Efficient Enzymatic Hydrolysis and Polyhydroxybutyrate Production from Non-Recyclable Fiber Rejects from Paper Mills by Recombinant Escherichia coli

Abstract

Non-recyclable fiber rejects from paper mills, particularly those from recycled linerboard mills, contain high levels of structural carbohydrates but are currently landfilled, causing financial and environmental burdens. The aim of this study was to develop efficient and sustainable bioprocess to upcycle these rejects into polyhydroxybutyrate (PHB), a biodegradable alternative to degradation-resistant petroleum-based plastics. To achieve high yields of PHB per unit biomass, the specific objective of the study was to investigate various approaches to enhance the hydrolysis yields of fiber rejects to maximize sugar recovery and evaluate the fermentation performance of these sugars using Escherichia coli LSBJ. The investigated approaches included size reduction, surfactant addition, and a chemical-free hydrothermal pretreatment process. A two-step hydrothermal pretreatment, involving a hot water pretreatment (150 °C and 15% solid loading for 10 min) followed by three cycles of disk refining, was found to be highly effective and resulted in an 83% cellulose conversion during hydrolysis. The hydrolysate obtained from pretreated biomass normally requires a detoxification step to enhance fermentation efficiency. However, the hydrolysate obtained from the pretreated biomass contained minimal to no inhibitory compounds, as indicated by the efficient sugar fermentation and high PHB yields, which were comparable to those from fermenting raw biomass hydrolysate. The structural and thermal properties of the extracted PHB were analyzed using various techniques and consistent with standard PHB.

1. Introduction

Plastic is a ubiquitous polymer derived from petroleum. Since the 1950s, global plastic production has increased at an average rate of around 9% annually. It is projected that worldwide production will reach 540 million metric tons by 2040 [1]. Its high rate of consumption and resistance to natural degradation would result in approximately 11,000 million metric tons of non-biodegradable plastics accumulating in landfills and in the natural environment by 2025, leading to plastic pollution and threatening land and marine ecosystems [2,3]. As environmental concerns escalate, there is a growing push to replace traditional petroleum-based plastics with non-toxic, biosynthetic, and biodegradable alternatives. Polyhydroxyalkanoates (PHAs) are regarded as one of the most promising alternatives to conventional, non-biodegradable petroleum-based plastics, with desirable eco-friendly characteristics.

PHAs are aliphatic polyesters that can be synthesized by a variety of microorganisms as intracellular carbon and energy reservoirs in the form of cytoplasmic granules under nutrient-limiting conditions and are fully biodegradable [1,4,5]. Polyhydroxybutyrate (PHB), the most prevalent and earliest recognized type of PHA, has garnered significant attention due to its promising physical and mechanical properties, making it a potential substitute for traditional plastics [6]. However, the relatively high cost of production of PHBs in comparison to petrochemical-derived plastics hinders their widespread production and commercialization. Since the cost of the feedstock contributes as much as 50% in the PHB production, utilizing zero/low-cost waste materials, especially agro-industrial waste, as a substrate is one method to lower production costs [1,7]. Most of these waste streams are rich in carbohydrates (mostly structural carbohydrates) that can be hydrolyzed to produce fermentable sugars and subsequently fermented to PHAs [8,9,10]. These waste streams provide a sustainable zero/negative-cost alternative to currently used food-based feedstock for PHB production and also address the waste management issue for agro-processing industries. Non-recyclable fiber rejects from paper mills, especially from recycled linerboard mills, contain high amounts of structural carbohydrates and are one such high potential feedstock for PHB production.

During the production of recycled paper, the bonding tendency between fibers may decrease in the repulping or de-inking unit operation. In order to compensate for the reduced bonding tendency, paper mills usually apply compression and shear forces to the pulp in the refining process. Although most recycled fibers have been refined, this measure generates a large number of short fibers [11]. Retention of these short fibers reduces the dewatering performance of the paper, impacting drying efficiency and productivity. The short fibers also contribute to decreased strength properties, such as tear strength and fold resistance. Therefore, these short fibers are unsuitable for recycling and are rejected by paper mills [12]. Due to the shortage of paper raw materials and the need to protect forests, the recycling and utilization of wastepaper have been strongly encouraged. In the past decades, the amount of recycled paper has increased, with the recovery rate rising from 33.5% in 1990 to 66.2% in 2019, peaking at 68% during this period. In 2014, recovered paper consumption amounted to 257 million tons worldwide (57.9% of the total paper production). Approximately 15–20% of the reused fiber becomes too short to be useful after the repulping process and is rejected from recycled paper mills, leading to about 34 to 46 million tons of fiber rejects produced very year. Therefore, the disposal of these fibers has placed a significant burden on the pulp and paper industry. Currently, these rejected fibers are sent to landfills. Landfill disposal can be expensive and contributes to land use challenges and methane emissions as organic materials decompose anaerobically [13]. Upcycling these fiber rejects into PHB could offer several benefits, including reducing PHB production costs, eliminating waste management issues, and promoting a circular economy within the paper industry.

Similar to any lignocellulosic biomass, utilization of fiber rejects for PHB production is constrained by low sugar recoveries during enzymatic hydrolysis [13]. This is primarily because of the recalcitrant structure of biomass consisting of cellulose, hemicellulose, and lignin, which are tightly intertwined and bound to each other by covalent or non-covalent bonds to form the lignocellulosic matrix [14]. The structural complexity of lignocellulosic biomass makes it highly resistant to enzymes, resulting in the relatively low digestibility of raw lignocellulosic feedstocks [15]. High ash content due to the presence of inorganics is another challenge in utilizing fiber rejects. To overcome this recalcitrance and make polysaccharides readily accessible for enzyme digestion, an appropriate pretreatment approach is needed, such as grinding, microwaving, steam explosion, hydrothermal pretreatment, ammonia fiber/freeze explosion, acid pretreatment, alkali pretreatment, organosolv processes, and biological pretreatment [16,17,18,19]. Although these pretreatments enhance the subsequent enzymatic hydrolysis efficiency, most of these suffer from the limitations of harsh operating conditions, need for expensive reactors, and production of compounds that are inhibitory to microbial fermentation [20].

This study intended to demonstrate the feasibility of using fiber reject hydrolysate as the sole carbon source for PHB production. To achieve high productivity from biomass, the first objective of this study was to investigate three approaches to enhance the hydrolysis yields of fiber rejects. Firstly, the effect of size reduction on the enzymatic hydrolysis efficiency was investigated because it is hypothesized that smaller fibers will be more labile to enzymatic degradation. Secondly, the addition of surfactants was explored to mitigate the negative effects of paper fillers, such as calcium carbonate, by preventing non-productive enzyme adsorption. The third approach to enhance hydrolysis efficiency focused on using a chemical-free two-step hydrothermal pretreatment process. This process uses a combination of hot water pretreatment and disk milling/refining to reduce biomass recalcitrance, solubilize hemicellulose, and increase the specific surface area of biomass predicted to lead to efficient enzymatic hydrolysis [20,21,22,23]. Various pretreatment conditions were investigated to determine the optimal conditions for maximum cellulose conversion. In the second objective of study, the hydrolysate obtained from the optimal process was used as the fermentation medium for PHB bioproduction using recombinant Escherichia coli LSBJ. It was hypothesized that the sugars obtained from the hydrolysis of pretreated biomass would ferment efficiently and yield PHB yields similar to those from the hydrolysate obtained from raw biomass. The fermentation performance (cell dry weight, PHB inclusion level in the bacterial cells, and PHB yields) of the optimized fiber reject hydrolysate was also compared to the fermentation yields from pure sugars. The structural and thermal properties of the PHB produced in this study were analyzed and compared with standard PHB.

2. Materials and Methods

2.1. Biomass Collection

Fiber reject samples were collected from a local paper mill (WestRock Paperboard Mill, Syracuse, NY, USA). The biomass was dried at 50 °C for about 48 h to reduce the moisture content below 10%. The samples were then stored in a refrigerator at 4 °C until needed for experimentation.

2.2. Chemical Composition

The chemical composition (cellulose, hemicellulose, lignin, extractives) of biomass was determined using the standard Laboratory Analytical Procedure (LAP) from the National Renewable Energy Laboratory (NREL) [24,25]. In the first step of the process, the extractives were removed by sequential water and ethanol-based extraction using the Soxhlet apparatus [24]. Carbohydrates and lignin in the extractive-free biomass were determined using a two-step acid hydrolysis method [25]. In brief, 0.3 g of biomass was mixed with 3 mL of 72% sulfuric acid and stirred in a water bath at 30 °C for 1 h. The acid concentration was then diluted to 4% by adding 84 mL of deionized water, and the slurry was autoclaved at 121 °C for 1 h. The hydrolyzed samples were vacuum filtered using the ashed Pyrex filter crucibles (ASTM 10–15 M) to separate the solid and liquid fractions. The acid-insoluble lignin (AIL) in the solid fractions was determined using gravimetric analysis. The filter crucible was dried in an oven at 105 °C for 24 h, followed by combusting the dry residues in a muffle furnace at 575 °C for 24 h. The AIL was calculated based on the weight of dry solids and ash weight after burning. The filtrate was spectrophotometrically analyzed for acid-soluble lignin (ASL) at a wavelength of 240 nm and for sugar content by high-performance liquid chromatography (HPLC). All experiments were performed in triplicate.

2.3. Hot Water Pretreatment

Hot water pretreatment was conducted in a 300-mL stainless-steel Parr reactor vessel (Parr Instrument Company, Moline, IL, USA) at a 15% solid loading and 150 mL working volume [20]. In each run, 22.5 g biomass was mixed with a pre-calculated amount of water (based on the moisture in biomass) to achieve 15% solids. The pretreatment was conducted at different temperatures, 140, 150, 160, 170, and 180 °C, for a residence time of 10 min. After the incubation time, the reactor was immersed in chilled water (2–3 min) to quickly reduce temperature and pressure [20,21]. The pretreated slurry was collected and carried over into the disk milling treatment/hydrolysis [20,21,22].

2.4. Disk Milling

The mechanical refining of biomass was conducted using a lab-scale disk mill (Quaker City Mill model 4E, Philadelphia, PA, USA). The mill was coupled with two nonporous disks, one stationary and the other rotating at a speed of 89 rpm [21]. The width between the two plates was reduced to nearly zero [21,22]. In the case of two-stage pretreatment, the recovered slurry from hot water pretreatment was directly processed in the disk mill, without any washing neutralization or separation. For the refining of untreated fiber rejects, a slurry was prepared by mixing 30 g biomass and 170 mL water, and the slurry was processed in the disk mill. Based on previous studies [20,21], milling was restricted to three consecutive cycles in all cases to optimize electricity usage, cost, and heat dissipation. The moisture content of the samples after disk refining was determined using gravimetric analysis.

2.5. Enzymatic Hydrolysis

Enzymatic hydrolysis of the untreated and pretreated biomass was performed using a modified NREL/TP-5100-63351 procedure [20,21]. The hydrolysis was performed at 10% solid loading at a 50 mL working volume in 125 mL flasks. The pH of the biomass slurry was adjusted to 5.0 by the addition of glacial acetic acid. A sodium acetate buffer (pH 5, 1 M) was added to obtain a final concentration of 50 mM to maintain pH 5.0 during the hydrolysis. The commercial cellulase and hemicellulase cocktails Cellic^®Ctec2 and Cellic^®Htec2 (Novozymes North America, Inc., Franklinton, NC, USA) were added to each flask at dosages of 0.17 mL/g biomass (15 FPU/g biomass) and 0.04 mL/g biomass (one-fourth the volume of cellulase). Sodium azide was added at a concentration of 0.002% sodium to inhibit microbial growth. The hydrolysis was carried out in a shaking incubator maintained at 50 °C controlled with an agitation rate of 200 rpm for 72 h. Samples were withdrawn at 0, 4, 8, 24, 48, and 72 h for determination of sugar release during hydrolysis. Samples were heated in a heating block at 95 °C for 6 min to deactivate the enzymes and centrifuged at 10,000× g for 5 min. The supernatant was filtered through a 0.22 μm nylon syringe filter into 2 mL microcentrifuge tubes for HPLC analysis. All experiments were performed in triplicate. In the case of hydrolysate generation for the PHB fermentation, hydrolysis was performed at a 100 mL working volume, and sodium azide was not added. At the end of the hydrolysis (72 h), the slurry was filtered with Wattman No. 4 cellulose filter paper to remove the solids, and the filtrate was sterile filtered and stored at 4 °C until used for fermentation. All experiments were performed in triplicate.

2.6. HPLC Analysis

The concentrations of sugars, organic acids, and furan compounds in the samples collected during composition analysis and hydrolysis experiments were analyzed by HPLC (LC-20AD; Shimadzu, Kyoto, Japan) coupled with a refractive index detector (RID-10A). The analyses were conducted using the Aminex HPX-87H column (Bio-Rad, Hercules, CA, USA) operating at 60 °C using 5 mM H₂SO₄ as mobile phase at a rate of 0.6 mL/min [20,21,22].

2.7. Cellulose Conversion Efficiency

Cellulose conversion efficiency is defined as the ratio of glucose produced from the biomass conversion (at the end of hydrolysis) to the theoretical maximum glucose based on the cellulose content in the biomass (Equation (1)).

$$\mathrm{Cellulose} \mathrm{Conversion} (\%)={\displaystyle \frac{\mathrm{G}\times \mathrm{V}}{M_b\times \mathrm{C}\times 1.11}}\ast 100$$

(1)

where “G” refers to adjusted glucose concentration (g/L) at the end of hydrolysis; “V” represents the volume of hydrolysate; “M_b” refers to the mass of the biomass; “C” represents the cellulose content in the biomass; and “1.11” represents the hydrolytic gain during cellulose to glucose conversion

It is important to note that the adjusted glucose concentration (G) value used in Equation (1) was calculated by subtracting the glucose concentration in the blank (because of sugars in enzymes) from the glucose concentration in the hydrolysate at the end of hydrolysis. To determine the hydrolysate volume (V), at the conclusion of the hydrolysis, the slurry was filtered using a pre-weighed Whatman No. 4 cellulose filter paper, and the filtrate density was calculated from the measured volume and weight. The solids retained on the filter were washed with 50 mL of water and then dried in an oven to determine the weight of the insoluble solids. The hydrolysate (liquid only) weight was determined by subtracting the solids’ weight from the slurry weight, and the hydrolysate volume was calculated by dividing this value by the density of hydrolysate [21].

2.8. PHB Production Using Hydrolysate by Recombinant E. coli LSBJ

The fermentation of pure sugars and hydrolysate was conducted using recombinant E. coli LSBJ containing PHB biosynthesis plasmid (pBBRSTQKAB) [20,26,27]. Information about the strain’s origin and its PHB synthesis pathway was already provided in previous chapters and published studies [27]. The growth medium was an LB medium (10 g/L tryptone, 5/L g yeast extract, and 5 g/L sodium chloride) adjusted to pH 7 and containing 50 mg/L of the antibiotic kanamycin (Km), when needed. For each experiment, cultures were freshly prepared by streaking the freezer stock onto solid LB agar plates (15 g/L) supplemented with 50 mg/L Km. These plates were incubated at 37 °C for 16 h to allow for colony formation. Individual colonies were then used to inoculate the seed culture in 10 mL test tubes containing the LB medium with Km. The seed culture was incubated in a shaking incubator at 37 °C and 200 RPM for 16 h to prepare the inoculum for subsequent shake flask experiments. Fermentation was carried out at a 100 mL scale in 500 mL baffled shake flasks containing the LB medium supplemented with 50 mg/L Km, the carbon source from hydrolysate or pure sugars, and a 1% (v/v) inoculation of seed culture. The carbon source was either hydrolysate from raw or pretreated biomass, glucose, xylose, and a mixture of glucose and xylose, all adjusted to a fixed total sugar concentration of 20 g/L at the start of fermentation. The flasks were cultivated in a shaking incubator at 30 °C, with an agitation rate of 200 rpm for 72 h. At the end of fermentation, the cells were harvested by centrifugation at 4000 rpm (3820× g relative centrifugal force) for 20 min. After decanting the supernatant, the cell pellets were washed with 35% ethanol followed by water and subsequently lyophilized for 48 h [27]. The cell dry weight was measured gravimetrically. All experiments were performed in triplicate.

In the case of kinetic experiments, to determine fermentation performance (biomass growth, PHB inclusion, substrate consumption, pH) at various time points, we employed the strategy of sacrificing flasks at every time point. Multiple flasks were set up initially, and at each specified time point during the fermentation, two flasks (as replicate) were dedicated solely for analysis purposes to determine pH, optical density (at 600 nm), cell dry weight, and PHB inclusions. Sampling was performed at 6 h, 12 h, 18 h, 24 h, 30 h, 36 h, 48 h, and 72 h.

2.9. Analytical Methods for PHB Extraction and Quantification

The PHB content within the cells was measured using a gas chromatograph equipped with a flame ionization detector (Shimadzu, Kyoto, Japan). In the first step of PHB extraction, about 15 mg of lyophilized cells were added to 2 mL of 15% (v/v) H₂SO₄ in methanol and 2 mL of chloroform in Kimble^TM KIMAX^TM tubes and vortexed. The tubes were heated to a temperature of 100 °C in a dry heating bath and incubated for 140–150 min. After cooling the tubes to room temperature, 1 mL of ultrapure water and 0.5 mL of 0.25% (v/v) methyl octanoate in chloroform (internal standard) were added to the tube. The tubes were briefly vortexed and then centrifuged at 700 rpm (equivalent to a relative centrifugal force of 85× g) for 5 min. The organic layer was filtered with 0.2 µm PTFE filters in GC vials. The samples were analyzed using GC with a Rtx^®-5 column and flame ionization detector in triplicate [20,27]. Then, 1 μL of samples was injected at 280 °C in split mode (40:1) and run with a detector temperature of 310 °C and the following oven heating profile: after holding at 100 °C for 7 min, ramp 8 °C/min to 280 °C holding for 2 min, and then ramp 20 °C/min to 310 °C holding for another 2 min. The data were analyzed using Shimadzu’s GCsolution Version 2.31 software [28].

2.10. PHB Characterization

2.10.1. Extraction and Purification of PHB

PHB purification was performed according to the protocol adapted from previous studies [29]. Briefly, 5–10 g of lyophilized cells were subjected to a Whatman cellulose extraction thimble and refluxed using 125 mL chloroform for 6 h by Soxhlet extraction. The polymer was concentrated via rotary evaporation and precipitated in ice-cold methanol (1:10, v/v CHCl3: MeOH) under continuous stirring. Methanol was removed by centrifuging at 7000× g for 10 min at 4 °C to collect the polymer pellet. The pellet was then redissolved in chloroform, and the methanol purification step was repeated two more times. The purified polymer was dried by rotor evaporation.

2.10.2. Nuclear Magnetic Resonance (NMR) Spectroscopy

Approximately 10 mg of the purified PHB was dissolved in 1 mL of deuterated chloroform and analyzed with a Bruker AVANCE III HD 600 MHz NMR (Bruker, Billerica, MA, USA) to perform ¹H and ¹³C NMR analyses [30]. Spectra were analyzed by Bruker TopSpin v4.4.1 software.

2.10.3. Fourier-Transform Infrared Spectroscopy with Attenuated Total Reflectance (FTIR-ATR)

FTIR spectra were recorded for the extracted PHB samples in the IR region between 4000 and 500 cm⁻¹ with a resolution of 4 cm⁻¹ and 32 scans using a spectrophotometer (IR Affinity-1, M/s Shimadzu, Kyoto, Japan). This setup was provided with an Attenuated Total Reflectance accessory, which allowed for the dried powders’ analysis by directly placing them over the probe window (crystal) [31].

2.10.4. Thermogravimetric Analysis (TGA)

TGA of extracted PHB samples was performed on a TGA Q500 (TA instruments, Waltham, MA, USA) from room temperature to 600 °C at a heating rate of 10 °C/min in a nitrogen atmosphere with a flow rate of 40 mL/min [32]. The reported decomposition temperature (onset) was obtained from the thermogravimetric curve using ASTM E2550.

2.10.5. Differential Scanning Calorimetry (DSC)

The thermal transitions of the purified polymer were determined with DSC Q200 equipment (TA Instruments, Waltham, MA, USA). The sample was packed in an aluminum pan and heated in the temperature range of −25 °C to 200 °C at a heating and cooling rate of 10 °C/min in a nitrogen environment and then heated again [33]. The crystallization temperature (T_c) and melting temperature (T_m) were obtained from thermograms using the TA universal analysis software.

2.11. Statistical Analysis

The collected data were analyzed using a one-way analysis of variance (ANOVA) followed by Tukey’s HSD test, with the analysis carried out using SPSS v29 software. A confidence level of 95.00% was set to evaluate differences between the means.

3. Results and Discussion

3.1. Composition of Fiber Rejects

The composition of fiber rejects is summarized in

Table 1. Fiber rejects composition.

Composition	Dry Matter (%)
Glucan	35.5 ± 1.4
Xylan	9.0 ± 0.3
Lignin	16.7 ± 0.3
Ash	25.0 ± 1.3
Extractives	9.7 ± 0.7
Others (by difference)	4.2

Mean values and standard deviations of three determinations.

. It was comprised of about 45% carbohydrates (35.5% glucan, 9.0% xylan) and 16.6% lignin. The high carbohydrate content, with about an 80% glucose fraction, indicated that the feedstock is a high potential candidate for fermentation sugar production. Ash and extractive contents were estimated to be 25.0% and 9.7%, respectively. The ash content was significantly high compared to that observed in the woody biomass but in agreement with the ash content reported in the waste streams from paper mills [13,34,35]. Min et al. observed more than 40% ash in the waste fines rejects from an old corrugated containerboard (OCC) paper mill. They reported that ash originated from the minerals as fillers in the original pulp, with calcium carbonate (CaCO₃) accounting for about half [34]. Similar results were reported by Rauzi and Tschirner, who conducted a characterization of various paper mill reject streams and reported that the ash content could be as high as 70% in some samples. The rejects from an OCC mill contained about 18% ash content [13]. Jeffries and Schartman characterized three types of recycled waste fines and reported 22–33% ash and 45–65% carbohydrates, which is in agreement with the current study [35].

3.2. Effects of Biomass Size Reduction on Enzymatic Hydrolysis

Size reduction is an effective physical pretreatment to improve substrate accessibility to enzymes for efficient hydrolysis. Biomass particle size reduction could boost the affinity between the cellulose and enzyme, leading to an increase in the hydrolysis rate [36]. To investigate the effect of size reduction, fiber rejects were ground using a Wiley mill to pass through 0.5 mm, 1 mm, or 2 mm screens. The enzymatic hydrolysis experiments for all samples were performed under the same process conditions as described in Section 2.5. The glucose concentration obtained at the end of hydrolysis (72h) from the 2 mm ground fiber rejects was 31.8 g/L, increased significantly by 22% compared to unground fiber rejects (26.0 g/L) (Figure 1). Correspondingly, the cellulose conversion was increased from 51.8% to 65.1%, which indicated that particle size reduction is necessary for fiber reject conversion.

However, further reduction in the particle sizes (1 mm and 0.5 mm) did not show any significant improvement in sugar production as shown in Figure 1. This type of size reduction was categorized as a Class I size reduction because it does not significantly compromise the structural integrity of the plant cell wall. Typically, reaching the level of the fibers and fiber bundles only increases the substrate’s external surface area without notably altering the fiber pore structure by breaking down the cell wall [37]. To hydrolyze the untreated lignocellulosic biomass efficiently, it is critical to improve the external surface since the majority of pores of the untreated lignocellulosic biomass are too small to be accessible to cellulase. However, a Class I size reduction is not able to increase saccharification by more than 20% based on the study, which is consistent with our experimental results [37].

3.3. Effects of Surfactant Addition on Enzymatic Hydrolysis

Non-ionic surfactants have been considered efficient additives to improve the mass transfer during enzymatic hydrolysis and are also believed to alter the lignocellulosic structure to promote cellulose accessibility, inhibit enzymes from adhering to lignin, and shield enzymes from denaturing by forming reverse micelles [38]. For this study, the surfactants Tween 80 and PEG 4000 were selected due to their proven effectiveness in enhancing enzymatic hydrolysis of lignocellulosic biomass, including waste from paper mills [34,39,40]. PEG 4000 and Tween 80 were added during the enzymatic hydrolysis of 2 mm grounded fiber rejects to evaluate their abilities to improve sugar production and hydrolysis efficiency. The addition of 3% Tween 80 did not significantly improve the glucose release, whereas the hydrolysis efficiency was improved by 15% with the addition of 2.5% PEG 4000 (Figure 2). Numerous studies have explored the mechanisms through which PEG enhances the efficiency of lignocellulosic materials’ hydrolysis. Huang et al. indicated that PEG 4000 has the capacity to diminish the non-productive adsorption of lignin on cellulase enzymes by occupying ineffective lignin adsorption sites through hydrophobic interactions [39]. Hou et al. proposed that PEG can augment enzymatic hydrolysis by mitigating cellulase deactivation, which is induced by shear forces and interactions at the air–liquid interface [40].

3.4. Effect of Hydrothermal-Mechanical Pretreatment on Enzymatic Hydrolysis

Pretreatment is a key factor to increase the efficiency of lignocellulosic hydrolysis by reducing the biomass recalcitrance. This work involved the investigation of three pretreatment approaches: disk milling of wet slurry, also known as wet disk milling; hot water pretreatment; and hot water pretreatment followed by disk milling (two-step hydrothermal pretreatment). The total sugar production and cellulose conversions after 72 h of enzymatic hydrolysis for all scenarios are shown in

Table 2. Effect of different pretreatment conditions on the sugar released during hydrolysis.

Pretreatment	Residence Time during Hot Water Pretreatment (min)	Sugar Concentration (g/L)		Cellulose Conversion (%)
		Glucose	Xylose
Untreated	NA	31.8 ± 0.4	8.2 ± 0.2	65.1 ± 0.6
Disk milling only (1 cycle)	NA	33.1 ± 0.2	7.5 ± 0.2	69.1 ± 0.7
Disk milling only (3 cycle)	NA	34.7 ± 0.1	7.7 ± 0.0	73.5 ± 0.9
Hot water pretreatment (180 °C)	10	33.7 ± 0.7	7.9 ± 0.1	67.8 ± 1.4
Hot water pretreatment 180 °C and disk milling	10	37.2 ± 0.3	9.0 ± 0.2	77.7 ± 1.8

Mean ± standard deviation.

. Compared to raw samples, glucose concentration increased from 31.8 to 33.1 g/L, resulting in an improvement by 4% after one cycle of wet disk milling. Wet disk milling pretreatment has proven to be an effective and promising pretreatment method to enhance sugar release during enzymatic hydrolysis [41]. Wet disk milling is mechanical grinding, which not only results in a size reduction of Class I but also a further size reduction of Class II. Shear pressures enhance fiber external surface area in class I size reduction by fiber separation, cutting, fragmentation, and external fibrillation (delamination). In Class II size reduction, microfibril cross-connections are broken, and internal fibrillation is generated to break down cell walls [42,43,44,45]. During the investigation of disk-milling-only pretreatment, the effect of increasing milling cycles to three was also evaluated. The milling cycle represents one pass through the machine. Increasing the milling cycles to three further enhanced the glucose yield to 34.7 g/L. Correspondingly, the cellulose conversion was estimated at 73.5%, which was 12.9% and 6.4% higher compared to that of untreated and single-cycle disk-milled treated biomass, respectively. The xylose concentration was found to be similar among the samples milled one time and three times. The number of cycles was limited to three because previous studies have reported that three disk milling cycles were optimal considering the constraints of process costs, energy use, moisture loss, and heat generation [23].

To investigate the effect of hot water pretreatment, biomass was pretreated in water at 15% solids at 180 °C for 10 min. The glucose concentration at the end of hydrolysis for the pretreated biomass was observed to be 5.9% higher (33.7 vs. 31.8 g/L) compared to the untreated biomass. Correspondingly, the cellulose conversion was estimated to be 67.8%. Hot water pretreatment enhances the enzyme accessibility to cellulose, with minimal to no generation of inhibitory compounds. Water serves as a catalyst and solvent in the pretreatment process together with organic acids generated from biomass to help break down the plant cell wall matrix. According to reports, the elimination of hemicellulose by hot water pretreatment causes cellulose microfibril bundles to develop microscopic holes, which can weaken the network structure of the polymer matrix, resulting in higher hydrolysis efficiency [23]. However, hot water pretreatment alone is not considered sufficient to significantly enhance cellulose conversion, but the combination of hot water pretreatment with a mechanical treatment (e.g., disk milling) could lead to high cellulose conversions [20,21,22]. Similar observations were made in the current study: the use of only hot water pretreatment increased cellulose conversion by 4.1% (65.1 to 67.8%), whereas the use of disk milling (three cycles) followed by hot water pretreatment (HWDM) resulted in a 16.2% higher cellulose conversion (77.7%) compared to untreated biomass. This synergistic effect of hydrothermal and mechanical pretreatment results in a higher degree of defibrillation and specific surface area, resulting in high sugar yields during hydrolysis, as has been reported by several studies for a variety of feedstocks [20,21,22,23]. Weiqi et al. reported a significant improvement in the enzymatic hydrolysis of eucalyptus by using the combination of liquid hot water pretreatment (180 °C for 20 min) and disk milling, yielding maximum xylose and glucose recovery rates of 91.6% and 88.1%, respectively [46]. Kim et al. reported that the use of the HWDM of corn stover resulted in an up to 89% glucose yield [23]. Most of the previous studies that investigated HWDM reported that temperature during hot water pretreatment plays a critical role in the final sugar concentrations at the end of hydrolysis. The use of high-severity pretreatment (high temperature) could result in a breakdown of sugars, especially xylose [20,21].

To investigate the effect of pretreatment temperature, fiber rejects were pretreated from 140 to 180 °C temperature and subsequently disk milled (three cycles), and the corresponding results are presented in Figure 3. It was interesting to observe that the combination of hot water pretreatment and disk milling significantly improved sugar yields even at low temperatures. Application of 140 °C temperature during HWDM resulted in 37.5 g/L glucose at the end of hydrolysis, compared to 31.8 g/L for untreated biomass, with the difference being statistically significant. Correspondingly, the cellulose conversion was increased significantly by about 22% (65.1 to 79.3%). With a change in pretreatment temperature from 140 °C to 150 °C, the glucose concentration was further enhanced to 39.6 g/L, and the cellulose conversion increased to 83.2%. With a further increase in temperature up to 170 °C, the sugar yield did not change significantly. However, glucose recovery was observed to be lower at 180 °C. This behavior has been reported in several previous studies where a high temperature led to the degradation of sugars [20,21,47]. Based on the results illustrated in Figure 3, it could be concluded that HWDM pretreatment at 150 °C would be the best choice for fiber rejects, resulting in 39.6 g/L glucose, 8.3 g/L xylose, and 83.2% cellulose conversions. The cellulose conversion in the current study was higher than the cellulose conversion (69%) observed from the hydrolysis of dilute acid (0.5% sulfuric acid) pretreated fiber rejects [13] Several other studies using various feedstock have reported similar observations of achieving the same or higher cellulose conversion using HWDM compared to chemical pretreatments, leading to significant environmental benefits [20,21,22].

The glucose production profile for untreated samples and those pretreated at 150 °C are shown in Figure 4. The production of sugars increased rapidly within the first four hours for both raw fiber rejects and pretreated fiber rejects; however, the hydrolysis rate of pretreated fiber rejects was higher. The difference is that the non-pretreated fiber rejects continued to release glucose from the 4th to 48th hours of the hydrolysis process and then leveled off. For the fiber rejects after hydrothermal pretreatment at 150 °C and disk milling pretreatment, the conversion of glucan to glucose stopped at 24 h. This result could be explained by the fact that the pretreated fiber rejects were more easily liquefied, leading to a rapid increase in sugar production [48]. The final sugar yield of pretreated biomass was statistically (24.44%) higher than the untreated biomass.

3.5. PHB Production

The hydrolysate obtained from enzymatic hydrolysis of fiber rejects pretreated using the HWDM process at 150 °C for 10 min was investigated as a sole carbon source to produce PHB by recombinant E. coli LSBJ. It is commonly observed that although the pretreatment process enhances the sugar recovery, the generation of some inhibitory compounds, such as furans, affects microbial fermentation processes and results in lower fermentation yields. The generation of inhibitory compounds is more frequently observed with the increase in severity (pretreatment temperature and time) or chemical use [22]. Nonetheless, to investigate the effect of pretreatment on the fermentation of hydrolysate, both raw and pretreated fiber reject hydrolysates were assessed for PHB production in this study. All fermentations were performed at an initial reducing sugar (glucose and xylose) concentration (20 g/L). Pure laboratory-grade glucose, xylose, and mixed glucose/xylose were used as pure carbon sources as reference values.

The PHB titers from the fermentation of raw and pretreated fiber reject hydrolysates were found to be 6.27 and 5.99 g/L, respectively (Figure 5). Although the PHB inclusion (PHB mass fraction in cell mass; 33.3 and 33.7%) was similar among both conditions, a relatively lower cell dry weight (17.8 vs. 18.83 g/L) resulted in a slightly lower (4.4%) PHB yield for pretreated biomass hydrolysate compared to raw biomass hydrolysate. However, this difference was not statistically significant and was much lower compared to the effect of pretreatment on fermentation yield observed in previous studies. In a study on PHB production from brewers’ spent grains using E. coli LSBJ, Thomas et al. [26] observed that while the sugars released increased by 27% (49.7 vs. 39.2 g/L reducing sugars) using hot water pretreatment, the PHA yield from the pretreated hydrolysate was 47% lower (2.39 g/L) compared to that from the untreated hydrolysate (3.53 g/L), potentially due to the presence of inhibitory compounds generated during pretreatment. The obtained PHB production from pretreated fiber rejects in this study is within the range of PHB yields obtained from various feedstocks in the literature [1,26,30]. The PHB yields in the current study were significantly higher compared to PHB yields reported for native producers Burkholderia sacchari (2.7 g/L) and Burkholderia cepacia (4.4 g/L), which were isolated from soil and utilized detoxified sugarcane bagasse hydrolysate as a substrate [49]. Compared to those native PHB-producing soil bacteria, E. coli LSBJ demonstrated greater efficiency in utilizing monosaccharide carbon sources. This capability may be attributed to the natural biological niche of E. coli in the mammalian intestinal gut, where it relies on carbohydrate metabolism for colonization and persistence [50]. Another advantage of E. coli LSBJ is its lack of PHA degradation pathways. This ensures that once the maximum PHB production is achieved, the PHB levels remain stable over time [26].

It is important to note that most of the studies utilizing lignocellulosic hydrolysate as a carbon source employed an additional detoxification process following the pretreatment process to reduce or eliminate inhibitory compounds such as hydroxymethylfurfural (HMF), furfural, and phenolic compounds [51,52]. It is noteworthy that even without any detoxification or neutralization step employed in the current study, the hydrolysate did not contain a detectable number of furans, common inhibitors found in the hydrolysates of pretreated biomass, and this was further validated by the high PHB yields during fermentation. High cellulose conversion (83%) was achieved along with similar PHB yields.

The yields of PHB from fiber reject hydrolysate were higher than any other pure sugar carbon source, despite having the same sugar concentrations (20 g/L) at the start of fermentation as shown in Figure 5. The PHB yields from the fermentation of pure glucose, xylose, and sugar mix were 2.6 g/L, 0.8 g/L, and 1.8 g/L, respectively. The corresponding PHB inclusions were 40.0%, 25.6%, and 38.1%, respectively. The PHB yields while using xylose as the carbon source were significantly lower compared to using glucose as the carbon source. Similar results were reported by Paul et al. [2]. Even though E. coli can convert C5 sugars such as xylose into Acetyl Co-A via the pentose phosphate pathway (PPP), the efficiency of the PPP in terms of xylose utilization is enhanced in the presence of glucose [20,53]. The findings of higher PHB yields from biomass hydrolysate compared to pure sugars were also reported in previous studies [26,54,55]. Scheel et al. reported that the PHB yield obtained from waste fines’ (a specific fraction of fiber rejects) hydrolysate was twice as high as those obtained from purified sugars using the same recombinant E. coli LSBJ. Similarly, Thomas et al. observed up to a 2.43-fold PHB yield (3.43 vs. 1.41 g/L) from the fermentation of hydrolysate compared to pure glucose at a 10 g/L initial sugar concentration [26]. As speculated by these authors, the increase in PHB yield when using hydrolysate as the carbon source is attributed to the presence of various important metal ions in the hydrolysate that promote bacterial growth [54]. Dietrich et al. suggested that the presence of supplementary nutrients in the hydrolysate, including polypeptides and amino acids obtained from denatured enzymes used during hydrolysis, extended the duration of acetyl-CoA flow into the citric acid cycle, consequently promoting enhanced cell growth [55]. Since our experiments used LB as the base medium for all conditions, the potential contribution of peptides present in LB is accounted for in the control conditions. The major reason for higher cell dry weight in the case of fermentation of fiber reject hydrolysate could be attributed to the high acetic acid concentration (~10 g/L) in the hydrolysate, which acts as an additional carbon source for E. coli LSBJ during fermentation. A high amount of acetic acid was observed because of the addition of acetic acid to change the pH to 5.0 at the start of enzymatic hydrolysis [56].

To gain a deeper understanding of the assimilation of the sugars originating from the hydrothermal-mechanical pretreated fiber rejects into PHB through microbial fermentation, biomass growth, PHB accumulation, and substrate consumption were monitored throughout the fermentation process, and the results are illustrated in Figure 6a,b. As per our knowledge, this is the first study to investigate the kinetics during fermentation of biomass hydrolysate using E. coli LSBJ.

The bacteria exhibited a lag phase from 0 to 12 h followed by an exponential growth phase starting at 12 h characterized by a swift increase in cell growth, PHB accumulation, and glucose consumption. The cell dry weight continued to increase until it reached a maximum of 17.6 g/L in 48 h, coinciding with the complete depletion of all carbon sources. Similarly, PHB yield followed a comparable trend to cell dry weight, steadily increasing and peaking at 48 h (5.9 g/L) before stabilizing. Xylose consumption increased at 24 h, while glucose concentrations approached zero. This trend was expected because glucose is the preferred carbon source of E. coli [57]. Only after the depletion of glucose does the consumption of xylose begin due to carbon catabolite repression (CCR), a phenomenon observed in many microorganisms [58]. PHB accumulation was observed to commence at 12 h, peaking at 51% of CDW at the 24 h mark, subsequently declining to 34% by 48 h when CDW increased, and remaining stable until the end of the 72 h fermentation process. This indicates that after glucose was depleted, the cells redirected their metabolic focus from PHB synthesis to biomass production, using xylose and acetate more for cell growth rather than for PHB biosynthesis. E. coli can metabolize both glucose and acetate when there is an excess of acetate in the medium [59]. Acetate is metabolized via the Pta-AckA pathway [59]. Once the glucose and xylose are consumed, residual acetate in the medium, albeit at lower concentrations, is taken up by the cells and converted to acetyl-CoA by acetyl-CoA synthetase, which is subsequently utilized in the TCA cycle to generate cellular energy in the form of reduced cofactors [5]. Additionally, acetyl-CoA can contribute to biomass synthesis in the absence of glucose through some of the carbon flux being directed through the glyoxylate shunt [60]. Sun et al. reported significantly high cell density with the supplementation of 5–15 g/L of acetate to a minimal medium culture containing 10 g/L of glucose compared to cultures with only glucose [57]. The decrease in PHB inclusions from 30 to 36 h indicates that after the depletion of glucose, the residual acetate was primarily utilized for cell growth rather than PHB accumulation (Figure 6a,b). To balance the use of precursor metabolites after glucose depletion, cells employ tight regulation mechanisms that prioritize sugars for cell growth, homeostasis, and maintenance. This tight regulation explains why the consumption of acetate and xylose primarily supports cell growth rather than PHB production [61].

Another observation of carbon flux regulation was that the consumption of glucose coincided with a significant accumulation of lactate (7.0 g/L), a metabolic by-product, in the culture medium (Figure 6b). It has been reported that during PHB synthesis, cells may produce excess reducing equivalents, which can result in the secretion of lactate to alleviate this stress [62]. This observation is consistent with findings reported by Sekar and Tyo, which suggest that enhanced PHB synthesis is associated with increased substrate consumption and lactate secretion, accompanied by decreased secretion of formate and acetate [62]. In our results, cell growth continued after 30 h, primarily supported by the metabolism of acetate and lactate, which contributed more to cell growth than PHB accumulation. As a result, an increase in CDW and a reduction in PHB content were observed, leading to no significant increase in PHB yield after that. The PHB production remained stable after it reached its maximum due to the lack of genes responsible for PHA depolymerization in E. coli LSBJ [26].

3.6. PHB Characterization

3.6.1. Nuclear Magnetic Resonance (NMR) Analysis

The chemical structures of the extracted PHB produced by Escherichia coli LSBJ pBBRSTQKAB using fiber reject hydrolysate were verified using ¹H and ¹³C NMR spectroscopy as shown in Figure 7a,b. The ¹H NMR spectrum revealed the doublet signals at 1.21 ppm corresponding to the methyl group (-CH₃). The multiple signals between 2.39 and 2.56 are attributed to the diastereotopic methylene group (-CH₂). Additionally, multiple signals between 5.16 and 5.21 ppm were assigned to the methane group (-CH) with chiral carbon [63]. The peak at 7.2 ppm corresponded to chloroform used as the solvent. The ¹³C NMR spectrum showed the methyl group (-CH₃) at 19.78 ppm, methylene group (-CH₂) at 40.8 ppm, methane group (-CH) at 67.62 ppm, and carbonyl group (-C=O) at 169.15 ppm. The triplet signals at 76.82–77.24 ppm are attributed to CDCl₃. The chemical shift signals of ¹H and ¹³C NMR spectroscopy obtained in this study agreed with the commercial PHB characterization reported by Oliveira et al. and Prabisha et al., confirming that the material synthesized by this recombinant E. coli strain was PHB [63,64].

3.6.2. Fourier Transform InfraRed (FTIR) Spectroscopy

The functional groups of the extracted PHB produced by Escherichia coli LSBJ pBBRSTQKAB using fiber reject hydrolysate were investigated by FTIR spectroscopy, as shown in Figure 8. The minor peak at 3431 cm⁻¹ corresponded to the terminal OH group of PHB [65]. The peaks at 2977 and 2930 cm⁻¹ represented the presence of C-H stretching of the methyl and methylene groups, respectively [66]. The sharp absorption peak at 1722 cm⁻¹ corresponded to carbonyl (C=O) stretching vibrations [65]. The peak at 1278 cm⁻¹ represented the carbonyl group (C–O), while the peak at 978 cm⁻¹ was attributed to the amorphous component of PHB [67]. The results correlated well with published reports [65,67]. Therefore, the obtained biopolymer from E. coli with FR was PHB.

3.6.3. Thermogravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC)

Thermal properties of PHB synthesis were investigated using DSC and TGA analyses (Figure 9). TGA was performed to observe the thermal stability of PHB synthesis. The material’s thermal stability was evaluated and measured by the degradation temperature (T_d) corresponding to the mass loss of 5%. As shown in Figure 9a, the Td of the extracted PHB in this study was 265 °C. This T_d value obtained denotes considerable thermal stability, making it suitable for applications requiring high-temperature resistance. The observed T_d is consistent with other studies where PHB was synthesized by microbial fermentation, reporting T_d values for PHB ranging from 250 °C to 275 °C [68,69].

DSC analysis was performed to elucidate the thermal transitions that a polymer undergoes as the sample is heated. The thermal transitions are manifested mainly in terms of melting point (T_m) and crystallization temperature (T_c). The T_c of extracted PHB was found to be 84 °C (Figure 9b). The extracted PHB exhibited dual melting temperatures (T_m) at 171 and 176 °C, which is close to that of polypropylene [70]. The dual melting point inferred that re-crystallization takes place followed by cross-linking isomerization reactions [71]. Overall, the combination of high degradation temperature and favorable DSC characteristics indicates that the PHB synthesized by E. coli LSBJ pBBRSTQKAB is a promising candidate for various industrial applications.

4. Conclusions

This study aimed to investigate the potential of utilizing fiber rejects, a waste stream from recycled paper mills, as a sustainable feedstock for cost-effective PHB production. This is the first comprehensive study to focus on developing efficient and sustainable bioprocess to maximize sugar recovery from this waste and utilize those sugars for PHB production. Three different approaches, including size reduction, the addition of surfactants, and the use of a chemical-free hydrothermal pretreatment, were investigated to improve the hydrolysis efficiency for high sugar recovery. Size reduction and the addition of PEG 4000 improved the enzymatic hydrolysis efficiency to some extent. A pretreatment including a hot water pretreatment followed by disk milling had synergistic effects on increasing sugar yields in enzymatic hydrolysis, achieving the highest reduction in sugar production (48 g/L) and cellulose conversion up to 83% without using any chemicals. Without performing any detoxification process (a major challenge associated with conventional thermo-chemical pretreatments), the hydrolysate obtained from pretreated biomass in this study fermented efficiently and yielded high PHB yields (6.0 g/L), which was similar to that observed (6.2 g/L) from the fermentation of the hydrolysate of raw biomass. The PHB yields were observed to be higher than those from the fermentation of pure sugars due to the presence of important metal ions and additional carbon sources in the hydrolysate. The fermentation kinetics provided important insights into the rates of substrate consumption, byproduct formation, and PHB synthesis, which are essential for scaling up the process and achieving consistent performance. The structural and thermal properties of the extracted PHB were consistent with standard PHB. Further research needs to be carried out to enhance PHB productivity through the optimization of fermentation parameters and to scale up this process in bioreactor systems.