AMR in YE broth for 72 h at (28 ± 2 ºC) with constant shaking (120 rpm), the cell was centrifugated, washed twice with NaCl 8.5 g/L (3000 rpm/ 20 min) and a volume corresponding to an initial count of 10⁸ CFU/mL was inoculated in fermentation media containing feather 10 g/L in 0.06 M Na₂HPO₄·7H₂O and 0.04 M KH₂PO₄, pH 8, the salts as presented in Table 1, and an additional carbon source. The two additional carbon sources tested were yeast extract 0.1 g/L or sucrose 0.5 g/L, which previously proved to be the best additional carbon sources for B. subtilis AMR (Mazotto et al. 2017). Cultivation was performed in 250 mL Erlenmeyer flasks containing 100 mL medium for 7 days at room temperature (28 ± 2 ºC) with constant shaking (300 rpm). At the end of cultivation, feather degradation and enzymatic activities were determined.

Table 1.

Fractional factorial design for evaluation of influence metallic salts on enzymatic activity and feather degradation

Run	Yeast extract 0.1 g/L			Sucrose 0.5 g/L
	X1	X2	X3	X1	X2	X3
1	− 1	− 1	− 1	− 1	− 1	− 1
2	− 1	− 1	+ 1	− 1	− 1	+ 1
3	− 1	+ 1	− 1	− 1	+ 1	− 1
4	− 1	+ 1	+ 1	− 1	+ 1	+ 1
5	0	0	0	0	0	0
6	0	0	0	0	0	0
7	0	0	0	0	0	0
8	+ 1	+ 1	+ 1	+ 1	+ 1	+ 1
9	+ 1	+ 1	− 1	+ 1	+ 1	− 1
10	+ 1	− 1	+ 1	+ 1	− 1	+ 1
11	+ 1	− 1	− 1	+ 1	− 1	− 1

X1: MgSO₄·7H₂O or FeSO₄·7H₂O; X2: CaCl₂·2H₂O or CuSO₄·5H₂O; X3: MnCl₂·4H₂O ou ZnS0₄·7H₂O). Decodified values are 0 mM (− 1), 0.5 mM (0) and 1 mM (1)

Effect of temperature, inoculum size, and feathers concentration on enzymatic activity and feather degradation

To evaluate the effect of inoculum size, temperature, and feathers concentration on the production of enzymes and feather hydrolysis, we use 2³ FFD; where the levels for variable temperature were 25 (− 1), 30 (0) and 35 ºC (1), the levels for inoculum size were 10⁶ (− 1), 10⁷ (0) and 10⁸ CFU/mL (1), and the levels for feather concentration were 0.5 (− 1), 1.0 (0) and 1.5% (1). The cultivation conditions were performed as described previously.

Effect of supplementary nitrogen source on enzymatic activity and feather degradation

Different supplemental nitrogen sources were tested to ascertain their effect on keratinase production using the OVAT approach. The medium used for this experiment was (g/L): MgSO₄·7H₂O 0.2, NaCl 0.5, CaCl₂·2H₂O 0.1, whole feather 10, sucrose 0.5, in 0.06 M Na₂HPO₄·7H₂O and 0.04 M KH₂PO₄ pH 8, supplemented with (NH₄)₂SO₄, NaNO₃, KNO₃ or urea at 0.6, 0.8 and 1.0%. The control did not receive an additional nitrogen source. Cultivation was performed with an inoculum of 10⁸ CFU/mL in 250 mL Erlenmeyer flasks containing 100 mL medium for 7 days at room temperature (28 ± 2 ºC) with constant shaking (300 rpm). At the end of cultivation, feather degradation and enzymatic activities were determined.

Keratinase activity determination

Keratinase activity was performed using feather keratin powder according to Wawrzkiewicz et al. (1987) protocol with modifications. Briefly, 10 g of feathers were submerged in 500 mL of DMSO, and the mixture was warmed with a reflux condenser at 100 ºC until the solubilization of the feather (80–120 min). Keratin was recovered by precipitation with 1 L of cold acetone and maintained at 4 ºC for 24–48 h. The keratin precipitates were collected by centrifugation (3000 rpm/15 min), washed twice with distilled water, and dried at 4 ºC. After drying, the keratin was titrated until a fine white powder. For keratinase activity, 250 µL of culture supernatant (enzymatic extract) was added to 375 µL of feather keratin solution (keratin 6.67% in phosphate buffer pH 7.4) and incubated for 1 h at 37 ºC. The reaction was stopped by adding 250 µL of 10% trichloroacetic acid (TCA), followed by refrigeration at 4 ºC for 30 min. The reaction mixture was centrifuged (15 min at 6000 rpm), and the supernatant absorbance was measured at 280 nm. For the blank, the TCA was added before incubation. One keratinolytic activity unit was defined as the amount of enzyme required to produce an increase of 0.01 in absorbance unit at 280 nm, under standard assay conditions (1 h at 37 ºC).

Peptidase activity determination

The activity of general peptidases was measured using gelatin substrate according to Jones et al. (1998) with modifications. Briefly, 100 µL of enzymatic extract and 900 µL of phosphate buffer pH 7.4 solution were added to 1.5 mL of gelatin 1%, and the mixture was incubated at 37 ºC for 30 min. A volume of 375 µL was removed from the reactional mixture and added to 500 µL of isopropanol before (blank) and after (test tube) the incubation time. All the tubes were centrifuged (10,000 rpm for 10 min) to collect the supernatant, and the liberated peptides were measured by the Lowry et al. (1951) method. One proteolytic activity unit was defined as the amount of enzyme required to produce an increase of 1 µg of protein under standard assay conditions (30 min at 37 ºC).

Feather hydrolysate determination

To evaluate the production of feather hydrolysate, the culture after fermentation time was harvested and centrifuged (10,000 rpm for 10 min). The supernatant was collected to measure the total solubilized protein from feather degradation by the Lowry et al. (1951) method. Readings were carried out at 660 nm, and a solution of bovine serum albumin (1.5 mg/mL) was used as standard.

Zymography

For zymographic analysis, culture supernatants were concentrated 20-fold by dialysis (cut off 9 kDa) against PEG 4000 overnight at 4 ºC and mixed with sample buffer for zymography (125 mM Tris, pH 6.8, 20% glycerol, 4% SDS, and 0.002% bromophenol blue) in a ratio 6:4 (sample: buffer). Keratinase and peptidase profiles were assayed in 12.5% SDS–PAGE co-polymerized with keratin or gelatin (0.1%). Gels were loaded with 30 µL of concentrated culture supernatant per slot for keratin-SDS–PAGE and 20 µL of concentrated culture supernatant per slot for gelatin-SDS–PAGE. The electrophoresis was conducted at 170 V for 2 h at 4 ºC, and the gels were washed with 2.5% Triton X-100 (1 h at 50 rpm) to remove the SDS. Afterward, the gels were incubated for 48 h at 37 ºC in proteolysis buffer Tris–HCl buffer, pH 7.4 (0.5 M Tris). Then, the gels were stained for 1 h with 0.2% Coomassie brilliant blue R- 250 in methanol-acetic acid–water (50:10:40) and destained in the same solution. Phosphorylase b (94 kDa), bovine serum albumin (67 kDa), ovalbumin (43 kDa), carbonic anhydrase (30 kDa), soybean trypsin inhibitor (20.1 kDa), and a-lactalbumin (14.4 kDa) were used as molecular mass standards (Amersham Pharmacia Biotech).

Statistical analysis

The experimental design and statistical analysis were performed using the statistical software Design-Expert® 9.0 (Stat-Ease, Minneapolis, MN, USA). ANOVA (analysis of variance) was used to perform the statistical analysis of the model. Fisher test value was used to determine the statistical significance of the model equation. The variance explained by the model was obtained by the multiple coefficients of determination.

Results and discussion

Effects of medium composition for inoculum production

Three different propagation media were tested to adapt (or to active) the B. subtilis AMR before fermentation. The main objective of inoculum activation was to decrease or suppress the adaptation phase of the microorganism in the fermentation medium. Three media were evaluated: a rich medium without feather (M1) and two media with 1% feathers as principal substrate supplemented with 0.01% or 0.1% yeast extract, namely M2 and M3. The presence of feathers in the inoculum production media did not accelerate enzymatic production in fermentation media (Fig. 1a) nor improve protein concentration or feather lysate production (Fig. 1b). The proteolytic activity was similar after 8 days of fermentation for all inoculum tested (Fig. 1c).

Fig.1.

Keratinolytic activity (a), protein concentration (b), and proteolytic activity of B. subtilis in fermentation broth during 8 days at 28 ± 2 °C from inoculum obtained in the media M1 (without feathers), M2 (with feathers and 0.1% of yeast extract), and M3 (with feathers and 0.01% of yeast extract)

Figure 2 shows the enzymatic profile expressed by strain AMR in inoculum propagation media after 72 h of growth. As expected, the zymography with gelatin and keratin substrates of the culture supernatants showed the production of more pronounced keratinases and peptidases in media containing feathers. The profiles of M2 and M3 were similar. Even expressing higher diversity of keratinases at the begging of fermentation, cells from M2 and M3 showed similar keratinolytic activity and feather lysate production (indirectly measured by protein soluble in the supernatant) between 6th and 7th days when compared with cells from M1 (Fig. 1).

Fig. 2.

Zymography analysis of general peptidases (a) and keratinases (b) producing by B. subtilis AMR in media M1 (without feathers), M2 (with feathers and 0.1% of yeast extract), and M3 (with feathers and 0.01% of yeast extract)

The profile expressed in M1 showed four bands with proteolytic activity (64, 72.8, 77, and 80 kDa), indicating that feathers do not induce the production of these peptidases. These bands also appear on the gels of samples from feather media. Although there is no keratin in the M1, a double band migrating in approximately 80 kDa with keratinolytic activity was observed. These keratinases could be expressed constitutively by B. subtilis AMR; however, more experimental data must affirm that. It has been widely described that the presence of keratinous material in a basal mineral medium induces the synthesis of extracellular microbial keratinases (Falco et al. 2019a, b; Gupta et al. 2013; Hossain et al. 2007), but our findings point to the production of keratinases even in the absence of substrates containing keratin. Similar results were observed by Gessesse et al. (2003), Manczinger et al. (2003), and Esawy (2007) that reported the constitutive production of keratinases by Nesterenkonia sp. AL20, Bacillus licheniformis, and Streptomyces albus AZA, respectively.

As the results indicated that the microbial keratinolytic activity does not depend on the propagation medium used, in the following experiments, yeast extract was used as a culture medium to obtain cell mass due to its simplicity of preparation. Although the feather medium is cheaper, it demands more time to reach enough cell quantity to use as inoculum (72 h), while in the rich medium, the cell concentration is reached in 24 h.

Effects of metallic salts on keratinase production and feather degradation by Bacillus subtilis AMR

One of the most critical challenges in any industrial grade enzyme production is process optimization since optimal cultivation conditions can differ significantly according to microorganism strain under investigation, and it interferes considerably in the production costs (Falco et al. 2019a, b; Sun et al. 2020; Chauhan and Devi 2020). Thus, several media components and fermentation conditions were investigated to establish those conditions which maximize keratinolytic production and feather degradation for B. subtilis AMR.

Two experiments were performed with three different salts in each one to evaluate the effects of ion salts in enzyme production and feather degradation. The first set tested the effect of FeSO₄, CuSO₄ and, ZnSO₄ using 2³ fractional factorial design (FFD). The statistical analysis of the results showed that CuSO₄ had a strong negative effect on the parameters evaluated. The interaction between these salts, assessed through analysis of variance (ANOVA), using a 95% significance limit, showed that the models were significant for the four responses—keratinolytic activity, protein concentration, feather degradation yield, and final count cells (Table 2). These three salts negatively affected bacterial growth, and the cupper inhibited enzyme production (Fig. 3).

Table 2.

Estimated effects and variance analysis of the tested variables on keratinolytic activity, concentration of soluble protein, feather degradation, and bacterial growth using 2³ fractional factorial design to analyses the variable FeSO₄, CuSO₄, and ZnSO₄

Source	Keratinolytic activity (U/ml)				Concentration of protein (mg/ml)				Feather degradation (%)				B. subtilis AMR growth (UFC/ml)
	Effect	Std errror	F value	p value	Effect	Std error	F value	p value	Effect	Std error	F value	p value	Effect	Std error	F value	p value
Model			67.76	0.0146*			41.56	0.0237*			112.18	0.0089*			3.54E + 10	< 0.0001*
Intercept	160.46	5.18			2.53	0.065			40.18	1.01			8.52	0.736
A-FeSO4	− 0.37	5.18	5.25	0.9488	− 0.095	0.065	2.09	0.2853	− 4.82	1.01	22.63	0.0415	− 5.51	0.736	5.60E + 10	< 0.0001
B-CuSO4	− 110.21	5.18	453.43	0.0022*	− 1.08	0.065	270.68	0.0037*	27.67	1.01	745.34	0.0013*	− 7.77	0.736	1.11E + 11	< 0.0001
C-ZnSO4	4.79	5.18	0.86	0.4522	− 0.024	0.065	0.13	0.7512	− 2.87	1.01	8.03	0.1053	− 1.17	0.736	2.52E + 09	< 0.0001
AB	− 12.04	5.18	5.41	0.1455	− 0.049	0.065	0.55	0.5355	0.75	1.01	0.55	0.5358	6.26	0.736	7.22E + 10	< 0.0001
AC	− 11.96	5.18	5.34	0.1470	− 0.17	0.065	6.45	0.1264	− 1.58	1.01	2.43	0.2591	− 1.54	0.736	4.40E + 09	< 0.0001
BC	-2.04	5.18	0.16	0.7312	0.026	0.065	0.16	0.7317	2.32	1.01	5.22	0.1496	0.42	0.736	3.24E + 08	< 0.0001
ABC	15.63	5.18	9.11	0.0944	0.22	0.065	10.86	0.0810	1.05	1.01	1.07	0.4104	0.79	0.736	1.16E + 09	< 0.0001
Center point	− 122.01	9.91			− 1.1	0.13			− 25.72	1.94			− 8.52	1409
Std. dev	14.64				0.18				2.87				2.08E-03
Mean	127.18				2.23				33.16				6.2
CV %	11.51				8.29				8.64				0.034
R2	0.9958				0.9932				0.9975				0.999
Adjusted R2	0.9811				0.9693				0.9886				0.99
Adeq prec	21.133				16.852				26.469				15,293.83

Prob > F values less than 0.05 indicate that the terms are significant

"Adeq Prec" (Adequate precision) measures the signal to noise ratio. A ratio greater than 4 is desirable

Fig. 3.

Zymography of peptidases produced by Bacillus subtilis AMR in feather media containing different ions composition

Similar results were observed for Bacillus licheniformis (Yoon and Shin 2010). This strain had cell growth and enzyme production completely inhibited by Cu²⁺. However, while Zn²⁺ also inhibited cell growth and enzyme production by that Bacillus strain, for B. subtilis AMR, Zn²⁺ improved the keratinolytic activity when tested without other ions and had a slightly negative effect on growth. In similar conditions, Zn²⁺ increased keratinase production of B. cereus N14 (Chauhan and Devi 2020). FeSO₄ negatively affected the keratinase production by AMR strain as observed for keratinase production by B. pumilus JYL (Sun et al. 2020) and B. cereus N14 (Chauhan and Devi 2020).

Table 3 shows the coded settings, the treatment combinations, and response values (observed and predicted). This design is represented by the regression model (Eqs. 1, 2, 3, 4) to calculate the predicted values. Furthermore, the parity plot showed a satisfactory correlation between the observed and predictive values, wherein the points clustered around the diagonal line confirmed the accurate fit of the model (Online Resource, Fig. S1).

$$Y_1=248.66+30.34\ast A-209.16\ast B+68.84\ast C-14.34\ast A\ast B-110.34\ast A\ast C-70.67\ast B\ast C+125\ast A\ast B\ast A$$ 1

$$Y_2=3.271+0.671\ast A-1.675\ast B+0.759\ast C-1.056\ast A\ast B-1.526\ast A\ast C-0.759\ast B\ast C+1.724\ast A\ast B\ast C$$ 2

$$Y_3=75.98-5.89\ast A-59.38\ast B-5.12\ast C-1.18\ast A\ast B-10.51\ast A\ast C+5.08\ast B\ast C+8.37\ast A\ast B\ast C$$ 3

$$Y_4=27.30-18.85\ast A-27.297\ast B+1.50\ast C+21.847\ast A\ast B-9.3475\ast A\ast C-1.50\ast B\ast C+6.3505\ast A\ast B\ast C$$ 4

where Y is the predicted response (Y1: keratinolytic activity, Y2: protein concentration, Y3: feather degradation, and Y4: bacterial growth) and A is the concentration of FeSO₄¸ B is the concentration of CuSO₄, and C the concentration of ZnSO₄. Considering the negative or not significative effect of CuSO_4, FeSO₄ e ZnSO₄ on most of the tested parameters, the use of these ions was disregarded.

Table 3.

Observed and predicted values by the experimental design using factorial design, with the variables FeSO₄, CuSO₄ and, ZnSO₄

Run	FeSO4	CuSO4	ZnSO4	Observed value				Predicted value
				Keratinolytic Activity (U/ml)	Bacterial Growth 107 UFC/ml (final)	Protein concentration (mg/ml)	Feather degradation (%)	Keratinolytic Activity (U/ml)	Bacterial Growth 107 UFC/ml (final)	Protein concentration (mg/ml)	Feather degradation (%)
1	− 1	− 1	− 1	248.67	27.3	3.271	75.98	248.66	27.32	3.271	75.98
2	− 1	− 1	1	317.5	28.8	4.030	70.86	317.5	28.8	4.03	70.86
3	− 1	1	− 1	39.5	2.36	1.421	16.82	39.5	0.003	1.596	16.6
4	− 1	1	1	37.67	0.003	1.596	16.56	37.67	0.003	1.596	11.48
5	1	− 1	− 1	279	8.45	3.942	70.09	234.32	8.44	3.942	74.8
6	1	− 1	1	237.5	0.6025	3.175	54.46	237.5	0.6	3.175	54.46
7	1	1	− 1	55.5	3	1.211	9.53	55.5	3	1.211	9.53
8	1	1	1	68.33	0.003	1.409	7.35	68.33	0.053	1.409	7.35
9	0	0	0	35.5	0.005	1.515	11.75	34.17	8.52	1.6935	23
10	0	0	0	54.34	0.001	1.564	14.16	34.17	8.52	1.693	23
11	0	0	0	25.5	0.002	1.222	17.56	34.17	8.52	1.693	23

Several works related the positive effect of Mn²⁺, Mg²⁺, and Ca²⁺ in Bacillus peptidases production (Chauhan and Devi 2020; Cai and Zheng 2009; Yoon and Shin 2010; Sahoo et al. 2012); thus, we evaluated the effect of these ions on keratinolytic activity, feather degradation, growth, and culture supernatant protein concentration of B. subtilis AMR in medium containing 1% of feather and 0.01% of yeast extract. The effect of the variables upon the responses and the summarized statistical analysis are reported in Table 4.

Table 4.

Estimated effects and variance analysis of the tested variables on keratinolytic activity, concentration of soluble protein, feather degradation, and bacterial growth using 2³ fractional factorial design to analyses the variable MgSO₄, CaCl₂, and MnCl₂ in feather media supplemented with yeast extract

Source	Keratinolytic activity (U/ml)				Concentration of protein (mg/ml)				Feather degradation (%)				B. subtilis AMR growth (UFC/ml)
	Effect	Std errror	F Value	p value	Effect	Std error	F Value	p value	Effect	Std error	F Value	p value	Effect	Std error	F Value	p value
Model			26.65	0.0366			4.66	0.1882			674.45	0.0015			600.33	0.0017
Intercept	300.84	3.20			3.44	0.050			71.89	0.11			21.94	0.38
A-MgSO4	5.87	3.20	3.36	0.2082	0.041	0.050	0.66	0.5013	− 0.041	0.11	0.15	0.7343	10.65	0.38	775.67	0.0013
B-CaCl2	− 4.83	3.20	2.28	0.2705	− 0.082	0.050	2.69	0.2427	3.18	0.11	904.69	0.0011	− 11.96	0.38	977.74	0.001
C-MnCl2	32.13	3.20	100.52	0.0098	0.039	0.050	0.61	0.5177	− 5.25	0.11	2458.85	0.0004	− 7.31	0.38	365.06	0.0027
AB	19.54	3.20	37.20	0.0258	0.066	0.050	1.77	0.3151	3.74	0.11	1249.26	0.0008	− 2.7	0.38	49.68	0.0195
AC	15.83	3.20	24.42	0.0386	0.18	0.050	13.09	0.0686	− 0.66	0.11	38.44	0.0250	− 6.28	0.38	269.84	0.0037
BC	13.63	3.20	18.08	0.0511	0.038	0.050	0.58	0.5249	− 0.85	0.11	63.92	0.0153	12.19	0.38	1015.71	0.001
ABC	− 2.67	3.20	0.69	0.4929	0.18	0.050	13.19	0.0681	0.26	0.11	5.86	0.1365	10.47	0.38	748.61	0.0013
Center Point	60.94	6.14	98.65	0.0100	0.29	0.096	9.38	0.0921	2.45	0.20	146.36	0.0068	− 15.91	0.73	472	0.0021
Std. dev	9.06				0.14				0.3				1.08
Mean	317.46				3.52				75.56				17.6
CV %	2.85				4.02				0.41				6.15
R2	0.9894				0.9422				0.9996				0.9995
Adjusted R2	0.9894				0.7398				0.9981				0.9979
Adq prec	17				7.343								83.972

Prob > F values less than 0.05 indicate that the terms are significant

"Adeq Prec" (adequate precision) measures the signal-to-noise ratio. A ratio greater than 4 is desirable

The variance analysis showed that the model is significant for all responses except protein concentration (p value for the model 0.1882). For this response, no term was significant. The addition of the three ions had a significant positive effect on keratinolytic activity, feather degradation, and bacterial counting. The regression analysis of the experimental data led to the obtainment of the following equations:

$$Y_1=319.34-64.34\ast A-81.34\ast B+1.14922E-013\ast C+88.84\ast A\ast B+74\ast A\ast C+65.7\ast B\ast C-21.33\ast A\ast B\ast C$$ 5

$$Y_2=75.98-5.74\ast A+1.09\ast B-6.98\ast C+13.94\ast A\ast B-3.65\ast A\ast C-4.41\ast B\ast C+2.05A\ast B\ast C$$ 6

$$Y_3=23.3+60.2\ast A-21.98\ast B-5.5\ast C-52.65\ast A\ast B-67\ast A\ast C+6.9\ast B\ast C+83.73\ast A\ast B\ast C$$ 7

$$Y_4=3.542-0.05\ast A-1.00000E-002\ast B+3.00000E-003\ast C-0.46\ast A\ast B-3.00000E-003\ast A\ast C-0.573\ast B\ast C+1.451A\ast B\ast C$$ 8

where Y is the predicted response (Y1: keratinolytic activity, Y2: protein concentration, Y3: feather degradation, and Y4: bacterial growth) and A is the concentration of MgSO₄¸ B is the concentration of CaCl₂, and C is the concentration of MnCl₂. The values obtained by the experimental analyzes and the values predicted by the model are in Table 5, wherein the points clustered around the diagonal line confirmed the accurate fit of the model (Online Resource, Fig. S2).

Table 5.

Values observed and predicted by the experimental design using factorial design, with the variables MgSO₄, CaCl₂, and MnCl₂ in media supplemented with yeast extract or sucrose

Run	MgSO4	CaCl2	MnCl2	Supplemented with yeast extract								Supplemented with sucrose
				Observed value				Predicted value				Observed value			Predicted value
				Bacterial Growth 107 UFC/ml (final)	Keratinolytic Activity (U/ml)	Protein concentration (mg/ml)	Feather degradation (%)	Bacterial Growth 107 UFC/ml (final)	Keratinolytic Activity (U/ml)	Protein concentration (mg/ml)	Feather degradation (%)	Keratinolytic Activity (U/ml)	Protein concentration (mg/ml)	Feather degradation (%)	Keratinolytic Activity (U/ml)	Protein concentration (mg/ml)	Feather degradation (%)
1	− 1	− 1	− 1	27.30	319.34	3.545	75.98	23.30	319.34	3.542	75.98	259.2	11.277	78.9	259.2	11.277	78.9
2	− 1	− 1	1	17.80	286.00	3.024	69.00	17.80	285.60	3.056	69.00	299.74	8.219	76.42	300.14	8.219	76.42
3	− 1	1	− 1	1.32	238.00	3.533	77.07	1.32	238.00	3.534	77.07	353	9.799	90.3	353	9.799	90.3
4	− 1	1	1	2.72	303.17	2.963	65.68	2.72	303.17	2.962	65.68	228.8	9.596	81	228.8	9.569	81
5	1	− 1	− 1	83.50	255.00	3.464	70.24	83.50	255.00	3.492	70.24	272	7.973	56.12	272	7.937	56.12
6	1	− 1	1	11.90	329.00	3.492	59.61	11.00	329.00	3.492	59.61	250.68	5.668	52.68	250	5.668	52.68
7	1	1	− 1	8.87	262.50	3.022	85.27	8.87	262.50	3.022	83.09	388.8	9.726	88.34	388.8	9.726	88.34
8	1	1	1	27.00	380.34	3.900	72.28	22.22	380.34	3.900	72.28	314.67	10.089	87.06	388.84	8.173	88.34
9	0	0	0	25.20	355.50	2.981	74.48	13.20	337.22	3.436	71.89	403.2	9.929	91.46	375.07	8.602	80.36
10	0	0	0	7.25	357.67	3.820	74.00	13.20	337.22	3.436	71.89	398.94	11.451	96.64	375.07	8.602	80.36
11	0	0	0	5.63	372.17	3.567	74.55	13.20	337.22	3.436	71.89	413.86	10.335	87.68	375.07	8.602	80.36

Previously, we observed that the supplementary carbon source could affect the keratinase activity and feather degradation. So, the effect of MgSO₄, CaCl₂, and MnCl₂ was investigated in the presence of sucrose instead of yeast extract. The responses analyzed were keratinolytic activity, protein concentration, and feather degradation. The cell growth did not present a significant difference when the supplementary carbon source was changed (data not shown). The interaction between these ions assessed using the ANOVA test showed that the model was significant for keratinolytic activity (Table 6). The model was not significant for protein concentration since the observed values for different combinations of these three salts in sucrose medium were close. However, comparing protein concentration results between media supplemented with yeast extract and sucrose, the concentration with sucrose was three times higher. Some studies have shown that supplementary carbon sources, other than the chicken feather, improve keratinase production and feather degradation. For instance, 0.1% of xylose, fructose, sucrose, and galactose increased the keratinase production significantly by Bacillus sp. FPF-1 (Nnolim et al. 2020a, b) and the keratinolytic activity of B. licheniformis ER-15 was higher with 1% of glucose, fructose, mannitol, and xylose (Tiwary and Gupta 2010). On the other hand, glucose, fructose, and sucrose inhibited the keratinase activity by B. licheniformis MKZ-3 (Hossain et al. 2007). The same three carbohydrates slightly inhibited the keratinase production by Chryseobacterium sp. RB (Gurav and Jadhav 2013b). The mechanisms for which some additional sources of carbon increase or decrease keratinase activity are uncertain; the increase in activity may be related to upregulation of the expression of keratinase genes, whereas inhibition may be associated with catabolic repression events.

Table 6.

Estimated effects and variance analysis of the tested variables on keratinolytic activity, concentration of soluble protein, and feather degradation using 2³ fractional factorial design to analyze the variables MgSO₄, CaCl₂, and MnCl₂ in feather media supplemented with sucrose

Source	Keratinolytic activity (U/ml)				Concentration of protein (mg/ml)				Feather degradation (%)
	Effect	Std errror	F value	p value	Effect	Std error	F value	p value	Effect	Std error	F value	p value
Model			68.17	0.0145			4.6	0.19			10.45	0.0901
Intercept	305.04	2.73			8.99	0.28			76.51	1.59
A-MnCl2	− 13.21	2.73	23.45	0.0401	− 0.69	0.28	6.21	0.1303	− 1.9	1.59	1.43	0.3542
B-MgSO4	19.86	2.73	53.01	0.0183	− 0.73	0.28	6.78	0.1213	− 5.14	1.59	10.46	0.0838
C-CaCl2	34.81	2.73	162.89	0.0061	0.71	0.28	6.58	0.1243	10.48	1.59	43.45	0.0223
AB	7.71	2.73	7.99	0.1057	0.13	0.28	0.21	0.6926	1.04	1.59	0.43	0.5795
AC	− 17.84	2.73	42.8	0.0226	0.64	0.28	5.22	0.1496	− 0.42	1.59	0.071	0.8154
BC	29.09	2.73	113.79	0.0087	0.75	0.28	7.18	0.1157	6.49	1.59	16.64	0.0552
ABC	23.34	2.73	73.25	0.0134	− 0.07	0.28	0.063	0.8255	1.28	1.59	0.65	0.5046
Center point	100.22	5.22	368.31	0.0027	1.58	0.53	8.77	0.0976	15.41	3.05	25.62	0.0369
Std. dev	7.71				0.79				4.5
Mean	332.38				9.42				80.72
CV %	2.32				8.37				5.57
R2	0.9958				0.9416				0.9734
Adjusted R2	0.9812				0.7371				0.8802
Adq prec	25.29				7.864				9.65

Prob > F values less than 0.05 indicate that the terms are significant

"Adeq Prec" (Adequate precision) measures the signal-to-noise ratio. A ratio greater than 4 is desirable

The following equations were obtained by regression analysis, and the predicted values are in Table 5.

$$Y_1=259.2+40.54\ast A+12.8\ast B+93.8\ast C-62.54000\ast A\ast B-164.74\ast A\ast C+23\ast B\ast C+186.74\ast A\ast B\ast C$$ 9

$$Y_2=11.277-3.058\ast A-3.34\ast B-1.478\ast C+0.789\ast A\ast B+2.828\ast A\ast C+3.267\ast B\ast C-0.559\ast A\ast B\ast C$$ 10

$$Y_3=78.9-2.48\ast A-22.78\ast B+11.4\ast C-0.96\ast A\ast B-6.82\ast A\ast C+20.82\ast B\ast C+10.26\ast A\ast B\ast C$$ 11

where Y is the predicted response (Y1: keratinolytic activity, Y2: protein concentration, and Y3: feather degradation), and A is the concentration of MnCl₂¸ B of MgSO₄, and C of CaCl₂. The plot representation of observed and predicted values is available in the Online Resource (Fig. S3).

At the concentration of 0.5 mM for the three ions tested, the highest keratinolytic activity was reached (405.34 U/ml), followed by 388.8 U/ml when the microorganism was grown in a medium containing 1 mM MgSO₄ and CaCl₂ (Table 5) without MnCl₂. In the absence of these salts, the activity was 259.2 U/ml. MnCl₂, when tested alone at a concentration of 1 mM, showed keratinolytic activity of 299.74 U/ml, a value slightly higher than in the absence of salts (259.2 U/ml); however, when tested with 1 mM CaCl₂ or 1 mM MgSO₄, keratinolytic activity dropped to 228.8 and 250.68 U/ml, respectively. These results suggested that Mn²⁺ should be added in medium only with Mg²⁺ and Ca²⁺.

Usually, these metal ions enhance the keratinolytic activity, but there is a wide variation in the effect of each ion on enzyme production. The effect of various metal ions on keratinase production by B. cereus N14 showed that Mn²⁺ and Mg²⁺ increased keratinase production, corroborating with our finds, but Ca²⁺ reduced the keratinolytic activity (Chauhan and Devi 2020). Similar results were observed for B. licheniformis ER-15; the presence of Ca²⁺ was inhibitory, while Mg²⁺ did not affect keratinase production (Tiwary and Gupta 2010). In contrast, Ca²⁺ was the most influential metal salt in increasing the keratinase activity of the Myceliophthora thermophila strain H49-1, while MgSO₄ inhibited the activity (Liang et al. 2011). For Stenotrophomonas maltophilia B6, the keratinolytic activity was enhanced with the addition of Ca²⁺ and Mg²⁺ (Mamangkey et al. 2020).

The effect of all ions tested on peptidase production was observed through zymography. The profile of extracellular peptidases showed that salts added to the culture medium did not significantly alter the band profile, except for CuSO₄, whose inhibition was marked. The most significant differences are in the 70 and 77 kDa bands range, presenting greater intensity in the presence of Mg²⁺, Ca²⁺, and Mn²⁺ (Fig. 3).

Effect of temperature, inoculum size, and feather concentration on keratinase production and feather hydrolysis

The effect of temperature ranging from 25 to 50 ºC as investigated on keratinolytic activity and the best activity was observed at 30 ºC (data not shown). Thus, to evaluate the effect of temperature in an integrated way with the feather concentration and inoculum size, we designed a 2³ fractionated factorial experiment. The model P-value implies the model was not significant for the range of the tested variable. However, the feather concentration was significant in model terms in all three responses (p < 0.05) (Table 7). The equations obtained by regression analysis and the values observed and predicted are in the Online Resource (Eqs. S1, S2, S3, Table S1, and Fig. S4).

Table 7.

Estimated effects and variance analysis of the tested variables on keratinolytic activity, concentration of soluble protein, and feather degradation using 2³ fractional factorial design to analyses the variable temperature, inoculum size, and feather concentration

Source	Keratinolytic activity (U/ml)				Concentration of protein (mg/ml)				Feather degradation (%)
	Effect	Std errror	F value	p value	Effect	Std error	F value	p value	Effect	Std error	F value	p value
Model			5.18	0.102			4.72	0.1148			8.8	0.0507
Intercept	348.67	7.19			5.41	0.39			85.42	1.19
A-temperature	− 18.08	8.2	4.87	0.1145	− 1.08	0.45	5.76	0.0959	0.12	1.36	0.00719	0.9378
B-inoculum	15.50	7.54	4.23	0.132	− 0.11	0.41	0.077	0.7995	1.82	1.25	2.14	0.2401
C-feather concentration	39.25	8.2	22.92	0.0173	2.13	0.45	22.32	0.018	− 10.37	1.36	58.45	0.0046
AB	− 8.25	8.2	1.01	0.3884	− 0.68	0.45	2.25	0.2307	0.41	1.36	0.091	0.7822
AC	− 1.42	8.2	0.03	0.8736	− 0.67	0.45	2.2	0.2343	− 0.35	1.36	0.067	0.8131
BC	3.08	8.2	0.14	0.7318	− 0.3	0.45	0.44	0.5545	− 1.15	1.36	0.71	0.4606
ABC	14.42	8.2	3.09	0.1769	0.062	0.45	0.019	0.8996	0.44	1.36	0.11	0.7644
Std. dev	23.19				1.27				3.84
Mean	345.22				5.44				85.02
CV %	6.72				23.41				4.51
R2	0.9236				0.9168				0.9535
Adjusted R2	0.7455				0.7228				0.8451
Adq prec	6.95				6.600				7.92

Prob > F values less than 0.05 indicate that the terms are significant

"Adeq Prec" (Adequate precision) measures the signal-to-noise ratio. A ratio greater than 4 is desirable

Comparing the data (Fig. 4), the keratinolytic performance was better at 25 ºC, 10⁸ CFU/mL of inoculum, and 1.5% feather concentration. The variation of temperature between 25 and 35 °C did not significantly affect the keratolytic productivity, a good characteristic for production at local room temperature, so there is no need for expenses for heating or cooling the system. This result is in line with that observed to produce keratinases by Bacillus sp. UPM-AAG1 (Gafar et al. 2020), B. cereus N14 (Chauhan and Devi 2020), B. subtilis KD-N2 (Cai and Zheng 2009), and B. aerius NSMk2 (Bhari et al. 2018), which had their optimum temperature between 25 and 35 °C.

Fig. 4.

Effect of different combinations (FFD 2³) of temperature, inoculum size, and feather concentration on keratinolytic activity (a), protein concentration (b), and feather degradation (c) by B. subtilis AMR

The concentration of feather increased the keratinase activity significantly and protein concentration; however, the percentage of feather degradation was better when the concentration of feather was lower (Fig. 5). Probably longer incubation time is necessary to degrade the excess of feathers in the culture medium. We observed that the highest activity (418.67 U/ml) and soluble protein concentration (9.423 mg/mL) occurred when the culture medium contained 1.5% of feathers and initial inoculum of 10⁸ CFU/mL. A similar result was observed for Bacillus aerius NSMk2 with optimum keratinase production with 1.375% feathers (Bhari et al. 2018) and Chryseobacterium aquifrigidense FANN1 with 1.5% feather (Bokveld et al. 2021). The increase of keratinase production and protein concentration was directly proportional to feather concentration, while the percentage of feather degradation was inversely proportional. Bacillus sp. UPM-AAG1 (Gafar et al. 2020) presented crescent keratinase activity up to 4.5% of feathers. The keratinase production by Bacillus sp. Nnolim-K1, whose enzyme production improved with increasing chicken feather concentration until 1.25%, decreased significantly beyond that value. B. subtilis AMR decreases the keratinolytic activity with feather concentration above 2%, incurring repression for excess substrate (data not shown). For B. cereus strains isolates by Almahasheer et al. (2022), this repression was observed in feather concentrations above 1%. Besides the downregulation of the keratinase-encoding gene in high keratin concentration, the high viscosity of the medium at higher concentration feathers can be another possible reason for decreased keratinase production (Nnolim et al. 2020a, b). Nnolim et al. (2020a, b) reported a discrepant result; the chicken feather concentration of 0.5% was optimum for Bacillus sp. FPF-1 keratinase activity in the fermentation medium containing about 10⁸ CFU/mL inoculum. A further increase in the concentration of the chicken feather leads to a decline in keratinase activity. However, the protein concentration in the fermentation broth increased with the feather concentration, as observed for us. The higher-than-expected protein concentration in higher feather concentration may be explained by breaking down the keratinous biomass faster than the bacteria consumption (Nnolim et al. 2020a, b).

Fig. 5.

Response surface of the interaction between temperature and feather concentration on keratinase activity (a), protein concentration (b), and feather degradation (c) of B. subtilis AMR

Inoculum size is the required initial bacterial mass for conducting fermentation. There were no statistically significant alterations in keratinolytic activity in the range tested from 10⁶ to 10⁸ CFU/mL. Most studies expressed the inoculum measures as percentages (v/v usually) concerning microbial growth volume for fermentation medium volume. However, this form of measurement, although very practical, is imprecise. In general, increasing the inoculum size increases keratinolytic activity up to a maximum quantity, after which no increase, or even reduction in the activity, can be observed (Chauhan and Devi 2020; Cai and Zheng 2009; Barman et al. 2017; Gafar et al. 2020; Verma et al. 2016).

Effect of additional nitrogen source on B. subtilis AMR keratinase production and feather hydrolysis

Different supplementary nitrogen sources in three concentrations were tested to ascertain their effect on keratinase production, the concentration of soluble proteins in the culture medium, and feather degradation. Potassium nitrate at 0.8 and 1.0% concentrations increased the keratinolytic activity by 129.8 and 150%, respectively, compared to the control (medium without an additional inorganic nitrogen source), as shown in Fig. 6. The effect of potassium nitrate on keratinase production of different keratinolytic microorganisms is variable. The activity of Bacillus cereus IIPK35 was drastically reduced by this nitrogen source (Jana et al. 2020), while for Arthrobacter sp. NFH5, the effect was the opposite (Barman et al. 2017).

Fig. 6.

Effect of different sources of additional nitrogen on keratinolytic activity (a), protein concentration (b), and feather degradation (c) of B. subtilis AMR

Figure 6 shows that 0.6% ammonium sulfate decreased keratinolytic activity, but when its concentration increases to 0.8%, activity increases, reaching a value (354.45 ± 11.67 U/ml) close to the value that found in control (348.8 ± 9.32 U/ml). The effect of ammonium salts on the production of keratinases is controversial. While for some strains, the addition leads to significant increases in the activity, for others, it promotes repression in keratinase production. For instance, the addition of ammonium sulfate in B. cereus culture medium and ammonium chloride in Chryseobacterium sp. RBT culture medium repressed their keratinase production (Gurav and Jadhav 2013b; Arokiyaraj et al. 2019) while supplementation of ammonium sulfate (0.2%) resulted in maximal keratinase production by B. weihenstephanensis PKD. For this strain, ammonium salts such as ammonium chloride and ammonium nitrate, and potassium nitrate reduced the keratinase activity (Sahoo et al. 2012), in contrast with the result of AMR for this last salt.

Urea inhibited the keratinolytic activity of strain AMR considerably. The value of keratinolytic activity found in the culture supernatant using 0.6% urea was 4.3 times lower than the control. A similar result was observed for the keratinase production by B. cereus N14 (Chauhan and Devi 2020). In contrast, urea was the best complemental nitrogen source for Myceliophthora thermophila strain H49-1 (Liang et al. 2011) and improved the keratinolytic activity of Bacillus subtilis KD-N2 (Cai and Zheng 2009).

Although some nitrogen sources increased the keratinolytic activity of B. subtilis AMR, the concentration of soluble proteins and the feather degradation performance did not follow this increase. The protein concentration in the medium containing 1.0% sodium nitrate corresponds to 66.59% of the medium without a source of inorganic nitrogen, while urea had a protein concentration 5.6 times lower than the control (Fig. 6). For Bacillus sp. Nnolim-K1 and Bacillus sp. FPF-1, none of the supplementary nitrogen sources increased keratinolytic activity or protein concentration (Nnolim et al. 2020a, b). Thus, no additional nitrogen source should be used when the fermentation with B. subtilis AMR is conducted to obtain keratin hydrolysates.

When the results were analyzed together (Fig. 7), the keratinolytic activity increased 85% when Ca²⁺, Mn²⁺, and Mg²⁺ were added to the culture medium, but the protein concentration and feather degradation did not significantly modify. When yeast extract was replaced by sucrose as an additional carbon source, the protein concentration and the degradation of feathers improved by 134% and 38% due to higher keratinolytic activity in the medium containing sucrose. Optimizing temperature, inoculum, and feather concentration increased the keratinolytic activity slightly. The change of supplementary carbon source was the variable that affects more positively the degradation of feathers and the protein concentration, maybe because yeast extract is also a protein source and sucrose not, making feathers the only protein source of bacteria, improving the expression of specific enzymes to consume the substrate.

Fig. 7.

Evolution of keratinolytic activity (black dot), protein concentration (white dot), and feather degradation (gray dot) along with improved medium and condition experiments

Each microorganism is unique in its metabolism; therefore, the degradation product of the feathers, as well as the time required for fermentation, may differ. All experiments were conducted with a fermentation time of 7 days to ensure maximum feather degradation. However, this is longer than described for other keratinolytic microorganisms in the literature. To improve the fermentation time, kinetic studies and tests under different controlled cultivation conditions in bioreactors will be carried out.

Conclusions

The verification of physical and chemical parameters for improving the productivity of fermentation processes is critical to guarantee the competitiveness and better performance of the production of bioproducts. In this work, it was observed that ions’ presence could affect enzyme activity and hydrolysate production. The ions Cu²⁺ and Fe²⁺ strongly inhibited keratinase activity, while Ca²⁺, Mg²⁺, and Mn²⁺ significantly increased the activity. The chosen carbon sources are also essential to ensure better yield. The replacement of yeast extract by sucrose was crucial to increase the protein concentration three times. The process can be conducted between 25 and 35 °C, a temperature range interesting for industrial production once does not require extensive heating or cooling. The best feather concentration was 1.5%, and the initial inoculum can variate from 10⁶ to 10⁸ CFU/mL without significantly reducing productivity. The microorganism keratinolytic activity does not depend on the propagation medium used; thus, this step could exploit a cheaper medium to obtain cellular mass in large-scale fermentation. No addition of a nitrogen source was necessary to produce keratin hydrolysate, representing an advantage of the system since the substrate used is enough to supply the nitrogen needs of the bacteria. The present investigation results suggest the promising potential of Bacillus subtilis AMR for industrial application in feathers degradation to obtain feather hydrolysate.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

ABV and SC contributed to the study conception and design. Material preparation, data collection, and analysis were performed by AMM, SMLC and, EPS. AMM wrote the first draft of the manuscript, and ABV reviewed the manuscript. All authors read and approved the final manuscript.

Declarations

Conflict of interest

The authors declare there are no conflicts of interest in the publication.