Bacillus velezensis LG37: Transcriptome profiling and functional verification of GlnK and MnrA in ammonia (NH4+) assimilation

Background In recent years, interest in Bacillus velezensis has increased significantly due to its role in many industrial water bioremediation processes, including the use of probiotics. In this study, we isolated and assessed the transcriptome of Bacillus velezensis LG37 (aquaculture pond) under different nitrogen sources. Since Bacillus species exhibit heterogeneity, it is worth investigating the molecular information of LG37 through ammonia nitrogen assimilation, where nitrogen in the form of ammonia is considered toxic to aquatic organisms. Results Here, a total of 812 differentially expressed genes (DEGs) from the transcriptomic sequencing of LG37 grown in minimal medium supplemented with ammonia (treatment) or glutamine (control) were obtained, from which 56 had Fold Change ≥ 2. BLAST-NCBI and UniProt databases revealed 27 out of the 56 DEGs were potentially involved in NH4+ assimilation. Among them, 8 DEGs together with the two-component regulatory system glnK/glnL were randomly selected for validation by quantitative real-time RT-PCR,and the results showed that expression of all the 8 DEGs are consistent with the RNA-seq data. Moreover, the transcriptome and relative expression analysis were consistent with the transporter (amtB) gene of LG37 and it is not involved in ammonia transport, even in the highest ammonia concentrations. Besides, CRISPR-Cas9 knockout and overexpression LG37 mutants of glnK further evidenced the exclusion of amtB regulation, suggesting the involvement of alternative transporter. Additionally, in the transcriptomic data, a novel ammonium transporter mnrA was expressed significantly in increased ammonia concentrations. Subsequently, OEmnrA and ΔmnrA LG37 strains showed unique expression pattern of specific genes compared to that of wild-LG37 strain. Conclusion Based on the transcriptome data, regulation of nitrogen related genes was determined in the newly isolated LG37 strain to analyse the key regulating factors during ammonia assimilation. Using genomics tools, the novel MnrA transporter of LG37 became apparent in ammonia transport instead of AmtB, which transports ammonium nitrogen in other Bacillus strains. Collectively, this study defines heterogeneity of B. velezensis LG37 through comprehensive transcriptome analysis

Corneybacterium, Desulfitobacterium, Flavobacterium, Geobacterium, Micrococcus, Mycobacterium, Nocardia, Pseudomonas, Rhodococcus, Sphingomonas, Vibrio, etc., [17][18][19]. In particular, Bacillus species demonstrate outstanding efficiency in the water restoration projects with multiple benefits, including distribution, easy isolation and cultivation, and endospores that can be stored for protracted periods [20]. However, Bacillus sp. respond to N-availability by displaying heterogeneity and autoregulation through both positive and negative feedback switching mechanisms in isogenic cell populations. Regulation of genes involved in signal perception, transmembrane transporter, transcriptional regulators, and key enzymes in N-metabolism of a whole population can be studied by applying phenotypic measurements and conventional molecular techniques [21,22].
Bacillus sp. take up NH 3

by diffusion under high pH and high NH 4 + concentration, while intracellular
In the present study, we isolated three pure strains of Bacillus sp., from the grass carp pond water.
Among them, a particular isolate, namely, Bacillus velezensis LG37, was successfully screened using minimal media with ammonia as the sole nitrogen source for the best ammonia nitrogen assimilation efficiency; thus, it was chosen for further analysis.

Growth characteristics of LG37 culture conditions
The bacterial growth kinetics of wild-type LG37 was analyzed by culturing in LB broth, and the growth was measured by spectrophotometry. From 25 to 37 °C, the LG37 showed an excellent growth trend.
The peak levels OD (≈7) were maintained for almost 30 h in the static cultivation environment (Additional File 1). As shown in Additional File 1, LG37 on 1 -6 % of salt concentrations had a sharp and stable increase in their growth exceeding OD 600 value of 5 at 28 h followed by a gradual decline.
Yet, the levels were maintained above 5 up at to 3% salinity, and the decrease in OD was proportional to salt concentration in the extended period. The LG37 growth was normal at pH 6-8.5, while slow growth was recorded at pH 9.5. The dissolved oxygen (DO) concentration available in the medium mainly influences the growth of the bacterial population [43]. The increasing growth values were observed in all the DO reactors from 0 to 40 h, and the LG37 growth was proportional to the level of DO. Peaked OD of 8.2, 6.1, and 4 -were obtained for LG37 with initial DO of 6.0, 4.2, and 3.0 mg/L, respectively, at 35 h of the culture period. Conversely, there was no substantial growth recorded in the 1.8 mg/L DO (Additional File 1).

Optimization of Glutamine and ammonia nitrogen concentrations for LG37
Growth characteristics of LG37 were determined at various levels of NH 4 + and Gln by OD 600 . The results showed, with the increase of Gln concentration, LG37 displayed an inclination of growth respect to the amount of Gln ( Fig. 2A). The growth curve of LG37 in minimal media with NH 4 + (5, 10, 15, 20, 25, and 30 mmol/L) as the sole nitrogen source showed a similar trend to that of Gln, but no significance was observed at 20 -30 mmol/L NH 4 + concentrations (Fig. 2B). However, the growth of LG37 was more significant in the increasing levels of nitrogen; the other nutrients in the minimal media might be a limiting factor for their growth above 20 mmol/L. Given that and to avoid overloading, we preferred an average volume (10 mmol/L) of N-source for both Gln-N and NH 4 + -N in our supplementary studies.

Transcriptome assembly and functional annotation
The mRNA of the LG37 cells acquired from the treatment group (NH 4 + -N) and control group (Gln-N), were sequenced by Illumina Hiseq TM 2500 to obtain the overview of the gene expression pattern.

RT-qPCR verification of selected genes
The Illumina sequence of LG37 expression profile data was verified through randomly selected 8 DEGs that including, 5 up-regulated and 3 down-regulated genes (Additional File 7) using RT-qPCR. The results exhibited a similar expression tendency but with slight variation in their levels, which confirmed the reliability of DEGs from the transcriptome sequencing results (Fig. 5).

GlnK is critical for NH 4 + assimilation
To understand the functions of the specific genes associated with NH 4 + assimilation pathway that short-listed by transcriptome data, especially the significantly upregulated glnK (4.26) and glnL (4.12), were primarily analyzed by RT-qPCR following cultivation of wild-LG37 at different concentrations of NH 4 + . The glnK and glnL showed a significant increase in their expression with the rise of NH 4 + concentration, and the glnK showed a greater tendency than the glnL (Fig. 6A). These results demonstrate that the two-component regulatory systems glnK/glnL play a synergistic regulation function in the LG37 NH 4 + metabolism.

Functional analysis of related genes using OEglnK and ΔglnK in NH 4 + metabolism
To demonstrate the functionality of the GlnK in NH 4 + assimilation pathway, glnK was knocked out (LG37-ΔglnK) by applying the CRISPR/cas9 technology (Additional File 8) using pHT1K plasmid that developed overexpression mutant (LG37-OEglnK) strain (Additional File 9). The mutants were analyzed for the regulation of ammonia assimilation related genes (glnL, amtB, and glnA) by RT-qPCR assays along with wild LG37 strain, following cultivation in 10 mmol/L NH 4 + containing minimal medium. The LG37-OEglnK strain increased the expression of about 2.5 folds and 1.8 folds, while the LG37-ΔglnK strain lead to a decrease in the expression to 0.25 and 0.7 folds for glnL and glnA, respectively. No notable changes were recorded in the amtB gene expression compared to that of control (Fig. 6B). The growth curve of both the mutated strains represented variations in growth with increased growth in LG37-OEglnK and decreased growth for LG37-ΔglnK strain when compared to the wild-LG37 strain (Fig. 6C). These results indicated that the GlnK plays a significant role in the NH 4 + assimilation of LG37.
Based on the above results, we noted that the GlnK senses the NH 4 + concentration and regulates glnL expression through signal transduction, and further, the GlnL promotes glnA and amtB in favor of NH 4 + assimilation. Besides, AmtB as an ammonium transporter, there was no significant expression among the 3 groups (Fig. 6B), which is similar to that of the transcriptome results of amtB between NH 4 + -N and Gln-N nitrogen groups. Hence, we analyzed the relative expression of AmtB in wild-LG37 at an increased NH 4 + concentration, and the results exhibited no difference in their expression pattern even at the highest NH 4 + -N levels (Fig. 6D). The above data suggest that the GlnK plays a vital role in regulating GlnL and GlnA, and still, the AmtB were not as a specific positive factor for NH 4 + transporter of LG37 in assimilation. Therefore, it has been ruled out that there might be other NH 4 + transporters in the case of LG37.

Determination of the NH 4 + transporter MnrA in LG37
Considering the relative expression of amtB indicated in Fig. 6B and D, we discovered that the AmtB is not specific in the NH 4 + transport of LG37, but this finding was only based on our transcriptomic and mRNA expression data and not direct assessment. Thus, we further analyzed our transcriptome data for another transporter, where we noticed mnrA with a 2.93-fold change in expression.
(Additional File 4). To verify whether MnrA plays a role in NH 4 + assimilation, we detected the expression of mnrA of LG37 after culturing with different NH 4 + -N concentrations by using RT-qPCR.
The expression of mnrA increased with an increase in NH 4 + -N concentrations (Fig. 7A). In the LG37-ΔglnK and LG37-OEmnrA mutant strains, the mnrA expression was increased by a factor of 1.67 and decreased to 0.35, respectively, compared to that of wild-LG37 strain (Fig. 7B). Similarly, we determined the growth of both strains showed increased growth pattern in LG37-OEmnrA, whereas LG37-ΔmnrA strain showed a decreased growth compared to that of wild-LG37 strain (Fig. 7C). These results demonstrated that MnrA is essential for the NH 4 + assimilation process in LG37. [23], B. subtilis, strain A1 [46], and B. subtilis [47] strains demonstrated substantial ammonia nitrate removal. All of those previous findings suggesting an encouraging sign, lead us to isolate the potentially efficient Bacillus strain for NH 4 + assimilation and to determine the mechanism of a complex metabolic network like signal inductors, transcriptional regulators, transporters, assimilation enzyme and so on. The increased heterogenetic gene-expression of the bacterial population could not only utilize the existing nutrients for their growth but also get a benefit for the survival of the bacterial population in extreme conditions. One needs to study the interacting molecules and their networks, contribute to understanding the biological system function [48].
In the present study, a Bacillus velezensis LG37 was isolated from the aquaculture pond that exhibited a greater nitrogenous compound removal between the three isolated strains. The LG37 strain illustrated good growth at a wide range of temperature, pH, and salinity. For the transcriptomic approach of nitrogenous substances, we determined the growth conditions using the minimal inorganic medium that lack rich nutrients incorporated with organic (Glutamine-N) and inorganic nitrogen (NH 4 + -N) for ammonia nitrogen metabolism at the molecular level in LG37. Undoubtedly, the LG37 could able to grow well in both N-substrates but to bypass overstressing and to offer other nutrients from the minimal media; the substrates were volume-averaged to 10 mmol/L. Mining the potential genes from the transcriptome data may shed light on the mechanism of N-metabolism of LG37.
To dissect the role of differential gene expression (DEGs) profiles, a global transcriptomic profiling method was employed for the LG37 genome that cultivated under two different N-sources. and further studies will address the features of the unreported functional genes involved during ammonia assimilation.
Considering the greater challenges related to the data derived from the transcriptomic study that have to be further verified by other approaches, including RT-qPCR [51,52]. In our study, to eliminate the effects of possible amplification bias, we validated the expression patterns of the 8 randomly selected genes using the RT-qPCR. The primary results displayed that the pattern distribution and gene-expression levels were highly correlated with Illumina sequencing data. Additionally, these comparative data presented that not all the functional genes contributed in the same manner but also proved the significance of the genes in the ammonia assimilation process. Although the preliminary results provide a theoretical basis, such information still required an in-depth analysis of potential functional genes in the NH 4 + assimilation metabolic network of LG37.
Following our transcriptome data, it was shown that there is no positive correlation between glnK and amtB. We speculate that the AmtB is not specific for NH 4 + transport, as well as there was not a direct interaction also between GlnK and AmtB. To test the role of LG37 specific genes in NH 4 + assimilation, mutant strains were constructed using the all-in-one CRISPR-Cas9 genome editing system to develop knockout (pJOE8999) [53] and overexpression (pHT1K) LG37 strains. It was again exposed that GlnK does not have a correlated mechanism with the transporter AmtB between the wild-LG37 and mutant however, little difference was observed between OEglnK and the wild-type strain (Fig. 6C). The reduced growth of the knockout cells might be due to the inability to utilize the N-compounds where the minimal media contained only necessary essential nutrients. These results indicated that although glnL and glnA was induced, the amtB gene was found to be non-essential for the transport of NH 4 + in LG37 strain. Thus, LG37 might likely adopted distinct NH 4 + transporter mechanism similar to that of AmtB in other Bacillus strains.
Conversely, based on the LG37 transcriptomic data, we found a distinct transporter (mnrA) was upregulated to 2.92 folds, and we speculate that MnrA might be an alternate NH 4 + transporter in LG37. The MnrA is a single-polypeptide secondary carrier transmembrane transporter protein belong to the Major Facilitator Superfamily (MFS) that promotes small solutes in response to chemiosmotic ion gradients [54,55]. The relative expression levels of mnrA showed remarkable expression levels when LG37 grown in the increasing NH 4 + concentrations, thus the same reflected in the OEglnK mutant strain. The ΔglnK mutant resulted in a reduction of mnrA expression compared to that of wild type.
Furthermore, to confirm this theoretical assumption, we constructed the ΔmnrA and OEmnrA LG37 strains, as mentioned earlier, and analysed for their growth and the expression of downstream genes (glnA and glnB) by RT-qPCR. Our analysis revealed that growth characteristics of the LG37 mnrA mutant strains were almost similar to the growth of glnK mutants and the contrasting expressions of glnA and glnB genes between the strains, compared to that of wild-LG37.

Conclusions
The newly isolated bacterium Bacillus velezensis was named as B. velezensis LG37, showed its ability of well growing in various basic parameters and its heterotrophic capacity to utilize both NH 4 + -N and Glutamine-N as a sole nitrogen source. The transcriptome sequence analysis under different nitrogenous sources proposed that LG37 up-regulated (18) or down-regulated (9) genes predominantly related to nitrogen metabolism. Moreover, 18 up-regulated DEGs associated with a few hypothetical proteins, transcriptional regulator, transporter, transporter permease, and GlnK/GlnL regulatory system for assimilating the N-compounds. Further, we applied genome editing technology to the genes whose expression was impacted in nitrogen assimilation; the glnK mutants led to uncovering that amtB was not associated with ammonium transport (Fig. 6). However, knock out of glnK in LG37 showed comparable growth to wild-type/OEglnK on ammonia as a sole nitrogen source, which demanded us to expose the unidentified ammonia transport system. In particular, the upregulation of mnrA transcriptional expression was shown to correlate with the ammonia transporter specifically and exhibited the regulation of related downstream genes (glnA and glnB). Altogether, the MnrA was proved as a novel ammonium transporter in Bacillus velezensis LG37 and our results provide a theoretical basis and new clues to the NH 4 + assimilation mechanism. Furthermore, we hypothesize that more than one ammonia transporter might have involved from the transcriptome of LG37 differential expression (Additional File 4), and this can be evaluated through subsequent experiments.

Bacterial Isolation
The Bacillus sp. was isolated among a collection of microbes from the aquaculture pond in Huazhong Agricultural University teaching practice base, Wuhan, Hubei Province, China, where grass carp was cultivated, following the standard operating protocol [56], with minor modifications. Briefly, a series of 10-fold diluted water samples in distilled water were inoculated (100 uL) on to the LB agar plates using a sterile triangle glass rod and incubated overnight at 32 °C. Three Bacillus isolates were obtained from the pond water and characterized for morphological, physiological, and biochemical characters extensively based on the Bergey's Manual of Determinative Bacteriology [57]. Among them, one Bacillus (wild) strain was identified with the highest nitrogen removal efficiency and named as Bacillus velezensis LG37.

Determination of optimal nitrogen concentration and other growth characteristics
To evaluate the growth of various nitrogen substrates and optimization of culture conditions for LG37 strains. The inoculum obtained from the late logarithmic phase was inoculated in 250 mL with 100 ml gram-positive bacterium that holds a thick peptidoglycan layer which resists such reagents for a while before acting upon, and the intended period or the mixtures might affect the nature of gene expression. The frozen bacteria were pulverized carefully in a mortar using a pestle with a constant supply of liquid nitrogen. The pulverized powder was uniformly (with high agitation) dissolved in 1 mL TRIzol Reagent (Invitrogen, CA, USA). Chloroform (200 µL) was added, and the solution further shaken thoroughly. It was then kept on ice for 5 minutes. Centrifugation (10,000×g, 15 minutes) was carried out before 1:1 (v/v) of supernatant, and isopropyl alcohol was uniformly and gently mixed with a pipette. The mixture was returned to the ice for 20 min, then centrifuged (10,000×g, 30 minutes).
The supernatant was carefully discarded, 1 ml of 70% ethanol pipetted into the tubes to wash the residual pellet by gently allowing the ethanol to flow over the pellets, while partially rotating the tubes. The ethanol was then discarded carefully. The wash-discard process repeated twice; then, the tubes were air-dried for 5 minutes. The RNA pellets were then dissolved in 80 µL of RNase-free water.
The to digest DNA remnants. RiboZero rRNA removal kit (Epicentre, USA) for gram-positive was used to eliminate ribosomal RNA before RNA sequencing analysis was done. One hundred nanograms of rRNAless RNA from each test were fractionated into 200~300 nucleotides (nts) and utilized as a format for random prepared PCR. Strand-specific cDNA libraries were generated by standard procedures for ensuing Illumina sequencing using the mRNA-seq Test Prep pack (Illumina, USA) system platform (Shanghai Oebiotech Co., Ltd., Shanghai, China).

Transcriptome data analysis
Duplicated sequences, ambiguous reads, and low-quality reads (Q > 20) were removed and assembled from the raw reads. Then the clean reads were mapped against the reference genome B.
velezensis FZB42 (GenBank accession code: NC_009725.1) using the Burrows-Wheeler transform (BWT) algorithm [59]. To analyse, differentially expressed genes (DEGs) levels with a different nitrogen source, the number of reads were calculated by using Reads Per Kilobase per Million Mapped The mutant ΔglnK strain with glnK deletion and ΔmnrA with mnrA deletion were constructed with chimeric single guide RNA (sgRNA) using the plasmid pJOE8999 and the homologous recombination methods as described [52]. The 20-nt protospacer adjacent motif (PAM) sequence was determined with 20bp upstream of 5′-NGG-3′ in glnK and mnrA sequences of B. velezensis LG37, and the primers were designed and synthesized by Shanghai Sangon Biotechnology Co., Ltd. (Additional File 11). The pJOE8999 plasmid was digested with BsaI restriction enzyme. Both products of primers sgRNA-glnK and sgRNA-mnrA were then ligated to generate pJOE8999-sgRNA-glnK and pJOE8999-mnrA plasmids, respectively.
The backbone of the vector and the spacer sequence was integrated by using 600bp of upstream and

Construction of glnK and mnrA overexpression LG37 strains
To investigate the exact role of glnK and mnrA in ammonia assimilation, the P xyl promoter DNA region from LG37 genomic DNA was amplified, and the PCR products were inserted into the pHT1K expression vector on NcoI and BamHI restriction endonuclease recognition site. Subsequently, PCR amplified products of LG37 genomic glnK and mnrA were cloned into the integrating plasmid pHT1K-P xyl at BamHI and KpnI restriction endonuclease recognition site to obtain pHT1K-P xyl -glnk and pHT1K-P xyl -mnrA recombinant plasmids, respectively. The DNA sequences were amplified using primers listed in (Additional File 11). The acquired plasmids were transformed into competent Escherichia coli DH5α cells for overnight cultivation in LB ampicillin plates at 37°C. Positive clones containing inserts of the expected size for P xyl , glnK, and mnrA fragments were verified by sequencing and by using specific restriction enzyme digestion analysis. Both recombinants were transformed separately into LG37 by electroporation, and successful transformants were screened with 25 μg/mL erythromycin in LB solid medium. Diagrammatic representation of a typical OE plasmid construction was shown in Additional File 9.

Statistical analysis
Statistical analyses were performed using a statistical package for social sciences (

Ethics approval and consent to participate
All the experimental procedures involved in the current research were followed by the appropriate guidelines of the Institutional Animal Care and Institute of Huazhong Agricultural University (IACIHZAU).

Consent for publication
Not applicable

Availability of data and materials
RNA-seq date have been submitted to GEO under the accession number GSE136178.

Competing interests
The authors declare that they have no competing interests. Revised the manuscript: BTA. All authors read, performed critical revisions and approved the manuscript.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.