Examination of the Borrelia burgdorferi Transcriptome in Ixodes scapularis during Feeding

Abstract

Borrelia burgdorferi gene expression within the guts of engorging Ixodes scapularis ticks was examined by use of differential immunoscreening and differential expression with a customized amplified library. Fourteen chromosomal genes involved in energy metabolism, substrate transport, and signal transduction and 10 (4 chromosomal and 6 plasmid) genes encoding putative lipoproteins and periplasmic proteins were preferentially expressed in engorging ticks. These data demonstrate a new approach to the global analysis of B. burgdorferi genes that are preferentially expressed within the vector during feeding.

Borrelia burgdorferi, the agent of Lyme disease, is transmitted when Ixodes ticks feed on mammals (5, 17). While the tick is engorging, the antigenic composition of B. burgdorferi changes dramatically (1, 8, 9, 20, 21). Analysis of differentially expressed genes will provide insights into the molecular changes that occur in spirochetes during tick feeding. Recent studies have focused on the expression profiles of only a few genes, such as ospA and ospC (12, 15, 20), erpT (10), bbk32 and bbk50 (11), and more recently rev, mlpA, erpa/i/n, and erpb/j/o (14), that are expressed in the tick vector. In order to systematically identify B. burgdorferi antigens that are differentially expressed during tick feeding, we have used an antibody-based approach, differential immunoscreening (24), and differential expression with a customized amplification library (DECAL), a new technique to examine the transcriptomes of prokaryotic genomes. DECAL was first developed and utilized for the global analysis of gene expression in Mycobacterium tuberculosis grown in vitro (2). The rationale for exploiting DECAL is based on the following factors: (i) the analysis can be performed with as little as 10 ng of total bacterial RNA and can detect differences in levels of gene expression that are as small as fourfold and (ii) the technique is ideally suited for investigations where contaminating host material is present.

Differential immunoscreening of B. burgdorferi.

Ixodes scapularis nymphs infected with B. burgdorferi strain N40 and either fed to repletion on C3H/HeN mice (9) or left unfed were dissected. The isolated guts were suspended in phosphate-buffered saline (10 μl/tick gut) and homogenized. The number of spirochetes in each preparation was measured by direct immunofluorescence (9). Equivalent doses of B. burgdorferi (10⁶ organisms) in complete Freund's adjuvant were used to immunize C3H/HeN mice after heat inactivating the spirochetes (1 h at 60°C). Control mice were immunized with gut extracts from uninfected I. scapularis nymphs. Ten days after the final booster injection, antisera against spirochetes from unfed and fed ticks were collected. A λ ZAP II-B. burgdorferi N40 expression library (10,000 plaques) was screened with antisera from unfed and fed ticks. Hybridizing plaques were visualized with a picoBlue immunoscreening kit (Stratagene, La Jolla, Calif.). About 100 clones hybridized to sera from both unfed and fed ticks. Seven positive clones that were preferentially recognized by the antisera from fed ticks were isolated, and the insert DNAs were sequenced. Homology searches performed against sequences in the B. burgdorferi genome database (www.tigr.org) revealed that four clones contained bba14 and three contained bb0129. To obtain a more global view of the differential expression profile of B. burgdorferi in engorging ticks, we exploited DECAL (2).

Construction of a Bb-CAL.

A B. burgdorferi customized amplified library (Bb-CAL) was constructed according to the protocol described by Alland et al. (2). Two thousand B. burgdorferi-ZAP II genomic clones were cored and suspended in phage dilution buffer in 96-well microtiter plates (18). B. burgdorferi rRNA genes were identified and removed from this library of 2,000 clones as described previously (2). The rest of the phage clones were pooled in groups of 50, the phagemids were excised, and the rescued phagemid DNA was digested with EcoRV and SmaI to release the inserts and electrophoresed on a 2% low-melting-point agarose gel. The gel region corresponding to 200 to 2,000 bp (a size range that can be efficiently amplified by PCR) was excised, and DNA was eluted from the gel, ligated with 2 pmol of Uniamp XhoI adapters, and PCR amplified with Uniamp primers (2). The amplified products represented the Bb-CAL encoding the proteome of B. burgdorferi.

Positive selection and amplification.

The guts dissected from B. burgdorferi-infected unfed and fed (72 h) nymphal ticks (9) were placed in RNAWIZ RNA isolation reagent (Ambion, Austin, Tex.), and RNA was isolated according to the manufacturer's protocol. Contaminating DNA was removed by use of a DNA-free kit (Ambion). The total RNA (∼1 μg) isolated from unfed and fed nymphs was reverse transcribed in the presence of biotin-labeled random hexamers and biotin dATP by use of the Superscript first-strand synthesis system for reverse transcription-PCR (Life Technologies, Inc., Gaithersburg, Md.) at 50°C for 1 h. Aliquots of the resultant cDNA pools were tested by PCR with B. burgdorferi flaB primers (25). This ensured the integrity of the cDNA and normalized the amounts of cDNA from unfed and fed ticks for use in the DECAL positive-selection step. Biotinylated cDNA was hybridized to the Bb-CAL, and the biotinylated B. burgdorferi cDNA-Bb-CAL hybrids were bound to streptavidin-coated magnetic beads (Dynal, Lake Success, N.Y.). The Bb-CAL DNA bound to the streptavidin-coated magnetic beads was eluted by boiling and PCR amplified by use of Uniamp primers and 20 μl of the eluted sample as described previously (2). The positively selected and amplified B. burgdorferi PCR probes prepared from unfed and fed I. scapularis ticks (Unfed Bb-DECAL and Fed Bb-DECAL, respectively) represent the B. burgdorferi transcriptome in the guts of unfed and fed nymphs, respectively.

Differential-expression analysis with a Bb-CAL.

Unfed Bb-DECAL and Fed Bb-DECAL were randomly labeled with [α-³²P]dATP, and the amounts were normalized based on hybridization intensity to the flaB PCR product prior to the probing of duplicate B. burgdorferi genomic arrays. A B. burgdorferi N40 genomic library constructed in the plasmid vector pHYB/LEX-ZEO (Invitrogen, Carlsbad, Calif.) was used to print Borrelia genomic arrays at Incyte Genomics (St. Louis, Mo.). The arrays were probed with Unfed Bb-DECAL and Fed Bb-DECAL. Hybridization was scored visually, and intensity values ranging from 0 to 3, with 0 being the weakest and 3 being the strongest hybridization signal observed by autoradiography, were assigned. Differentially hybridizing clones with spot intensity scores of 2 and 3 on each array were selected for the initial analysis. Sixteen clones were preferentially expressed in unfed ticks, and 32 were preferentially expressed in fed ticks. The B. burgdorferi genes in these clones were identified by DNA sequencing and homology searches as described above (Table 1). The list of genes identified by DECAL appears skewed towards chromosome-borne (bb0) genes. B. burgdorferi replicates rapidly in the guts of ticks during feeding (9). It is therefore reasonable to observe the recruitment of functions involved in cellular processes such as (i) DNA metabolism, (ii) energy metabolism, and (iii) the uptake of essential substrates, the majority of which are indeed chromosome encoded (13). Virulence genes predominantly carried by the extrachromosomal elements of B. burgdorferi (16, 22) are perhaps more selectively expressed than genes involved in cellular processes. When we extended our efforts to the sequencing of differentially hybridizing signals in a low intensity range (intensity 1), we observed a preponderance of differentially expressed plasmid-borne genes (data not shown).

TABLE 1.

Primary DECAL screening results

Tick group	Clone	Gene(s) present in the clonesa	Predicted function(s) and/or gene product(s)
Fed	A-7-11b	bb0237-bb0241c	Central intermediary metabolism; hypothetical protein
	K-4-5	bb0329-bb0335c	Oligopeptide ABC transporters
	A-7-17	bb0787-bb790c	Translation, cellular process, unknown functions
	C-16-9	bbg24-bbg27c	Conserved hypothetical proteins
	P-15-2	bb0420	Response regulator/sensory transduction histidine kinase
	O-16-14b	bb0435-bb0436	DNA gyrase subunits A and B
	L-19-14	bb0835-bb0838	Intermediary metabolism, cellular process, unknown function
	C-10-16	bba11-bba15c	Conserved hypothetical proteins, OspA
	A-4-4	bb0600-bb0603c	Unknown function, cellular process; integrin-binding protein
	J-17-4	bb0154	Preprotein translocase
	A-11-9	bb0374-bb0378	Regulatory and unknown functions
Unfed	O-19-2	bb0193-bb0181c	Energy metabolism, regulatory function, replication, unknown function
	F-15-2	bbl17 and bbo17	Unknown function
	I-18-24	bb0009-bb0013	Energy metabolism
	I-21-24	bb0146-bb0151c	Energy metabolism, central intermediary metabolism; transport proteins
	E-11-18	bb0420-bb0421c	Sensory transduction histidine kinase, hydrolase
	F-3-2	bb0434c	Stage O sporulation protein J (chromosome segregation)
	L-9-22	bbg10	Hypothetical protein
	O-6-15	bbg12-bbg14c	Conserved hypothetical proteins
	E-8-14	bb0635	Conserved hypothetical protein

^aSpirochete genes upregulated in both fed and unfed ticks.

^bIdentified five times. All other clones were identified more than once.

^cSouthern analysis was performed to identify the hybridization of the gene(s) to Bb-DECAL probes.

Southern analysis.

The DECAL hybridization profile of each of the genes carried by the clones with more than one gene (Table 1) was investigated by Southern analysis. PCR products corresponding to the individual genes present in these clones were electrophoresed and transferred to nylon membranes. The membranes were probed with the [α-³²P]dATP-labeled Unfed Bb-DECAL or Fed Bb-DECAL probe and exposed to autoradiography. The Southern hybridizations identified individual genes contributing to DECAL differential hybridization (Fig. 1). Of physiological significance is the hybridization of Fed Bb-DECAL to amplicons corresponding to bb0240 (glycerol uptake facilitator) and bb0241 (glycerol kinase). It is likely that bb0240 and bb0241 are upregulated to enable the uptake of glycerol as an energy source (23). The Southern analysis (Fig. 1A) also suggests that only a subset of the oligopeptide ABC transporter genes contained in the opp operon corresponding to bb0329, bb0330, bb0333, and bb0335 are upregulated on spirochetes during the vector's feeding. The hybridization of amplicons corresponding to bbk32 and bbk50, but not to bba24 and bbf01 (Fig. 1B), to Fed Bb-DECAL and the lack of hybridization of Unfed Bb-DECAL to ospC are consistent with earlier observations on the expression profiles of these genes (6, 11, 21). The hybridization of Fed Bb-DECAL to the PCR product corresponding to bba14 and bb0129 (Fig. 1B) is also consistent with the results obtained by differential immunoscreening. These data emphasize the utility and validity of DECAL.

FIG. 1.

Southern analysis of the individual genes carried by the differentially hybridizing B. burgdorferi clones. The genes in selected differentially hybridizing clones were PCR amplified by use of gene-specific primers. The PCR products were Southern blotted to nitrocellulose membranes. Shown are Southern blots hybridized to the radioactively labeled Fed Bb-DECAL probe (A and B) and to the radioactively labeled Unfed Bb-DECAL probe (C).

Temporal analysis of expression by RT-PCR.

The results obtained by DECAL and differential immunoscreening were confirmed by RT-PCR analysis. RNA isolated from the guts of unfed and fed B. burgdorferi-infected I. scapularis ticks was reverse transcribed. The resultant cDNA was used as a template to amplify PCR products corresponding to several B. burgdorferi genes identified by DECAL and differential immunoscreening. The expression profiles observed (Fig. 2) confirmed the results obtained by DECAL (Table 1) and differential immunoscreening. The presence of spirochetes in the guts of engorging ticks increased the expression of bb0420 and bb0377 (Fig. 2). The bb0420 gene encodes a sensory transduction histidine kinase, and bb0377 encodes a homolog of LuxS (Table 2). LuxS has been shown to be a key enzyme in the biosynthesis of a quorum-sensing signal molecule in several gram-negative bacteria (19). We speculate that BB0377 and BB0420 may be components of signaling circuits operating in B. burgdorferi that presumably mediate intracellular responses to environmental stimuli. Gene bb0603 encodes an integrin receptor that facilitates binding to mammalian host tissue (4, 7). Its upregulation in spirochetes upon tick feeding (Fig. 2) perhaps signifies a preparation for entry into the mammalian host. The observation that bb0151 (nagA, N-acetylglucosamine-6-phosphate deacetylase) is upregulated on spirochetes in unfed ticks (Fig. 2) supports the hypothesis that N-acetylglucosamine (a constituent of chitin in tick midguts) may serve the spirochete as a substrate for energy and cell wall biosynthesis (13, 23). To glean information on the potential biochemical functions of some of the genes identified by DECAL that encode proteins with unknown functions, we compared their protein sequences with those in the National Center for Biotechnology Information (NCBI) nonredundant protein database by use of the BLAST-P program (3). The best matches obtained are listed in Table 2. In future experiments, analysis will be conducted on potential genes to follow up on the pathways and mechanisms implicated by the DECAL results.

FIG. 2.

Confirmation of DECAL and differential immunoscreening results by RT-PCR. cDNA prepared from B. burgdorferi-infected unfed and fed I. scapularis nymphs were used as templates in PCRs with gene-specific primers.

TABLE 2.

Homologs of differentially expressed genes that were identified as hypothetical proteins in the B. burgdorferi genome database

Protein	NCBI best matchb	NCBI E valueb
BB0129a	Pseudouridine synthase	3.00E−58
BB0788a	GMP synthetase	1.00E−54
BB0377a	Autoinducer 2 synthase (LuxS)	1.00E−47
BBA14a	B. burgdorferi Mlp lipoprotein family (ORF D)	6.00E−14
BBA11a	B. burgdorferi ORF X upstream of blyA and blyB	1.00E−08
BBG25a	B. burgdorferi Mlp lipoprotein family (ORF D)	1.10E+00
BBG10	Putative tail protein (Streptococcus pneumoniae bacteriophage MM1)	2.00E−06
BB0635	Nicotinic acid phosphoribosyl transferase	1.00E−50

^aProtein whose corresponding gene was preferentially expressed on spirochetes in fed nymphs. ORF, open reading frame.

^bSee www.ncbi.nlm.nih.gov. E value is a statistical parameter to estimate sequence similarity scores.

We have for the first time demonstrated the utility of DECAL for the examination of the Borrelia transcriptome in ticks during feeding. DECAL is a powerful tool that overcomes the limitations of the differential immunoscreening technique and can potentially be exploited to examine the in vivo dynamics of B. burgdorferi gene expression in any tissue.