ZntR-mediated transcription of zntA responds to nanomolar intracellular free zinc

Abstract

In E. coli, ZitB and ZntA are important metal exporters that enhance cell viability under high environmental zinc. To understand their functions in maintaining zinc homeostasis, we applied a novel genetically-encoded fluorescent zinc sensor to monitor the intracellular free zinc changes in wild type, ∆zitB and ∆zntA E. coli cells upon sudden exposure to toxic levels of zinc ("zinc shock"). The intracellular readily exchangeable zinc concentration (or "free" zinc) increases transiently from picomolar to nanomolar levels, accelerating zinc-activated gene transcription. After zinc shock, the zitB mRNA level is constant while the zntA mRNA increases substantially in a zinc-dependent manner. In the ∆zitB E. coli strain the free zinc concentration rises more rapidly after zinc shock compared to wild type cells while a prolonged accumulation of free zinc is observed in the ∆zntA strain. Based on these results, we propose that ZitB functions as a constitutive, first-line defense against toxic zinc influx, while ZntA is up-regulated to efficiently lower the free zinc concentration. Furthermore, the ZntR-mediated transcription of zntA exhibits an apparent K(1/2) for zinc activation in the nanomolar range in vivo, significantly higher than the femtomolar affinity for zinc binding and transcription activation previously measured in vitro. A kinetically-controlled transcription model is sufficient to explain the observed regulation of intracellular free zinc concentration by ZntR and ZntA after zinc shock.

Fig. 1. Carbonic anhydrase-based ratiometric zinc sensors

Selective binding of dapoxyl sulfonamide (DPS) to the active site of zinc-bound CA leads to a dramatic increase in quantum yield of this fluorophore. TagRFP fused to CA serves as a FRET acceptor to CA-bound DPS as well as an internal reference. The fluorescence intensity ratio I_FRET/I_FP represents the zinc-bound fraction of the sensor and is calibrated in E. coli cells equilibrated with a series of Zn-NTA buffers.

Fig. 2. Growth curves of WT, ΔzntA and ΔzitB strains

Overnight E. coli cultures were diluted into MOPS minimal medium with either 100 μM ZnSO₄ or no zinc added. The optical density at 600 nm was measured as a function of time. The doubling times (T_G) calculated from a fit of Eq. (1) to the growth curves are listed in the table on the right. The ΔzntA strain exhibits significant growth defects under high zinc, while the growth of the ΔzitB strain is similar to that of WT.

Fig. 3. Changes in total intracellular zinc after zinc shock

50 μM ZnSO₄ was added to the E. coli cell culture at log phase and samples were prepared in duplicate as described in the Experimental section. The total zinc in the cells from 1 ml of culture was measured by ICP-MS, and adjusted by the total cell volume, which was calculated by multiplying the number of cells by the average volume of an E. coli cell.

Fig. 4. Measurement of intracellular free zinc concentrations using wild type CA_TagRFP sensor

E. coli strains BW25113, ΔzntA and ΔzitB transformed with pTH_CA_TagRFP (K_Zn ~ 30 pM) were grown in MOPS minimal medium and the cells were imaged at log phase after addition of DPS to measure the fluorescence intensity ratio I_FRET/I_FP. The sensor occupancy and the free intracellular zinc concentration were determined by comparison to an in situ calibration curve of wt CA_TagRFP, as described previously [24]. The intracellular free zinc as an average of ~100 cells in each sample and the error range are illustrated as follows: WT (filled circle, solid gray), ΔzntA (open square, dash), ΔzitB (open circle, cross hatch).

Fig. 5. Changes in intracellular free zinc in E. coli WT, ΔzitB and ΔzntA strains after zinc shock

E. coli cells expressing the H94N CA_TagRFP (K_Zn ~ 9 nM) sensor were grown in MOPS minimal medium, and imaging samples were prepared by incubation with DPS as described in the Experimental section. Various concentrations of ZnSO₄ (0 – 100 μM) were added to the imaging medium and the cells were imaged to measure the fluorescence intensity ratio I_FRET/I_FP as a function of time. The sensor occupancy and the free intracellular zinc concentration were determined by comparison to an in situ calibration curve of H94N CA_TagRFP, as described previously [24]. Free zinc concentrations for cells grown in MOPS minimal medium without zinc (as measured in Fig. 4) are used as the baseline level at time zero. (a).(b).(c). The intracellular free zinc changes over time after addition of varying external zinc concentrations in wild type

Fig. 6. The de-coupler FCCP and protein synthesis inhibitors alter the intracellular free zinc fluctuations after zinc shock

E. coli cells expressing the H94N CA_TagRFP sensor were grown in MOPS minimal medium and imaging samples were prepared as described in Experimental 2.3. (a). The de-coupler FCCP inhibits the function of the ZitB exporter. 50 μM ZnSO₄ was added to WT (solid circle, solid line) and ΔzitB (open circle, solid line) cells, or 50 μM ZnSO₄ and 10 μM FCCP were added to WT cells (solid square, dashed line) in imaging medium at time zero. The effects of FCCP on the free zinc fluctuations in E. coli after zinc shock mimic the effects of the zitB knock-out. (b). Incubation of E. coli cells with protein synthesis inhibitors lead to prolonged high zinc levels after zinc shock. WT E. coli cells were incubated either with 170 μg/ml chloramphenicol and 50 μg/ml tetracycline for 10 min (solid square, dashed line), or without inhibitors (solid circle, solid line), then 50 μM ZnSO₄ was added and the fluorescence intensity ratio was measured. The time dependence of the free zinc levels in the ΔzntA (open circle, solid line) strain after zinc shock was measured in the absence of inhibitors.

Fig. 7. zitB mRNA level is up-regulated in ΔzntA but remains relatively constant upon zinc shock

Relative mRNA levels were measured by RT-PCR, and normalized by the level of the housekeeping gene transcript, rrsD, a 16S ribosomal RNA of the rrnD operon. (a). The mRNA level in WT is set as 1 (black bar), and the relative levels of zupT, znuC and zitB transcripts in ΔzntA are shown (gray bar). While transcription levels of zupT and znuC stay relatively constant, the mRNA level of zitB is increased more than 6-fold in the ΔzntA strain compared to that in WT. (b). E. coli cells (WT, solid circle; ΔzntA, open circle) were grown in MOPS minimal medium and 100 μM ZnSO₄ was added at OD₆₀₀ ~ 0.3, samples were taken at various time points thereafter. The zitB transcript level in WT before adding zinc is set as the base level 1. The mRNA level of zitB in ΔzntA is normalized against the baseline (before adding zinc) zitB mRNA level in WT. The level of the zitB transcript remains relatively constant over 1 hr after zinc shock with less than two-fold changes.

Fig. 8. Transcription of zntA is strongly induced upon zinc shock

Relative mRNA levels were measured by RT-PCR, and normalized by comparison to rrsD. (a). mRNA level of zntA in wild type (100 μM Zn, open circle; 50 μM Zn, open square; 10 μM Zn, solid circle) and in ΔzntR (100 μM Zn, solid square) at various extracellular zinc concentrations. The transcription levels increase as the zinc concentration increases. (b). mRNA level of zntA in wild type (solid circle) and ΔzitB (open square) strains at 50 μM extracellular zinc. The level of zntA mRNA in the ΔzitB strain is significantly higher than in the WT strain due to the higher intracellular zinc concentrations upon zinc shock (shown in Fig. 5).

Fig. 9. Correlating time course between intracellular free zinc changes and the transcription level of zntA

The changes in the transcript level of zntA mRNA largely follow the changes of intracellular free zinc with a 5 – 10 min time lag across various zinc concentrations. This correlation suggests that zntA transcription is regulated kinetically by the intracellular free zinc changes, and that ZntR has a nanomolar intracellular zinc affinity.