Sustainable Endoglucanase Production from Lignocellulosic Waste Through Fungal Co-Culture Technology: A Step Towards Circular Economy

Simple Summary

Large amounts of agricultural waste rich in lignocellulose are produced every year and often remain underutilized. This study demonstrates a sustainable way to convert such waste into a valuable industrial enzyme, endoglucanase, using a fungal co-culture of Aspergillus fumigatus and Rhizopus arrhizus. Among several tested residues, alkaline-pretreated pea hulls proved to be the most effective substrate due to their high cellulose and low lignin content after treatment. Process conditions were statistically optimized using Design of Experiments approaches, which significantly enhanced enzyme production. The optimized enzyme system was further applied to hydrolyze pretreated biomass, releasing a high amount of fermentable sugars. These findings highlight the dual role of pea hulls as both a substrate for enzyme production and a feedstock for biomass hydrolysis, supporting waste valorization and circular bioeconomy concepts.

Abstract

This study focused on optimizing endoglucanase production using a peculiar fungal co-culture comprising Rhizopus arrhizus and Aspergillus fumigatus, identified through morphological and 18S rDNA analyses. The co-culture achieved the highest enzyme production after 72 h of fermentation with alkaline-treated substrates. Scanning Electron Microscopy (SEM) revealed substantial structural disruption in pretreated biomass, enhancing enzyme accessibility. Among the tested substrates, pea hulls proved to be the most effective for enzyme production. Optimization of physical and nutritional parameters was performed using Design of Experiments (DOE) approaches, specifically Plackett–Burman Design (PBD) for screening and Central Composite Design (CCD) for fine optimization. The maximum endoglucanase activity of 119.58 U/mL/min was obtained under the optimized conditions of 27.5 °C, pH 5.5, inoculum age 3.5 days, and supplementation with 1.5% fructose, 1.25% yeast extract, 1.25% sodium nitrate, and 1.25% Tween 80. Analysis of Variance (ANOVA) confirmed the significance of these parameters and their interactions at a 95% confidence level, with a strong model fit (R² = 0.9052). This study demonstrates the potential of waste pea hulls as a cost-effective substrate for enzyme production, supporting waste valorization and contributing to a circular bioeconomy through sustainable biomass utilization.

1. Introduction

The circular economy concept, introduced in 2015, transformed waste management by shifting linear resource–product–waste models toward sustainable circular systems based on reducing, reusing, and recycling waste, thereby minimizing raw material costs and dependence on primary resources [1]. This approach enhances resource efficiency and waste reduction while addressing global challenges such as climate change, population growth, urbanization, environmental degradation, and resource depletion, making a paradigm shift toward sustainable and circular economies essential [1,2]. Large quantities of lignocellulosic waste generated from forestry biomass, municipal solid waste, and agricultural residues pose serious environmental and economic challenges; however, this waste can be converted into valuable resources through enzyme production for industrial applications [3]. Agricultural lignocellulosic waste thus represents a sustainable source of value-added products, and recent advancements in fungal co-culture technology have opened new possibilities for producing sustainable endoglucanase from such waste while promoting closed-loop production systems and reducing environmental degradation [3].

Endoglucanase (EC 3.2.1.4) degrades cellulose by hydrolyzing β-1,4-glycosidic linkages in the amorphous regions of cellulose, generating shorter oligosaccharides and glucose units [4,5,6,7]. Due to its extensive industrial applications, the global demand for endoglucanase continues to rise; however, its production remains costly, highlighting the need for affordable production using inexpensive substrates and efficient technologies [8]. Microorganisms, particularly fungi, are preferred for endoglucanase production because of their ability to grow on diverse lignocellulosic wastes and secrete multiple extracellular cellulolytic enzymes that act synergistically to degrade cellulose [9]. Filamentous fungi such as Aspergillus and Rhizopus employ non-complex secreted enzyme systems, and their extensive hyphal networks enhance substrate penetration, enzyme–substrate interactions, and hydrolytic efficiency. Although oxidative enzymes such as LPMOs may assist cellulose depolymerization in some systems, endoglucanases primarily function through hydrolytic mechanisms [10].

The conversion of highly ordered polymers into monomers is inefficient in monoculture systems, whereas microbial consortia enable synergistic interactions that enhance product yield, substrate utilization, and process stability while reducing contamination risks [11]. Fungal co-culture, defined as the simultaneous growth of compatible fungal strains, offers advantages over single cultures by improving enzyme productivity, environmental adaptability, and substrate utilization, thereby increasing enzyme yield with reduced contamination [12]. Integrating fungal co-culture technology with lignocellulosic waste management can significantly advance the bio-based sector and promote environmentally friendly manufacturing within the circular economy framework [13].

A major barrier to enzyme production is the recalcitrant nature of lignocellulosic substrates, which can be overcome through pretreatment methods that enhance cellulose hydrolysis by primarily removing lignin [12,14]. Common pretreatment approaches include steam, acid, grinding, alkali, and hot water treatments, each yielding different outcomes due to their distinct physicochemical characteristics [15]. Among these, alkali pretreatment is considered the most promising because it employs mild, cost-effective, and non-polluting chemicals that directly react with lignin, facilitating its efficient removal [16,17]. Another major limitation in enzyme production is high cost, which can be addressed by selecting appropriate fermentation strategies that utilize inexpensive lignocellulosic residues. Although both submerged fermentation and solid-state fermentation (SSF) are effective, SSF is preferred due to lower energy and water requirements, minimal waste generation, and improved substrate utilization [18,19,20].

Response Surface Methodology (RSM) is widely used for bioprocess optimization because it enables the analysis of complex interactions among multiple variables using advanced statistical modeling, allowing accurate identification of optimal conditions that improve yield, productivity, and efficiency while reducing experimental trials and process variability [21]. Consequently, RSM offers significant economic and environmental benefits and has become an essential tool for bioprocess innovation across disciplines [22,23].

In this study, the potential of fungal co-culture technology for sustainable endoglucanase production from lignocellulosic waste is investigated as a cost-effective alternative to conventional methods, supporting eco-friendly manufacturing and circular bioeconomy development [13]. The novelty of this work lies in the utilization of pea hulls, an underexploited agro-industrial residue, as both a production substrate under solid-state fermentation and a hydrolysis feedstock through a synergistic co-culture of Aspergillus fumigatus and Rhizopus arrhizus. Unlike earlier studies based on conventional substrates or monocultures, this work integrates statistical optimization (PBD and CCD) with downstream applications, demonstrating an innovative and industrially relevant approach. Despite advances in enzyme production, studies on the simultaneous production of cellulase components using potent fungal co-cultures remain limited; therefore, this work addresses a critical knowledge gap by hypothesizing that strain selection and fermentation optimization can significantly enhance endoglucanase production and industrial applicability.

2. Materials and Methods

2.1. Isolation and Screening of Fungal Strains

Fungal strains having cellulolytic potential were isolated from various sources, such as soil, compost, etc., from different areas of Lahore, Pakistan. Isolation of cellulolytic fungi was carried out following ref [24] on Berg’s media plates. The zone of cellulose hydrolysis of colonies was determined by the Congo red method. Colonies showing the biggest zone of cellulose hydrolysis were picked and grown in Berg’s media slants for storing and maintaining the culture. The isolated strains were identified morphologically [25].

2.2. Compatibility Test

A compatibility assessment was performed to investigate the interactive effects among the isolated fungal strains. Strains displaying synergistic growth patterns, indicative of cooperative relationships, were selected for further study. Conversely, strains exhibiting antagonistic effects, hindering each other’s growth, were discarded [13]. The outcomes of the isolation, screening, and compatibility assays are summarized in Section 3.1.

2.3. Molecular Characterization of Efficient Fungal Co-Culture

The fungal strains constituting the co-culture having maximum cellulolytic potential were sent to Macrogen Korea for molecular characterization. The obtained sequences of the fungal strain were compared to reference sequences in NCBI-GenBank. Then, MEGA11 was utilized to construct a phylogenetic tree based on Internal Transcribed Spacer (ITS) regions [26].

2.4. Lignocellulosic Substrates and Pretreatment

Different agro-industrial, water and kitchen-based waste materials were used as a lignocellulosic substrate. These include wood, sawdust, newspaper, cardboard, sugarcane bagasse, algae, pea hulls, peanut shells, rice straw, and rice husk. The algal substrate consisted of locally collected mixed freshwater algal biomass, while the wood substrate comprised sawdust obtained from mixed hardwood species. All the lignocellulosic substrates were pretreated with an alkaline treatment method [2].

2.5. Estimation of Lignin and Cellulose

Both treated and untreated substrates were subjected to cellulose and lignin determination tests [27,28].

2.6. Inoculum Preparation

A 3-day-old fungal culture was used to prepare the inoculum. Saline solution (10 mL) was added to the slants, and the fungal spores were gently scraped off to create a homogeneous suspension [29].

2.7. Solid State Fermentation (SSF)

Solid-state fermentation was carried out in 250 mL Erlenmeyer flasks using 10 g (dry weight) of each lignocellulosic substrate as the solid matrix. The substrates were moistened with 10 mL of mineral salt medium to achieve the desired initial moisture content, as defined in the experimental design. The mineral salt medium consisted of (g/L): KH₂PO₄, 2.0; K₂HPO₄, 1.0; MgSO₄·7H₂O, 0.5; CaCl₂, 0.1; FeSO₄·7H₂O, 0.01; and NH₄NO₃, 1.0. The pH of the medium was adjusted prior to sterilization according to the experimental design. After sterilization, 0.5 mL (5 × 10⁵ spores/mL) of each fungal inoculum was added to the flasks. The contents were mixed thoroughly and incubated under static conditions at 30 °C for 72 h unless otherwise specified by the Plackett–Burman or Central Composite Design experiments. Subsequently, 100 mL of 0.1 M phosphate buffer, pH 7, was added to each flask and agitated in a shaking incubator for 1 h at 120 rpm and 30 °C. The fermented broth was then centrifuged at 6000 rpm for 10 min. The resulting supernatant was used for enzyme activity analysis [30]. All SSF experiments were performed in triplicate.

Endoglucanase Assay

The activity of CMCase was estimated by adding crude enzyme (0.5 mL) to a test tube. 0.5 mL of 1% CMC solution prepared in 0.1 M citrate buffer (pH 5) was added to the same test tubes. Blank was also run in parallel, to which 0.5 mL of distilled water was added instead of the enzyme. The test tubes were placed in a water bath for incubation at 60 °C for 30 min [31]. After incubation, the (DNS) reagent in a quantity of 2 mL was added to all test tubes. The solution was incubated in boiling water for 5 min. After cooling down, the volume of the reaction mixture was raised to 10. The absorbance of the reaction mixture was analyzed at 546 nm in a UV-visible spectrophotometer [32].

One unit (U) of CMCase activity was defined as the amount of enzyme required to release 1 µmol of glucose equivalent per minute under the assay conditions [31]. Enzyme activity was expressed as U/mL/min to explicitly denote the rate of enzymatic hydrolysis per unit volume and time.

2.8. Design of Experiments Methodology

2.8.1. Optimization of Physical Factors in Plackett-Burman Design

Optimization and screening of physical factors, along with their statistical analysis, were carried out in a Plackett-Burman design (PBD). A total of six factors, temperature, inoculum size, pH, inoculum age, substrate concentration, and moisture level, were analyzed (

Table 1. Physical factors for endoglucanase production through PBD.

Name	Units	Low Level	High Level
Temperature	°C	25	30
pH		4	7
Inoculum size	mL	1	2.5
Inoculum age (growth of fungal co-culture)	days	2	5
Substrate concentration	g	5	20
Moisture level	%	60	100

). A set of 12 trial runs was designed with a possible combination of 6 variables in Design Expert 11 software [32].

PDB model generated on the first- order polynomial equation was expressed in Equation (1).

Y = β₀ + Σ β_iX_i

(1)

where Y represents endoglucanase activity, β₀ shows model interception, β_i shows linear coefficient, and X_i represents the coded levels of the independent factors, where −1, 0, and +1 correspond to the low, center, and high levels, respectively.

2.8.2. Optimization of Significant Physical Factors in Central Composite Design

The optimum concentration of all the significant factors obtained through PBD was analyzed through a Central composite design (CCD). Significant factors, temperature, pH, and inoculum age were analyzed through Design Expert software version 11 (

Table 2. Physical factors for endoglucanase production through CCD.

Name	Units	Low Level	High Level
Temperature	°C	25	30
pH		4	7
Inoculum age	days	2	5

). A total of 18 trial runs were analyzed [32].

The effect and interaction of all three variables were analyzed through a second-degree polynomial equation (Equation (2)).

Y = β₀ + Σ_i β_iX_i + Σ_ii β_iiX_i² + Σ_ij β_ijX_iX_j

(2)

where Y represents the predicted response, β₀ shows the coefficients of regression, β_i shows the linear coefficient, β_ii shows the quadratic coefficient, β_ij shows the interaction coefficients, and X_i represents coded levels of independent variables. ANOVA was performed to assess model significance (p ≤ 0.05) and lack-of-fit. Response surface and contour plots were generated to visualize factor interactions. Optimum conditions predicted by the model were validated experimentally, and the actual yield was compared to the predicted value to determine model accuracy.

2.9. One Factor at a Time (OFAT) Analysis of Nutritional Factors

Influence of Supplementary Carbon, Nitrogen Sources and Surfactants

The influence of supplementary carbon and nitrogen sources, and surfactants on the yield of endoglucanase enzyme was evaluated by adding 1% concentration of all the above-mentioned sources individually into the fermentation media [32,33,34].

2.10. Optimization of Significant Nutritional Factors in CCD

The CCD of RSM was also used to optimize the significant nutritional factors obtained through OFAT analysis. A total of 4 factors, fructose, yeast extract, sodium nitrate, and Tween 80, were selected in CCD, which then generated a set of 28 run trials. The concentration range of these factors is shown in

Table 3. Nutritional factors for endoglucanase production through CCD.

Name	Units	Low Level	High Level
Fructose	%	0.5	2.5
Yeast extract	%	0.5	2
Sodium nitrate	%	0.5	2
Tween 80	%	0.5	2

. The effect and interaction of all four variables were analyzed through a second-degree polynomial equation [32]. For this part of the analysis, the same quadratic regression equation of the model (Equation (2)) was utilized.

2.11. Enzymatic Hydrolysis of Pretreated Biomass

To evaluate the industrial relevance of the optimized enzyme system, enzymatic hydrolysis experiments were carried out using pea hulls and rice straw as representative lignocellulosic feedstocks. Pea hulls were selected because they were identified as the best substrate for endoglucanase production in this study, showing high cellulose content and reduced lignin levels after alkaline pretreatment [35,36]. Rice straw, one of the most abundant agricultural residues globally, was chosen as a comparative biomass due to its widespread use in enzyme and bioethanol research [11,37]. Hydrolysis reactions were performed in 100 mL Erlenmeyer flasks containing 5% (w/v) substrate suspended in 50 mM citrate buffer (pH 5.0). Optimized enzyme extract, obtained under the best fermentation conditions, was added at a loading of 30 U/g of dry substrate. Reactions were incubated at 50 °C, 150 rpm for 72 [10]. Samples were collected at 0, 24, 48, and 72 h, boiled for 5 min to stop enzymatic activity, and centrifuged at 6000 rpm for 10 min. Reducing sugars released were quantified using the DNS method [32].

2.12. Calculation of Reducing Sugars and Hydrolysis Yield

Sugar concentrations were determined using a glucose calibration curve. The total sugars released were calculated as:

Total sugars (mg) = C × V

where C = sugar concentration (mg/mL) and V = hydrolysate volume (mL).

Sugar yield was expressed per unit of dry substrate:

Reducing sugars (mg/g dry substrate) = Total sugars (mg)\Substrate dry weight (g)

The hydrolysis yield was calculated as:

Hydrolysis yield (%) = Reducing sugars released (mg/g)\Theoretical glucose from cellulose (mg/g) × 100

Theoretical glucose values were estimated based on the cellulose composition of the pretreated substrates [16,28,37].

3. Results and Discussion

3.1. Isolation and Screening of Fungal Strains

Isolation of fungi is vital for endoglucanase production as it enables the selection of potent strains with high enzyme yields, specificity, and activity, while facilitating the identification of novel endoglucanase-producing species and the development of sustainable, scalable, and cost-effective production processes, contributing to a circular bioeconomy [12]. In this study, fungal strains with cellulolytic potential were isolated from different sources and qualitatively screened based on clear zones of cellulose hydrolysis. To enhance enzyme production, synergistic interactions among fungal strains were evaluated, as fungal co-culture systems mimic natural symbiosis and significantly improve enzyme yield, diversity, and substrate utilization efficiency [13].

The fungal strain pairs selected for co-cultivation were compiled following a systematic compatibility-based screening approach. Initially, individual isolates were screened for cellulolytic potential (

Table 4. (a): Monoculture CMCase activity (U/mL/min) of individual fungal strains. (b): Screening of fungal co-cultures having cellulolytic potential.

(a)
Monoculture No.	Fungal Strain	CMCase Activity (U/mL/min)
Monoculture 1	Aspergillus fumigatus	17.76 ± 0.2
Monoculture 2	Alternaria alternata	13.16 ± 0.1
Monoculture 3	Fusarium oxysporum	10.95 ± 0.2
Monoculture 4	Rhizopus arrhizus	19.17 ± 0.2
Monoculture 5	Phytophthora inundata	8.74 ± 0.1
(b)
Strains		CMCase U/mL/min
Co-culture 1 Aspergillus fumigatus + Alternaria alternata		45.86 ± 0.2
Co-culture 2 Aspergillus fumigatus + Fusarium oxysporum		22.76 ± 0.2
Co-culture 3 Aspergillus fumigatus + Phytophthora inundata		44.23 ± 0.2
Co-culture 4 Aspergillus fumigatus + Rhizopus arrhizus		50.63 ± 0.2
Co-culture 5 Fusarium oxysporum + Rhizopus arrhizus		44.34 ± 0.1
Co-culture 6 Fusarium oxysporum + Phytophthora inundata		11.76 ± 0.1
Co-culture 7 Fusarium oxysporum + Alternaria alternata		32.87 ± 0.2
Co-culture 8 Rhizopus arrhizus + Phytophthora inundata		30.78 ± 0.2
Co-culture 9 Rhizopus arrhizus + Alternaria alternata		30.27 ± 0.1
Co-culture 10 Alternaria alternata + Phytophthora inundata		23.67 ± 0.1

The mean difference of co-cultures is significant at p ≤ 0.05 level; ± indicates the standard deviation among each replicate.

a). Strains exhibiting pronounced cellulolytic activity were then evaluated in pairwise combinations to assess growth compatibility, absence of antagonistic interactions, and complementary enzyme-producing behavior under solid-state fermentation conditions. Only those strain combinations that showed stable co-growth and enhanced CMCase activity relative to their respective monocultures were selected. Co-culturing provides a cost-effective alternative to monocultures, with improved substrate utilization (30–40%) that reduces waste, pollution, and environmental impact, while enhancing tolerance to environmental stress [11,12,35]. Screening of compatible co-cultures revealed that co-culture 4 (LBT-4) exhibited the highest CMCase activity (50.63 U/mL/min) under solid-state fermentation conditions (Table 4b). The best-performing strains were identified based on morphological characteristics (Figure 1a–d) and were macroscopically identified as Rhizopus arrhizus and Aspergillus fumigatus. R. arrhizus colonies appeared cottony, white to grayish-white, irregular in shape, and possessed coenocytic hyphae, whereas A. fumigatus formed greenish-gray, velvety colonies (2–5 cm diameter) with septate, hyaline hyphae, dense columnar conidial heads, spherical smooth conidia arranged in chains, and globose to sub-globose vesicles bearing metulae and phialides.

The CMCase activity obtained from the co-culture substantially exceeded typical Rhizopus monoculture values and fell within the mid-to-high range of reported Aspergillus monocultures, highlighting the advantage of synergistic interactions. In this strategy, A. fumigatus contributes diverse extracellular cellulolytic enzymes, while R. arrhizus enhances substrate accessibility through extensive hyphal penetration, resulting in a 30–40% improvement in substrate utilization efficiency [11,12]. Similar enhancements have been reported for microbial consortia, where complementary metabolic roles improve enzyme induction and productivity [35]. The elevated CMCase activity and reduced fermentation time observed in co-culture 4 underscore its industrial relevance and suitability for low-cost lignocellulosic substrates under solid-state fermentation conditions [35].

SEM analysis revealed detailed structural features of fungal mycelia and sporangia, aiding accurate visualization and identification (Figure 1e,f) [38]. However, as morphological methods alone may lead to uncertain classification [39], molecular characterization was performed. Following morphological identification, co-culture 4 was subjected to 18S rDNA sequencing, and ITS region comparison using the GenBank database showed 99.84% similarity with Aspergillus fumigatus and 99.71% similarity with Rhizopus arrhizus. The sequences were submitted to GenBank under accession numbers PP943377 and PP940183, respectively. Phylogenetic trees were constructed using MEGA 11 by the Maximum Composite Likelihood method (Figure 1g,h), which is preferred due to its high probability rate and accurate estimation of minimal evolutionary changes [24].

3.2. Alkaline Pretreatment Effects on Lignin–Cellulose Composition and Morphology

Using lignocellulosic substrates for enzyme production is a cost-effective and sustainable approach that converts waste biomass into valuable products. Pretreated lignocellulosic substrates undergo enhanced hydrolytic breakdown and become more amenable to enzymatic hydrolysis by reducing cellulose crystallinity and increasing substrate surface area [37,40]. Among pretreatment techniques, alkaline pretreatment is widely preferred due to its simple procedure and strong delignification effect, as selective lignin removal increases surface area without damaging the existing cellulose content [16]. In this study, alkali pretreatment was applied to several lignocellulosic substrates to delignify and expose cellulose, with pea hulls exhibiting the highest cellulose content owing to their fibrous nature and high initial lignin content, which, when removed, revealed cellulose-rich fiber surfaces [16,36]. Enhanced cellulose accessibility after alkaline pretreatment improves enzyme–substrate interactions, accelerates hydrolysis rates, and reduces inhibitory compounds, ensuring higher endoglucanase yields during fermentation [41]. Lignin and cellulose contents determined before and after treatment showed that pea hulls had the lowest lignin content (18%) and the highest cellulose content (above 80%) after pretreatment (Figure 2a,b), likely due to efficient delignification and lignin solubilization that enriched cellulose availability and improved hydrolytic efficiency [40]. When these alkali-pretreated substrates were combined with the optimized fungal co-culture system of A. fumigatus and R. arrhizus, a synergistic effect was observed, resulting in significantly higher CMCase production and reduced fermentation time, demonstrating a scalable and industrially relevant strategy [42]. SEM analysis further confirmed major morphological and topographical alterations in alkali-treated substrates, clearly visualizing increased surface area and structural disruption compared to untreated materials (Figure 2c,d), supporting improved substrate accessibility after pretreatment [41].

3.3. Screening of Substrates for Endoglucanase Production

Appropriate substrate selection is crucial for enzyme production, influencing enzyme yield, specificity, and productivity. Optimal substrates provide essential nutrients, moisture, and structural support, enhancing microbial growth and enzyme synthesis [43]. In the current study, different substrates, including wheat straw, pea hull, algae, cardboard, newspaper, sugarcane bagasse, rice husk, rice straw, and peanut shell, were screened for endoglucanase production. Among all tested substrates, the treated pea hull exhibited maximum endoglucanase activity (Figure 3a). Pea hulls showed a post-treatment cellulose content exceeding 80% and lignin content reduced to 18%, enabling superior enzyme–substrate interaction. This optimized composition enabled efficient enzyme–substrate interactions, sustained robust microbial proliferation, and maximized enzyme biosynthesis. The high cellulose-to-lignin ratio of treated pea hull likely supports enhanced microbial metabolism and prolonged enzyme secretion during fermentation. Furthermore, the combination of structural accessibility, nutrient availability, and moisture retention creates a highly favorable microenvironment for fungal growth, explaining its superior performance over other substrates. Additionally, pea hull’s favorable nutrient profile and moisture retention properties contributed to its superior performance. This makes pea hull an ideal substrate for endoglucanase production [35,36,37].

To statistically confirm the superiority of pea hulls as a substrate for endoglucanase production, a one-way ANOVA was performed to compare enzyme activities among the four tested lignocellulosic substrates: pea hulls, rice husk, sugarcane bagasse, and cardboard. The analysis revealed a highly significant difference in enzyme activity among the substrates (F = 1241.10, p = 5.2 × 10⁻¹¹) (

Table 5. One-way ANOVA comparing endoglucanase activity among different substrates.

Source	Sum of Squares	df	Mean Square	F-Value	p-Value
Substrate	536.776	3	178.93	1241.10	5.2 × 10−11
Residual	1.153	8	0.144	—	—

). Pairwise Student’s t-tests further demonstrated that pea hulls yielded significantly higher endoglucanase activity than rice husk (t = 37.26, p = 3.1 × 10⁻⁶), sugarcane bagasse (t = 47.32, p = 1.2 × 10⁻⁶), and cardboard (t = 58.89, p = 5.0 × 10⁻⁷) (

Table 6. Pairwise Student’s t-tests comparing pea hulls with other substrates.

Comparison	t-Value	p-Value	Significance Level
Pea hulls vs. Rice husk	37.26	3.1 × 10−6	p < 0.001
Pea hulls vs. Sugarcane bagasse	47.32	1.2 × 10−6	p < 0.001
Pea hulls vs. Cardboard	58.89	5.0 × 10−7	p < 0.001

). These analyses were performed using IBM SPSS Statistics version 26.0 to ensure statistical robustness. The results confirm that the observed variation in enzyme activity was statistically significant and not due to random error, validating pea hulls as the most favorable substrate for fungal enzyme synthesis. The superior performance of pea hulls may be attributed to their favorable carbon–nitrogen ratio and porous cell wall structure following alkaline pretreatment, which enhances enzyme induction and substrate accessibility. Overall, these findings provide quantitative evidence supporting the selection of pea hulls for further optimization studies.

3.4. Effect of Time of Incubation

Time of incubation plays a vital role in enzyme production. The Optimum time of incubation in fermentation can boost the yield of the enzyme. Insufficient incubation restricts the yield of the enzyme, while incubation for a prolonged period can destroy the enzyme and fungal co-culture [44]. The effect of different incubation periods (0–120 h) was analyzed on endoglucanase production by using fungal co-culture (LBT 4) (Figure 3b). Enzyme production increased gradually up to 72 h. At this time of incubation, the best activity of CMCase was 56.80 U/mL/min. There was a decrease in the production of the enzyme after 72 h of incubation. This might be during initial times; the fermentation media is enriched with essential micro and macro nutrients, which enhance the fungal growth and endoglucanase production. During the early hours, mycelia production is high as the spore germinates rapidly. With a gradual increase in time of incubation, mycelia production also increases until it reaches an optimal time. Thus, extended incubation time negatively impacts enzyme production because nutrient depletion forces fungi into their stationary phase, where metabolic priorities shift from biosynthesis to maintenance and survival. Additional factors such as product inhibition, proteolytic degradation of enzymes, and unfavorable changes in culture conditions further contribute to reduced yields [45]. Our findings are also aligned with ref [46], who obtained maximum endoglucanase production from the fungus through SSF by incubating it for 72 h. However, ref [47] disagrees with our results as they obtained a maximum yield of enzyme by fermentation after 120 h of incubation.

3.5. Screening of Physical Factors Using Plackett-Burman Design (PBD)

In this study, a Plackett-Burman Design (PBD) with 12 trial runs was employed to investigate the effects of six key variables: temperature, pH, inoculum size, inoculum age, substrate concentration, and moisture content. The results revealed that the combination of factors in trial run 10 yielded the highest response value (69.54 U/mL/min), indicating optimal conditions. Conversely, trial run six resulted in the lowest production value (45.73 U/mL/min) (

Table 7. (a): Physical factors used in PBD for endoglucanase production. (b): ANOVA for CMCase activity. (c): Summary of ANOVA of CMCase activity.

(a)
Run	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	Factor 6	Actual Value	Predicted Value
	A: Temperature	B: pH	C: Inoculum Size	D: Inoculum Age	E: Substrate Conc.	F: Moisture Level	CMCase	CMCase
	C	-	mL	days	g	%	U/mL/min	U/mL/min
1	30	4	1	2	20	60	58.34	58.00
2	25	4	2.5	2	20	100	50.67	49.09
3	25	7	1	5	20	60	61.87	63.09
4	30	4	2.5	5	5	100	57.86	60.27
5	25	7	2.5	2	20	100	54.97	56.55
6	25	4	1	2	5	60	45.73	47.26
7	30	7	2.5	2	5	60	66.79	64.73
8	25	4	1	5	5	100	52.48	50.27
9	25	7	2.5	5	5	60	62.89	62.35
10	30	7	1	5	20	100	69.54	68.47
11	30	4	2.5	5	20	60	65.43	65.62
12	30	7	1	2	5	100	59.24	60.11
(b)
Source		Sum of squares		df	Mean square	F-value	p-value
Model		515.99		6	86.00	16.46	0.0037	significant
A-Temperature		196.75		1	196.75	37.65	0.0017
B-pH		167.18		1	167.18	31.99	0.0024
C-Inoculum size		10.85		1	10.85	2.08	0.2092
D-Inoculum age		98.21		1	98.21	18.79	0.0075
E-Substrate concentration		20.88		1	20.88	4.00	0.1021
F-Moisture level		22.11		1	22.11	4.23	0.0948
Residual		26.13		5	5.23
Cor total		542.12		11
(c)
Std. dev.		2.29		R2				0.9518
Mean		58.82		Adjusted R2				0.8940
C.V. %		3.89		Predicted R2				0.7224
				Adeq precision				12.1467

a). The Pareto chart displays all tested factors, including statistically insignificant ones, A (temperature), B (pH), and D (inoculum age), as part of the design output. (Figure 4a). The endoglucanase production showed a significant increase at 30 °C temperature, 7 pH, and 5 days of inoculum age. Polynomial model (Equation (3)) regression analysis was used to predict the production of CMcase in terms of coded factors.

CMCase activity = +58.82 + 4.05A + 3.73B + 0.9508C + 2.86D + 1.32E − 1.36F

(3)

where A, B, C, D, E, and F represent the coded levels of temperature, pH, inoculum size, inoculum age, substrate concentration, and moisture level.

The statistical analysis revealed that the model was highly significant (p = 0.0037), as evidenced by a substantial F-value. Moreover, temperature (p = 0.0017), pH (p = 0.0024), and inoculum age (p = 0.0075) were the key factors influencing CMCase activity (Table 7b). The high coefficient of determination (R² = 0.954) indicated a strong correlation between actual and predicted values, explaining 95.40% of the response variability (Table 7c). Our findings align with ref [48], who utilized Plackett-Burman Design for optimizing cultural parameters. However, the significance of specific factors differed, likely due to variations in substrate and fermentation factor ranges, as noted by ref [21].

3.6. Optimization of Significant Physical Factors in CCD

The screening of significant factors, i.e., temperature, pH, and inoculum age, was further carried out through CCD, and the regression equation was selected that relates the response variables to independent variables. For this design, a rotatable axial distance (α) of 1.682 was applied. A set of 18 runs with a synergistic effect of independent variables was selected (

Table 8. (a): Optimization of significant physical factors in CCD. (b): ANOVA of quadratic model for CMCase activity. (c): Statistic summary of CMCase activity.

(a)
Run	Factor 1	Factor 2	Factor 3		Actual Value	Predicted Value
	A: Temperature	B: pH	C: Inoculum Age		CMCase	CMCase
	°C		days		U/mL/min	U/mL/min
1	27.5	5.5	3.5		81.29	80.84
2	30	4	5		70.56	70.31
3	31.7045	5.5	3.5		64.87	64.78
4	27.5	5.5	3.5		79.98	80.84
5	27.5	5.5	3.5		77.43	80.84
6	27.5	8.02269	3.5		57.87	54.74
7	23.295	5.5	3.5		39.78	38.69
8	25	4	5		53.86	54.93
9	27.5	2.97731	3.5		59.76	61.71
10	30	7	5		62.98	65.77
11	25	7	2		42.59	43.67
12	25	4	2		49.38	47.43
13	27.5	5.5	6.02269		75.78	72.66
14	27.5	5.5	3.5		84.47	80.84
15	27.5	5.5	0.977311		58.98	60.92
16	25	7	5		50.98	53.29
17	30	7	2		59.56	59.32
18	30	4	2		67.45	65.97
(b)
Source	Sum of squares	df	Mean square	F-value	p-value
Model	2893.50	9	321.50	33.99	<0.0001	significant
A-Temperature	821.56	1	821.56	86.86	<0.0001
B-pH	58.72	1	58.72	6.21	0.0374
C-Inoculum age	166.28	1	166.28	17.58	0.0030
AB	4.20	1	4.20	0.4446	0.5237
AC	5.02	1	5.02	0.5312	0.4869
BC	2.23	1	2.23	0.2353	0.6406
A2	1339.57	1	1339.57	141.62	<0.0001
B2	808.91	1	808.91	85.52	<0.0001
C2	312.28	1	312.28	33.02	0.0004
Residual	75.67	8	9.46
Lack of fit	49.93	5	9.99	1.16	0.4805	not significant
Pure error	25.74	3	8.58
Cor total	2969.17	17
(c)
Std. dev.		3.08	R2			0.9745
Mean		63.20	Adjusted R2			0.9458
C.V. %		4.87	Predicted R2			0.8560
			Adeq precision			18.3903

a). Polynomial equation (Equation (4)) of second order was generated from this analysis.

CMCase activity = +80.84 + 7.76A − 2.07B + 3.49C − 0.7250AB − 0.7925AC + 0.5275BC − 10.29A² − 8.00B² − 4.97C²

(4)

where A, B, and C represented the coded levels of factors: temperature, pH, and inoculum age.

Statistical analysis of the model was also carried out (Table 8b), which showed a highly significant model (p < 0.0001) with temperature (p < 0.0001), pH (p = 0.0374), and inoculum age (p = 0.0030) as significant individual factors influencing CMCase activity. The F-value of the model was 33.99, and the coefficient of determination (R²) was close to 1, which indicated that the experimental values and predicted values of the responses were in a good fit and confirmed its robustness and predictive capability (Table 8c). 3D surface graphs represent the interaction of factors individually with each other (Figure 4b–d). The cool color in image represents lower response values, the warm color represents moderate response values and dark color represents high response values. Using the optimizer tool in Design-Expert 11, the model predicted a maximum CMCase activity of 84.47 U/mL/min at 27.5 °C, pH 5.5, and an inoculum age of 3.5 days. Validation experiments at these settings gave 81.29 U/mL/min, confirming the model’s accuracy (error < 4%). This represents a 20–25% improvement over the best result from the initial screening. Building on this, the nutritional optimization in Section 3.8 boosted the yield further to 119.58 U/mL/min, giving an overall gain of nearly 48% compared to the unoptimized baseline [21].

3.7. OFAT Analysis of Nutritional Factors

Using a one-factor-at-a-time (OFAT) approach, the effects of nutritional components and surfactants on endoglucanase production were evaluated. Among the tested carbon sources, fructose resulted in the highest enzyme activity (Figure 5a), owing to its rapid assimilation, high energy efficiency, minimal catabolite repression, and induction of cellulase gene expression [49,50,51,52]. Nitrogen source selection significantly influenced enzyme biosynthesis, with yeast extract identified as the most effective organic nitrogen source (Figure 5b) due to its balanced nutritional composition and growth-promoting factors [53,54,55]. Among inorganic nitrogen sources, sodium nitrate supported maximum endoglucanase production (97.43 U/mL/min; Figure 5c), likely due to its gradual nitrogen release, pH-buffering capacity, and low toxicity compared to ammonium salts [51,56,57]. Surfactant supplementation further enhanced enzyme yield, with Tween 80 producing the highest endoglucanase activity (Figure 5d) by improving substrate accessibility and membrane permeability while minimizing cellular toxicity [21,58,59]. Overall, optimization of carbon, nitrogen, and surfactant components significantly improved endoglucanase production under OFAT conditions.

3.8. Optimization of Significant Nutritional Factors in CCD

The important nutritional components that affect the formation of endoglucanase, such as fructose, yeast extract, sodium nitrate, and Tween 80, were examined using a central composite design (CCD). The regression analysis yielded a second-order polynomial equation (Equation (5)) that correlates the response variable to the independent variables.

CMCase activity = +105.78 − 12.60A − 1.73B + 1.23C − 0.6033D + 3.94AB −
0.8050AC − 4.93AD + 0.0413BC + 0.2500BD + 0.5538CD − 5.31A² − 2.89B² −
3.37C² 2.45D²

(5)

where A, B, C, and D represented the coded level of fructose, yeast extract, sodium nitrate, and Tween 80 concentrations, respectively.

A total of 28 trial runs were selected, revealing a synergistic effect among the variables (

Table 9. (a): Optimization of significant nutritional factors in CCD. (b): ANOVA of quadratic model for CMCase activity. (c): Statistic summary of CMCase activity.

(a)
Run	Factor 1	Factor 2	Factor 3	Factor 4	Actual Value	Predicted Value
	A: Fructose	B: Yeast Extract	C: Sodium Nitrate	D: Tween 80	CMCase	CMCase
	%	%	%	%	U/mL/min	U/mL/min
1	1.5	1.25	−0.25	1.25	83.45	89.83
2	2.5	0.5	0.5	2	67.34	70.24
3	1.5	2.75	1.25	1.25	91.45	90.78
4	1.5	1.25	1.25	1.25	103.67	106.62
5	2.5	2	2	0.5	86.46	86.58
6	0.5	0.5	2	0.5	108.56	107.40
7	2.5	0.5	2	0.5	79.36	82.57
8	0.5	2	2	2	105.98	105.90
9	−0.5	1.25	1.25	1.25	107.67	109.74
10	0.5	2	0.5	0.5	93.78	92.60
11	2.5	0.5	2	2	73.48	72.11
12	0.5	2	0.5	2	100.86	100.64
13	2.5	2	0.5	0.5	89.67	86.76
14	3.5	1.25	1.25	1.25	61.87	59.36
15	2.5	2	0.5	2	76.48	75.09
16	2.5	2	2	2	69.78	77.13
17	0.5	0.5	0.5	0.5	108.89	104.52
18	1.5	1.25	1.25	1.25	106.57	106.62
19	0.5	0.5	0.5	2	114.23	111.56
20	1.5	1.25	1.25	1.25	96.67	106.62
21	1.5	−0.25	1.25	1.25	97.45	97.69
22	2.5	0.5	0.5	0.5	85.37	82.90
23	0.5	2	2	0.5	95.56	95.64
24	0.5	0.5	2	2	110.76	116.65
25	1.5	1.25	2.75	1.25	101.56	94.75
26	1.5	1.25	1.25	1.25	119.58	106.62
27	1.5	1.25	1.25	−0.25	92.65	97.21
28	1.5	1.25	1.25	2.75	99.78	94.79
(b)
Source	Sum of squares	df	Mean square	F-value	p-value
Model	5486.33	14	391.88	8.87	0.0002	significant
A-Fructose	3807.22	1	3807.22	86.18	<0.0001
B-Yeast extract	71.48	1	71.48	1.62	0.2256
C-Sodium nitrate	36.36	1	36.36	0.8230	0.3808
D-Tween 80	8.74	1	8.74	0.1977	0.6639
AB	248.85	1	248.85	5.63	0.0337
AC	10.37	1	10.37	0.2347	0.6361
AD	388.29	1	388.29	8.79	0.0110
BC	0.0272	1	0.0272	0.0006	0.9806
BD	1.0000	1	1.0000	0.0226	0.8827
CD	4.91	1	4.91	0.1111	0.7443
A2	730.57	1	730.57	16.54	0.0013
B2	230.24	1	230.24	5.21	0.0399
C2	308.20	1	308.20	6.98	0.0203
D2	169.31	1	169.31	3.83	0.0721
Residual	574.32	13	44.18
Lack of fit	298.65	10	29.87	0.3250	0.9227	not significant
Pure error	275.67	3	91.89
Cor total	6060.65	27
(c)
Std. dev.		6.65		R2		0.9052
Mean		93.89		Adjusted R2		0.8032
C.V. %		7.08		Predicted R2		0.6353
				Adeq precision		11.7764

a). The maximum production of endoglucanase was obtained at the following concentration of nutritional factors: 1.5% fructose, 1.25% yeast extract, 1.25% sodium nitrate and 1.25% Tween 80. Further increase or decrease resulted in a decline in endoglucanase production. The decrease in enzyme production could be due to inhibitory effects, where excess nutrients repress cellulase gene expression or compete with cellulose for enzyme binding sites. Additionally, catabolite repression, osmotic stress, and nitrogen overload may also contribute to this decline [60]. Furthermore, surfactant inhibition caused by excessive Tween 80 may disrupt fungal cell membrane function and impair enzyme activity, ultimately reducing endoglucanase yields [61].

The optimum nutritional conditions were re-evaluated using the optimizer tool in Design-Expert 11 rather than relying solely on the highest single experimental. The model predicted a maximum CMCase activity of 119.58 U/mL/min at 1.52% fructose, 1.26% yeast extract, 1.24% sodium nitrate, and 1.27% Tween 80. Validation under these settings confirmed the model’s accuracy, with experimental activity closely matching the prediction (error < 3%). The optimum nutritional conditions predicted by the optimization tool were experimentally validated under the previously optimized physical parameters, including an incubation time of 72 h as determined in Section 3.5. Under these conditions, a maximum endoglucanase activity of 119.58 U/mL/min was obtained, closely matching the predicted value with a deviation of less than 3%, thereby confirming the accuracy and reliability of the optimization model. These optimized nutritional parameters, when combined with the previously optimized physical conditions, delivered the highest overall yield, representing an approximate 48% increase compared to the unoptimized baseline.

Statistical analysis of the model was also carried out (Table 9b), which indicated a high F value and a low p value with fructose concentration (p < 0.0001) as the most significant nutritional factor. Interaction effects such as fructose × yeast extract (p = 0.0337) and fructose × Tween 80 (p = 0.0110) were also significant, indicating synergistic effects on enzyme production. The coefficient of determination (R²) was close to 1 in the model, which indicated that the experimental values and predicted values of the responses were in a good fit (Table 9c). The 3D surface graphs represent the effect and interaction of factors individually and with each other (Figure 6a–f). The cool color in image represents lower response vlues, the warm color represents moderate response values and dark color represents high response values.

3.9. Hydrolytic Efficiency of Optimized Enzyme

Hydrolysis of pretreated biomass showed a progressive increase in reducing sugars over 72 h (

Table 10. Reducing sugar yield from enzymatic hydrolysis of lignocellulosic biomass.

Substrate (5% w/v)	Pretreatment	Reducing Sugars (mg/g Dry Substrate, 72 h)	Hydrolysis Yield (%)
Pea hulls	Alkali	412.3 ± 8.6	68.5
Rice straw	Alkali	356.8 ± 7.4	59.7
Pea hulls	Untreated	143.2 ± 5.1	23.8
Rice straw	Untreated	129.4 ± 4.8	21.6

), with pea hulls releasing 412.3 mg/g (calculated as glucose equivalents released per gram of dry substrate) dry substrate (68.5% yield, and rice straw 356.8 mg/g (59.7%) (Figure 7). In contrast, untreated substrates produced markedly lower sugars, demonstrating that alkaline pretreatment enhanced delignification and enzyme accessibility, resulting in nearly threefold higher release [16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37]. The high saccharification efficiency of pea hulls emphasizes their dual role as a substrate for enzyme production in solid-state fermentation and as a feedstock for hydrolysis [35,36]. The fermentable sugars generated can be utilized in biofuels and other value-added products [5,10], reinforcing the industrial relevance of the optimized enzyme system and its alignment with circular economy and waste valorization concepts.

3.10. Biosafety Considerations for Aspergillus Fumigatus

Aspergillus fumigatus is an opportunistic pathogen and requires careful handling under appropriate laboratory containment. All experiments in this study were conducted strictly at laboratory scale using standard microbiological safety procedures and controlled fermentation conditions. Any application of this co-culture system beyond the laboratory should take into account potential health risks, regulatory limitations, and the possible need to substitute non-pathogenic or industrial Aspergillus strains for large-scale enzyme manufacturing. These considerations are important when evaluating the practical extension of the process to industrial or field environments.

4. Conclusions

This study highlights the strong potential of lignocellulosic biomass, particularly agricultural residues, in advancing sustainable enzyme production and promoting a circular bioeconomy. The successful isolation and identification of Aspergillus fumigatus and Rhizopus arrhizus, combined with the effective utilization of alkaline-treated pea hulls, provide a promising approach for maximizing endoglucanase yield. Optimization of process parameters using Plackett–Burman Design (PBD) and Central Composite Design (CCD) offered significant insights into factors influencing enzyme production. The optimum conditions determined were 27.5 °C, pH 5.5, and an inoculum age of 3.5 days, with supplementation of 1.5% fructose, 1.25% yeast extract, 1.25% sodium nitrate, and 1.25% Tween 80, resulting in a maximum endoglucanase activity of 119.58 U/mL/min. These optimized parameters establish a clear roadmap for achieving high yield, cost-effective enzyme production. The strong correlation between experimental and predicted data (R² = 0.9052) confirms the robustness of the optimization approach. Furthermore, the evaluation of multiple substrates, including wheat straw, algae, cardboard, and sugarcane bagasse, underscores the versatility of lignocellulosic feedstocks. Overall, this work demonstrates the dual valorization potential of pea hulls as both a substrate for enzyme synthesis and a hydrolysis feedstock, reinforcing their industrial applicability and contribution to sustainable bioprocessing.