Riboflavin salvage by Borrelia burgdorferi supports carbon metabolism and is essential for survival in the tick vector

Abstract

The Lyme disease agent, Borrelia burgdorferi, harbors a significantly reduced genome and relies on the scavenging of critical nutrients from its tick and mammalian hosts for survival. Riboflavin salvage has been shown to be important for B. burgdorferi infection of mice, yet the contributions of riboflavin to B. burgdorferi metabolism and survival in the tick remain unknown. Using a targeted mass spectrometry approach, we confirmed the importance of bb0318, the putative ATPase component of an ABC-type riboflavin transporter, for riboflavin salvage and the production of FMN and FAD. This analysis further revealed that Δbb0318 B. burgdorferi displayed increased levels of glycerol 3-phosphate compared to the wild-type. The glycerol 3-phosphate dehydrogenase activity of GlpD was found to be FAD-dependent and the transcription and translation of glpD were significantly decreased in Δbb0318 B. burgdorferi. Finally, gene bb0318 was found to be important for maximal spirochete burden in unfed larvae and essential for survival in feeding ticks. Together, these data demonstrate the importance of riboflavin salvage for B. burgdorferi carbon metabolism and survival in ticks.

Keywords: Borrelia burgdorferi; Ixodes scapularis; Lyme disease; carbon metabolism; glycerol 3-phosphate dehydrogenase; riboflavin.

FIGURE 1

Borrelia burgdorferi lacking bb0318 have reduced intracellular levels of riboflavin, FMN, and FAD, and increased levels of glycerol 3-phosphate. Wild-type (WT) and Δbb0318 B. burgdorferi were grown to mid-logarithmic phase in BSK-II and targeted metabolomics was performed via LCMS. (a) Graph displaying the log₂ fold change of Δbb0318 B. burgdorferi versus WT in intracellular levels of riboflavin, FMN, and FAD. p-values from an unpaired t-test for each metabolite are indicated across the top. (b) Volcano plot of the metabolite data set with the log₂ fold change of the Δbb0318 mutant versus WT on the x-axis and the −log₁₀ p-value from an unpaired t-test on the y-axis. A horizontal line at 1.5 (p-value = .032) corresponds to the 20% false discovery rate (FDR) cut off for the dataset by a Benjamini-Hochberg correction and vertical lines indicate a positive or negative twofold change. n = 4. Metabolites that passed the 20% FDR filter and had a log₂ fold change of ≤−1 or ≥1 between Δbb0318 and WT are indicated by blue circles. Metabolites that passed the 20% FDR filter but not the fold change cut-off between Δbb0318 and WT are indicated by black circles and metabolites that fell below both thresholds are indicated by open circles. FAD, flavin adenine dinucleotide; FMN, flavin mononucleotide; cdGMP, cyclic di-GMP; GlcN6P, glucosamine 6-phosphate; 6PG, 6-phosphogluconate; F1,6BP, fructose 1,6-bisphosphate; G1P, glucose 1-phosphate; Gro3P, glycerol 3-phosphate.

FIGURE 2

Borrelia burgdorferi GlpD is an FAD-dependent, peripheral membrane glycerol 3-phosphate dehydrogenase. (a) The reducing equivalents produced by increasing amounts of recombinant GlpD protein or the Nus protein control in the presence (+FAD) or absence (−FAD) of 10 μM FAD. Error bars represent standard deviation from the mean of triplicate measures. Statistical significance was determined via an unpaired t-test, GraphPad Prism 9.0.0. (***p ≤ .005,***p < .0001, ns: not significant). (b) The glycerol 3-phosphate dehydrogenase activities of recombinant GlpD or the Nus control protein in the presence (+FAD) or absence (−FAD) of 10 μM FAD were calculated as reducing equivalents produced per minute (min) per milligram (mg) of recombinant protein (reducing equivalents/min/mg). Statistical significance was determined via a one-way ANOVA with Dunnett’s multiple comparisons test to the Nus protein control in the absence of FAD, GraphPad Prism 9.0.0. (***p < .0001). (c) Total protein lysate (T) of wild-type B. burgdorferi was fractionated into the membrane (M) and soluble (S) components using ultracentrifugation. Protein fractions from equivalent numbers of spirochetes were analyzed by immunoblot with antibodies against GlpD (αGlpD), VlsE (membrane protein) (αVlsE), or SodA (cytoplasmic protein) (αSodA). (d) The B. burgdorferi membrane fraction was washed with 500 mM NaCl (high NaCl), 3.6 M urea (Urea) or 2% Triton X-100 (Triton X-100). The samples were separated into the membrane (M) and soluble (S) components using ultracentrifugation and analyzed by immunoblot with antibodies against GlpD (αGlpD). Molecular weights are shown in kilodaltons (kDa).

FIGURE 3

Borrelia burgdorferi lacking bb0318 demonstrates decreased expression of the glp operon. (a) RNA was extracted from wild-type (WT), Δbb0318, and Δbb0318/bb0318+ B. burgdorferi grown in BSK-II for RT-qPCR analysis of the glp operon genes and gpsA. Expression of the target genes was normalized to that of enolase, bb0337. Data are presented as relative gene expression to WT and are the average of biological triplicate samples ± standard deviation. Statistical significance was determined via a one-way ANOVA with Dunnett’s multiple comparisons test to WT (*p < .05, ** p < .01, ***p < .001, ****p < .0001). Protein lysates were generated from WT, Δbb0318, and Δbb0318/bb0318+ B. burgdorferi and analyzed by immunoblot with antibodies against GlpD (αGlpD, green) and FlaB (αFlaB, red). (b) A representative image of three biological replicates is shown. (c) Relative GlpD protein levels determined by normalization of the GlpD protein signal intensity to that of the FlaB protein. Data represent the average of biological triplicate measures ± standard deviation. Statistical analyses were performed using the one-way ANOVA with Dunnett’s multiple comparisons test to WT, GraphPad Prism 9.0.0 (**p < .01).

FIGURE 4

Borrelia burgdorferi lacking bb0318 demonstrates wild-type levels of rrp1 expression. (a) RNA was extracted from wild-type (WT), Δbb0318, wild-type clone 5A4 (5A4-WT) and Δrrp1 (5A4-Δrrp1) grown in BSK-II for RT-qPCR analysis of rrp1. Expression of the target gene was normalized to that of enolase, bb0337. Data are presented as relative gene expression to the respective wild-type and are the average of biological triplicate samples ± standard deviation. Statistical significance was determined via an unpaired t-test, GraphPad Prism 9.0.0. (*p < .05, **p < .01, ***p < .0001, ns: not significant). (b) Protein lysates were generated from WT, Δbb0318, and 5A4-Δrrp1 B. burgdorferi and analyzed by immunoblot with antibodies against Rrp1 (αRrp1, green) and FlaB (αFlaB, red). A representative image of three biological replicates is shown. Relative Rrp1 protein levels were determined by normalization of the Rrp1 protein signal intensity to that of the FlaB protein. Data represent the average of biological triplicate measures ± standard deviation. Statistical analyses were performed using the one-way ANOVA with Dunnett’s multiple comparisons test to WT, GraphPad Prism 9.0.0 (**p < .01, ns: not significant).

FIGURE 5

Gene bb0318 is important for B. burgdorferi to achieve maximal spirochete burden in unfed larvae and essential for B. burgdorferi survival in feeding ticks. Ixodes scapularis larvae were artificially infected by immersion with wild-type (WT), Δbb0318 and Δbb0318/bb0318+ B. burgdorferi. (a) B. burgdorferi were quantified by qPCR at various time points post-immersion over a 37-day time course by qPCR. Each data point represents a group of 20 larvae. (b) B. burgdorferi were quantified by qPCR 3 days post-immersion. Each data point represents a group of 10 larvae. The mean is indicated by a horizontal black line. (c) 10 days post-immersion, larvae were fed to repletion on mice. Seven days post-feeding to repletion individual fed larva were homogenized and the number of B. burgdorferi per larva determined by colony forming units counts in solid medium. The mean is indicated by a horizontal black line. Statistical significance was evaluated by one-way ANOVA with Dunnett’s multiple comparisons test to WT (*p < .05, **p < .01, ***p < .0001).