Bioremediation of Cu and Zn by guinea fowl feather wastes and biodegradation of Cu- and Zn-polluted feathers by free and immobilized heavy-metal-tolerant bacterium Pseudochrobactrum sp. IY-BUK1

Sample collection and processing

GFF wastes were collected from a chicken abattoir in the Nigerian state of Kano. The collected GFFs were spread on clean polythene bags, and non-feather parts such as peak, nails, and excreta were handpicked and removed. GFFs were thoroughly washed with tap water and detergent, and then their fat remnants were removed by soaking in a solution containing 200 ml of deionized water, methanol, and chloroform in the ratio of 1 : 1 : 10. The cleaned GFFs were further rinsed in distilled water and oven-dried at 80°C for 6 h until a constant weight was achieved (Yusuf et al., 2016, 2019).

Chemical and physical pretreatment of GFFs for adsorption of Cu and Zn

Both physical (steam/autoclaving and boiling) and chemical (NaOH, tannic acid, and acetone) methods were used in the pretreatment of GFFs. Steam (thermal) pretreatment was carried out by autoclaving 5 g of the whole GFFs suspended in 100 ml of distilled water at 121°C and 21 psi for 30 min. GFFs preheated with boiled water for 30 min were also used. In chemical pretreatments, three chemicals were used, viz NaOH, acetone, and tannic acid. NaOH pretreatment of GFFs was carried out according to the method of Sun et al. (2009) using 0.01 mol/l NaOH. In brief, GFFs (10 g) were suspended in 100 ml solution of NaOH and stirred at room temperature for 8 h (Sun et al., 2009). For pretreatment with tannic acid, Ratnakumari and Kota’s (2012) method was used, where 10 g GFFs was suspended in 5% tannic acid solution for 1 h. GFFs in both pretreatments were then separated, precipitated (only in the case of tannic acid), and washed several times with distilled water. In the case of acetone pretreatment, 10 g GFFs was added to 100 ml of 0.6% acetone and incubated at room temperature for 90 min (Mrajji et al., 2019).

Preparation of Zn and Cu stock solutions

As reported in an earlier study, 1000 parts per million (ppm) solutions of Cu and Zn were prepared by dissolving 3.392 g and 4.42 g of CuSO₄ and ZnSO₄, respectively, in 1000 ml of distilled water (Na deem et al., 2016). From these stock solutions, working solutions were prepared by pipetting an appropriate volume of the stock solution into a volume of distilled water.

Adsorption of Cu and Zn by pretreated and untreated GFFs

This experiment was performed in a batch mode in distilled water (dH₂O). Briefly, 100 ml of Cu and Zn working solutions of 15, 25, and 50 ppm final concentrations was taken in appropriately labeled 250 ml Erlenmeyer flasks. In separate flasks containing different concentrations of Cu and Zn, 5 g of each of the pretreated and untreated GFFs was placed. Untreated GFFs kept in Cu and Zn solutions and those in flasks without heavy metals served as control. The flask setup was then kept on a rotary shaker at 150 rpm for 6 h (Tan et al., 2003; Guo et al., 2016). GFFs were separated from metal solutions using No. 1 Whatman filter paper. Using an Atomic Absorption Spectrometer (AAS) (CG-AA-7000 equipment, Shimadzu, Japan), the residual concentrations of Cu and Zn in each flask were determined (Carvalho et al., 2003). To determine the removal percentage of Cu and Zn by GFFs, the difference between the initial and residual Cu and Zn concentrations was used, as shown in Equation 1.

where Co is the initial metal ion concentration (mg/l), Ct is the metal ion concentration at time t (mg/l), and %R is the percentage removal.

Optimization of physical conditions affecting Cu and Zn adsorption by GFFs

Using a one factor at a time (OFAT) approach, the effects of pH, temperature, and amount of GFFs on Cu and Zn adsorption by GFFs were determined. Various amounts of GFFs measured in g/l (1, 2, 3, 4, and 5) were placed in separate flasks containing 100 ml of 25 ppm of each Cu and Zn solution, and the incubation temperature and the initial pH of aqueous solutions were maintained at 30°C and 7.5, respectively. The effects of pH on Cu and Zn adsorption by GFFs were determined by adjusting the initial pH of the aqueous solutions to 2, 4, 6, 8, and 10 while keeping GFF concentrations at the optimum amount determined earlier and temperature at 30°C. The incubation temperature was then varied (25, 30, 35, 40, 45, 50, 55, and 60°C) while maintaining other parameters at the optimum levels. The residual Cu and Zn in all flasks were determined using an AAS, and the optimum conditions were used in the subsequent experiments.

Isolation of GFF-degrading bacteria (GFF-DB) and preparation of culture media

Seven GFF-DB were obtained from two sources: 1) three from previous collections of feather-degrading bacteria (FDB) in the laboratory and 2) four from soil samples collected from three guinea fowl slaughter and farm locations. These samples were collected and processed as previously described (Yusuf et al., 2016, 2019). In brief, 1 g of the soil samples was suspended in 0.9% saline solution, vortexed for 2–3 min intermittently, and centrifuged at 1000 rpm for 5 min. The supernatant from each tube was inoculated in a 250 ml Erlenmeyer flask containing 100 ml feather meal broth (FMB) and supplemented with GFFs as a sole source of carbon and nitrogen. FMB was prepared as described earlier by Yusuf et al. (2016, 2019, 2020) by dissolving 0.5 g/l of NaCl, 0.7 g/l of K₂HPO₄, 1.4 g/l of KH₂PO₄, and 0.001 g/l of MgSO₄ ∙ 6H₂O in 1000 ml of distilled water, and the pH of the mixture was adjusted to 7.5. The inoculated FMB were incubated on an orbital shaker operating at 150 rpm at 30°C. Bacteria from flasks that showed high degradation of GFFs were isolated through inoculation of 1 ml of FMB on feather meal agar (FMA), pH 7.5, which contained 1.0 g/l feather, 0.5 g/l NaCl, 0.7 g/l K₂HPO₄, 1.4 g/l KH₂PO₄, 0.001 g/l MgSO₄ ∙ 6H₂O, and 1.5 g/l agar. Isolates from the previous collections were inoculated into FMB containing GFFs, and their ability to degrade GFFs was determined as previously described (Yusuf et al., 2020).

Screening GFF-DB for Cu and Zn tolerance

One part per million solutions of Cu and Zn were prepared from their respective stock solutions in FMB, and the bacterial growth in two sets of FMB (FMB with metal and without metal) was compared. One percent suspensions of the collected and the isolated GFF-DB were inoculated into 100 ml FMB containing 5 g/l of GFFs and incubated at 30°C. Any bacteria that grew, degraded feathers faster, and showed a higher keratinolytic activity in the presence of both Cu and Zn were selected and inoculated onto fresh FMA.

Morphological, biochemical, and molecular identification of GFF-DB

Gram staining and Bergey’s manual ofsystematic bacteriology were used to identify the morphological and biochemical characteristics of the bacterium, respectively (Cappuccino and Sherman, 1996; Brenner et al., 2005). Molecular identification was carried out using 16S rRNA sequence analysis, as previously described (Yusuf et al., 2020).

Biodegradation of Cu- and Zn-polluted GFFs

Biodegradation of Cu- and Zn-laden GFFs by free cells of strain IY-BUK1 was carried out in FMB containing 5 g/l of Cu- and Zn-laden GFFs as substrates. Cu- and Zn-free GFFs placed in another FMB and flasks that contained Cu- and Zn-laden GFFs but without GFF-DB served as control. The flasks were incubated on a rotatory shaker at 150 rpm for 7 days at 30°C. At the end of the incubation period, the percentage of feather degradation was calculated, as previously described (Yusuf et al., 2016). The growth of the GFF-DB in each flask was estimated in duplicate in colony-forming units (CFU).

Optimization of factors influencing keratinase production and degradation of Cu- and Zn-laden-GFFs by selected GFF-DB

To determine how pH, temperature, inoculum size, and extra carbon and nitrogen influence keratinase production and degradation of Cu- and Zn-laden GFFs, OFAT and response surface methodology (RSM) were employed. Effects of pH (2, 4, 6, 8, and 10), temperature (25, 30, 35, 40, 45, and 50°C), inoculum size (1, 2, 3, 4, and 5%), extra carbon source (glucose, sucrose, and maltose), and extra nitrogen source (ammonium chloride (AMC), ammonium sulfate (AMS), urea, skim milk, and yeast extract) were optimized using OFAT, as described earlier (Yusuf et al., 2016).

For RSM, Plackett–Burman factorial design (PBFD) was used to select the major factors required for the maximum keratinase yield. The effects of GFF concentration, pH, temperature, and inoculum size on keratinase production were investigated at two levels (minimum and maximum). Design Expert software, version 6.0 (Stat Ease, Minneapolis, USA), was used to generate 12 sets of experiments that were conducted in triplicate. Effects of each factor on keratinase production were determined by the analysis of variance (ANOVA) using the calculated p-value of each factor.

To design and check the interactions between the factors selected with PBFD, central composite design and RSM were used. Three factors coded as A, B, and C were studied at four levels in triplicate. The factors included temperature (35–40°C), inoculum size (2–4%), and GFF concentration (2–4 g/l). Twenty experimental runs were included in the design. The response (y ), which is the average keratinase production for each run, was fitted to a second-order polynomial equation, and a model equation that relates to the independent factors (A–C) was generated as shown in Equation 2.

where y is the predicted response of keratinase production), βo is the intercept, β _i is the linear coefficient, β _ij is the quadratic coefficient, and X represents the factors tested (in this case, A, B, and C). The P -values and confidence levels of the data obtained were calculated using the ANOVA tool embedded in Design Expert software, version 6.0.6. The quality of the model was ascertained statistically by the coefficient of determination R ², and its statistical significance was determined using F -test values. Using 3D plots, optimal values from the linear and interactions of factors were estimated. The predicted values from RSM were validated by experiments conducted in triplicate.

Immobilization of GFF-DB in gellan gum

According to the method employed in a previous study, immobilization of the selected GFF-DB bacterium in the gellan gum matrix was carried out using melanized black feathers (Yusuf et al., 2020). By dissolving 0.75 g of gellan gum powder in 100 ml of distilled water, a 0.75% (w/v) suspension of gellan gum powder (Sigma) was prepared, which was then heated to 75°C on a hot plate for 5 min. Then 0.06 g of calcium chloride was added to the gum suspension, and the mixture was stirred continuously until the suspension cooled down to 45°C. By adding drops of 0.1 M NaOH solution, the pH of the gellan gum + CaCl₂ solution was adjusted to 7, and then, 3.5 g of GFF-DB pellet was added, and the mixture was continuously stirred. Using a peristaltic pump connected to a narrow tubule, beads with an average size of 2–3 mm were produced and allowed to drop in 500 ml of canola oil containing 0.1% Span 80 (Sigma). The beads were separated from the canola oil using a 0.5-mm mesh, kept in distilled water, and refrigerated for 2 h. The refrigerated beads were then washed several times with 0.1% (v/v) Tween 80 solution and distilled water and stored at 4°C until future use.

Determination of keratinase activities and bacterial growth

Every 24 h, an aliquot of 10 ml hydrolysate was collected from FMB and centrifuged at 10 000 rpm for 15 min. For measuring keratinase activities, the cell-free supernatant was used and soluble keratin was used as a substrate as previously described (Wawrzkiewicz, 1987). Bacterial growth was estimated using the platecounting method in CFU/ml (Yusuf et al., 2016).

Statistical analysis

A completely randomized design was used in this study. All experiments were run in triplicate, except where indicated. To check for errors and variations, means and standard deviations (SD) were determined from the repeated experiments, and the values were represented as mean ± SD. Under different conditions, differences in the rate of degradation and keratinase production between the groups were analyzed by Student’s t-test and one-way ANOVA at 95% confidence level using the Minitab (version 16) software package (State College, Pennsylvania). To check the differences in keratinase production and feather degradation under the influence of different heavy metals, Tukey’s post hoc analysis after ANOVA was used.

Optimization of parameters affecting the ability of the bacteria to degrade GFFs and produce keratinase

The ability of Pseudochrobactrum sp. IY-BUK1 to produce keratinase required for degrading GFFs laden with a mixture of Cu and Zn was influenced by the initial pH of FMB, temperature of incubation, the amount of polluted GFFs, bacterial inoculum size, and extra carbon and nitrogen. When the pH of FMB increased from 7.0 to 8.0, keratinase production increased from 80 to 119 U/ml (Fig. 2A). Complete degradation of 1 g/l of Cu- and Zn-laden GFFs was achieved in 4 days at pH 8 as against 5 days when the pH was 7 (result not shown). A gradual decrease in keratinase activity and GFF degradation was observed above pH 8. The decrease in keratinase production and GFF degradation was observed at pH below 8 and above 9, which indicated that the bacteria were slightly inhibited under higher acidic and alkaline conditions. Similar observation was reported for Alcaligenes faecalis AQ005-01 and Bacillus sp., which degraded white unpolluted feathers and melanized heavy-metal-polluted chicken feathers, respectively, at pH 8 (Yusuf et al., 2016, 2020). There are many reports on FDB species indicating that slight acidic to alkaline pH enhanced keratinase activity and feather degradation (Jeong et al., 2010; Pandian et al., 2012; Bhari et al., 2018; Łaba et al., 2018). Similarly, the strain’s capacity to degrade Cu- and Zn-laden GFFs and produce high keratinase was optimal at 40°C (Fig. 2B). Even though Pseudochrobactrum sp. IY-BUK1 was previously reported to selectively degrade black chicken feathers optimally at temperatures slightly above 40°C (Yusuf et al., 2020), FDB have been reported to degrade non-melanized chicken feathers optimally at room temperature (Sangali and Brandelli, 2000; Lo et al., 2012; Tamreihao et al., 2019; NoerHidayat et al., 2019).

Fig. 2

Optimization of pH, temperature, inoculums size, amount of polluted GFF, extra carbon and nitrogen sources for enhanced production of keratinase by Pseudochrobactrum sp. IYBUK-1

Optimum keratinase activities were achieved within the 3–4% inoculum size (Fig. 2C), which is in agreement with the report of Cai et al. (2008) and contradicts other findings in which 2% inoculum size was the optimum (El-Refai et al., 2005; Cai et al., 2008; Sahoo et al., 2012; Yusuf et al., 2016). The amount of keratinase produced by GFF-DB was also influenced by the amount of heavy-metal-polluted GFFs placed in FMB (Fig. 2D). Even though degradation of Cu- and Zn-polluted GFFs decreased with an increase in the GFF concentration from 2 g/l, keratinase activity was optimal at 4 g/l and decreased further. A lower keratinase production and feather degradation at GFF concentrations above 4 g/l might be attributable to the increase in the Cu and Zn concentration in the FMB, which inhibited bacterial growth and/or keratinase production by GFF-DB. Previous studies have reported feather concentration as an influential parameter for keratinase production by many FDB (Bach et al., 2012; Patinvoh et al., 2016); however, the toxicity of the accumulated heavy metals adsorbed onto GFFs in the present study might have resulted in reduced keratinase production and feather degradation. The effects of 5 g/l extra elemental sources of carbon and 1 g/l extra nitrogen sources on keratinase production and GFF degradation by strain IY-BUK1 were also determined. Findings of this study indicated that the extra amount of different carbon and nitrogen sources in FMB showed varying effects on keratinase yield depending on the types of FDB and substrates used (Cao et al., 2009; Lo et al., 2012). The addition of extra sucrose, glucose, maltose, and mannose suppressed the keratinase activity and degradation of heavy-metal-laden GFFs. This contradicts the previous report in which the strain’s keratinase yield was stimulated by sucrose in the degradation of melanized chicken feathers (Yusuf et al., 2020). This suggests that the strain’s nutritional requirements vary from one substrate to another. Except for the yeast extract, which did not influence the keratinase activity (Fig. 2F), the addition of extra nitrogen sources of urea, AMC, AMS, and skim milk to FMB containing heavy-metal-laden GFFs suppressed keratinase yield and feather degradation. Findings from previous studies have suggested supplementation of FMB with yeast extracts for enhanced FDB growth and keratinase yield (Kainoor and Naik, 2010; Villa et al., 2013).

The parameters that enhanced the strain’s ability to yield higher keratinase and degrade polluted GFFs faster were subjected to Plackett–Burman design to obtain the significant factors among them. Keratinase yields (response) from 12 experimental runs are shown in Table 3A. The ANOVA and p-value for each parameter are presented in Table 3B. Parameters with a P-value of < 0.05 were considered to have a significant effect on the response and were selected for RSM studies. Two of the significant factors—pH and GFF concentration—exerted a positive effect on keratinase production, whereas inoculum size exerted a negative effect. Using RSM, the effects of the interaction between individual significant parameters on keratinase activities with IY-BUK1 were evaluated in 20 experimental runs.

As shown in Table 4, the average keratinase activities (responses) obtained from each experimental run and the predicted responses were much closer to each other. The maximum keratinase activity of 191 U/ml was recorded in 48 h with 3 g/l GFF concentration, 3% inoculum size, and at 35°C (run 17, Table 4). p-values for the model (< 0.0001) and for “lack of fit” (0.6271) indicated that the obtained experimental data were a good fit with the model (Table 5). Regression equation coefficients were calculated, and the data were fitted to a second-order polynomial equation, where A, B, and C represented pH, GFF concentration, and inoculum size (coded values), respectively. P-values of interactive terms (AB and BC) of the model indicated that high keratinase production depended on the interactions of the pair factors.

RSM results showed that the obtained data fit well with the model. To measure the strength of the model, multiple correlation coefficients (R ²) and F-values were used. The higher accuracy of the model was indicated by the R ² value of 0.9756, an adjusted R ² value of 0.9537, and a high F value of 44.47. The closeness of R ² and adjusted R ² values to each other and to 1 further implied that only 2.19% of the total variation was not explained by the model and that there was a good correlation between the observed and predicted values. A 3D interactive plot showing keratinase activities with respect to pH and amount of GFFs is depicted Figure 3. The interaction response of inoculum and temperature showed a curvature, which suggests that keratinase activities were maximum at the middle points (3% inoculum size and 35°C temperature) – Figure 3A; however, keratinase as a function of the inoculum size and GFF concentration (Fig. 3B) showed that higher keratinase yield was achieved at the inoculum size of 3% and GFF concentration of 4 g/l. The maximum keratinase yield of 184 U/ml was achieved from the following optimum conditions: GFF concentrations (4.0 g/l), temperature (35°C), and inoculum size (3%). The data predicted from the model were verified by experiments performed in triplicate, and the average keratinase produced (182.5 U/ml) matched the predicted yield of 184.1 U/ml.

Fig. 3

Response surface plots showing the interaction of factors affecting keratinase production (A) inoculum size and temperature (B) GFF concentration and inoculum size

Degradation of heavy-metal-polluted GFFs by gellan-gum-immobilized GFF-DB

Cu- and Zn-polluted GFFs were degraded by gellangum-immobilized GFF-DB under optimized conditions, and the rate of degradation and the amount of heavy metal residue in the hydrolysates were compared with those of free-living GFF-DB. Immobilization of bacterial whole cells or their enzymes is a novel technique used in bioprocess and possesses various advantages such as stability, high efficiency, and reusability over the use of free cells (Ahmad et al., 2012). Even though previous studies involving melanized feathers have reported keratinase activities by immobilized bacteria (Yusuf et al., 2019), the present study was the first immobilization study involving GFFs. Immobilization of IY-BUK1 cells resulted in higher keratinase production and faster degradation of heavy-metal-polluted GFFs than that of free-living cells. Complete degradation of 5 g/l of polluted GFFs was achieved within 72 h with a higher amount of keratinase (214 U/ml) when 250 beads of immobilized GFF-DB were used as against 120 h and 191 U/ml when 3.5% of free cells were used (Table 6). A significant difference (P < 0.05) in the degradation of heavy-metalladen GFFs was observed between free and immobilized IY-BUK1. Key metabolic features of the immobilized cells, such as cell growth, might have been modified by the faster degradation of Cu- and Zn-laden GFFs (Moslemy et al., 2002). The residual Cu and Zn ions in the resultant protein hydrolysate produced by immobilized cells were analyzed and compared with those of free cells. The amount of residual Cu and Zn in the hydrolysates after complete degradation of Cu- and Zn-polluted GFFs was significantly lower than that of free cells, which agrees with the strain’s capacity to accumulate Cd and Pb following the degradation of melanized feathers (Abba et al., 2020). These findings indicate that Pseudochrobactrum sp. IY-BUK1 possesses the additional capacity to absorb/accumulate Cu and Zn besides featherdegrading properties.

Similarly, the immobilized IY-BUK1 maintained the degradation of 5 g/l of different heavy-metal-laden GFFs for seven cycles of 72 h. The number of cycles decreased to four when the concentration of Cu and Zn adsorbed onto the GFFs increased to 33.2 ppm. This result is consistent with an earlier report in which immobilized Alcaligenes sp. AQ05-001 degraded feathers in FMB containing a mixture of Ag, Co, and Cu for eight cycles (Yusuf et al., 2019).