The Lon-1 Protease Is Required by Borrelia burgdorferi To Infect the Mammalian Host

Borrelia burgdorferi encodes a functional homolog of canonical Lon protease termed Lon-2. In addition, B. burgdorferi encodes a second Lon homolog called Lon-1. Recent studies suggest that Lon-1 may function differently from the prototypical Lon protease. However, the function of Lon-1 in B. burgdorferi biology remains virtually unknown. Particularly, the contribution of Lon-1 to B. burgdorferi fitness and infection remains hitherto unexplored.

ABSTRACT

Borrelia burgdorferi encodes a functional homolog of canonical Lon protease termed Lon-2. In addition, B. burgdorferi encodes a second Lon homolog called Lon-1. Recent studies suggest that Lon-1 may function differently from the prototypical Lon protease. However, the function of Lon-1 in B. burgdorferi biology remains virtually unknown. Particularly, the contribution of Lon-1 to B. burgdorferi fitness and infection remains hitherto unexplored. Herein, we show that Lon-1 plays a critical role for the infection of B. burgdorferi in a mammalian host. We found that lon-1 was highly expressed during animal infection, implying an important function of this protein in bacterial infection. We further generated a lon-1 deletion mutant and an isogenic complemented strain. Relative to that of the wild-type strain, the infectivity of the mutant was severely attenuated in a murine infection model. Our data also showed that the mutant displayed growth defects in regular BSK-II medium. Furthermore, bacterial resistance to osmotic stress was markedly reduced when lon-1 was inactivated. When exposed to tert-butyl hydroperoxide, survival of the lon-1 mutant was impaired. In addition, production of several virulence factors, such as BosR, RpoS, and OspC, was elevated in the mutant. These phenotypes were restored when the lon-1 mutation was complemented. Finally, we created a lon-1(S714A) mutant and found that this mutant failed to infect mice, suggesting that the proteolytic activity of Lon-1 is essential for bacterial infection. Taken together, these results demonstrate that Lon-1 is required by B. burgdorferi to infect animal hosts and to cope with environmental stresses.

INTRODUCTION

The ATP-dependent Lon protease is a member of the AAA+ (ATPases associated with a variety of cellular activities) superfamily (1–3). Since first recognition of Lon protease in Escherichia coli, homologs of Lon have been reported in all forms of life. Based on sequence and structure information, Lon proteases are divided into two subfamilies (4). LonA proteases are found mainly in bacteria and eukaryotes, whereas LonB proteases are found only in Archaea. Like prototypical E. coli Lon, LonA members consist of three functional domains, including an N-terminal domain for substrate binding and recognition, a central AAA+ module for ATP binding and hydrolysis, and a C-terminal peptidase domain for proteolytic cleavage (5, 6). In contrast, LonB members contain only the AAA+ module and the proteolytic domain but lack the N-terminal domain. Different from cytosolic LonA proteases, LonB proteases are anchored to cytoplasmic membranes via one or two transmembrane segments within the AAA+ module. Lon degrades misfolded proteins and prevents abnormal protein accumulation and, thus, is crucial for protein quality control and cellular homeostasis (1–3). Lon also plays an important regulatory role by degrading key regulatory proteins. In addition, Lon has been reported to be involved in quorum sensing, biofilm formation, motility, stress response, and pathogenesis in a variety of bacteria (see reviews [1–3]).

Unlike most bacteria encoding only one Lon homolog, Borrelia burgdorferi, the etiological agent of Lyme disease, encodes two Lon homologs: Lon-1 (BB0253) and Lon-2 (BB0613) (7–9). These two proteases are conserved among Lyme disease Borrelia species (e.g., B. burgdorferi, B. afzelii, and B. garinii) and relapsing fever Borrelia species (e.g., B. recurrentis, B. duttonii, and B. hermsii). Lon-1 (∼90.7 kDa), Lon-2 (∼90.3 kDa), and the E. coli Lon protease (∼87.4 kDa) have similar molecular masses. When B. burgdorferi was exposed to human blood, expression of lon-1 was upregulated ∼5.2-fold, but lon-2 expression remained unchanged (10). By using a functional complementation assay, Coleman et al. reported that introduction of B. burgdorferi Lon-2 into an E. coli lon mutant corrected phenotypes associated with lon deficiency, such as cell filamentation and cps dysregulation, suggesting that Lon-2 is a functional homolog of the E. coli Lon protease (8). Recently, we created a lon-2 deletion mutant and found that the infectivity of the mutant in mice was significantly attenuated, suggesting that Lon-2 is critical for infection of B. burgdorferi in the mammalian host (11). In contrast, B. burgdorferi Lon-1 was incapable of complementing the E. coli lon mutant. Furthermore, unlike the E. coli Lon, B. burgdorferi Lon-1 does not degrade SsrA-tagged proteins, despite recombinant Lon-1 possessing ATP- and Mg²⁺-dependent proteolytic activities (8). These data suggest that Lon-1 may function differently from prototypical Lon protease. Despite these important findings, the contribution of Lon-1 to the parasitic strategy of B. burgdorferi remains unknown. Particularly, the role of Lon-1 in B. burgdorferi infection and pathogenesis hitherto remains unexplored. Herein, we report the construction of a lon-1-deficient mutant in virulent B. burgdorferi. Characterization of the mutant and its isogenic complemented counterparts demonstrates a critical role of this protease in the infectivity and stress response of B. burgdorferi.

RESULTS

lon-1 is highly expressed during animal infection.

To garner information regarding the biological role of Lon-1 in B. burgdorferi, we examined the effects of various environmental factors on gene expression via reverse transcription-quantitative PCR (qRT-PCR) using the primers listed in Table 1. First, the kinetics of lon-1 expression was profiled throughout all phases of growth. Comparable levels of lon-1 expression were observed at early logarithmic phase and stationary phase, whereas slightly lower levels (∼0.57-fold relative to gene expression at early logarithmic phase) of lon-1 transcripts were detected at mid-logarithmic phase (Fig. 1A). Higher levels (∼1.5-fold) of lon-1 expression were found at 37°C than at 23°C, but the difference was not statistically significant (Fig. 1B). Moreover, lon-1 expression was downregulated ∼3-fold in B. burgdorferi cultivated at pH 6.8 compared with gene expression at pH 7.6 (Fig. 1C). In addition, the expression of lon-1 was assessed in infected animals. As described previously (12, 13), RNA was isolated from infected mouse skin, joints, and heart, and gene copy numbers were determined through qRT-PCR via the absolute quantification method. As shown in Fig. 1D, relative to gene expression in spirochetes cultivated in vitro, lon-1 was highly expressed in infected animals. Approximately 564, ∼1,101, and ∼646 copies of lon-1 transcripts per 100 flaB transcripts were detected in the murine skin, joint, and heart samples, respectively, whereas ∼6 copies of lon-1 transcripts per 100 flaB transcripts were detected in stationary-phase spirochetes cultivated in BSK-II at 37°C/pH 7.6 (Fig. 1D). This information suggests that Lon-1 is important for the infection of B. burgdorferi.

TABLE 1.

Oligonucleotide primers used in this study

Primer	Sequence (5′–3′)a
PCR and cloning
68F	TATGCCTCTTCCGACCATCAAGCA
68R	AGGCAGTTCCATAGGATGGCAAGA
290F	GATTACAAGGACCACGACGG
666F	GTGTAAAGCCTGGGGTGCC
666R	AGGTGCCTCACTGATTAAGCATTG
667F	AGATCTCAGCTTTTTTTTGA
667R	TCATGAGGCGCGCCG
671R	CGGCGCGCCTCATGATCACTTGTCATCGTCATCC
798F	CTGAAGCCACACAGTGATATT
798R	CTACCAAGGCAACGCTATG
853F	ATCAGTGAGGCACCTCCGCAAGAGTGCCTTTAATA
853R	CGGCGCGCCTCATGACGCATTCCACCAACAGAA
854F	AAAAAGCTGAGATCTGCCTTCTGCAGGAATTACA
854R	CCCCAGGCTTTACACAAAGCGATTGGACTAAGGAC
855F	GGAGAGCCTTTAGAGGTAGT
855R	AGTTCTTCTGTGTCCCTTAATC
900.2F	ATCAGTGAGGCACCTCTCTGATTCTGATATGAAAGCAATCG
900R	CCGTCGTGGTCCTTGTAATCAAACAATAATTTAATGACCTCGCGCA
902F	CCCGGGCTCTGATTCTGATATGAAAGCAATCG
902R	TAATGGCGCGCCAGAGTAAAGCTAGCCTCAAACAA
904F	TAATGGCGCGCCAGATAATATGCGCGAGGTCAT
904R	TATGCGCGCTTGAAAGACTTGCCAAACTAAATAC
lon-1 site-directed mutagenesis
968F	CCCCAAAAGATGGGCCTGCTGCAGGAATTACAATA
968R	TATTGTAATTCCTGCAGCAGGCCCATCTTTTGGGG
qPCR primers
flaB-Forward	TGATTAGCCTGCGCAATCATT
flaB-Reverse	AATGACAGATGAGGTTGTAGCAGC
lon-1-Forward	GAGTTGAAGGGAGAGCCTTTAG
lon-1-Reverse	CGCCAAGGAAGCTCAGTAATA
β-actin-Forward	AGAGGGAAATCGTGCGTGAC
β-actin-Reverse	CAATAGTGATGACCTGGCCGT
ospC-Forward	TGGTACTAAAACTAAAGGTGCTGAAGAA
ospC-Reverse	GCATCTCTTTAGCTGCTTTTGACA
rpoS-Forward	CTGGACAAAGAAATAGAGGGATCTG
rpoS-Reverse	CAAGGGTAATTTCAGGGTTAAAAGAA
bosR-Forward	TCCACCCTATTCAACTTGACGATATTA
bosR-Reverse	CCCTGAGTAAATGATTTCAATAGATTTTG
rrp2-Forward	ACAGCCCACGGAACAGTTG
rrp2-Reverse	AAAAAATCATAAGCACCCTCTCTCA
rpoN-Forward	CGAATCCCTACGATATAGAGACGAA
rpoN-Reverse	TGCTCAAGTTCATTGGCCTTAA

^aRestriction enzyme sites are underlined.

FIG 1.

Analyses of lon-1 expression in B. burgdorferi by qRT-PCR. (A) B. burgdorferi was cultivated in BSK-II medium at 37°C, and spirochetes were collected when bacterial growth reached early logarithmic phase (E), mid-logarithmic phase (M), or stationary phase (S). (B) B. burgdorferi was cultured in BSK-II medium at 23°C or 37°C, and spirochetes were collected when bacterial growth reached stationary phase. (C) B. burgdorferi was cultured at 37°C in BSK-II medium at pH 7.6 or pH 6.8, and spirochetes were collected when bacterial growth reached stationary phase. (D) Mouse samples, including skin, heart, and joints, were isolated from mice infected with B. burgdorferi at 3 weeks postinfection. In panels A to C, gene expression was analyzed by qRT-PCR via relative quantification. Data were normalized by using B. burgdorferi flaB gene as an internal control. In panel D, gene expression was measured by qRT-PCR via absolute quantification. The values represent the average copy number of lon-1 transcripts normalized per 100 copies of flaB transcripts. The number of lon-1 transcript was also determined in B. burgdorferi grown in BSK-II medium (37°C/pH 7.6) and harvested at stationary phase. All data were collected from three independent experiments and the bars represent the mean values ± standard deviation. The asterisk indicates statistical significance using Student’s t test (P < 0.05).

Construction of lon-1 deletion mutant and complemented strain.

To study the role of Lon-1 in B. burgdorferi infection, we generated a lon-1 deletion mutant. The mutant was created via homologous recombination (Fig. 2A). Through allelic exchange, an 880-bp internal fragment of the 2,421-bp open reading frame of lon-1 was replaced with the PflgB-Kan cassette, yielding the kanamycin-resistant strain OY457. To complement the lon-1 mutant, the suicide vector pOY699 was created by linking lon-1 to the PflgB-aadA cassette (Fig. 2A). After electroporation of pOY699 into OY457, the corresponding streptomycin-resistant complemented strain OY475 was obtained. Plasmid profiling analysis revealed that both OY457 and OY475 retained all plasmids contained in the wild-type (WT) strain CE162. The inactivation and complementation of lon-1 in these strains were confirmed by the use of PCR. Using the lon-1-specific primers listed in Table 1, a fragment was amplified in both WT and the complemented strain, but not in the mutant (Fig. 2B). The kanamycin resistance gene was amplified only in the mutant, whereas the streptomycin resistance gene was amplified only in the complemented strain (Fig. 2B). When gene expression in these strains was analyzed by using reverse transcription-PCR (RT-PCR), lon-1 transcripts were detected in both WT and the complemented strain as expected, but not in the mutant (Fig. 2C).

FIG 2.

Construction of the lon-1 deletion mutant and its isogenic complemented strain. (A) Schematic representation of the lon-1 locus in WT CE162, the mutant OY457, and the relevant complemented strain OY475. Genes are shown as thick arrows circumscribing the respective gene numbers. (B and C) PCR (B) and RT-PCR (C) analyses of WT CE162, the lon-1 mutant, and the complemented strain. The target gene names are indicated on the right. Lane 1, WT CE162; lanes 2 and 3, two representative clones (M1 and M2) of the mutant; lanes 4 and 5, two representative clones (Com1 and Com2) of the complemented strain.

Inactivation of lon-1 impaired the growth of B. burgdorferi in BSK-II medium.

To assess the effect of lon-1 mutation on bacterial growth, WT CE162, two independent clones (M1 and M2) of the mutant OY457, and two independent clones (Com1 and Com2) representing the complemented strain OY475 were cultivated in BSK-II medium and growth curves were determined. Relative to WT CE162 and the complemented strain, growth of the mutant M1 and M2 was impaired (Fig. 3). Both WT CE162 and the complemented strain reached the stationary phase at day 7 postinoculation, with a maximum cell density of ∼3 × 10⁸ spirochetes per ml. However, both the M1 and M2 clones of the mutant displayed a longer lag phase and attained a cell density of ∼1.5 × 10⁸ spirochetes per ml at day 7 postinoculation. At day 9 postinoculation, both M1 and M2 reached a cell density of ∼3 × 10⁸ spirochetes per ml. No obvious differences were observed among WT CE162, the mutant, and the complemented strain when spirochete morphology was examined using dark-field microscopy. These data suggest that Lon-1 is important for the growth of B. burgdorferi.

FIG 3.

Lon-2 is important for the growth of B. burgdorferi in vitro. WT CE162, two lon-1 mutant clones (M1 and M2), and two clones for the complemented strains (Com1 and Com2) were grown in BSK-II medium at 35°C. Spirochetes were enumerated daily using dark-field microscopy. Data are presented as the mean values ± standard deviation from three biological replicates.

Lon-1 is critical for B. burgdorferi infection in mice.

To investigate the role of Lon-1 in mammalian infection, C3H mice were injected intradermally with WT CE162, the lon-1 mutant, or the complemented strain. After 3 weeks, skin, heart, and joint tissues were harvested, and infection was quantitatively analyzed by comparing the spirochete load. Specifically, DNA was isolated from tissue samples and the copy number of B. burgdorferi flaB was determined through quantitative PCR (qPCR) and normalized to mouse β-actin copies. Compared with bacterial number in samples collected from mice infected with WT CE162 or the complemented strain, significantly lower spirochete loads were detected in skin, heart, and joints from mice inoculated with the mutant strain (Fig. 4). Infection was also determined using the cultivation method. Specifically, animal tissues were cultured in BSK-II medium for the outgrowth of spirochetes. Motile spirochetes were recovered from tissues of all mice inoculated with either WT CE162 or the complemented strain OY475 (Table 2). In contrast, no spirochetes were recovered from mice infected with the lon-1 mutant. These results demonstrate that Lon-1 is crucial for the infectivity of B. burgdorferi.

FIG 4.

Lon-1 is critical for infection of B. burgdorferi in mice. Mice were injected with WT CE162, the lon-1 mutant (Δlon-1), or the complemented strain (lon-1-com). Mice were sacrificed at 3 weeks postinoculation and spirochete burdens were quantified using qPCR via absolute quantification. The values represent the average copy number of B. burgdorferi flaB normalized per 10⁵ mouse β-actin gene copies. Data were obtained from mouse tissue samples indicated in Table 2, and the bars represent the mean values ± standard deviation. The asterisk indicates statistical significance using Student’s t test (P < 0.05).

TABLE 2.

Infectivity of B. burgdorferi in mice^a

Strainb	Dose	No. of cultures positive/total no. of specimens examined				No. of mice infected/total no. of mice
		Skin	Heart	Joint	All sites
CE162	104	9/9	8/9	9/9	26/27	9/9
OY457	104	0/15	0/15	0/15	0/45	0/15
OY475	104	14/14	14/14	11/14	39/42c	15/15
OY477	104	6/6	6/6	6/6	18/18	6/6
OY487	104	0/6	0/6	0/6	0/18	0/6

^aData were collected from four independent experiments.

^bCE162, wild-type; OY457, Δlon-1; OY475, lon-1-com; OY477, lon-1-3×FLAG; OY487, lon-1-3×FLAG(S714A).

^cIn total 15 mice were infected with OY475 and 45 samples were collected. Culture data from several samples were not determined due to contamination.

The lon-1 mutant is more sensitive to osmotic stress.

To discern the mechanisms underlying the importance of Lon-1 during mammalian infection, we examined the osmotic stress tolerance of B. burgdorferi by cultivating spirochetes in modified BSK-II medium supplemented with various concentrations of NaCl. As shown in Fig. 5A and B, relative to WT CE162 and the complemented strains, both mutant clones M1 and M2 grew much more slowly in medium containing 150 or 180 mM NaCl. In medium containing 200 mM NaCl, the replication and growth of M1 and M2 were completely inhibited, whereas WT CE162 and the complemented strains were able to grow (Fig. 5C). Additionally, the resistance of B. burgdorferi to osmotic stress was assessed using an osmotic shock assay (14, 15). Upon exposure to 1 M NaCl for 30 min, ∼19% of WT CE162 and ∼16% of Com1 and Com2 spirochetes remained viable (Fig. 5D). However, none of the lon-1 mutant spirochetes survived. All strains failed to survive when exposed to 1 M NaCl for 60 min or 90 min. These combined data suggest that Lon-1 contributes to the survival of B. burgdorferi under high osmolarity conditions.

FIG 5.

Lon-1 is required for the tolerance of B. burgdorferi to osmotic stress. B. burgdorferi was inoculated at 10,000 spirochetes per ml into modified BSK-II medium with 150 mM (A), 180 mM (B), or 200 mM (C) NaCl. Cultures were grown at 35°C and spirochetes were enumerated daily by dark-field microscopy. (D) Osmotic shock assay. B. burgdorferi culture at stationary phase was exposed to 1 M NaCl for 30, 60, or 90 min. Spirochete viability was determined, and percent survival was calculated by dividing the number of spirochetes recovered from the treated group by the number of spirochetes recovered from the control group. Data are presented as the mean values ± standard deviation from three biological replicates. The asterisk indicates statistical significance using Student’s t test (P < 0.05). WT, CE162; M1 and M2, two representative clones of the mutant; Com1 and Com2, two representative clones of the complemented strain.

The lon-1 mutant is more sensitive to tert-butyl hydroperoxide.

We also assessed the susceptibility of B. burgdorferi to an oxidizing agent. To this end, WT CE162, the lon-1 mutant, and the complemented strain were exposed to various concentrations of tert-butyl hydroperoxide. Approximately 57% or 38%, respectively, of WT cells survived exposure to 5 mM or 10 mM tert-butyl hydroperoxide (Fig. 6). In contrast, the lon-1 mutant was significantly more sensitive to tert-butyl hydroperoxide. When clones M1 and M2 were exposed to 5 mM tert-butyl hydroperoxide, approximately 12% or 2%, respectively, of the cells survived. When exposed to 10 mM tert-butyl hydroperoxide, approximately 6% of M1 and none of M2 survived (Fig. 6). The survival rates of the complemented strains (Com1 and Com2) were comparable to those of WT CE162. These data suggest that Lon-1 is involved in the oxidative stress response in B. burgdorferi.

FIG 6.

The lon-1 mutant is sensitive to treatment with tert-butyl hydroperoxide. WT CE162, two representative clones of the mutant M1 and M2, and two representative clones of the complemented strain Com1 and Com2 were cultivated in BSK-II medium. When growth reached stationary phase, ∼5 × 10⁷ cells were treated with various amounts of tert-butyl hydroperoxide. Spirochete viability was determined, and percent survival was calculated. Data are presented as the mean values ± standard deviation from three biological replicates. The asterisk denotes statistical significance (P < 0.05).

Expression of bosR, rpoS, and ospC was changed in the lon-1 mutant.

The alternative sigma factor RpoS is a master regulator for virulence gene expression in B. burgdorferi (see reviews [16, 17]). More specifically, RpoS activates expression of many outer surface lipoproteins associated with mammalian infection, such as OspC. To unravel the effects of lon-1 mutation on virulence gene expression, we compared the levels of RpoS and RpoS-dependent OspC between WT and the mutant by immunoblot analyses. As shown in Fig. 7A, relative to protein levels in WT CE162, the levels of RpoS and OspC were increased in the lon-1 mutant. When mutation was complemented, protein expression was restored to WT or near-WT levels, suggesting that changes in protein expression were solely due to the mutation of lon-1. In agreement with immunoblot results, transcription of rpoS and ospC was significantly increased in the mutant when gene expression was analyzed using qRT-PCR (Fig. 7B).

FIG 7.

Expression levels of bosR, rpoS, and ospC were changed in the lon-1 mutant. B. burgdorferi was cultivated at 37°C in BSK-II medium, and spirochetes were collected at the stationary phase. Protein and gene expression was assessed via immunoblot (A) or qRT-PCR (B), respectively. In immunoblot (A), specific antibodies, denoted as α, are indicated on the left. Lane 1, WT CE162; lane 2, the mutant OY457; lane 3, the complemented strain OY475. In qRT-PCR analyses (B), data were obtained from three biological replicates and normalized using flaB as an internal control. Bars represent the mean measurements ± standard deviations. The asterisk indicates statistical significance using Student’s t test (P < 0.05). WT, CE162; Mut, the mutant OY457; Com, the complemented strain OY475.

Transcription of rpoS in B. burgdorferi is directly controlled by the other alternative sigma factor RpoN (i.e., σ⁵⁴) via a classical −24/−12 promoter, and requires an AAA+ activator ATPase called Rrp2 and a PerR/Fur homolog called BosR (see reviews [16, 17]). To understand why rpoS transcription is affected in the lon-1 mutant, we analyzed the expression of these regulators via immunoblot and qRT-PCR. As shown in Fig. 7A, the level of BosR was markedly increased in the mutant, whereas the protein level of Rrp2 remained unchanged. Genetic complementation restored protein expression. These data were further confirmed by qRT-PCR (Fig. 7B). We also examined the transcription of rpoN via qRT-PCR and found that rpoN transcription was not significantly changed in the lon-1 mutant (Fig. 7B).

Generation and characterization of the lon-1(S714A) mutant.

Lon-1 contains a conserved S714 residue in its C-terminal peptidase domain. This residue has been reported to be essential for the protein’s proteolytic activity (8). More importantly, mutation of S714 did not affect the chaperone-like activity of Lon-1. To examine the contribution of the proteolytic activity of Lon-1 to animal infection, we created a lon-1(S714A) mutant. To this end, we first created a suicide vector pOY696 containing three essential components: a DNA fragment harboring lon-1-3×FLAG, the streptomycin-resistant cassette PflgB-aadA, and a DNA fragment downstream of lon-1. Next, we introduced the S714A point mutation into pOY696 via site-directed mutagenesis, resulting in the construct pOY723. When the plasmids pOY696 and pOY723 were transformed into the lon-1 mutant OY457, the streptomycin-resistant strain OY477 (lon-1-3×FLAG) or the lon-1(S714A) mutant OY487 [lon-1-3×FLAG (S714A)], respectively, was obtained. A fragment was amplified in these strains by PCR using lon-1 specific primers and DNA sequencing analysis confirmed the presence or absence of the S714A point mutation in pertinent amplicons. In addition, plasmid profiling confirmed that these two strains retained all plasmids contained in CE162.

The growth curves of OY477 and OY487 were determined when B. burgdorferi was cultured in BSK-II medium containing various amounts of NaCl. As shown in Fig. 8A, in regular BSK-II media, the lon-1(S714A) mutant OY487 exhibited a delayed growth phenotype compared with OY477. Moreover, when grown in medium containing 150, 180, or 200 mM NaCl, growth of the lon-1(S714A) mutant was much slower than that of OY477 (Fig. 8B to D). These data suggest that the S714 residue and consequently the proteolytic activity of Lon-1 are involved in the tolerance of B. burgdorferi to osmotic stress.

FIG 8.

Growth of the lon-1(S714A) mutant. B. burgdorferi was inoculated at 10,000 spirochetes per ml into regular (A) or modified BSK-II medium with 150 mM (B), 180 mM (C), or 200 mM (D) NaCl. Spirochetes were enumerated daily by dark-field microscopy. Data are presented as the mean values ± standard deviation from three biological replicates. OY477, lon-1-3×FLAG; OY487, lon-1-3×FLAG (S714A).

The infectivity of the lon-1(S714A) mutant was analyzed by using the murine model. In brief, C3H mice were injected with OY477 or the lon-1(S714A) mutant OY487 at a dose of 10,000 spirochetes per animal. Mice were sacrificed at 3 weeks postinfection and skin, heart, and joint samples were taken. Infection in tissue samples was determined by using the cultivation method. As shown in Table 2, motile spirochetes were recovered from all mice injected with OY477. In contrast, no spirochetes were recovered from samples taken from mice injected with the lon-1(S714A) mutant (i.e., OY487), suggesting that the proteolytic activity of Lon-1 is essential for the function of this protein during animal infection.

DISCUSSION

B. burgdorferi is maintained in nature through a complex enzootic life cycle involving an arthropod vector (e.g., the Ixodes tick) and mammalian hosts (e.g., small rodents) (16, 18–21). Interaction with these two diverse environments is critical for B. burgdorferi not only to survive in nature, but also to establish infection and to cause disease. To this end, B. burgdorferi continuously changes its metabolome, transcriptome, proteome, and surface architecture during its transit between ticks and mammals. Global proteome change is usually the outcome of two independent processes: protein synthesis and degradation. Our lab has a long-standing interest in understanding how these two processes are regulated to achieve a balance in the Lyme disease pathogen. Thus, we investigated the machinery dedicated to tunable protein degradation, particularly the contributions of proteolytic enzymes, to the parasitic strategy of B. burgdorferi. In this study, the role of the Lon-1 protease in B. burgdorferi infection was investigated. Our results demonstrate that Lon-1 is required by B. burgdorferi to infect a mammalian host. First, relative to gene expression in spirochetes cultivated in vitro, lon-1 expression was highly induced in infected animals. Second, compared with that of the WT strain, the infectivity of the lon-1 mutant was significantly attenuated. When mutation was complemented, the infectivity phenotype was restored.

During infection, B. burgdorferi encounters a variety of hostile conditions under which the bacterium needs to cope with environmental stresses such as osmotic pressure and oxidative stress (see reviews [16, 17, 22, 23]). To understand how Lon-1 contributes to B. burgdorferi infection, we analyzed the ability of the lon-1 mutant to survive reactive oxygen species (ROS) and hyperosmotic stress. Our results revealed that inactivation of lon-1 impaired the resistance of B. burgdorferi to oxidizing agents. Compared with WT CE162 and the complemented strain, the lon-1 mutant was much less tolerant to osmotic stress. These combined results support that Lon-1 is critical for the fitness and survival of B. burgdorferi under stressful conditions, which provides mechanistic insight into the involvement of Lon-1 in the virulence of B. burgdorferi. Moreover, the lon-1 mutant displayed a growth defect in BSK-II medium. One possible explanation for this growth defect is the hypersensitivity of the mutant to osmotic pressure, as regular BSK-II contains ∼120 mM NaCl (24). The growth phenotype of the mutant may also be attributed to its defect in utilizing nutrients to support spirochete replication. Therefore, Lon-1 may also modulate the levels of proteins associated with the growth kinetics of B. burgdorferi. The Lon protease in Yersinia pestis has been reported to indirectly activate the expression of the important type III secretion system (TTSS) (25). Furthermore, we recently found that the protein levels of RpoS and OspC in B. burgdorferi were reduced in a lon-2 deficient mutant (11). As such, Lon-1 may directly or indirectly impact the expression of key virulence determinants. To test this hypothesis, we measured the expression of several key virulence determinants in the lon-1 mutant. Our data showed that expression of rpoS and RpoS-dependent ospC was upregulated in the mutant at both protein and mRNA levels. Upon analyses of several known factors regulating rpoS transcription, we found that the mutant showed increased levels of BosR, whereas expression of rpoN and rrp2 remained unchanged. Collectively, our data indicate that inactivation of lon-1 elevates the expression of rpoS, probably by affecting the expression of bosR. Our results also suggest that Lon-1 may impact the syntheses of other unidentified factors associated with stress response and mammalian infection.

Previously, Coleman et al. reported that recombinant Lon-1 protein (rLon-1) degraded casein in an ATP-dependent manner (8). This proteolytic activity of Lon-1 was abolished when the catalytic serine residue S714 was mutated. Moreover, rLon-1(S714A) inhibited the aggregation of insulin B-chain. Based on these observations, Lon-1 was proposed to possess dual functions in vitro: (i) a proteolytic activity involved in protein degradation and (ii) a chaperone-like activity involved in protein stabilization (8). To determine whether the chaperone-like activity alone was responsible for the phenotypes of the lon-1 deletion mutant, we complemented the lon-1 mutant using a suicide construct containing an S714A point mutation and obtained the lon-1(S714A) mutant. When the tolerance of this mutant to osmotic stress was measured, our results showed that growth of the lon-1(S714A) mutant in medium containing 150, 180, or 200 mM NaCl was much slower than that of the control strain. We also examined the infectivity of the lon-1(S714A) mutant using the murine infection model and found that this mutant, like the lon-1 deletion mutant, was incapable of infecting mice. These data unequivocally support that S714 and thus the proteolytic activity of Lon-1 are essential for the function of this protein in animal infection and stress response of B. burgdorferi.

To the best of our knowledge, this work represents the first study to demonstrate the importance of the Lon-1 protease in B. burgdorferi infection. Despite the presence of Lon homologs in numerous bacterial species, the link between Lon and bacterial infection has been established in only a few pathogens (26–30). The current study thus expands our understanding about the role of this protease in bacterial pathogenesis. Recently, a series of studies from different research groups have revealed that the serine protease HtrA is crucial for the infection of B. burgdorferi in animal hosts (22, 31–37). Moreover, the ATP-dependent protease FtsH has been reported to be essential for B. burgdorferi infection (38). Additionally, we recently found that the Lon-2 protease is critical for B. burgdorferi to infect the mammalian host (11). This current study on Lon-1 thus provides further evidence for the implication of proteases as a previously unappreciated layer of control over virulence and pathogenesis in B. burgdorferi. Our study also opens new avenues for understanding how Lon-1 controls key cellular processes in this important human pathogen. Continued efforts are warranted to elucidate the role of this protein in the tick-phase infection as well as the tick-mammal transmission of the Lyme disease spirochete. Further research is also needed to identify the substrates of Lon-1, which will uncover the precise molecular details underlying the essential attributes of this protease.

MATERIALS AND METHODS

Ethics statement.

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of South Florida in compliance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health.

Bacterial strains and culture conditions.

Infectious B. burgdorferi strain 297 isolate CE162 (14) was used as the wild-type (WT) strain in this study. B. burgdorferi was routinely cultivated at 35°C and 5% CO₂ in BSK-II medium (pH 7.6) (39) supplemented with 6% rabbit serum (Pel-Freeze, Rogers, AR). To examine the effects of temperature and pH on gene expression, B. burgdorferi was cultured at 23°C or 37°C and the pH value of BSK-II medium was adjusted to pH 6.8. When appropriate, BSK-II medium was supplemented with 160 μg/ml kanamycin or 100 μg/ml streptomycin. E. coli strain TOP10 (Thermo Fisher Scientific, Grand Island, NY) was used as the host for molecular cloning and plasmid propagation. E. coli was maintained in lysogeny broth (LB) medium supplemented with the appropriate antibiotics as follows: ampicillin, 100 μg/ml; kanamycin, 50 μg/ml; or spectinomycin, 100 μg/ml. All plasmid constructs used in this study were confirmed by using PCR amplification, restriction digestion, and sequence analysis.

Construction of the lon-1 mutant and complemented strains.

The lon-1 deletion mutant OY457 was generated through homologous recombination. In brief, an 1,174-bp 5′ coding region of lon-1 and an 1,102-bp downstream region of lon-1 were amplified from CE162 by use of the primers listed in Table 1. Additionally, the pMB1 origin of replication (ORI) was amplified from pUC19 (Thermo Fisher Scientific), while the kanamycin resistance cassette PflgB-Kan was amplified from our bmtA mutant (40). These four fragments were assembled using the GeneArt Seamless Cloning and Assembly kit (Thermo Fisher Scientific). In the resultant construct pOY664, the PflgB-Kan cassette was oriented in the opposite direction of lon-1. Purified pOY664 was then transformed into CE162 as previously described (41), generating the kanamycin-resistant lon-1 mutant OY457.

For genetic complementation of the lon-1 mutant, a suicide plasmid (pOY699) was constructed. Briefly, a 2,303-bp 3′ coding region of lon-1 and an 1,198-bp downstream region of lon-1 were amplified from CE162. These two fragments were digested with AscI and fused together through ligation. The fused DNA was then cloned into pSC-B-amp/kan (Agilent Technologies, Santa Clara, CA), generating pOY698. Next, the streptomycin-resistant cassette PflgB-aadA was cloned into pOY698 at the AscI site. In the resulting suicide vector pOY699, the PflgB-aadA was oriented in the opposite direction as lon-1 transcription. Purified pOY699 was transformed into the lon-1 mutant OY457 through electroporation, which generated the complemented strain OY475. Plasmid profiling for B. burgdorferi was performed via multiplex PCR as previously described (42).

Generation of the B. burgdorferi lon-1(S714A) mutant.

We first created a suicide vector pOY696 by using the GeneArt Seamless Cloning and Assembly kit. To this end, a 2,263-bp 3′ coding region of lon-1 was amplified from CE162, while the 3×FLAG was amplified from our strain OY266 (43) that expresses Rrp2-3×FLAG. These two fragments, together with the pMB1 ORI fragment, the streptomycin-resistant cassette PflgB-aadA, and the 1,198-bp fragment downstream of lon-1, were assembled in a reaction, resulting in the construct pOY696. Purified pOY696 was then transformed into the lon-1 mutant OY457, generating the streptomycin-resistant strain OY477 harboring 3×FLAG-tagged lon-1 at the native chromosomal locus.

To create a lon-1(S714A) mutant, a S714A point mutation was introduced into pOY696 through site-directed mutagenesis by using the QuikChange II XL site-directed mutagenesis kit (Agilent Technologies, Santa Clara, CA). The primers used for site-directed mutagenesis are listed in Table 1. The introduced mutation in the resultant construct, pOY723, was confirmed by DNA sequencing. pOY723 was then transformed into the lon-1 mutant OY457, resulting in the streptomycin-resistant strain OY487.

Osmotic stress tolerance assay.

B. burgdorferi was cultivated in BSK-II until late logarithmic phase (∼1 × 10⁸ cells per ml). Spirochetes were then inoculated at 1 × 10⁴ cells per ml into BSK-II medium containing various concentrations of NaCl. Bacterial growth was monitored daily by using dark-field microscopy.

Osmotic shock assay.

This assay was performed as previously described (14, 15). When bacterial growth reached stationary phase, NaCl was added to the culture to increase osmolarity by 1 M, whereas BSK-II was added to another culture as a corresponding control. Samples were collected at 30-, 60-, and 90-min post treatment, and spirochete viability was assayed by growth endpoint determinations as previously described (14, 15, 40, 44). Percent survival was calculated by dividing the number of spirochetes recovered from the treated group by the number of spirochetes recovered from the control group.

Sensitivity of B. burgdorferi to tert-butyl hydroperoxide.

This test was performed as previously described, with modifications (14, 40, 44). In brief, when B. burgdorferi growth reached stationary phase, spirochetes were collected by centrifugation at 14,000 × g for 5 min, washed thrice with sterile 0.9% saline water, and resuspended in modified BSK-II medium that did not contain sodium pyruvate. The culture was then treated with various concentrations of tert-butyl hydroperoxide at 35°C for 4 h. After incubation, cells were harvested by centrifugation, washed thrice with saline water and resuspended in BSK-II medium. Spirochete viability was assayed by growth endpoint determinations (14, 15, 40, 44).

Mouse infection studies.

The infectivity of B. burgdorferi was assessed by using the murine needle-challenge model of Lyme borreliosis (45–47). C3H/HeN mice (Charles River Laboratories) were inoculated with 10⁴ of spirochetes per animal via intradermal injection. At 3 weeks postinoculation, mice were sacrificed and mouse specimens, including skin, heart, and joints, were harvested and cultured in BSK-II medium supplemented with 1 × Borrelia antibiotic mixture (BAM). The outgrowth of spirochetes was assessed by using dark-field microscopy.

Quantification of spirochete burdens in mouse tissues.

Quantitative PCR (qPCR) was employed to calculate spirochete load in mouse tissue samples. Briefly, mouse specimens were homogenized using the Bead Mill 4 homogenizer (Thermo Fisher Scientific). DNA was extracted using the Thermo GeneJET Genomic DNA purification kit (Thermo Fisher Scientific) according to the manufacturer’s protocol. qPCR was performed using the PowerUP SYBR green Master Mix (Thermo Fisher Scientific) on the Applied Biosystems QuantStudio 3 real-time PCR system. The copy numbers of murine β-actin gene and B. burgdorferi flaB gene were determined via the absolute quantification method using standard curves (12, 34, 48). To this end, β-actin gene was amplified from mouse DNA and cloned into pSC-B-amp/kan (Agilent Technologies) via blunt cloning. The resultant plasmid pOY707 was then serially 10-fold diluted and used in qPCR as the template to generate the standard curve for murine β-actin. Plasmid pOY16 (pGEM-Teasy-flaB) (12) was used to generate the standard curve for flaB quantification. Gene copy number was then calculated by using the Absolute Quantification Analysis program (Applied Biosystems). Specific primers for qPCR are listed in Table 1.

Total RNA isolation.

To measure gene expression in in vitro-cultivated spirochetes, B. burgdorferi was grown under various conditions, harvested by centrifugation, and stored at −80°C until RNA isolation. To analyze gene expression in infected animals, mice were infected with B. burgdorferi CE162 as described above. At 2 weeks postinfection, ear punch biopsy samples were taken and cultured in BSK-II medium with 1 × BAM to confirm bacterial infection. Mice were sacrificed at 3 weeks postinfection and tissue samples, including skin, heart, and joints, were collected and homogenized using the Bead Mill 4 homogenizer. Total RNA was isolated by using TRIzol (Thermo Fisher Scientific) and further purified by using the RNeasy minikit (Qiagen, Valencia, CA) according to the instructions. Genomic DNA was removed using Turbo DNase (Thermo Fisher Scientific). RNA was finally purified using the GeneJET RNA Cleanup and Concentration microkit (Thermo Fisher Scientific) and quantified using a Nanodrop OneC spectrophotometer (Thermo Fisher Scientific).

Gene expression analysis via qRT-PCR.

cDNA was generated using the SuperScript IV reverse transcriptase (Thermo Fisher Scientific) and then used as the template in PCR and qPCR. As described above, qPCR was performed using the PowerUP SYBR green Master Mix. Variation in gene expression in in vitro-grown spirochetes was calculated using the relative quantification method (ΔΔC_T) as previously described (13, 49–51), in which B. burgdorferi flaB gene was used as an endogenous control. To measure transcript copies of lon-1 and flaB present in mouse specimens, the absolute quantification test was performed. Standard curves for flaB and lon-1 were created by using 10-fold serial dilutions of pOY16 (pGEM-Teasy-flaB) or pOY699, respectively, as the template in qPCR, and transcript copy number was calculated by using the Absolute Quantification Analysis program.

SDS-PAGE and immunoblot analysis.

SDS-PAGE and immunoblot analysis were carried out as previously described (40, 49, 52). Briefly, equal amounts of whole-cell lysates prepared from each strain were loaded on a 12.5% acrylamide gel. After proteins were separated, they were transferred to a nitrocellulose membrane and probed with specific antibodies against FlaB, BosR, Rrp2, RpoS, and OspC. These antibodies have been described in our previous studies. Immunoblots were developed by chemiluminescence using the ECL Plus Western Blotting Detection system (Thermo Fisher Scientific).

Statistical analysis.

An unpaired Student’s t test was employed to compare the mean values, in which statistical significance was determined when P < 0.05.