Materials

The activated sludge used in this study was collected from a municipal wastewater treatment plant located at Shanghai (120°51′E, 30°40′N), treating 93 % municipal wastewater and 7 % industrial wastewater through anoxia–anaerobic–aerobic process. The sludge had a total solids (TS) content of 3 % wt and a volatile solids (VS) content of 68 % dw. The content (dry weight basis) of polysaccharides, proteins, and DNA content was 11.7, 37, and 3 % dw, respectively.

Five kinds of hydrolytic enzymes were used separately to specifically target polysaccharides, proteins, or DNA. Three enzymes were used to attack different polysaccharides: α-amylase (Sigma, Product No.: 10070-10G, about 50 U/mg derived from Bacillus subtilis, to be used at pH 6.9 and 25 °C), cellulase (Sigma, Product No.: C2605-50 mL, at least 1000 U/g derived from Aspergillus sp., to be used at pH 8.0–8.2 and 40 °C), and polygalacturonase (Sigma, Product No.: 17389-10G, greater than 1 U/mg derived from Aspergillus niger, to be used at pH 4.0 and 50 °C). Proteins were targeted using proteinase (Yinggong Trading Co. Ltd., Shanghai, China, Product No.: LG1036, at least 200 units from Bacillus licheniformis, to be used at pH 6.0–8.5 and 50–60 °C). Lastly, DNA was attacked using DNase (Sigma, Product No.: DN25-10MG; at least 85% protein and at least 400 Kunitz units/mg protein derived from bovine pancreas, to be used at pH 7–8).

Enzymatic treatment

EPS extraction and stratification

Based on its strength of binding to cells, EPS was extracted and stratified through centrifugation and ultrasonication into four fractions: “supernatant,” “slime,” “loosely bound EPS” (LB-EPS) and “tightly bound EPS” (TB-EPS). The final pellet that remained after EPS extraction was mainly composed of microbial cells and unextractable EPS. The extraction procedure suggested by Yu et al. [16] was followed with some modification. An additional file describes the detailed procedure (see Additional file 1).

The aqueous suspended solutions of the enzymes themselves were stratified according to the same method, and every content was determined as sludge samples. Therefore, the reported values were subtracted from the enzyme background.

Determination of organic components

The TS and VS contents of raw sludge were measured as the mass loss in samples following heating at 105 and 600 °C, respectively. The polysaccharides content was measured using Anthrone colorimetry [17] on sludge samples that were mixed with 50 mL 0.5 M H₂SO₄ and autoclaved at 121 °C for 1 h. Protein content was measured using the Lowry method [18] the same on sludge samples that were mixed with 50 mL 0.5 M H₂SO₄ and autoclaved at 121 °C for 1 h. DNA content was measured using diphenylamine colorimetry with 1 h incubation at 60 °C on sludge samples that had been subjected to two cycles of bead grinding at 6300 rpm for 2.5 min each.

After extraction, each EPS fraction was tested for polysaccharides, proteins, external DNA (eDNA), dissolved organic carbon (DOC), and dissolved nitrogen (DN). The determination of polysaccharides, protein, and eDNA followed the Anthrone colorimetry, Lowry method, and diphenylamine colorimetry, respectively. DOC and DN were analyzed using a Total Carbon/Total Nitrogen analyzer (TOC-V CPN, TNM-1, Shimadzu, Kyoto, Japan).

Raw sludge and pellets remaining after EPS extraction were first lyophilized and then tested for carbon, nitrogen, and hydrogen content using an element analyzer (Vario EL III, Elementar, Hanau, Germany).

Fluorescent excitation–emission matrix analysis

Three-dimensional fluorescent excitation–emission matrix (EEM) analysis was performed using a Cary Eclipse Fluorescence Spectrophotometer (VARIAN, Santa Clara, CA, USA) to obtain the relative content of fluorescent organics in each fraction. The fluorescence was collected using an excitation wavelength of 200–550 nm and an emission wavelength of 250–600 nm. Parallel factor analysis (PARAFAC) was applied to the EEM data to qualitatively and quantitatively analyze the main fluorescent substances. The detailed procedure for EEM and PARAFAC analyses is given by Lü et al. [19].

Environmental scanning electron microscopy (ESEM)

Environmental scanning electron microscopy (ESEM) was performed using a Quanta 250 (FEI, Hillsboro, OR, USA) to observe structural changes of biological samples due to enzyme treatment. Duplicate images of each sample were made at 500×, 1000×, 2000×, 5000× and 10,000× magnifications.

Determination of sludge dewaterability

The dewaterability of sludge samples was evaluated according to the TS content of cakes obtained by centrifuging sludge at 2000×g for 10 min.

Statistic analysis

In order to distinguish the effect of control group and experiment group at the same pH, analysis of variance (i.e., ANOVA analysis) was applied on the contents of polysaccharides, proteins, and eDNA in supernatant and TB-EPS fraction. pH 4.0 group includes pH 4.0 control and polygalacturonase-added group; pH 6.9 group includes pH 6.9 control, amylase-added group, and proteinase-added group; pH 8.0 group includes pH 8.0 control, cellulase-added group, and DNase-added group. Supernatant and TB-EPS are the two representatives of the four fractions.

To distinguish the significance of difference among samples after different treatments, PCA was applied. Four factors (the concentration of each category of organic matters in the supernatant, slime, LB-EPS, and TB-EPS fractions) were analyzed and converted to two principal uncorrelated variables, thus formed a map in which samples distributed according to their influence by two dimensions, and four factors’ contributions to principal variables were shown. The percentage of how much each principal component can explain the variance can also be calculated.

Reallocation of different organic matters

Organic matters in the raw sludge

Proteins, polysaccharides, and DNA comprised 70.5 % of sludge organics on a VS basis and were the dominant components in EPS. Figure 1a–c indicates the content of proteins, polysaccharides, and eDNA in each EPS fraction. To confirm the observed changes in organic substances that contained carbon or nitrogen, DOC (Fig. 1d) and DN (Fig. 1e) were assessed correspondingly. Raw sludge had 33.2 % of proteins, 10 % of polysaccharides, and 27.3 % of the eDNA in EPS, and of these, 97, 85, and 95.3%, respectively, were located in TB fraction. After treatments, the content in EPS was 11.5–40 % for proteins and 24–43.5 % for eDNA. For polysaccharides, the average total content of EPS was 2.7–15 %, but for the polygalacturonase treatment group, the EPS polysaccharide content was 55–73 %. And the location of each component significantly shifted from TB-EPS to LB-EPS, slime, and supernatant, in concern with the increase in treatment duration. Both DOC and DN followed the same trend. This may be owing to continuing enzymatic hydrolysis, sludge adsorption, and release of enzymes as they were still active.

PCA analysis on organic distribution profiles

PCA results on each indicator for organics are shown in Fig. 2, which obviously observed the different effect of diverse treatments. From Fig. 2a–e, that the vector of “supernatant” is orthogonal to the vector “TB-EPS,” which represents totally different distribution levels of organic matters. Temperature showed strong impact on EPS fractions distributing to supernatant and TB-EPS, with samples under 35 °C treatments close to “TB-EPS,” and samples under 55 °C treatments closer to “supernatant.” And samples showed patterns to gather according to different pH adjustments, too. Basically, the relative position of samples in the PCA results acted in coordination with content analysis above, especially on alkaline environment improving organic matters’ release. Moreover, polygalacturonase’s improvement on polysaccharides distribution to supernatant and cellulase promoting most kinds of organic matters release from TB-EPS are quite clear.

Redistribution of dissolved fluorescent substances

Fluorescent excitation–emission matrix analysis (EEM) was used to understand the extracellular distribution of recalcitrant organic matter (e.g., humic substances) which might contribute to the EPS biodegradability or dewaterability. The PARAFAC analysis of EEM data extracted two components that contributed 90.6 % of spectrum data variation. Component I, which peaked at excitation (Ex) wavelengths of 220/280 nm and emission (Em) wavelengths of 306/356 nm (Fig. 3a), represented proteins and soluble microbial products. Component II, which peaked at Ex 220/240/290 nm and Em 364 nm (Fig. 3b), represented organic substances that had a higher degree of humification, according to the classification by Chen et al. [20] and Lü et al. [19]. The distribution of fluorescent substances is shown in Fig. 3c. It shows that the score of Component I was twice that of Component II in the amount among the groups, including raw sludge. In raw sludge, fluorescent substances were mostly distributed in the TB-EPS fraction and rarely in supernatant. The concentration ratio of Component I substances in supernatant to those in the TB-EPS fraction increased from 0.014 in raw sludge to 1.4 and 19.0 following the 35 and 55 °C treatments, respectively. Similarly, the ratio of Component II substances in the supernatant to those in the TB-EPS fraction increased from 0.02 (raw sludge) to 1.48 and 19.8 following treatment at 35 and 55 °C, respectively. Both the extent of redistribution and the total content of each component in supernatant were greater under the 55 treatment than in the 35 °C treatment. Like temperature, pH also played a role in the redistribution of fluorescent substances. The total content and extent of redistribution into supernatant were quite low in the pH 4.0 control group, but both measures increased in the pH 6.9 control group and increased further in the pH 8.0 control group.

In the polygalacturonase treatment (at pH 4.0), the two treatment temperatures (35 and 55 °C) had very similar effects on the total content and redistribution of fluorescent substances in the two EPS fractions. The amylase (at pH 6.9) and proteinase (at pH 6.9) enzymes had similar effects. Compared to raw sludge, both enzymes resulted in a lower total content of the substances in the TB-EPS, but a greater redistribution to supernatant under the 35 °C treatment, and almost the same total content and an even greater redistribution under the 55 °C treatment. However, in the 35 °C treatment, amylase and proteinase caused the lowest redistribution of fluorescent substances from the TB-EPS fraction to supernatant among all enzymatic treatments. The DNase treatment (at pH 8.0 and 35 °C) produced the greatest increases in the total content of fluorescent substances and in the redistribution of both components of fluorescent substances. At 35 °C, the cellulase treatment (at pH 8.0) produced changes like those in the pH 8.0 control group; however, at 55 °C cellulase increased, the total content of Component I and II substances by 48 and 100 %, respectively, compared to the content in raw sludge. In addition, among all enzyme treatments, cellulase caused the highest migration of fluorescent substances from the TB-EPS fraction to supernatant.

Morphology changes of pellets

Scanning electron microscopy was used to visualize the morphology of raw sludge and pellets after different treatments; the ESEM images are shown in Fig. 4. These images show differences between the microstructures of raw sludge and those of the pellets resulting from various treatments. Both raw sludge and the pellets demonstrated a rich spatial structure composed of porous “backbones,” abundant EPS flocs, and thick superficial microbes, similar to what was reported by [21].

After enzymatic treatment and pH adjustment, a massive number of free particles were observed in samples. And the surficial microstructure abundance decreased more or less, especially in samples to which polygalacturonase, proteinase, and cellulase had been added. In the ESEM pictures of these three enzyme groups, “backbone,” inorganic particle, and aggregate are seldom observed, and they appeared plainer and more homogenous. The pores inside individual particles became more dispersed and smaller. These structural changes could have reflected EPS disintegration. In fact, different treatments had distinctive morphological effects, which indicated different spatial distributions or roles of the hydrolyzed substrates in EPS formation.

Four types of changes in microstructure were observed. The first type of microstructure change was associated with the pH 6.9 treatment, pH 4.0 treatment, and polygalacturonase treatment (at pH 4.0) in which the surface of the bulk raw sludge appeared smoother and less abundant of arches and filamentous structures than did the sludge pellet. The second type of microstructure change resulted from amylase treatment (at pH 6.9), pH 8.0 treatment, and DNase treatment (at pH 8.0). The whole field of vision in ESEM images displayed rich superficial microstructures similar to the sludge pellet observation, but no apparent backbone could be observed. The third type of microstructure change was visible in the cellulase treatment (at pH 8.0). This type of change was characterized by a much plainer and simpler structure with only an accumulation of granules. The fourth type of microstructure change resulted from the proteinase treatment (at pH 6.9). The most distinctive feature of samples in this treatment was a vertical spatial structure, similar to that of a cave with inclined sleek walls tier upon tier.

From results observed in ESEM, some groups with different pH had similar morphology changes. It is clear that these differences did not caused by pH changes. Hence, the specific effect of each enzyme led to diverse types of EPS disintegration.

Dewaterability of treated sludge

The TS content of cake from raw sludge was 8.5 % wt and increased after treatment (Fig. 5). Within any treatment group, sludge that was treated at 55 °C showed higher dewaterability than when treated at 35 °C. Furthermore, cake TS decreased as pH increased. The TS of cake from pH 4.0 control samples and from polygalacturonase (at pH 4.0) treatment samples reached 15 % wt (the highest of all samples) quickly and remained high for all treatment durations, while the TS in cake from other treatments started at lower concentrations and increased as the duration of treatment increased. At the end of treatment, most treatments produced cakes that had a TS content over 12 %wt in the 55 °C treatment and over 10 %wt in the 35 °C treatment. In contrast, the TS content of cake from the DNase treatment decreased significantly to 6 %wt, followed by increasing dewaterability; however, the dewaterability of DNase-treated sludge was not better than that of raw sludge.

Role of temperature, pH and different hydrolases

In this study, the quantity of organic substances released from the interior fraction (TB-EPS) of activated sludge to the exterior fraction (supernatant) was greater at higher pH and higher temperature, even without added enzymes. These results demonstrate that activated sludge flocs are very unstable in an alkaline environment and at high temperature (i.e., 55 °C). Alkaline environment was widely applied to improve sludge’s hydrolysis degree [14, 15], and it is reported that higher temperature could promote the abiotic dissolution of proteins [22]. And higher temperature could speed up the enzymatic hydrolysis so as to demonstrate larger impact on disintegrating sludge floc.

As shown in Figs. 1, 2 and 3, enzymes affected the release of organic matter from the inner fraction (TB-EPS) to the most exterior fraction (supernatant) to different extents. Polygalacturonase was especially effective in mobilizing polysaccharides, with increases up to 700 % in total EPS polysaccharides amount, compared with raw sludge. However, proteins and DNA were not released in large amounts in accordance with polysaccharides, which meant these three components were not from the same source (e.g., resulted from bacteria cell lysis). Hence, it is unlikely that massive polysaccharides came out of cells lysis. Furthermore, the DN content of sludge treated using polygalacturonase was released to a much less extent than by other enzyme treatments, and the C/N ratio of polygalacturonase-treated samples decreased as incubation time increased. Combined with the ESEM results, which in this treatment were slightly different from those from the sludge pellet, the analytical results indicate that the polygalacturonic acid hydrolyzed by polygalacturonase has an important role in EPS floc formation (i.e., linkages of small flocs) but has no direct links with proteins or other organic matter that would be liberated when the linkages are broken. Hence, under polygalacturonase treatment, the polysaccharides content of sludge changed hugely, yet without a massive release of proteins or eDNA. Furthermore, compared with the ESEM observation of the pH 6.9 control group, there was no obvious relationship with the surfaces of floc from the polygalacturonase-treated sludge, suggesting the substrate that is acted upon by polygalacturonase is too minor to introduce significant morphological change.

Cellulase treatment caused the release of all organic species at a relatively high level, as indicated by DOC (Fig. 1d), DN (Fig. 1e), and humic-like substances (Fig. 3). Since cellulase acts on cellulose, or the so-called beta-carbohydrate, which comprises 1.2–38.9 %dw [23] of active sludge, the plain microstructure may indicate that beta-carbohydrate is the adhesive of the organic matter within the floc. When this adhesive was hydrolyzed by cellulase, the floc was dramatically broken down. As a result, the ESEM image of cellulose-treated sludge (Fig. 4) revealed a homogenous appearance that lacked structure.

A proposed hypothesis on EPS biochemical distribution model

Many conceptual models have been developed to elucidate the biological flocculation process and compositional distribution in the floc. The existing conceptual models can be classified into two types. In the first type, the whole floc is considered to be a gel-like matrix containing cell particles and organic colloids homogeneously bridged by various polymers and usually described using the Sierpinski carpet model [24–28]. In contrast, the second type of model considers the spatial distribution of biomass cells to be inhomogeneous, an interpretation mostly concluded from mathematical simulation of sludge floc growth, calculated parameters such as porosity, and morphological changes [29, 30]. Proteins, polysaccharides, and glycoproteins are considered to be the adhesive substances among microbial cells or between cells and attachable particles in the floc-forming process and are considered to be capable of influencing the subsequent treatable characteristics of activated sludge. These structural models normally consider the integrity of individual microbial cells or exopolymer particulates but rarely specifically describe the inner organic composition of cells and particulates; in addition, floc structures are usually described using particle size and porosity to distinguish among cells, fibers, and exopolymer.

However, asynchronous migration of carbon- and nitrogen-containing matter was observed in the present enzymatic study, as well as the unexpected hydrolysis efficiency of different enzymes. For example, polygalacturonase treatment changed the polysaccharides content of EPS significantly, yet without massive release of proteins or eDNA. The polysaccharide-releasing efficiency of polygalacturonase was far higher than that of the carbohydrases targeting amylum and cellulose, which were regarded as the major polysaccharides in EPS [5, 6]. In addition, the proteinase that targeted proteins, which comprised the majority of EPS organic material, was almost incapable of releasing protein; instead, DNase showed higher protein-releasing efficiency.

Therefore, the various objective polymers (e.g., proteins and polysaccharides) must have different spatial distributions within EPS. Much research on floc formation [5, 6, 31] has accepted that cellulose might be the backbone for the whole floc structure. In the present study, when cellulose was hydrolyzed by cellulase, both proteins and polysaccharides appeared in very high amounts in all treatment conditions and were highly redistributed among EPS fractions. These results indicated that cellulose is the constructive substance for EPS, such that when this “backbone” was destroyed, the organic contents of the sludge were readily released.

However, the role of polygalacturonase cannot be discounted. Polygalacturonase is an enzyme that hydrolyses the alpha-1, 4 glycosidic bonds between galacturonic acid residues. Galacturonic acid originating from bacterial source, as reported in the references [32, 33], is a component of bacterial polysaccharide and is important for cell structure. Frirdich and Whitfield [32] pointed out that Klebsiella pneumoniae l contains galacturonic acid residues that are proposed to serve a function in outer membrane stability. Ugalde et al [33] pointed out that a cell wall polymer containing galactose and galacturonic acid affects the bacterial strain’s agglutinability with lectin and virus sensitivity. Hence, it is quite possible that polygalacturonic acid-like matter becomes a component of the adhesives that bind sludge floc. The elimination of polygalacturonic acid-like matter could result in the disintegration of the whole extracellular polymer substances matrix. As confirmed in the present study, when polygalacturonic acid-like matter was hydrolyzed, the content of polysaccharides in supernatant soared up extremely high, while the content of proteins did not increase (Fig. 1). Furthermore, polygalacturonase will not only target on the polygalacturonic acid of bacterial source abundant in sludge floc, but it can also act on the polygalacturon of plant source, i.e., pectin. Since municipal wastewater treatment plant accepts sewage containing plant fiber [34], the recalcitrantly biodegradable pectin in sewage will finally accumulate in sludge floc. As a result, the added polygalacturonase can also take some effect by acting on this component.

Hence, a plausible description of EPS (illustrated in Fig. 6) is that it consists of a large amount of small flocs in which major microbes and most proteins and other nitrogen-containing substances are firmly embedded. In addition, polygalacturonic acid-like matter might be a minor, but key, adhesive among small flocs, such that several small flocs together with fiber made of cellulose and other polysaccharides further construct the EPS. As a result, when polygalacturonic acid-like matter is hydrolyzed by polygalacturonase, cellulose and other polysaccharides are liberated, while proteins remain in the small flocs. If proteinase cannot access the protein aggregates and cells in small flocs easily, its efficiency in releasing proteins from sludge will be limited.

Nonetheless, the conceptual model needs experimental evidence to directly visualize the spatial distribution of exopolymers at nanometer scale. This requirement highlights the utility of high magnification microscopy combined with fluorescent probing or functional group probing techniques to revisit the structure of cell exopolymer, so as to better understand the relationship between sludge biochemical properties and sludge treatability. Currently, there are few published studies on the absolute content of polygalacturonic acid-like matter in sludge and EPS. Hence, the need is acute for research on this special minor organic species, as well as for an accurate determination method for the sludge matrix.