Draft genome sequence of a new carotenoid-producing strain Brevibacterium sp. XU54, isolated from radioactive soil in Xinjiang, China

Abstract

Some species of the genus Brevibacterium are orange bacteria involved in cheese ripening, synthesis of odoriferous compounds, and carotenoids with aromatic end groups. Here, we report the genome sequence of Brevibacterium sp. XU54, isolated from radioactive soil in Xinjiang, China. The genome of XU54 consists of 4,899,099 base pairs with a GC content of 62.2%. The genome sequence was annotated with 4453 genes, encoding 4260 proteins, 13 rRNAs, and 49 tRNAs. 16S rRNA BLAST and comparative genomic analysis both indicated that XU54 may be a new species of Brevibacterium. In addition, compared to the type strains, some enzymes related to sulfur metabolism showed a low similarity of 66.85, 79.53 and 14.61%, respectively. The carotenoids biosynthesis gene cluster was identified and analyzed according to the genomic data, which revealed relatively low identity (5–85%) with existing strains. The optimum conditions for its growth and carotenoid production were then discussed. The whole-genome sequence of Brevibacterium sp. XU54 will be beneficial for utilizing these newly identified genes in carotenoid biosynthesis and regulation of sulfur metabolism pathway to promote the production of novel carotenoids and other structurally diverse compounds through combinatorial biosynthesis, which facilitates cheese ripening and coloration.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13205-022-03366-1.

Introduction

The genus Brevibacterium can be isolated from permafrost sediments, rice, and contaminated soils (Maizel et al. 2015; Kumari et al. 2021). Brevibacterium aurantiacum and Brevibacterium antiquum are newly discovered species belonging to this genus, which is orange pigmented (Anast et al. 2019; Levesque et al. 2019; Gavrish et al. 2004). They were first assigned to the species of Brevibacterium linens, which is the major component of the cheese microflora. Then they were classified as the new species, Brevibacterium aurantiacum sp. nov. and Brevibacterium antiquum sp. nov. (Gavrish et al. 2004). These two species, as well as B. linens, are usually found on the surface of cheese, which turns red or orange attributing to the carotenoids produced by these bacteria (Giuffrida et al. 2020). It has been reported that they can produce carotenoids with aromatic end groups, such as hydrocarbon isorenieratene, 3-hydroxyisorenieratene, and 3,3¢-dihydroxyisorenieratene (Krubasik and Sandmann 2000). These carotenoid biosynthetic pathways have also been reconstructed, redesigned, and extended for producing novel carotenoids (Kim et al. 2010). Besides, B. aurantiacum and B. antiquum are involved in the synthesis of volatile sulfur compounds (VSCs), which are the key aromas of cheese flavor (Forquin et al. 2011). However, in addition to the above two reported articles, there are few detailed studies on Brevibacterium genus, whereas little is known about carotenoids and VSCs synthesis pathways and related key genes. Due to genus of Brevibacterium playing an essential role in cheese ripening and flavor, new types of Brevibacterium should be found and studied for its secondary metabolites in color development and flavor compounds during cheese ripening in detail.

We recently isolated a new bacterial species, Brevibacterium sp. XU54 from the extreme environment of irradiated area in Xinjiang Uyghur Autonomous Region of north-west China instead of from cheese. In this study, the genome of XU54 was sequenced and analyzed. The 16S rRNA sequence of XU54 showed 98.22% and 97.95% similarity with B. aurantiacum NCDO 739^T and B. antiquum VKM Ac-2118^T, respectively. The phylogenetic and comparative genomic analysis between XU54 and other type strains of Brevibacterium genus was analyzed and compared. The predicted genes in the genome were annotated and the gene clusters of secondary metabolites were identified. Genes involved in VSCs metabolism process were also found in the sequenced genome. The carotenoids from XU54 were extracted and investigated for their production. These predicted functional genes and secondary metabolites can provide information of genes and proteins that are different from other genera, to explore their special mechanisms in cheese ripening, flavor production and carotenoid synthesis.

Materials and methods

Culture conditions and DNA extraction

The strain XU54 was isolated from soil samples 5–10 cm underground in the extreme radioactive environment in Xinjiang Uygur Autonomous Region of northwest China. 0.5 g of the soil sample was weighed into a flask containing 20 mL of sterile saline and shaken at room temperature for 10 min. The suspension was spread onto LB solid medium plates after gradient dilution and incubated at 30 °C for 1–7 days. For pure culture, XU54 was grown in LB medium (1% peptone, 0.5% yeast extract, 1% sodium chloride) at 25 °C and 200 rpm for 28–48 h. Genomic DNA of XU54 was extracted using a genomic DNA extraction kit (Takara, China). Cell growth curves were measured periodically by OD₆₀₀.

Genome sequencing, assembly, and annotation

The whole genome was sequenced by Majorbio (Shanghai, China) with a paired-end library, using the Illumina HiSeq 2000 sequencing platform. Quality control assessment was performed using Trimmomatic (Bolger et al. 2014), SeqPrep (https://github.com/jstjohn/SeqPrep), Sickle (https://github.com/najoshi/sickle) and Fastqc (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) to filter out adapters in reads, ambiguous reads, reads with more than 10% “N” base, reads having average quality score lower than 20 in paired-end reads, and reads with length less than 25 bp after quality-trimming. Illumina sequencing data was assembled using SOAPdenovo software (Version 2.04) (Luo et al. 2012). The rRNA and tRNA genes were predicted using barrnap (Version 0.8) and tRNA-scan-SE (Version 2.0), respectively. Coding DNA sequences (CDS) were predicted and annotated using Glimmer (Version 3.02, https://ccb.jhu.edu/software/glimm er/index.shtml) (Delcher et al. 2007). The functional annotation of proteins was compared using NR database in GenBank (ftp://ftp.ncbi.nlm.nih.gov/blast/db/), Blast2go (Version 2.5, https://www.blast2go.com/), Kyoto Encyclopedia of Genes and Genomes (KEGG) database (http://www.genome.jp/kegg/) (Kanehisa and Goto 2000), eggNOG database (Version 4.5.1, http://eggnogdb.embl.de/#/app/home) (Muller et al. 2010), Pfam database (Version 31.0, http://pfam.xfam.org/), and Swiss-Prot database (https://web.expasy.org/docs/swiss-prot). The secondary metabolite gene clusters were predicted using the antiSMASH (Antibiotics and Secondary Metabolite Analysis Shell) online tool (https://antismash.secondarymetabolites.org/) (Weber et al. 2015). The phylogenetic tree of the carotenoid biosynthesis gene clusters from different sources was built using MEGA 7.0. Default parameters were used for all software unless otherwise noted.

Phylogenetic and comparative genomic analysis

The 16S rRNA gene was amplified using universal primers (27F and 1492R) and then sequenced. Type strains were selected through BLAST and compared by ClustalW of MEGA 7.0 software. The neighbor-joining phylogenetic tree was constructed with partial 16S rRNA gene sequences of closely related type strains showing the phylogenetic position of strain XU54 in the genus Brevibacterium using MEGA 7.0. Kocuria rhizophila TA68^T was used as an outgroup to root the tree. Bootstrap values (expressed as percentages of 1000 replications) above 50% are given at the nodes. Average nucleotide identity (ANI) and in silico DNA–DNA hybridization (isDDH) were used to assess the genetic relationships between species at the genomic level using ANI Calculator of EZ BioCloud platform (http://www.ezbiocloud.net/tools/ani) and genome-to-genome distance calculator 3.0 (GGDC) (https://ggdc.dsmz.de/ggdc.php) (Yoon et al. 2017; Meier-Kolthoff et al. 2013). Type (Strain) Genome Server (TYGS) was used for accurate genome-based taxonomy (https://tygs.dsmz.de) (Meier-Kolthoff et al. 2022; Meier-Kolthoff and Göker 2019). Default parameters were used for all software unless otherwise noted.

Extraction and analysis of carotenoids

Brevibacterium sp. XU54 cells were collected by centrifugation at 5000g for 10 min. Carotenoids were extracted as described previously (Guyomarc’h et al. 2000). Then 8 mL of methanol was added and incubated at 25 °C for 2 h until the cell debris turned colorless, followed by centrifuging at 8000g for 15 min. The resulting supernatant was filtered through a 0.22 μm pore-size organic membrane and measured at 450 nm using a UV/visible spectrophotometer. Total carotenoid yield was calculated according to previous studies (Guyomarc’h et al. 2000; Tian et al. 2019).

Nucleotide sequence accession number

This whole-genome shotgun project has been deposited in DDBJ/ENA/GenBank under the accession number VLTK00000000. The BioProject ID in GenBank is PRJNA555436. The partial 16S rRNA gene sequence is available at GenBank under accession number OP159406.

Results and discussion

Characteristics and 16S rRNA sequences of strain XU54

Brevibacterium sp. XU54 was isolated from radiation-contaminated soil in the Xinjiang Uyghur Autonomous Region of northwest China. The strain was cultured at 25 °C in LB medium for 48 h and the final culture was orange. The strain appeared white for the first 12 h of incubation of XU54 in LB medium, then gradually changed to orange after 12 h of incubation and gradually deepened with time (Fig. 1). The 16S rRNA gene of XU54 was sequenced (1417 bp) and submitted to NCBI GenBank with the accession number OP159406 (Fig. S1). It was the closest to B. aurantiacum NCDO 739^T (GenBank no. X76566) with a similarity of 98.22%, and also showed 97.95% similarity with that of B. antiquum VKM Ac-2118^T (GenBank no. AY243344). Notably, the similarity with other species was even lower. The BLAST results were all lower than the threshold (98.65%) of 16S rRNA identification for differentiating two species (Kim et al. 2014). The phylogenetic analysis of XU54 was performed by the neighbor-joining method between related type strains of the Brevibacterium species, which showed the phylogenetic position of strain XU54 (marked in bold) in the genus Brevibacterium (Fig. 2). In the phylogenetic tree, XU54 and Brevibacterium antiquum VKM Ac-2118^T clustered in the same branch, which showed the closest genetic relationship. Based on phylogenetic analysis and 16S rRNA sequence similarity, it is obvious that strain XU54 may be a new species of Brevibacterium.

Fig. 1.

Color change of Brevibacterium aurantiacum XU54 on LB medium. XU54 was grown in LB medium at 25 °C and 200 rpm for 28–48 h and the final culture was orange. a XU54 on the plate for 12 h at 25 °C. b XU54 on the plate for 48 h

Fig. 2.

Phylogenetic tree of 16S rRNA sequence comparison from XU54 and other type Brevibacterium species. Neighbor-joining tree based on partial 16S rRNA gene sequences, showing the phylogenetic position of strain XU54 (marked in bold) in the genus Brevibacterium. Kocuria rhizophila TA68^T was used as an outgroup to root the tree. Bootstrap values (expressed as percentages of 1000 replications) above 50 % are given at the nodes. Scale bar, 0.005 substitutions per nucleotide position. The GenBank accession numbers for the sequences are given in parentheses

Genome sequence information and comparative genomic analysis of XU54

The Illumina HiSeq 2000 sequencing platform generated 9,131,436 raw reads with a length of 150 bp. Low-quality sequences in the raw data were quality-trimmed and filtered to obtain high-quality reads (clean data) of 7,862,187 pair reads. A genome of 4,899,099 bp was assembled into 64 scaffolds with a G + C content of 62.20%. A total of 4453 gene products were predicted, including 4260 proteins, 13 rRNAs and 49 tRNAs (Table 1). Among the predicted genes, 4164 (93.51%) genes were annotated by matching with the non-redundant protein database (Nr), while only 3041 (68.29%) genes were annotated by Swiss-Prot database. In addition, 61.46, 83.85 and 45.88% were assigned to GO, COG and KEGG databases, respectively. Eight CRISPR-Cas repeats were predicted. The classified genes are represented in a circular format in Fig. 3. Furthermore, annotation and mapping of predicted genes in different databases revealed that many genes in XU54 were considered unique, without any functions matching in the databases mentioned above.

Table 1.

General features and minimum information of Brevibacterium aurantiacum XU54 genome (MIGS)

Items	Values
BioProject accession	PRJNA555436
BioSample accession RefSeq assembly accession WGS project	SAMN12318922 GCF_007558825.1 VLTK01
Geographical location	Xinjiang, China
Longitude and latitude	41° 18′ 27.90″ N 86° 22′ 22.26″ E
Sampling date	April 2018
Sequencing method	HiSeq Illumina 2000
Assembly	SOAPdenovo v. 2.04
Coverage	562.9 ×
Finishing strategy	High-quality draft genome
Genome features
Genome size (bp)	4,899,099
Total number of scaffolds	64
Scaffold N50 (bp)	344, 754
G + C content (%)	62.20
Total number of genes	4453
Protein coding genes (CDSs)	4388
rRNA genes (5S, 16S, 23S)	1, 6, 6
tRNA genes	49
TandemRepeat	223
Genes with function prediction	4260
Annotated in Nr	4164 (93.51%)
Annotated in SwissProt	3041 (68.29%)
Annotated in GO	2737 (61.46%)
Annotated in COG	3734 (83.85%)
Annotated in KEGG	2043 (45.88%)
CRISPRs	8

Fig. 3.

Circular representation of the chromosome of Brevibacterium aurantiacum XU54. From outside to inside: the identifier of the genome size (Mb); predicted protein-coding sequences [colored according to clusters of orthologous groups (COG) categories] on the plus strand; predicted protein-coding sequences (colored according to COG categories) on the minus strand; rRNA and tRNA; genomic GC content; and GC skew (the relative content of GC calculated by G − C/G + C, green is the positive part of GC skew, orange is the negative part of GC skew)

Taxonomic relationships between XU54 and related Brevibacterium species were further assessed by comparative genomic analysis of ANI and isDDH methods (Table 2). The ANI values between XU54 and closely related B. antiquum and B. aurantiacum were 92.68 and 86.44%, respectively. The ANI values with other Brevibacterium species were 86.39–69.40%. The corresponding isDDH values ranged from 18.5 to 48.9%. The analysis results of ANI and isDDH were all lower than the species-delineating threshold (95% and 70%) (Colston et al. 2014; Richter and Rosselló-Móra 2009), indicating that strain XU54 does not belong to an existing species and would be a unique new species in the Brevibacterium genus. According to the TYGS taxonomic classification, XU54 does not belong to any species found in TYGS database and was detected as a potential new species. As shown in Fig. 4, the whole genome-based phylogenomic tree of XU54 and related type strains of Brevibacterium genus showed the monoclade position of strain XU54 with B. antiquum, suggesting the closest genetic relationship between them, which was similar to the results of 16S rRNA phylogeny. Combined with the results of comparative genomic analysis, XU54 would be a novel candidate species in Brevibacterium genus.

Table 2.

ANI and IsDDH values between XU54 and phylogenetic-related type strains of the genus Brevibacterium

Query genome	Reference genome	ANI (%)	dDDH (d4, in %)	C.I (d4, in %)	G + C content difference (in %)	Genbank accession no.
XU54	Brevibacterium antiquum DSM 21545	92.68	48.9	[46.3–51.5]	0.52	JABASX000000000
XU54	Brevibacterium aurantiacum NBRC 12171	86.44	31.4	[29.0–33.9]	0.61	BJME00000000
XU54	Brevibacterium aurantiacum ATCC 9175	86.39	31.4	[29.0–33.9]	0.54	FXZB00000000
XU54	Brevibacterium sandarakinum DSM 22082	85.56	30.1	[27.7–32.6]	1.11	LT629739.1
XU54	Brevibacterium marinum DSM 18964	83.01	26.8	[24.4–29.3]	2.22	JAATJN000000000
XU54	Brevibacterium oceani BBH7	77.84	22.2	[19.9–24.6]	3.24	JABKDE000000000
XU54	Brevibacterium atlanticum WO024	77.96	22	[19.8–24.5]	3.22	CP050152.1
XU54	Brevibacterium renqingii REN4-1	78.28	21.9	[19.6–24.3]	3.65	JADBHG000000000
XU54	Brevibacterium permense VKM Ac-2280	77.94	21.9	[19.6–24.3]	2.35	JABSSX000000000
XU54	Brevibacterium renqingii REN4	78.10	21.8	[19.6–24.3]	3.65	JADBHF000000000
XU54	Brevibacterium limosum o2	77.87	21.8	[19.6–24.3]	2.65	CP050154.1
XU54	Brevibacterium mcbrellneri ATCC 49030	69.81	21.8	[19.5–24.2]	4.17	ADNU00000000
XU54	Brevibacterium iodinum ATCC 49514	77.85	21.8	[19.6–24.3]	2.3	FXYX00000000
XU54	Brevibacterium sediminis CGMCC 1.15472	77.60	21.7	[19.5–24.2]	2.43	BMJG00000000
XU54	Brevibacterium linens ATCC 9172	78.01	21.7	[19.4–24.1]	2.57	FXYY00000000
XU54	Brevibacterium sediminis FXJ8.269	77.57	21.7	[19.5–24.1]	2.43	JABKDD000000000
XU54	Brevibacterium pigmentatum YB235	77.88	21.7	[19.5–24.2]	2.46	CP050153.1
XU54	Brevibacterium epidermidis NBRC 14811	77.71	21.5	[19.2–23.9]	2.16	BCSJ00000000
XU54	Brevibacterium siliguriense DSM 23676	77.55	21.4	[19.1–23.8]	1.99	LT629766.1
XU54	Brevibacterium casei CIP 102111	77.74	21.2	[19.0–23.7]	5.85	FXZC00000000
XU54	Brevibacterium paucivorans DSM 13657	69.40	21.1	[18.8–23.5]	3.87	JAFBCP000000000
XU54	Brevibacterium jeotgali DSM 29217	71.19	20.2	[18.0–22.6]	6.04	VIUB00000000
XU54	Brevibacterium ihuae cv3	72.82	19.5	[17.3–21.9]	8.55	FXWZ00000000
XU54	Brevibacterium yomogidense DSM 24850	71.12	19.4	[17.2–21.8]	5.93	JABASY000000000
XU54	Brevibacterium rongguiense 5221	71.77	19.1	[16.9–21.4]	10.22	WWEQ00000000
XU54	Brevibacterium album DSM 18261	71.62	19	[16.8–21.4]	8.76	AUFJ00000000
XU54	Brevibacterium luteolum DSM 15022	72.26	18.9	[16.7–21.2]	4.81	JAFBBX000000000
XU54	Brevibacterium daeguense DSM 27938	72.14	18.7	[16.5–21.0]	5.73	JAJTWW00000000
XU54	Brevibacterium ravenspurgense CCUG 56047	70.29	18.5	[16.3–20.8]	0.19	LQQC00000000
XU54	Brevibacterium senegalense JC43	71.12	18.5	[16.3–20.9]	7.79	CAHK00000000

Fig. 4.

The phylogenomic tree based on whole-genome sequences of XU54 and type members of the genus Brevibacterium. 30 closely related type strains were selected from TYGS database. The Genome BLAST Distance Phylogeny (GBDP) tree inferred with FastME 2.1.6.1 from GBDP distances calculated from genome sequences. The branch lengths are scaled in terms of GBDP distance formula d5. The numbers above branches are GBDP pseudo-bootstrap support values > 60 % from 100 replications, with average branch support of 75.2 %. The tree was rooted at the midpoint

Gene function annotation of XU54

Figures S2–S4 showed the distribution of genes into functional categories of GO, COG, and KEGG databases. Many genes related to the metabolic pathways of secondary metabolites and terpenoids were found. There were 92 genes clustered to the function of secondary metabolites biosynthesis, transport, and catabolism in COG database. According to the annotation results of KEGG database, 49 genes were classified as genes related to the terpenoid and polyketides metabolic pathways. These results indicated that the XU54 genome is predicted to encode genes related to the synthesis of secondary metabolites, such as terpenes, polyketides, carotenoids, and so on.

Prediction and analysis of sulfur metabolism-related genes of XU54

The genus of B. linens, B. antiquum, and B. aurantiacum can produce VSCs which result in the aroma of cheese during the ripening process. Casein is hydrolyzed to release amino acids such as methionine, which is then degraded to methanethiol by a methionine γ-lyase to produce VSCs (Delcher et al. 2007; Forquin et al. 2011). However, there are few studies on the metabolism of VSCs in Brevibacterium genus. Annotation analysis of the XU54 genome sequence will help us understand the process of sulfur metabolism and regulate VSCs synthesis. Table 3 showed the comparison of related genes in sulfur metabolism from the annotated genome of XU54 and other reported strains B. aurantiacum, B. antiquum and B. linens. There are many genes encoding enzymes related to sulfur metabolism in XU54 genome, such as cysteine synthase A, methionine synthase, methionine adenosyltransferase, and serine O-acetyltransferase. They showed some differences from other three types of strains with the amino acid similarity of 14.61–99.82%, and the similarity compared with B. aurantiacum was higher than that with B. linens. Interestingly, compared with other enzymes, the similarity of cystathionine beta-lyase in XU54 was relatively low (14.61, 15.30 and 15.40%, respectively). The similarity of other enzyme sequences was less than 98% compared to Brevibacterium genus. These comparisons indicated that these related genes in XU54 genome are different from other strains, which would be helpful to discover new functional genes and regulate the process of sulfur metabolism, to make fermented cheese better.

Table 3.

Enzymes comparisons between XU54 and Brevibacterium antiquum, Brevibacterium aurantiacum and Brevibacterium linens in sulfur metabolism pathway

Enzyme name	Gene name	Length (bp)	Location	Genbank accession no.	Brevibacterium antiquum	Brevibacterium aurantiacum	Brevibacterium linens
Methionine adenosyltransferase	metK	1221	Scaffold3	TSI17872.1	99.01% (WP_198396284.1)	97.29% (WP_035277957.1)	91.75% (WP_039207435.1)
Cystathionine gamma-synthase	metB	1158	Scaffold1	TSI19474.1	89.87% (SMX93533.1)	91.95% (SMY00176.1)	66.85% (SMX99304.1)
Serine O-acetyltransferase	cysE	681	Scaffold5	TSI15974.1	96.46% (WP_198396997.1)	92.04% (WP_096161181.1)	84.96% (SMX66268.1)
Cysteine synthase A	cysK	927	Scaffold5	TSI15973.1	87.99% (WP_101619629.1)	93.83% (WP_096161715.1)	87.66% (WP_039211070.1)
Sulfite reductase	cysJ	1737	Scaffold28	TSI11668.1	93.08% (WP_232488811.1)	99.82% (WP_143924661.1)	86.93% (WP_101596823.1)
Sulfate adenylyltransferase	cysN	1329	Scaffold28	TSI11658.1	97.27% (WP_198398593.1)	97.49% (WP_069599759.1)	79.53% (WP_127363732.1)
Sulfate adenylyltransferase subunit 2	cysD	924	Scaffold28	TSI11659.1	98.34% (WP_198398594.1)	95.53% (AZL08744.1)	86.58% (WP_039209115.1)
Phosphoadenosine phosphosulfate reductase	cysH	774	Scaffold28	TSI11660.1	97.28% (WP_198398595.1)	94.94% (WP_096163003.1)	81.58% (WP_127363731.1)
Methionine synthase	metE	1230	Scaffold10	TSI13831.1	98.29% (WP_198398571.1)	97.56% (WP_069599479.1)	89.00% (WP_139906980.1)
Cystathionine beta-lyase	qcrC	1194	Scaffold6	TSI15383.1	14.61% (SMY04390.1)	15.30% (SMX89954.1)	15.40% (SMX92641.1)
Cysteine desulfurase	sufS	1251	Scaffold3	TSI17821.1	98.08% (WP_198396234.1)	92.79% (WP_193078184.1)	85.54% (WP_039207521.1)

Prediction and identification of carotenoid biosynthetic gene cluster from the genome of XU54

Brevibacterium sp. XU54 is yellow, indicating that it can produce secondary metabolites such as pigments and carotenoids (Giuffrida et al. 2020; Krubasik and Sandmann 2000; Dufossé and De Echanove 2005). The gene clusters of secondary metabolites were predicted by antiSMASH (Table S1), and five gene clusters were identified, which were associated with the pathways of terpene (on scaffold 1), ectoine (on scaffold 7), NRPS (on scaffold 8), siderophore (on scaffold 16) and others (on scaffold 11), respectively. These five gene clusters showed similarities of 85%, 75%, 11%, 5% and 75% with the previously reported gene clusters. In addition, the XU54 genome contains some key genes involved in carotenoid biosynthesis, such as genes encoding isopentenyl-diphosphate Delta-isomerase, geranylgeranyl pyrophosphate synthase, lycopene cyclase, phytoene desaturase, and phytoene synthase (Table 4), which are all located on scaffold 1 of the sequenced genome. These carotenoid biosynthesis genes from XU54 were blasted with the strain B. aurantiacum in GenBank database, which showed 86.70, 82.84, 95.20, 95.09, and 95.50% of sequence similarity compared to B. aurantiacum (CP025331.1). Alignment of amino acid sequences showed 85.56, 88.70, 94.87, 95.61 and 94.22% of sequence similarity to the corresponding proteins of B. aurantiacum in Genbank database, respectively.

Table 4.

Gene sequences and amino acid comparisons between XU54 and Brevibacterium aurantiacum in carotenoid biosynthesis

Gene ID	Location	Genbank accession no.	Enzyme name	Length (aa)	Amino acid similarity (%)	Gene similarity (%)
gene0004	Scaffold 1	TSI19405.1	Isopentenyl-diphosphate Delta-isomerase	180	85.56	86.70
gene0005	Scaffold 1	TSI19406.1	Geranylgeranyl pyrophosphate synthase	354	88.70	82.84
gene0007	Scaffold 1	TSI19407.1	Lycopene cyclase	117	94.87	95.20
gene0012	Scaffold 1	TSI19412.1	Phytoene desaturase	534	95.61	95.09
gene0013	Scaffold 1	TSI19413.1	Phytoene synthase	295	94.22	95.50

In carotenoid biosynthetic pathway, geranylgeranyl pyrophosphate synthase, phytoene synthase and phytoene desaturase are used for lycopene synthesis, which was also investigated and verified in our previous study (Xu et al. 2018; Jin et al. 2015). Phytoene synthase encoded by crtB from extreme microorganisms is very difficult to express in prokaryotes (e.g., Escherichia coli). Furthermore, the pigment extracted from XU54 is orange-yellow, which is speculated to contain β-carotene (Giuffrida et al. 2020). This also explains why the gene encoding lycopene cyclase (crtLm) exists in the genome, so it can further catalyze the cyclization of lycopene to form β-carotene. Therefore, we performed multiple sequence alignment of proteins encoded by crtB and crtLm, and analyzed between XU54 and other Brevibacterium genera (Fig. S5). The protein encoded by crtB gene was the closest to that of B. aurantiacum, with a high similarity of 94.22%. Although the amino acid sequence similarity was high, there are still some differences. As shown in Fig. S5a, the protein from XU54 had a loss of 22 amino acids at positions 125–130 and a difference of 4 amino acids at positions 1–4 compared with other proteins. Protein sequence alignment showed that the similarity between the protein encoded by crtLm of XU54 and B. aurantiacum was 94.87%. An alignment of protein sequences encoded by crtLm that revealed clearly conserved regions in all proteins was presented in Fig. S5b, but deletions and differences occurred at the end of the sequence. These results indicated that the proteins in XU54 were different from those of other strains to some extent, especially B. aurantiacum and type strain B. linens.

In addition, many putative proteins were found in the results of annotation and antiSMASH prediction to the sequenced genome of XU54, and most of these sequences had only sequence-based genomic annotations without experimental verification of gene function. There are some differences in the genes and coding proteins in carotenoid biosynthesis compared with B. aurantiacum, which indicates many unknown proteins or genes in secondary metabolite synthesis remain to be developed and studied. The genome sequence of strain XU54 provides a genomic basis for in-depth comparison of genomes and analysis of the specific mechanisms of secondary metabolite biosynthesis pathways.

Cell growth and carotenoid production of XU54

XU54 can grow in a low-temperature range of 10–37 °C, probably because it was isolated from radioactive frozen soil (Fig. 5). XU54 can grow at relatively low temperatures such as 10 and 15 °C, and can still produce carotenoids at such low temperatures despite the low yield of carotenoids. The maximum cell biomass of OD₆₀₀ reached 4.682 at 25 °C after 24 h with a maximal carotenoid content of 2.935 mg/L, indicating that XU54 grew rapidly in the first 24 h and could accumulate more carotenoids. Compared with other temperatures, 25 °C is the optimum temperature for XU54 growing and producing carotenoids. Carotenoids from B. linens were isolated and characterized to the aryl carotenoid family, including new types of diversified carotenoids except β-carotene, confirming the effective role of strains in cheese coloring (Giuffrida et al. 2020; Guyomarc’h et al. 2000; Galaup et al. 2015). A carotenogenic gene cluster and a novel type of lycopene cyclase from B. linens were identified based on sequence homologies and cloned for the synthesis of aromatic carotenoids (Krubasik and Sandmann 2000). The carotenoid biosynthetic pathways of B. linens were redesigned and reconstructed in E. coli. Extension with enzymes from B. linens and other key enzymes in carotenoid pathways demonstrated the potential ability for generating unexpected structurally diverse carotenoids (Kim et al. 2010). Cytochrome P450-dependent monooxygenase related to carotenoids biosynthesis in B. linens was investigated for hydroxy-isorenieratene synthesis in the cheese ripening (Dufossé and De Echanove 2005). There have been some studies on carotenoids extracted from Brevibacterium genus, but few studies were reported about carotenoids from B. aurantiacum. The composition of carotenoids, and the mechanism in carotenoid biosynthesis and coloring in cheese ripening in Brevibacterium genus is poorly understood, which should be considered and studied in future work.

Fig. 5.

The growth curves and carotenoid contents of XU54 at different temperatures. XU54 cells were collected by centrifugation at 5000g for 10 min. Carotenoids were extracted by adding methanol and incubated at 25 °C for 2 h. The supernatant was measured at 450 nm. Cell growth curves were measured by OD₆₀₀ at regular intervals. a Cell biomass of XU54 in LB medium at 10, 15, 20, 25, 30 and 37 °C. b Extracted carotenoid content of XU54 in LB medium at 10, 15, 20, 25, 30 and 37 °C

Conclusions

In conclusion, the draft genome sequence of Brevibacterium sp. XU54 was reported, annotated, and analyzed. Through genome annotation, the genes involved in sulfur metabolism and carotenoid biosynthetic operon were identified and compared. Then the maximal OD₆₀₀ of 4.682 and carotenoid content of 2.935 mg/L were both obtained at 25 °C after culturing for 24 h. The sequenced and annotated complete genome of XU54 may shed light on its promising potential for cheese ripening. These sequencing results will provide useful genomic information for genetic engineering and synthetic biology aimed at producing carotenoids with different structures as food additives. Meanwhile, the availability of the strain XU54 and its genome sequence may facilitate studies on better understanding the function and mechanism involved in VSCs metabolism pathway, carotenoid synthesis pathway, and strain growth on the cheese surface.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

XX and ST designed the experiments. ZZ, CH and BD performed the research. CX and ZZ collected and analyzed the data. ZZ and LJ supervised the research. ZZ wrote the paper. ST, XX and LJ polished the paper. All authors read and approved the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (21978136, 32060004), the National Natural Science Foundation of China Youth Fund (31922070), the Natural Science Foundation of Jiangsu Province (BK20211268, BK20180038), and the Xinjiang Natural Science Foundation for Distinguished Young Scholars (2022D01E19).

Declarations

Conflict of interest

The authors declare that they have no conflict of interest in the publication.

Ethical statements

This article does not contain any studies with human participants or animals performed by any of the authors.