Fluorescein diacetate hydrolytic activity as a sensitive tool to quantify nitrogen/sulfur gene content in urban river sediments in China

The relative abundance of functional genes used to quantify the abundance of functional genes in communities is controversial. Quantitative PCR (qPCR) technology offers a powerful tool for quantifying functional gene abundance. However, humic substances can inhibit qPCR in sediment/soil samples. Therefore, finding a convenient and effective quantitative analysis method for sediment/soil samples is necessary. The functional genes and physicochemical properties in sediments with different-level pollutions were analyzed in this study. Correlations between physicochemical properties and the relative abundance of functional genes were used to test whether relative abundance in gene prediction quantifies the abundance of functional genes. The abundance of functional genes could be corrected by multiplying the fluorescein diacetate (FDA) hydrolytic rates by the relative abundance of functional genes since the FDA assay has been widely used as a rapid and sensitive method for quantifying microbial activity in sediments. Redundancy analysis showed significant interrelations between the functional genes and the physicochemical properties of sediments. The relative abundance of functional genes is unreliable for quantifying the abundance of functional genes because of the weak correlation (R < 0.5, P < 0.05) between different pollutants and the relative abundance of functional genes. However, a significant positive correlation between concentrations of different pollutants and the activities of associated enzymes was obtained (R > 0.933, P < 0.05), which revealed that the abundance of functional genes could be reliably quantified by the relative abundance and FDA hydrolytic rate. This study proposed an alternative method besides qPCR to quantify the absolute abundance of functional genes, which overcomes the problem of humic interference in the quantitative analysis of sediment/soil samples.


Introduction
Human activities have altered and will continue to alter nitrogen and sulfur cycling by increasing the amount of reactive nitrogen and sulfur in urban river sediments (Nelson et al. 2016a). Nitrogen and sulfur cycling in sediment can lead to the formation of odorous thioether and black-colored particles, which are accountable for forming black-odorous rivers (He et al. 2022;Li et al. 2020). Environmental changes are altering the distribution of biodiversity, which in turn is a crucial driver of biogeochemical cycles Yang et al. 2022). Meanwhile, functional gene analysis is perfectly suited to link microbiology with the concentrations of different pollutants (Kandeler et al. 2006;Sun et al. 2022). Therefore, a quantitative analysis of functional genes is essential to predict how biogeochemical cycles will be changed with environmental conditions (Louca et al. 2016).
Quantitative real-time PCR assays are particularly promising because they can increase sensitivity, speed, and automation (Gentry-Shields et al. 2013). However, humic and fulvic acids are frequently cited as PCR Responsible Editor: Robert Duran * Jingmei Sun jmsun_group@163.com 1 inhibitors in sediment/soil samples, increasing the sample's detection limit and skewing calculated marker concentrations (Gentry-Shields et al. 2013;Kim et al. 2011). It was reported that the humic impurities (< 0.1 μg/ mL) interfere with the interaction between target DNA and Taq polymerase, a key enzyme in PCR amplification (Kim et al. 2011). Thus, measures must be taken to reduce the inhibitory effect of humic acid on quantitative real-time PCR, given that humic substances are significant constituents of organic matter in soils and sediments (Xiao et al. 2022). Removal of inhibitors has been reported using density gradient centrifugation using cesium chloride, hexadecyltrimethylammonium bromide, polyvinylpolypyrrolidone, and Sephadex G-100 and G-200 columns (Gentry-Shields et al. 2013). However, many of these methods are costly, labor-intensive, lengthy, or can result in significant, or even complete, loss of DNA during recovery procedures (Gentry-Shields et al. 2013). Therefore, finding a more economical and straightforward method for sediment/soil samples is necessary.
The fluorescein diacetate (FDA) assays have long been considered the standard method for determining enzyme activity, such as protease, lipase, and esterase (Serafini et al. 2022). FDA can be used to assess the level of substrate stimulation and heavy metal toxicity in river sediment 2019a). The CO 2 emission is also regulated by FDA hydrolase (Jaiswal and Pandey 2019b). The application of FDA hydrolytic rates includes biomass characterization and ecosystem functioning assessment (Jaiswal and Pandey 2021;Zhao et al. 2021). In a recent field trial , we found that the activity of FDA hydrolase, an extracellular microbial enzyme, is considered a proximate of microbial activity in urban river sediments. Besides, it is well known that the enzymes coded by functional genes catalyze each reaction step in the biogeochemical cycle of elements (Zhang et al. 2020). However, the current results obtained by Tax4Fun are all the relative abundance of functional genes . Therefore, we hypothesized that coupling FDA hydrolytic rates with the relative abundance of functional genes may quantitatively express functional genes.
This study investigated a simple alternative to PCR for quantitative gene expression in humic-rich soil and sediment samples. The main objectives of this study were as follows: (1) to analyze the prediction of microbial metabolic functions in urban river sediments with different-level pollution; (2) to test the reliability of the quantitative analysis of relative abundance in gene prediction; and (3) to correct the relative abundance of functional genes in amplicon sequencing results to quantify the abundance of functional genes reliably.

Sample collection
A total of 38 sampling sites in 9 urban rivers in northern China were investigated, and 9 of these sampling sites with different levels of contamination were selected for the highthroughput sequencing. The distribution of sampling sites is shown in Fig. 1. The 9 urban rivers were mainly influenced by fishpond pollution and rainwater runoff. The selected urban rivers represent the typical characteristics of northern Chinese cities, i.e., slow flow, severe pollution, poor selfpurification ability, and frequently malodorous before remediation. The surface layer of sediments (0-10 cm depth) was collected with a Petersen grab. The sediments and overlying water were sealed and quickly transported to the laboratory. These samples were put through a sieve with a 0.5-cm opening so that solids like stumps and shells would not mess up the experiment.

Physicochemical analysis
Sediments were pretreated by freeze-drying. The alkaline potassium persulfate oxidation method determined the sediments' total nitrogen (TN). The organic matter (OM) in sediments was measured using the potassium dichromate titration method (Liu et al. 2017). The acidified volatile sulfides (AVS) in sediments were determined by methylene blue spectrophotometry (Hsieh and Shieh 1997). A portable multifunctional instrument measured sediments' oxidation-reduction potential (ORP) values (HQ30d, HACH). The physicochemical properties of sediments are summarized in Table S1. We defined TN and AVS content in the sediments of sampling sites exceeding the mean (TN, 4.08 g/kg; AVS, 2.06 g/kg) as high-pollution sediment (HPS, including S1, S2, and S3), otherwise defined as low-pollution sediment (LPS S23, S17, S27, and etc.) ( Table S1). The FDA determination adopted the revised method described by Jiang et al. (2016).

FDA assay
In particular, 1.8 g of fresh sediment was mixed with 15 mL of 50 mM phosphate buffer solution (pH: 7.2) in a 50-mL conical flask. A total of 0.2 mL of 3 g/L FDA concentration was adopted. In control, the substrate solution was replaced by 0.2 mL acetone. The sediments were incubated at 30 °C for 0.5 h in the shaker at 50 rpm. Subsequently, 8 mL acetone was added to terminate the reaction. The sediment suspension was then centrifuged at 5000 × g for 5 min. The supernatant was then filtered through a 0.22-μm filtration membrane, and the fluorescence intensity of the liquid was determined by spectrophotometry at a wavelength of 490 nm.

High-throughput sequencing
Samples were collected twice from each system, and genomic DNA was extracted using the FastDNA® SPIN Kit for Soil (MP Biomedicals, USA) following manufacturer instructions. Subsequently, the DNA samples were amplified using the primers 338F (ACT CCT ACG GGA GGC AGC AG) and 806R (GGA CTA CHVGGG TWT CTAAT), targeting the V3-V4 region of the bacterial 16S rDNA genes. Highthroughput sequencing was sequenced using the Illumina Miseq PE300 platform (Illumina, Inc., CA, USA) by Beijing Auwigene Tech Ltd. (Beijing, China) ).
The heatmap analysis was performed using the gplots package of R. The redundancy analysis (RDA) was carried out using the vegan package of R. The downstream analysis used a non-metric multidimensional scaling (NMDS) method to compare the similarity of these 9 types of sediments with different levels of contamination. The NMDS data were created using the R software project using OTUs from various genera, with the analysis of similarities (Anosim) technique used to see whether the differences between groups were significant ).

Functional gene
In order to study the metabolic pathways of nitrogen and sulfur in urban river sediments, 9 types of sediments with different-level pollution were used to predict functional genes. The abundance of functional genes explored the elemental metabolic pathways in different ecosystems (Nelson et al. 2016b;Shu et al. 2016). Tax4Fun has become a popular and robust tool for analyzing complex communities' functional genes . The normalized bacterial OTU table was imported to Tax4Fun to predict microbial metabolic function (Aßhauer et al. 2015). Tax4Fun function predictions based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) metabolic pathway database analysis have been used to predict functional genes ). The main pathways, related genes, and relative nitrogen and sulfur metabolism abundance are listed in Table S2 and  Table S3.

Statistical analysis
Samples were evaluated in three sets of concurrent experiments to quantify the random error of measurement. For all outcomes, the covariation coefficient was less than 5%. The experimental data were also statistically examined using one-way ANOVA (SPSS 21.0) with a significance threshold of P < 0.05. SPSS Statistics 21.0 was also used for statistical analysis.

Sulfur functional genes
The main pathways, related genes, and relative abundance of sulfur metabolism are listed in Table S3. The biogeochemical cycle of sulfur integrates multiple metabolic pathways such as dissimilatory sulfate reduction, assimilatory sulfate reduction, sulfur transport system, and sulfide oxidation (Bradley et al. 2011;Fike et al. 2015). Metabolic function predictions showed that Sox, hdr, and iscS genes encode the sulfide oxidation enzymes. Sulfur transport enzymes are encoded by cysA, cysU, cysW, and cysP genes. The genes sat, cysH, cysI, cysNC, and sir are detected for assimilatory sulfate reduction. The genes dsrA, dsrB, dsrC, aprA, and aprB are detected for dissimilatory sulfate reduction. As shown in Fig. 2 (b), most of the genes associated with sulfur oxidation in HPS showed downregulation, while those showed upregulation in LPS. At the same time, the genes for the dissimilatory sulfate reduction showed upregulation in HPS, while those in LPS showed downregulation. Besides, the relative abundance of dissimilatory sulfate reduction genes in S2, S30, S31, S28, S20, S19, S23, S17, and S27 was 0.10%, 0.07%, 0.07%, 0.08%, 0.06%, 0.07%, 0.06%, 0.04%, and 0.07%, respectively (Table S3). The result indicated that the dissimilatory sulfate reduction enzymes were stimulated in HPS, which could be attributed to the high organic matter content of HPS in comparison to that of LPS (Nakayama et al. 2000). Meanwhile, the relative abundance of sulfatereducing bacteria (e.g., Desulfobulbus, Desulfobacca, Desulfatirhabdium, Desulfovirga, Desulfatiglans) in HPS was higher than that of LPS (Fig. S1). These results indicated that the relative abundance of sulfur function genes might qualitatively reflect gene abundance as a general trend exists between these two variables.

Linkages between physicochemical parameters and functional genes
To further investigate the correlation between the physicochemical parameters and functional genes, redundancy analysis (RDA) was conducted using ORP, OM, TN, AVS, and FDA in all samples as key variables (Fig. 3). The PC1 and PC2 explained the maximum variation of 98.43% and 1.02%, respectively, indicating significant interrelations between the physicochemical parameters and functional genes in all samples. The samples of LPS were well separated from HPS along the PC1 vector, indicating that the metabolic pathways in HPS differed from those in LPS. The NMDS result is consistent with the RDA results and showed clear distinctions in the functional genes between LPS and HPS (Fig. S2). As shown in Fig. 3, RDA analysis revealed that dissimilatory sulfate reduction, sulfur oxidation, nitrogen fixation, nitrogen mineralization, and denitrification were related to AVS, FDA, TN, and OM. In contrast, assimilatory sulfate reduction was assigned to ORP. These results showed a high correlation between the physicochemical parameters and functional genes. Although the relative abundance of function genes can qualitatively reflect gene abundance, an extra test needs to be investigated when using relative abundance to make quantitative inferences regarding functional genes.

Test of quantitative analysis of relative abundance
As a simple tool for characterization of biomass and total enzyme activity , FDA was used to test the reliability of the quantitative analysis of relative abundance in gene prediction. The correlation between the concentration of different pollutants and FDA hydrolytic rates (Fig. 4) was compared with the correlation between the concentration of different pollutants and the relative abundance of functional genes (Fig. 5) to test the quantitative analysis of relative abundance.
As shown in Fig. 4 a, the correlation between the concentrations of TN and FDA hydrolysis rates in sediments was significantly positive (R = 0.704, p < 0.05), indicating that the concentrations of TN in sediments as a driver of microbial community structure have a more   (Han et al. 2020). There was also a significant positive correlation between the concentrations of AVS and FDA hydrolytic rates in the sediments (R = 0.773, p < 0.05) (Fig. 4 b).
The result was primarily attributable to the high activity of sulfate-reducing bacteria in the sediments. The values of ORP in sediments ranged from 100 to − 300 mV (Table S1), which were the favorable ORP values of sulfate-reducing bacteria (Zhang et al. 2018). Meanwhile, sulfate-reducing bacteria (e.g., Desulfobulbus, Desulfobacca, Desulfatirhabdium, Desulfovirga, Desulfatiglans) were detected in the present study (Fig. S1). Therefore, AVS in sediments has a more significant impact on FDA hydrolytic rates. As shown in Fig. 5 a, the correlation between the concentrations of TN and the relative abundance of nitrogen fixation enzymes showed a positive correlation (R = 0.775, p < 0.05), which was consistent with the positive correlation between FDA hydrolytic rates and the concentrations of TN in the sediments (R = 0.704, p < 0.05). The result illustrated that the relative abundance of nitrogen fixation genes could be a quantified reflection of gene abundance. In contrast, there was a weak correlation (R = 0.412) between the concentrations of AVS and the relative abundance of sulfur reduction genes, which was lower than the correlation between FDA hydrolytic rates and the concentrations of AVS in the sediments (R = 0.773, p < 0.05). These results demonstrated that quantifying gene content based on gene prediction through mapping 16S rRNA genes to homologous taxa with fully sequenced genomes is unreliable.
In addition, the relative abundance of nitrogen fixation genes in HPS was 1.2-1.5 times higher than in LPS (Table S2). However, the TN concentrations in HPS were 2.6-11.4 times higher than in LPS (Table S1). Furthermore, the relative abundance of dissimilatory sulfate reduction genes in HPS was 1.3-2.5 times higher than in LPS (Table S3), and the concentrations of AVS in HPS were 2.2-11.8 times higher than in other LPS (Table S1). These results indicated that mapping 16S rRNA genes to homologous taxa with fully sequenced genomes potentially underestimates gene frequencies of genes involved in biogeochemical cycles.

Correction of quantitative analysis of relative abundance
Microbial enzyme activities in sediments are expressed by functional genes (Sun et al. 2019). Meanwhile, the FDA hydrolytic rates can characterize sediment microbial activity (Serafini et al. 2022). Thus, higher FDA hydrolytic rates reflect higher gene expression (Zhang et al. 2020). As a result, coupling FDA hydrolytic rates with the relative abundance of functional genes might be a reliable tool for quantitatively analyzing the abundance of functional genes.

3
The functional genes can be reliably quantified by FDA hydrolytic rate. After correction, the higher Pearson product-moment correlation coefficient indicated that the FDA hydrolytic rates in the sediments multiplied by the relative abundance of functional genes could quantify the functional genes in community analysis.

Conclusion
The functional genes in 9 types of sediments with different-level pollutions were analyzed in this study. A general trend was identified whereby the pollutant concentrations strongly influenced the relative abundance of genes. The unreliability of quantifying the abundance of functional genes in community analysis was further verified by constructing the correlation between the concentrations of different pollutants and the relative abundance of functional genes. Subsequently, a significant positive correlation between concentrations of different pollutants and activities of associated enzymes was obtained (R > 0.933, P < 0.05), which confirmed that the FDA hydrolytic rates in the sediments multiplied by the relative abundance of functional genes could reliably quantify the abundance of functional genes. This study has expanded the application of FDA hydrolytic rates in nitrogen/sulfur gene prediction analysis.