Comparison of methods for detecting protein extracted from excess activated sludge

The protein contents of hydrolyzed sludge supernatant are commonly determined with the Kjeldahl method, but this method suffers from complicated operations, long process times, and large quantities of chemicals consumed. In this paper, the Lowry, bicinchoninic acid (BCA), and Bradford methods were used to test the precision and spiked recovery of proteins from sludge supernatants hydrolyzed by alkaline-thermal hydrolysis (ATH), enzymatic hydrolysis (EH), and ultrasound-assisted enzymatic hydrolysis (UEH), and the results were compared with those obtained with the Kjeldahl method. For all the hydrolytic processes, the sludge protein values determined with the three tested methods were within 0.05 of each other, which met the experimental requirement for accuracy. Both the Lowry and BCA methods had recovery rates of 95–105%, while the Bradford method showed large deviations and was not highly reliable. The three protein determination methods showed significant differences with the Kjeldahl method (P<0.05). However, the relative deviation between the Kjeldahl and BCA methods was the smallest (3–5%), followed by those between the Kjeldahl and the Lowry (11–21%) and Bradford methods (21–90%), and the causes of the deviations were analyzed based on the protein hydrolysate components and the mechanisms for the different detection methods. On the basis of these results, the BCA method was chosen as the most appropriate quantification method for use with sludge protein extraction, and it was used to analyze the protein contents extracted from residual sludge samples obtained from two sewage treatment plants. The reliability of the method was verified, and this lays a foundation for the extraction and reclamation of sludge proteins.


Introduction
The excess sludge produced by municipal wastewater treatment plants is increasing and has become an urgent problem to be addressed. Excess sludge contains 30-60% protein (equivalent to soybean protein), which can be used as a foaming agent (Collivignarelli et al. 2017), liquid fertilizer (Liu et al. 2009), animal feed (Hwang et al. 2008), and so on after extraction to realize the recycling of high value-added resources. Currently, the recovery and utilization of sludge protein have attracted much attention, but there is no uniform method for the determination of sludge protein, which directly affects the ability to compare different extraction processes and is detrimental to the further development and application of sludge protein recycling techniques.
At present, the determination of sludge protein mainly utilizes protein detection methods used in the food industry, including the Kjeldahl method, Lowry method, bicinchoninic acid (BCA) method, and Bradford method. Among them, the Kjeldahl method is the standard method for the determination of food protein, and it is also the most widely used method in the detection of sludge protein (Marcó et al. 2002;Li et al. 2012;Hall and Schönfeldt 2013). The principle is that under catalytic conditions, organic nitrogen is converted into inorganic ammonium salts by digestion with concentrated sulfuric acid, and the ammonium salts are then converted into ammonia under alkaline conditions. The ammonium salts are distilled by steam, absorbed into a solution of excess boric acid, 1 3 and then titrated with standard hydrochloric acid to calculate the nitrogen content of the sample. Since the nitrogen content of protein is relatively constant, the protein content can be calculated from the nitrogen content; thus, this method is a classical protein quantification method. However, this method is complicated to operate, has a long experimental duration and requires a large consumption of reagents, so it is inconvenient to apply.
The Lowry method is based on protein complexation with copper ions in alkaline copper solution, resulting in the extension of peptide bonds; then, the exposed tyrosine and tryptophan react with the Folin reagent under the alkaline copper conditions to produce a blue color. Within a certain concentration range, the color depth is proportional to the content of tyrosine and tryptophan in the protein (Jiang et al. 2013). The BCA method is an improved version of the Lowry method; its principle is that under alkaline conditions, Cu 2+ forms complexes with protein and is reduced to Cu + , which easily combines with BCA to form a purple-blue complex with a maximum absorbance at 562 nm, which is proportional to the protein concentration (Zhou et al. 2013). The principle of the Bradford method is that the hydrophobic region of protein under acidic conditions readily combines to form a blue protein-dye complex with a maximum absorbance at 595 nm. In a certain range of protein concentrations, the absorbance is proportional to the protein content (Chen et al. 2015) and thus can be used in the determination of protein content.
As seen from the above analysis, the Lowry, BCA, and Bradford methods are considerably simpler to operate than the Kjeldahl method. Application of the former methods in the determination of sludge protein will be highly convenient for researchers and greatly reduce experimental effort. However, unlike food or medicine, excess sludge has a complex composition and numerous interfering substances, which easily interfere with protein content determination methods. Moreover, different hydrolysis methods are used to extract sludge protein, and the obtained protein composition is quite different. It is unknown whether the above methods are suitable for the determination of sludge protein after hydrolysis by different hydrolysis methods.
Among the chemical methods, alkaline-thermal hydrolysis (ATH) has realized industrial application with high extraction efficiency and mature technology (Xue et al. 2014). Biological methods mainly involve enzymatic hydrolysis (EH) with alkaline protease, which has mild extraction conditions without secondary pollution, but the extraction rate is low (Li et al. 2011;Zhang et al. 2012). Among combined methods, ultrasonic-assisted enzymatic hydrolysis (UEH) can achieve a relatively high extraction rate with low energy consumption . These three methods are representative methods of sludge protein extraction.
In this study, the Lowry, BCA, and Bradford methods were compared to explore the feasibility of their application in the determination of sludge protein extracted from excess sludge by ATH, EH, and UEH. We mainly carried out the following work: (i) the precision, spiked recovery rate, and deviation of each detection method were analyzed to judge their reliability; (ii) the reasons for the deviation were explored by measuring the main hydrolyzed products in sludge, such as polypeptides, amino acids, and polysaccharides; and (iii) the BCA method was selected to measure the protein content of hydrolysate obtained from excess sludge from two different wastewater treatment plants to verify the feasibility of this approach. The findings of this study are expected to lay the foundation for the future exploitation and application of sludge protein reclamation.

Excess sludge
The sludge used in the experiment was taken from the secondary sedimentation tank of an urban sewage treatment plant in Zhengzhou. The main biological treatment process of the plant was the modified Carrousel oxidation ditch process and the sludge moisture content was 99.2%. According to previous optimization results (Gao et al. 2020b), a centrifuge was used to concentrate and adjust the suitable moisture content to approximately 95%. The characteristics of the experimental sludge are summarized in Table 1.

Hydrolysis method of excess sludge
According to the results of the previous sludge protein extraction test, ATH (Li 2017), EH , and UEH (Miao 2017) under optimized conditions were used to extract sludge protein. The specific operation process is shown in Fig. 1. ATH of sludge was carried out in a homemade hydrolysis reactor (Gao et al. 2020a), and the ultrasonic pretreatment step of UEH was carried out in an ultrasonic crusher (Ningbo Xinzhi Biotechnology Co., Ltd., model JY92-11N).

Determination method of protein content
The protein solution obtained from the process in Fig. 1 was analyzed by the BCA method, Lowry method and Bradford method. By calculating the precision and recovery rate of each method and comparing them with the Kjeldahl results, the most suitable method for the detection of sludge protein in each hydrolysis process was selected.
(1) Lowry method BSA was selected as the protein standard sample, and the concentrations of diluted protein solution were 0, 25, 50, 100, 150, 200, and 250 μg/mL. A total of 4 mL of the alkaline copper solution was added to 0.5 mL of the diluted standard protein solution, and the mixture was held at 20-25℃ for 10 min. Then, 0.5 mL of Folin reagent was added, the sample was held at 20-25℃ for 30 min, the absorbance at 650 nm was measured, and the standard curve was drawn. The samples were diluted to a certain proportion and assessed with the same operation process as that used to prepare the standard curve. The protein concentration was calculated according to the standard curve (Jiang et al. 2013).
(2) BCA method The BCA kit method was used (Smith et al. 1985). BSA was used as the standard, and the diluted protein solution concentrations were 0, 25, 50, 100, 200, 400, 600, 800, and 1600 μg/mL. A total of 50 μL of protein liquid was added to 1 mL of BCA reagent and incubated at 60°C for 30 min. The absorbance at 562 nm was measured using a UV-visible spectrophotometer, and then the concentration was calculated according to the standard curve. (

3) Bradford method
The standard BSA protein samples were diluted to give concentrations of 0, 50, 100, 150, 200, and 250 mg/L. One milliliter of diluted standard solution was added to 5 mL of Coomassie staining solution and held at room temperature for 10 min. The absorbance was measured at 595 nm with a spectrophotometer, and a standard curve was drawn. The samples to be tested were diluted to a certain proportion and assessed using the same operation process as that used to prepare the standard curve, and the protein concentration of each sample was calculated according to the standard curve (Xu et al. 2022).
(4) Kjeldahl method The separated supernatant protein was digested with a mixture of concentrated sulfuric acid and catalyst to decompose the protein, and the decomposed ammonia combined with sulfuric acid to form ammonium sulfate. NaOH distillation was then performed to release ammonia, which was absorbed in boric acid solution and titrated, and the amount of acid consumed was multiplied by a conversion factor (6.25) to obtain the protein content (Lu and Li 2015).

Precision test
Prepared sludge supernatant was placed in a vessel, and the absorbance of the sample was determined using the BCA method, Bradford method, and Lowry method. The protein concentrations were calculated following the standard curve. The same samples were analyzed three times within 1 day to investigate the precision (Lu et al. 2010).

Spiked recovery test
Spiked recovery rate is measured by adding a certain amount of standard material (BSA) to the sample to be tested, and is used to judge the systematic and random errors of the measurement methods (Navarro et al. 1989;Dobrowolska-Iwanek 2015). BSA protein standard solution was prepared with different concentrations (the concentrations were selected according to the measurement ranges of the different protein determination methods). Two samples were prepared: one was mixed with different concentrations of standard protein solution and the other was diluted to appropriate degrees without standard protein solution. The protein content of the samples was then determined by the BCA method, Bradford method, and Lowry method; three parallel experiments were performed to calculate the spiked recovery rate (Zhu et al. 2011).

Comparison with the Kjeldahl method
The same batch of ten sludge samples hydrolyzed by ATH, EH, and UEH was measured using the Lowry, BCA, and Bradford methods. The results were compared with the Kjeldahl method. To evaluate the results, statistical analysis was performed with SPSS 26.0. Analysis of variance (ANOVA) was used to test whether there was a significant difference between the results of different protein determination methods and the Kjeldahl method. If p<0.05, it was considered that there was a significant difference.

Determination of sludge composition
The supernatant of sludge hydrolyzed by different hydrolysis methods was assessed, and the protein, polypeptide, amino acid, and polysaccharide contents were analyzed. The protein content was determined by the BCA method (Smith et al. 1985), the polypeptide content was measured with the biuret method (Shrivastaw et al. 1995), the amino acid content was determined using the ninhydrin chromogenic method (Rosen 1957), and the polysaccharide content was determined using the anthrone-sulfuric acid method (He and Zhang 2019). All indicators were measured three times in parallel, and the results were averaged.

Determination of actual sludge hydrolyzed protein
The excess sludge used in the tests was obtained from the secondary sedimentation tanks at two municipal sewage treatment plants in Zhengzhou. The sludge of S1 was from the secondary sedimentation tank (carrousel oxidation ditch process), and the sludge of S2 was from the mechanical dewatering (centrifugation) facility (A 2 /O process). After the sludge moisture content of S1 and S2 was adjusted to 95% by centrifugation, the protein was extracted by ATH, EH, and UEH, and then the protein content of the hydrolyzed sludge supernatant was determined by the BCA method.

Standard curves
The standard curve equations, correlation coefficients and reaction conditions for the Lowry, BCA, and Bradford methods are shown in Table 2.
All three methods had a standard curve equation with a correlation coefficient (R 2 ) above 0.99, indicating that the standard curves of the three methods had good feasibility and linearity. In terms of reaction temperature and time, the Bradford method is fast and easy to operate, and the reaction can be completed in 10 min, while the Lowry and BCA methods require more steps, accurate temperature control, and a reaction duration of 30 min.

Precision tests
The protein content of the sludge supernatant treated by ATH, EH and UEH was measured by the Lowry, BCA, and Bradford methods. For the detailed results of the precision tests, see the Supplementary Information and table A.1.
The level of precision is usually expressed as the relative standard deviation (RSD). The smaller the RSD is, the higher the accuracy of the method. The precisions of the three determination methods for the sludge protein obtained by different hydrolysis processes were all lower than 0.05, which meets the experimental requirements of precision (Li and Zhang 2000).

Spiked recovery tests
BSA was used as the standard protein, and standard addition recovery tests were performed to calculate the recovery rate. Studies have shown that a recovery rate of standard addition between 95 and 105% meets the test requirements (Navarro et al. 1989;Li and Zhang 2000;Dobrowolska-Iwanek 2015). As seen from the Supplementary Information and table A.2, the recovery rates of both the Lowry and BCA methods were between 95 and 105% for the protein solutions obtained by all hydrolysis methods, indicating that the accuracy and reliability of the Lowry and BCA methods were excellent; in contrast, the recovery rate of the Bradford method had a large deviation and did not indicate high credibility of the data for the protein content in sludge.

Comparison with the Kjeldahl method
The protein contents determined by the Lowry, BCA, and Bradford methods after ATH, EH, and UEH were compared with the results of the most commonly used Kjeldahl method, as shown in Fig. 2. Figure 2 shows that the protein content varied widely after the excess sludge was hydrolyzed by the three methods. The average protein contents of ATH, EH and UEH were 7497 mg/L, 3542 mg/L, and 5387 mg/L, respectively, as determined by the Kjeldahl method. The protein content of sludge hydrolysis supernatant mainly depends on the degree of hydrolysis of excess sludge by each process (Gao et al. 2020b). As seen from the analysis of Table 3, the results of the Lowry, BCA, and Bradford protein determination methods were all significantly different from the Kjeldahl measurements. However, the BCA method had the smallest relative deviation from the Kjeldahl method regardless of which hydrolysis method was adopted, followed by the Lowry and Bradford methods. Additionally, the BCA method performed well in the precision and spiked recovery tests. Therefore, the BCA method is suitable and feasible for the determination of protein content in sludge.

Analysis of sludge components after hydrolysis by different methods
To further investigate the reasons for the deviation between the Lowry, BCA, Bradford, and Kjeldahl methods, the specific components, such as proteins, polypeptides, amino acids, and polysaccharides, in the supernatant of sludge hydrolyzed by different hydrolysis methods were analyzed. The results are shown in Fig. 3.
As shown in Fig. 3, the content of proteins and polypeptides in the supernatant was significantly higher than that of amino acids and polysaccharides. ATH had a polypeptide content 224.6% and 50.6% higher than those of EH and UEH, respectively. However, the two biological enzymatic methods of hydrolysis had higher amino acid contents, with EH and UEH achieving 2.8 and 3.6 times higher amino acid contents than ATH, respectively.

Discussion
Regardless of the hydrolysis method used, the precision of the protein content measured by the Lowry, BCA, and Bradford methods met the requirements (for details, see Supplementary Information and table A.1), but the reliability . The principle of the Bradford method is mainly based on the fact that the hydrophobic group in Coomassie Brilliant Blue G-250 has an affinity with the hydrophobic region of protein under acidic conditions and easily combines to form a blue protein-dye complex with a high extinction coefficient and high sensitivity in protein measurement (Yan et al. 2006). Studies suggest that Coomassie dye is most likely to combine with arginine and lysine residues in proteins. Due to the different contents of arginine and lysine residues in various proteins, the binding strength varies greatly, resulting in a large deviation in the calculated content of different proteins (Silvério et al. 2012). It has also been shown that Coomassie brilliant blue does not form complexes with peptides with molecular weights lower than 3000 Da (Le et al. 2016). As shown in Fig. 3, a large portion of the protein obtained by ATH will be further hydrolyzed to polypeptides (molecular weight below 10,000 Da), while macromolecular peptides could be further  hydrolyzed into small molecular amino acids (the average molecular weight of an amino acid is approximately 100 Da) by alkaline proteases in EH and UEH. Additionally, it can be seen in Fig. 2 that the protein content measured by the Bradford method was 31-36% lower than that measured by the Kjeldahl method when ATH was adopted, while the difference was as high as 94-95% when EH and UEH were employed, indicating that the error of the Bradford method was greater because of the large amount of small-molecule amino acids in sludge protein extracts obtained by EH and UEH.
Both the Lowry and BCA methods had good precision and spiked recovery rates, but Fig. 2 shows that the Lowry method had a higher measured value than the Kjeldahl method with all hydrolysis processes. This may be due to the poor specificity of the Lowry method, which is affected by interference from phenols, citric acid, ammonium sulfate, potassium, and magnesium (ions usually lead to precipitation), EDTA, Tris buffer, thiol compounds, glycine, sugars, glycerol, carbohydrates, etc. (Lucarini and Kilikian 1999;Kumar et al. 2005). The interference effects are manifested by an enhancement or weakening of the chromogenic effect, inhibition of the reaction with the Folin reagent or production of a precipitate. The three hydrolysis processes destroy the flocculation structure of sludge and the lipid components in the microbial cell wall, resulting in the release of organic matter into the liquid phase and an increase in protein and polysaccharide contents (Gao et al. 2020b). However, the chemical components of cells, such as sugars and lipids, can be detected together in the supernatant by the Lowry method. In addition, the Lowry method is particularly sensitive to the presence of amino acids and peptides, which will increase the determined values (Lucarini and Kilikian 1999). In addition, the Folin reagent has a chromogenic effect on humic acid (Vakondios et al. 2014), and it will also cause a higher measured value (Mei et al. 2018) since the content of humic acid in sludge is relatively high. Therefore, it is not appropriate to use the Lowry method to determine the hydrolyzed protein content of sludge solution.
Although there were significant differences between the BCA and Kjeldahl methods, their results were the most similar. The Kjeldahl method is mainly based on nitrogen-containing compounds in sludge, including all ammonia nitrogen and organic nitrogen. In addition to protein, there are small amounts of other nitrogen-containing substances, such as nucleic acids and uric acid. Because the measured value is almost unaffected by other complex components in sludge (Hao 2015), the Kjeldahl method has been widely used in the determination of sludge protein content. As shown in Fig. 2, the mean value of the BCA method was the closest to the result of the Kjeldahl method, being only 0.4-15% lower. From the measurement principles, the BCA value was very close to the actual sludge protein content. This is consistent with the conclusion of M. Ras (Ras et al. 2008). However, the BCA method requires strict control of the cooling time. During the experiment, the color depth was enhanced by 10% for each increase in cooling time of 10 min, so the cooling time should be tightly controlled at 30 min.

Actual sludge determination results
The protein extracted from the sludge of the two wastewater treatment plants using ATH, EH, and UEH was determined by the BCA method, and the results are shown in Table 4.
As seen from Table 4, the protein extraction efficiency of S2 was higher than that of S1 regardless of which hydrolysis method was used. Typically, the sludge load of an A 2 /O system (S2) is higher than that of an oxidation ditch (S1) (Peng et al. 2012). Therefore, the growth rate of activated sludge and the content of nutrients, such as proteins, in S2 should be higher than those in S1. The results in Table 4 indicate that the measured values were consistent with the theory of the activated sludge process. In addition, ATH showed the best extraction effect, while EH had a poor extraction effect; after ultrasonic pretreatment, the EH process could be effectively strengthened. The measured protein extract contents obtained by the three different methods indicated that the determination results were in accordance with the theory of hydrolysis.

Conclusion
In this study, the Lowry, BCA, and Bradford methods were used to detect the protein obtained by ATH, EH, and UEH. The Bradford method had a large deviation in spiked recovery rates and hence low credibility. The Lowry method is subject Fig. 3 Protein content of sludge after different hydrolysis methods to many interference factors that readily result in larger measured values. The BCA method had a satisfactory precision and recovery rate and was close to the Kjeldahl method in protein determination. Therefore, the BCA method is suitable as a replacement for the Kjeldahl method for the determination of sludge protein due to its simple reagents and convenient operation. Experiments with sludge from two different sewage treatment plants showed that the measured values obtained with the BCA method were consistent with the theoretical trends. It should be noted that the cooling time should be strictly controlled during measurement.
Author contribution Yixin Yan: conceptualization, formal analysis, writing-review and editing, resources, supervision, funding acquisition, project administration. Mengnan Zhang: investigation, data analysis, visualization, writing-original draft. Jianlei Gao: methodology, supervision, validation. Lei Qin: data curation. Xi Fu: software, writing-original draft. Junfeng Wan: writing-review and editing. Data availability All data and materials will be available upon reasonable request.
Ethical approval Not applicable.

Declarations
Consent to participate All authors agreed to participate in this research and in writing the manuscript.

Consent for publication
All authors approved this manuscript for publication.

Competing interests
The authors declare no competing interests.