The Interaction between Enterobacteriaceae and Calcium Oxalate Deposits

Abstract

Background: The role of calcium oxalate crystals and deposits in UTI pathogenesis has not been established. The objectives of this study were to identify bacteria present in pediatric urolithiasis and, using in vitro and in vivo models, to determine the relevance of calcium oxalate deposits during experimental pyelonephritis.

Methods: Pediatric kidney stones and urine were collected and both cultured and sequenced for bacteria. Bacterial adhesion to calcium oxalate was compared. Murine kidney calcium oxalate deposits were induced by intraperitoneal glyoxalate injection and kidneys were transurethrally inoculated with uropathogenic Escherichia coli to induce pyelonephritis.

Results: E. coli of the family Enterobacteriaceae was identified in patients by calcium oxalate stone culture. Additionally Enterobacteriaceae DNA was sequenced from multiple calcium oxalate kidney stones. E. coli selectively aggregated on and around calcium oxalate monohydrate crystals. Mice inoculated with glyoxalate and uropathogenic E. coli had higher bacterial burdens, increased kidney calcium oxalate deposits and an increased kidney innate immune response compared to mice with only calcium oxalate deposits or only pyelonephritis.

Conclusions: In a murine model, the presence of calcium oxalate deposits increases pyelonephritis risk, likely due to preferential aggregation of bacteria on and around calcium oxalate crystals. When both calcium oxalate deposits and uropathogenic bacteria were present, calcium oxalate deposit number increased along with renal gene transcription of inner stone core matrix proteins increased. Therefore renal calcium oxalate deposits may be a modifiable risk factor for infections of the kidney and urinary tract. Furthermore, bacteria may be present in calcium oxalate deposits and potentially contribute to calcium oxalate renal disease.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Fig 1. Induction of CaOx deposits and experimental pyelonephritis in mice.

Fig 2. Kidney stones contain bacteria and/or DNA.

(A) Bacteria key. (B) All stones contained multiple bacterial taxa on sequencing (left column) Bacteria identified included the family Enterobacteriaceae (which includes E. coli), and the genera Pseudomonas, Gardnerella, Lactobacillus, Brucella, Phyllobacterium and Bradyrhizobium. Bacteria were cultured (right column) from 2 stones and represented the most abundant bacteria identified by sequencing. (C) Bacteria were only sequenced and cultured from upper tract urine in 1 patient. (D) When detected in the bladder urine, the taxa were similar to those observed in the stones, although the ratios were often dissimilar. To allow low percentage organisms to be visualized, the data was graphed on a logarithmic Y-axis.

Fig 3. UPEC selectively aggregate around CaOx monohydrate crystals: following incubation with GFP labeled UPEC, bacteria (green) could be seen aggregating around CaOx monohydrate (A, arrows) but not CaOx dihydrate (A, arrowheads) or silicon dioxide (B, arrows) crystals.

At 12 hours, significantly more bacteria per crystal surface area were seen with CaOx monohydrate crystals than with CaOx dihydrate crystals, silicon dioxide crystals or background. There were no other significant differences between groups. Magnification 40X right panels, 100X left panels Scale bars = 20 microns. Incubation time = 6 hours for left panels and 12 hours for right panels.

Fig 4. Bacteria increases the murine CaOx deposit burden.

A. The mean right (A), left (B) and combined mean (C) kidney bacterial burdens were lower with UPEC inoculation alone compared to kidneys with CaOx deposits and UPEC inoculation at 56,157±1.68X105 versus 2.14X106±3.34X106; 24,843±69,527 versus 4.50X106±7.83X106 and 40,500±1.12X105 versus 5.28X106±1.78X106 respectively. To present on a log scale graph, but not during statistical analysis, 0 values were assigned a value of 0.01 (D). Following glyoxalate injection, CaOx deposits are seen around the corticomedullary junction (arrows) (E) Following UPEC inoculation and glyoxalate injection, an increased distribution and number of CaOx deposits (arrows) is noted, extending into the medulla (arrowheads). D-E, representative 4X magnification image stitches, background cropped for clarity, scale bar = 1000μm. (F) CaOx deposit number per mean 4X imaging stitch cross-section area was significantly higher in the CaOx deposits and UPEC inoculation group compared to kidney stones alone. There was a higher percentage of CaOx deposit/total kidney cross section area (H) with CaOx deposits and UPEC than CaOx deposits alone. The location on the scatterplot for the representative images (D) and (E) are indicated by an arrowhead and arrow respectively.

Fig 5. PCR results.

(A) At peak murine CaOx deposit formation (6 days), the RT2 Bacterial Response Array revealed that 12/83 genes were significantly up-regulated > 4 fold. These genes included the inflammatory genes Chemokine (C-C motif) ligand 5 (Ccl5), Cluster of differentiation 14 (Cd14), Interleukin 1β (Il1b), Interleukin 6 (Il6), Lysozyme 2 (Lyz2), Mediterranean fever (Mefv), and Tumor necrosis factor (Tnf); the toll-like signaling genes, Interleukin–1 receptor-associated kinase 3 (Irak3) Toll-like receptor adaptor molecule 1 (Ticam1) and Toll-like receptor 2 (Tlr2) along with the inflammasome genes NLR family, apoptosis inhibitory protein 1 (Naip1) and NOD-like receptor family, pyrin domain containing 3 (Nlrp3). (B) In female C57Bl/6 mice, targeted RT-PCR revealed that at day 3 the inflammatory genes CD14 and Il6 are not increased in UPEC alone inoculated mice or CaOx alone inoculated mice compared to saline control, but are when both CaOx deposits and UPEC inoculation are present. (C) Toll-like signaling genes: Tlr2 followed a similar pattern to the inflammatory genes while Irak3 was not increased at day 3. (D) The inflammasome genes were not up-regulated at 3 days in any of the groups (E) Stone matrix protein inner core components were markedly up-regulated, but only when both CaOx deposits and UPEC inoculation were present.

Fig 6. Speculated mechanisms for bacterial contribution to CaOx stones.

(A) Figure key. (B) Bacteria bind to CaOx crystals that may provide a nidus for pyelonephritis or remain persist in a subclinical state (1) and bacterial communities form a biofilm (2). The biofilm results in crystal aggregation (3). (C) The bacterial enzymes citrate lyase splits citrate resulting in increased CaOx supersaturation (1). CaOx crystals form providing a key element of lithogenesis. (D) Bacteria bind to the urothelium (1) that results in secretion of innate immune proteins from recruited inflammatory cells (2) and the urothelium (3) The innate immune proteins are incorporated as stone matrix proteins.