Characterization and Modification of the Zinc Sensor

ZapCY2 for Improved Nickel Binding

by

Farah Badr

A thesis submitted in conformity with the requirements

for the degree of Master of Science

Graduate Department of Chemistry

University of Toronto

© Copyright by Farah Badr 2017

ii

Characterization and Modification of the Zinc Sensor ZapCY2 for

Improved Nickel Binding

Farah Badr

Master of Science

Graduate Department of Chemistry

University of Toronto

2017

Abstract

Nickel is an important element needed for the function of many microorganisms, including

several strains of pathogenic bacteria such as Escherichia coli and Helicobacter pylori.

The free intracellular nickel concentration has not been quantified for any microorganism,

and such information would be very beneficial for understanding the biochemistry of

pathogenic bacteria. This research aims to develop an intracellular FRET-based nickel

sensor for this purpose. A previously designed zinc sensor, ZapCY2, was selected to

undergo characterization and modification in an effort to tune its selectivity to favor nickel.

Cysteines in the metal-binding sites were substituted with histidine or aspartic acid.

Characterization of the variants revealed a 15-30 fold reduction in the affinity for zinc, and

an unchanged affinity for nickel. Additional modifications will be required to further

improve the sensor’s response towards nickel.

iii

Acknowledgements

I would first like to thank my supervisor Dr. Deborah Zamble for giving me the valuable

opportunity of being part of her lab, and for her continuous support, encouragement, and

patience. I would also like to thank my lab mates who have made it easy for me to look

forward to going to the lab everyday, particularly Thomas Panagiotou, who was there to

help with the project during the bumpy beginning.

I must also thank my parents and my grandparents without whom I would not have made

it this far.

Finally, I would like to thank Dr. Andrew Woolley for providing comments and

suggestions during the writing of this thesis.

iv

Table of Contents Abstract ......................................................................................................................................................... ii

Acknowledgements ...................................................................................................................................... iii

List of Abbreviations .................................................................................................................................... v

List of Tables ............................................................................................................................................... vi

List of Figures ............................................................................................................................................. vii

1. Introduction ......................................................................................................................................... 1

1.1 Nickel in bacteria ................................................................................................................................. 1

1.2 Nickel sensors ...................................................................................................................................... 3

1.3 Genetically encoded FRET-based metal sensors ................................................................................. 5

1.4 ZapCY2................................................................................................................................................ 6

1.5 Purpose of study................................................................................................................................... 9

2. Experimental ...................................................................................................................................... 11

2.1 Materials ............................................................................................................................................ 11

2.2 Methods ............................................................................................................................................. 11

3. Results and Discussion ...................................................................................................................... 21

3.1 Protein purification ............................................................................................................................ 21

3.1.1 Initial characterization ................................................................................................................ 21

3.1.2 Metal removal ............................................................................................................................. 24

3.1.3 Protection from oxidation ........................................................................................................... 26

3.2 Response to zinc ................................................................................................................................ 28

3.2.1 ZapCY2 ...................................................................................................................................... 28

3.2.2 Mutant sensors ............................................................................................................................ 29

3.3 Response to nickel ............................................................................................................................. 31

3.3.1 ZapCY2 ...................................................................................................................................... 31

3.3.2 Mutant sensors ............................................................................................................................ 33

3.4 High metal concentrations ................................................................................................................. 35

3.5 Size exclusion chromatography on ZapCY2...................................................................................... 37

3.6 Response to magnesium ..................................................................................................................... 38

4. Conclusion and Future Directions ................................................................................................... 39

5. Appendix ............................................................................................................................................ 41

5.1 Calculating half-life of binding .......................................................................................................... 41

5.2 PCR mutagenesis primers .................................................................................................................. 41

5.3 ESI-MS spectrum of PMB treated ZapCY2....................................................................................... 42

5.4 ZapCY2 sequence .............................................................................................................................. 43

6. References .......................................................................................................................................... 44

v

List of Abbreviations

DTNB: 5,5’-dithio-bis-(2-nitrobenzoic acid)

DTT: Dithiothreitol

eCFP: Enhanced cyan fluorescent protein

EDTA: Ethylenediaminetetraacetic acid

EGTA: Ethylene glycol tetraacetic acid

ESI-MS: Electrospray ionization mass spectrometry

FP: Fluorescent protein

FRET: Förster Resonance Energy Transfer

HEPES: 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

ICP-MS: Inductively couple plasma mass spectrometry

IPTG: Isopropyl β-D-1-thiogalactopyranoside

MOPS: 3-(N-morpholino)propanesulfonic acid

PAR: 4-(2-pyridylazo)resorcinol

PDC: Pyridine-2,6-dicarboxylic acid (dipicolinic acid)

PMB: p-Hydroxymercuribenzoic acid

R: FRET ratio = I525/I475

Rmax: FRET ratio at saturation

Rmin: FRET ratio in the absence of metal

TCEP: Tris(2-carboxyethyl)phosphine

TRIS: Tris(hydroxymethyl)aminomethane

YFP: Yellow fluorescent protein

vi

List of Tables

Table 1. Summary of nickel sensors reported in literature

Table 2. Comparison of ZapCY1 and ZapCY2

Table 3. Optical features of eCFP and YFP

Table 4. Summary of response to zinc of 4 sensor variants

Table 5. Phusion PCR mutagenesis primers

vii

List of Figures

Figure 1. Chemical structure of NS1

Figure 2. Representation of the FRET event of ZapCY2

Figure 3. ZapCY2 first binding site

Figure 4. ZapCY2 second binding site

Figure 5. Alignment of linker sequence of ZapCY2 and mutant sensors

Figure 6. SDS-PAGE of purified sensor

Figure 7. Electronic absorption spectrum of ZapCY2

Figure 8. Fluorescence emission spectrum of ZapCY2 with zinc addition

Figure 9. Relationship between sensor concentration and fluorescent intensity

Figure 10. Effect of progression of protein purification on R and Rmin

Figure 11. Effect of metal removal methods on Rmin

Figure 12. Chemical structure of PDC and PMB

Figure 13. Effect of TCEP on Rmin

Figure 14. Effect of TCEP on R-Rmin of C2D with zinc addition.

Figure 15. ZapCY2 zinc competition assay

Figure 16. C2D zinc competition assay

Figure 17. C2H zinc competition assay

Figure 18. WF-H zinc competition assay

Figure 19. ZapCY2 nickel titration

Figure 20. C2D nickel titration

Figure 21. C2H nickel titration

Figure 22. WF-H nickel titration

Figure 23. ZapCY2 zinc and nickel titration with millimolar metal concentrations

Figure 24. Size exclusion chromatography spectrum of ZapCY2

Figure 25. Size exclusion chromatography spectrum of ZapCY2 in the presence of zinc

Figure 26. ZapCY2 magnesium titration

1

1. Introduction

1.1 Nickel in bacteria

Nickel is a first-row transition metal extensively used in multiple plant, fungal and

notably pathogenic bacterial species,1-3 where it is used as a catalytic centre in many crucial

enzymes. A group of these enzymes such as [NiFe]-hydrogenase,4 carbon monoxide

dehydrogenase (CODH),5 and methyl-CoM reductase,6 allow various anaerobic and

methanotrophic bacteria and archaebacteria to extract energy from hydrogen, carbon

dioxide, and methane gas respectively. Other enzymes, such as urease and superoxide

dismutase, are used to control the microorganisms’ external and internal environment. For

example, urease is used by the pathogenic bacteria Helicobacter pylori to neutralize the

local pH during colonization of the human stomach,7 while superoxide dismutase is used

by many bacterial species for the detoxification of harmful superoxide radicals.8 Nickel

also affects the activity of various transcriptional regulators by serving as an activating

effector, and -in turn- regulating gene expression in the cell. One noteworthy transcription

factor is NikR in H. pylori, which plays a role in the activation and also repression of genes

responsible for the uptake, metabolism and storage of nickel,9 as well as iron metabolism.10

In addition to serving as an essential nutrient, nickel can also inflict cell damage in multiple

ways. It can replace different transition metals in non-nickel centred enzymes, altering

their activity or rendering them inactive. It can also bind to the catalytic and allosteric sites

of non-metalloenzymes, inhibiting their activity.11 Finally, nickel can also participate in

the modification of biological molecules in the cell, such as DNA cleavage, inter-strand

crosslinking, base modification, as well as lipid and protein side-chain oxidation.12-13

2

With the amounts of nickel in a constant flux due to its involvement in many processes,

monitoring the intracellular concentrations of nickel ions would be very beneficial for

understanding the biochemistry of pathogenic bacteria, as well as other organisms that use

nickel, or are impacted by environmental exposure to nickel. In addition, it would be very

illuminating to examine how nickel concentrations are influenced by changes in the

environment. This might also aid with the development of potential therapeutics, with

nickel pathways posing as an attractive target because of nickel’s absence in human

biochemistry.2

Total nickel concentrations have been reported for a number of organisms, for example the

total nickel concentration in E. coli is as high as 30 μM.11 However, the free intracellular

concentration has not been determined for any microorganism. It has been speculated that

free nickel concentrations in human cells might resemble those of zinc, lying in the

picomolar range.14 This value however might not be representative of bacterial species,

which depend on nickel for cell function, in contrast to mammalian cells. The likely range

of free intracellular nickel concentrations in an organism might be estimated by measuring

the dissociation constants of major nickel-dependent transcriptional factors. H. pylori and

E. coli utilize the nickel-dependent transcription factor NikR, which regulates the nickel

uptake and possesses a Kd in the low pM range.10, 15-16 RcnR as another example, is a

transcription factor that controls the E. coli nickel efflux transporter RcnA, where apo-

RcnR represses the expression of RcnA in the absence of nickel. When nickel is available,

the metal ion binds to RcnR and relieves the repression. The nickel Kd of RcnR is in the

mid nM range.17 The two Kd’s of NikR and RcnR together might reflect the range of free

nickel concentration in these bacterial species.

3

1.2 Nickel sensors

A small number of nickel sensors have been reported in the literature; these include

small-molecule, peptide, protein, and polymer based sensors (Table 1), all of which report

on nickel levels through changes in fluorescence, and possess affinities in the high

micromolar range. So far, only one of these sensors by the name of NS1 can detect

intracellular nickel. This sensor is a small molecule based on a boron-dipyrromethene

(BODIPY) dye that can penetrate the cell membrane, and binds to one nickel ion. Nickel

binding results in an increased emission at 507 nm with an apparent Kd of 193 ± 5 μM.18

This weak Kd is incompatible with bacterial systems of interest that are speculated to have

free nickel concentrations in the picomolar range, so it is not suitable for real-time in vivo

nickel detection in bacteria. The development of a sensor with a Kd lying in the picomolar

range will provide a much-needed understanding of nickel’s involvement in the

biochemistry of pathogenic bacteria, and perhaps other organisms.

Figure 1. The only intracellular nickel sensor reported in literature, NS1, is based on a BODIPY dye that

binds to nickel resulting in an increased fluorescence emission at 507 nm with increasing nickel

concentration. The sensor possesses a Kd of 193 ± 5 μM.18

4

Sensor Components Mechanism Apparent Kd Intracellular Ref. NS1 BODYPI dye and N/O/S

receptor

Increased fluorescence

with nickel binding

193 ± 5 μM

Yes 18

Dioxotetraamine

fluorenyl ligand

Fluorenyl and

dioxotetraaza derivatives

Fluorescence quenching

of fluorenyl with nickel

binding

n/a No 19

NBP-MDCC E. coli nickel binding

protein (NBP)

conjugated to a

coumarin derivative

(MDCC)

Fluorescence quenching

of coumarin derivative

due to NBP

conformational change

resulting from nickel

binding

≈ 10 μM No 20

dH3w Histidine-rich

glycoproteins (HRP’s)

conjugated to a dansyl-

amide group and a C-

terminus tryptophan

Nickel binding induces

a conformation change

resulting in FRET

between the dansyl-

amide and tryptophan

≈ 50 μM No 21

ATCUN

peptides

Amino terminal Cu(II)

and Ni(II) binding motif

(ATCUN) conjugated to

coumarin and rhodamine

derivatives

Nickel binding induces

a conformation change

resulting in FRET

between coumarin and

rhodamine derivatives

n/a No 22

poly(Ac-HMPC-

co-AM)

Coumarin acrylic

polymer

Increased fluorescence

with nickel binding to

polymer

≈ 15 μM No 23

poly(Ac-HMPC-

co-VP)

Coumarin and vinyl-

pyrrolidone acrylic

polymer

≈ 30 μM No 24

P1 and P2 Benzochalcogendiazole

and triazole derivatives

polymer

≈ 1 μM No 25

Table 1. Summary of nickel sensors reported in the literature, including name, structural components,

mechanism, and apparent affinity. FRET-based sensors are highlighted in grey. Only one reported sensor,

NS1, has been tested in vivo and found to be suitable for intracellular measurements.

5

1.3 Genetically encoded FRET-based metal sensors

A popular strategy for detecting intracellular metal levels is to use FRET-based

genetically encoded sensors. These sensors are composed of a fluorescent protein (FP)

pair, a donor and an acceptor, that engage in Förster Resonance Energy Transfer (FRET).

This phenomenon involves an emission event by an acceptor FP in the presence of an

excited donor FP, located at a physical distance of 10 nm or less, where the fluorescence

emission intensity depends on the separation distance.26-27 The sensor is designed such that

this separation distance is controlled by a ligand-induced structural change. FRET-based

systems offer the advantage of ratiometric detection, i.e. fluorescence changes are

monitored as a ratio of the donor’s emission and the acceptor’s emission detected upon

excitation of the donor. This feature eliminates the difficult task of controlling intracellular

sensor concentration,28 since the sensor concentration can be calculated from the

fluorescence intensity of the donor, or the direct excitation of the acceptor, which is

independent of the concentration of the ligand. This design also provides an internal

reference for overcoming environmentally induced fluctuations in fluorescence, as well as

independence from optical path length and excitation intensity.29

These types of systems have been successfully designed to detect intracellular zinc, copper,

iron, cobalt, cadmium, calcium and magnesium, but so far, none have been designed

exclusively for nickel.28 A great number of these sensors target zinc with varying

affinities,28, 30 and some have been reported to be weakly responsive to nickel.21, 31 With

free nickel concentrations predicted to be at the picomolar level, a nickel sensor should

have an affinity in that range. It should also have a high sensitivity to resolve the slight

concentration changes to be expected in vivo. It should also have a higher selectivity for

nickel than other transition metals.

6

1.4 ZapCY2

A particularly successful series of FRET-based zinc sensors was designed by

Palmer et. al., the most recent being ZapCY1 and ZapCY2.32 These sensors are inspired

by the first and second zinc fingers, ZF1 and ZF2, in the zinc-binding domain of Zap1, a

zinc-regulated transcription factor from Saccharomyces cerevisiae.33 Zap1 binds zinc in a

tetrahedral coordination in two Cys2His2 binding sites (in ZapCY1). ZapCY2 was created

by mutating a total of two cysteines, one on each site, to two histidines (Figure 3 and 4).

The binding region is flanked by two fluorescent proteins, an enhanced cyan fluorescent

protein, eCFP, on the N-terminus, and a yellow fluorescent protein, YFP, on the C-

terminus. The binding of two zinc ions decreases the separation distance between the

fluorescent proteins, resulting in increased FRET (Figure 2). These sensors displayed

resistance against photobleaching and pH changes.32

ZapCY2 has a lower selectivity, dynamic range and affinity for zinc than ZapCY1 (Table

2).32 For this reason, ZapCY2 was deemed as a better starting point for developing a sensor

with a higher selectivity for nickel.

Zinc addition

FRET

525 nm 434 nm 475 nm 434 nm

Figure 2. Representation of the ZapCY2 FRET event. In the absence of zinc, exciting the CFP/donor at

434 nm leads to CFP emission at 475. With zinc addition, the two chromophores are closer, and

excitation of the CFP can lead to emission at 525 nm by the YFP/acceptor.

eCFP YFP eCFP YFP

7

FRET is conventionally measured by comparing the emission of the acceptor upon

excitation of the donor, with the emission upon direct excitation of the acceptor. The

excitation wavelength of the YFP/acceptor is 515 nm while its emission is at 529 nm. In

practice, this small difference between the excitation and emission wavelength means that

measuring YFP/acceptor emission as a result of direct excitation is very difficult due to the

low signal to noise ratio.

For that reason, the FRET response is instead measured using the intensities at the two

emission maxima, of the donor and acceptor, coinciding with 475 nm and 525 nm

respectively. This ratio equates to IFRET/ICFP, where ICFP represents the emission intensity

of eCFP/donor at 475 nm upon its excitation at 434 nm, while IFRET represents the emission

intensity of YFP/acceptor at 525 nm upon the excitation of eCFP at 434 nm followed by

energy transfer (Figure 2). The ratio is thus calculated as I525/I475, where I is the

fluorescence intensity in relative fluorescent units, RFU’s.32 The emission wavelength of

525 nm was used in this study instead of the literature reported value of 529 nm because it

appeared to result in a higher fluorescence intensity. This ratio is referred to as the “FRET

ratio” or “R”.

Another assessed parameter of the system is what was labelled by Palmer et. al. as the

“dynamic range”. The dynamic range here refers to Rmax/Rmin, where Rmax represents the

maximum FRET ratio, which is reached upon saturation of the zinc binding sites, and Rmin,

which represents the FRET ratio in the absence of any metal.

Sensor Kd (zinc binding) Cooperativity Dynamic range

ZapCY1 2.51 pM 1.37 1.65

ZapCY2 811 pM 0.44 1.25

Table 2. Reported affinity, cooperativity, and dynamic range of zinc binding for the two sensors developed

by Palmer et. al,32 ZapCY1 and ZapCY2.

8

Figure 3. Pymol® visualized structure of zinc bound to the first Cys1His3 zinc finger of the ZapCY2 zinc binding/linker

region. The crystal structure was retrieved from PDB code 1ZW8 representing the Cys2His2 ZapCY1,34 and modified

through Pymol® to represent ZapCY2 by changing Cys238 to His238, displayed as an arbitrary rotamer. Residues

mutated to generate the sensor variants in this study are outlined.

Figure 4. Pymol® visualized structure of zinc bound to the second Cys1His3 zinc finger of the ZapCY2 zinc

binding/linker region. The crystal structure was retrieved from PDB code 1ZW8 representing the Cys2His2 ZapCY1,34

and modified through Pymol® to represent ZapCY2 by changing Cys275 to His275, displayed as an arbitrary rotamer.

Residues mutated to generate the sensor variants in this study are outlined.

W240

H261

H256

C243

H238

H293

F282

H298

C280

H275

9

1.5 Purpose of study

The initial aim of this study was to characterize the ZapCY2 sensor, and quantify

its response to both zinc and nickel in terms of affinity (through Kd determination), and

Rmax/Rmin, which is indicative of sensitivity. This stage was followed with mutation of the

binding residues and further characterization. The ultimate goal is to design a mutant

protein that has a notably reduced affinity for zinc, and an increased affinity for nickel

when compared to the original version.

As a starting point, efforts were focused on changing the two cysteine residues in the

binding motif, Cys243, and Cys280. Since nickel is a slightly harder acid than zinc,

according to the Hard Soft Acid Base (HSAB) theory, one strategy would be to increase

the overall hardness of the residues involved in binding. Since cysteine is a softer ligand

than histidine, the cysteine could be substituted for harder amino acids such as histidine,

glutamic acid, and aspartic acid. A survey of nickel binding proteins in bacteria has shown

that these proteins exclusively use these four amino acids for nickel binding, particularly

the latter two.1, 35

Other target sites for mutation are two strategically positioned aromatic residues in the

two zinc fingers; a tryptophan neighbouring the first histidine in the first metal binding

site, Trp240, and a phenylalanine neighbouring the second histidine of the second binding

site, Phe282 (Figures 3 and 4). Altering these residues would serve two purposes. Given

that these residues are thought to be involved in π-stacking with one of the histidines

involved in metal binding, as suggested by the crystal structure of the native Zap1, it would

be useful to study the effect of disrupting this interaction. Secondly, converting these

residues to amino acids capable of binding metal might alter the coordination, perhaps

10

allowing a square planar coordination in place of the existing tetrahedral one, which will

in turn favor nickel over zinc binding. A histidine was chosen to replace these two aromatic

residues. The modifications were initially introduced into both zinc fingers for ease of

assessment (Figure 5).

The characterization of the response of these mutant variants to zinc and nickel has

revealed that the site modifications slightly weakened the affinity for zinc but did not affect

the affinity for nickel. There were also several challenges faced in terms of purification

and post-purification treatments needed to achieve a reduced and metal free (apo) sensor.

ZapCY2: LKHKWKECPESCCSSLFDLQRHLLKDHVSQDFKHPMEPLAHNWEDCDFLGDDTCSIVNHINCQH

C2D: LKHKWKEDPESCCSSLFDLQRHLLKDHVSQDFKHPMEPLAHNWEDDDFLGDDTCSIVNHINCQH

C2H: LKHKWKEHPESCCSSLFDLQRHLLKDHVSQDFKHPMEPLAHNWEDHDFLGDDTCSIVNHINCQH

WF-H: LKHKHKECPESCCSSLFDLQRHLLKDHVSQDFKHPMEPLAHNWEDCDHLGDDTCSIVNHINCQH

1st zinc finger 2nd zinc finger

Figure 5. Alignment of ZapCY2 and the 3 variants made in this study. The 4 residues involved in metal

coordination are marked by an arrow, and the residues replaced through mutagenesis are underlined.

11

2. Experimental

2.1 Materials

HEPES, MOPS, glycerol, NaCl, KCl, EDTA, LB media, kanamycin, and IPTG

were obtained from BioShop. Protease inhibitor tablets were obtained from Roche™.

Bl21DE3® and NEB turbo® competent cells were obtained from New England Biolabs.

Plasmid prep kit and Phusion® PCR components were obtained from Thermo Scientific,

and rapid ligation buffer from Fermentas. Chelex 100® resin and gel filtration molecular

weight standard were obtained from BioRad. Streptactin® resin was purchased from IBA

Lifesciences, and HiTrapQ™ and 200 30/100 GL Superdex™ columns from GE Healthcare.

Protein concentrators were obtained from Amicon, and dialysis tubing from

SpectrumLabs. Black and white NUNC plates were obtained from Thermo Scientific.

Sterilizing 0.45 um filters were obtained from Millipore. PMB, PAR, EGTA,

desthiobiotin, and metal salts for metal stocks were obtained from Sigma Aldrich, and the

AA multi-element metal standards from PlasmaCal. Milli-Q filtration system was used to

prepare 18.2 ΜΩ.cm resistance water which was used in all experiments.

2.2 Methods

Plasmids

The sensor sequence obtained from the Palmer lab at the University of Boulder, Colorado,

was cloned into a pET24-b vector using BamHI and HindIII restriction sites, along with

an N terminus/upstream strep tag that was inserted using BamHI and NdeI restriction sites

(see DNA sequence in section 5.4). The heat shock method as described by NEB® was

used to transform 50 μl of BL21DE3® or NEB turbo® cells using around 50 ng of purified

plasmid, for protein and plasmid purification respectively.

12

Protein expression and purification

Overnight cultures were started using 50 ml of LB media supplemented with 50 μg/ml

kanamycin, and were inocculated with a single colony from an overnight plate of

transformed BL21DE3® cells. The 50 ml overnight cultures were mixed into 1.5 l of LB

media with 50 μg/ml kanamycin. The cells were grown to an O.D.600 of 0.6-0.8 at 37 °C,

after which they were moved to room temperature for overnight incubation with shaking

after the addition of 1 ml of 1 M IPTG (to a final concentration of 0.67 mM) for each 1.5

l of culture. Cell culture was divided into 2 x 750 ml bottles, and centrifuged at around

3400 g for 1 hour, and pellets from 750 ml of culture were suspended in 30 ml of 20 mM

Tris-HCl, 100 mM KCl, 1 mM TCEP, pH 7.4 resuspension buffer supplemented with ½

Roche™ EDTA free protease tablet.

Resuspended cells were sonicated on ice using a Branson Sonifier Cell Disruptor 185 on

setting 7, for 10 minutes of alternating 1 minute on/off rounds. The solution was further

centrifuged at around 16000 g for a minimum of 1 hour to pellet the cell debris. This was

done using a Sorvall RC6 plus centrifuge. The supernatant was filtered through a 0.45 um

sterilizing syringe filter. The pH of the filtered supernatant can be slightly increased to pH

8 before resuming purification by adding around 0.2 ml of 1 M NaOH to 30 ml of

supernatant.

The supernatant was then loaded onto a column packed with 5 ml of IBA streptactin® resin

equilibrated in a wash buffer containing 50 mM Tris-HCl, 200 mM NaCl, 1mM TCEP, at

pH 8. Instructions provided with the resin were followed. Elution buffer was made by

supplementing the wash buffer with 2.5 mM desthiobiotin. Elution with 6-12 ml of elution

buffer was sufficient to recover the protein.

13

The protein-containing fractions were combined and either dialyzed against buffer

containing 20 mM Tris-HCl, 1 mM TCEP, 1 mM EDTA, pH 7.5, overnight. Alternatively,

it can diluted at least 4X in the buffer. This is to reduce the NaCl concentration before

further purification. The protein was further purified by using a 5 ml GE Healthcare

HiTrapQ™ column with a linear NaCl gradient protocol. The FPLC running buffer (buffer

A) used can be Tris based (20 mM Tris-HCl, 1 mM TCEP, 1 mM EDTA, pH 7.5). MOPS

based buffer is not recommended in the operating manual. Buffer B was made by adding

1 M NaCl to buffer A. Buffers were initially chelexed to remove metal using 2.5-5g resin

per 100 ml buffer, and mixing for 30 minutes before filtering. Protein usually elutes at

around 35% buffer B or 0.35 M NaCl. The HiTrapQ™ column can be initially manually

rinsed to remove built up common contaminants using: 15 ml 2 M NaCl, followed by 25

ml NaOH, 15 ml 2 M NaCl, 15 ml water, then finally 25 ml buffer A.

Collected fractions were pooled and concentrated using a spin concentrator with a MWCO

of 3K to a concentration of around 30 μM, which was determined using electronic

absorption spectroscopy as described below. Concentrated protein was dialyzed for 3 hours

in 100 mM MOPS, 100 mM KCl, 300 mM DTT, and 100 mM EDTA. The protein was

then dialyzed for 3 hours, then overnight in two changes of MOPS buffer containing 100

mM MOPS, 100 mM KCl, and 0.5 mM TCEP.

All processes, except cell culturing, were done at 4 °C.

Protein concentration determination

Protein concentration was determined using the absorbance of the protein at 280 nm and

515 nm, which reflects the absorbance of the tryptophans and the YFP/acceptor with molar

absorptivities of 64,500 and 77,000 l mol-1 cm-1 respectively (Table 3). MOPS buffer was

14

used as the blank. A 515/280 absorbance ratio less than 1.194 indicates protein with

underdeveloped chromophores, possibly due to insufficient oxygen levels needed for

maturation. It can also indicate contamination with other proteins. A ratio higher than

1.194 indicated free chromophores which could result from possible cleavage by proteases.

Protein

(Acronym)

Excitation

Maximum

(nm)

Emission

Maximum

(nm)

Molar

absorptivity, ε

(l mol-1cm-1)

Quantum

Yield, Φ

In vivo

Structure

Ref.

eCFP 434 476 29,000 0.37 Weak dimer 36-37

YFP(Citrine) 516 529 77,000 0.76 Monomer 38

SDS-PAGE gels

Protein samples were mixed with urea dye in a ratio of 2:1 by volume, usually 13.3 μl of

around 10 μM protein with 6.7 μl dye per sample. The urea dye was made by mixing 4X

loading dye containing 4 ml upper tris buffer, 3.4 ml 10 % SDS, 2 ml 100 % glycerol,

0.01 g bromophenol blue, and 0.58 g DTT with 8 M urea in a 1:3 ratio by volume.

Immediately before addition to the protein, the urea dye was mixed with 100% β-

mercaptoethanol (BME) in a 10:1 ratio by volume before addition to the protein. The

samples were ran on a 10% SDS-PAGE gel. The gels were then conventionally stained

with Coomassie dye then de-stained.

Oxidation level determination

With the presence of a total of nine cysteines in the protein, two of which are directly

involved in binding metal, the protein is highly prone to oxidation. DTNB (5,5’-dithio-

bis-(2-nitrobenzoic acid) assays are traditionally used for the determination of oxidation

state. This assay however requires the absence of reducing agents such as TCEP.

Table 3. Optical features of eCFP and YFP, the two chromophores involved in the ZapCY1 and ZapCY2

structure. Tryptophan and YFP absorbance is used for the determination of protein concentration.

15

The removal of TCEP from the protein using a desalting PD10 column leads to inevitable

protein oxidation (Figure 13). This can be avoided by removing the TCEP under anaerobic

conditions (in the glovebox), however the PD10 column results in substantial loss of

protein. For that reason, DTNB assays were usually avoided, and instead, the FRET ratio

“R” in the absence of metal, also referred to as Rmin, provides a qualitative indication of

whether the protein is oxidized. Sufficiently reduced ZapCY2 protein that displayed the

literature reported dynamic range (Table 2), was consistently observed to have Rmin’s

ranging from 1.7-1.8, while oxidized protein had Rmin value of around 2 and higher, which

remained unchanged upon the addition of high concentrations of zinc. Due to the presence

of cysteines in the binding sites, as well as additional cysteines in the linker region

(uninvolved in metal binding), oxidation involving disulfide bridge formation might lock

the linker in a permanent position that is unfavorable to a successful FRET response.

Measuring FRET response

Fluorescence was measured by incubating 1 μM protein in 120 μl of 150 mM HEPES, 100

mM NaCl, 10% glycerol, 1 mM TCEP, pH 7.4 “FRET” buffer chelexed with 5 g of resin

per 100 ml of buffer, along with the appropriate volume of metal stock. The mixtures were

divided into three wells of a black NUNC 384 plate. Fluorescence was measured through

a spectral emission scan from 465 to 545 nm with a bandwidth of 10 nm after excitation at

434 nm with a bandwidth of 16 nm, with the resolution set to 5 nm using a ClarioSTAR

plate reader. The focal length and gain were initially adjusted for the full plate.

Fluorescence intensities at 475 and 525 nm were used to calculate the FRET ratio, where:

FRET ratio “R” = I525/I475

16

Metal titrations

Metal titrations involved the addition of ZnSO4 or NiSO4 metal stocks into the FRET

buffer and mixing well before the addition of the protein, followed by mixing by pipetting

to avoid damaging the protein. For zinc and nickel titrations, the mixtures were incubated

for 2.5 and 1 hour respectively at room temperature before measuring fluorescence. These

times coincide with roughly 10 half lives of binding (based on koff), derived through the

measured affinities (See section 5.1 for calculation). FRET buffer was always kept on ice

prior to use since TCEP is temperature sensitive.

Metal titrations in the presence of a competitor

Chelator-buffered metal mixes were made by incubating 3000 or 4500 μM of EGTA and

3-2700 μM zinc in FRET buffer for 20-30 minutes, before diluting 1:2 in more FRET

buffer, mixing well, then adding the protein. This ensures good equilibration of metal and

competitor/chelator. The protein was incubated for 2.5 hours with the mixture before

measuring fluorescence.

Zinc Kd determination

Titration with metal and competitor (zinc and EGTA) was used for Kd determination. The

data were fit using the software Prism® using the Hill coefficient equation:

[𝑅𝐿]

[𝑅]𝑇=

[𝐿]𝑛

𝐾𝑑𝑛 + [𝐿]𝑛

(𝑒𝑞. 1)

Here, “Kd” is the dissociation constant in molar concentration, M, “n” is the Hill

coefficient, and [RL]/[R]T is bound protein concentration over total protein concentration.

17

An alternative is the one-site binding model, where Kd’ is the apparent dissociation

constant.

[𝑅𝐿]

[𝑅]𝑇=

[𝐿]

𝐾𝑑′ + [𝐿] (𝑒𝑞. 2)

The one binding site model was found to produce a weaker fit of data than the Hill equation

as show by Palmer et. al.32

Modifying equation 1 to represent the FRET ratio, the term [RL]/[R]T is substituted for:

(R-Rmin)/(Rmax-Rmin)

Since the free zinc concentration is plotted against R-Rmin, equation 1 becomes:

(R − Rmin) = (Rmax − Rmin).[L]n

Kdn + [L]n

(𝑒𝑞. 3)

The x-axis represents the free ligand concentration [L], or free metal concentration.

The free zinc concentration in the EGTA buffered metal-chelator mix is calculated as:

Kd′ =[EGTA][Zn]

[EGTA. Zn] (𝑒𝑞. 4)

[EGTA.Zn]= x [EGTA]=0.001-x (if 1000 μM EGTA was used) [Zn]= [Znadded]-x

Kd’= 8.13 x 10-11 M at 25 °C, and 0.1 N ionic strength. This value was calculated as

described by Fahrni et. al.39

Equation 4 is solved as a quadratic equation to calculate [Zn], or free zinc:

(Kd’)(x)= (0.001-x)([Znadded]-x)

18

Replicates and Error bars

For all graphs presenting the FRET response or fluorescence intensity, the number of

replicates is denoted as “N”. Data points represent an average of readings per fluorescence

plate well since all samples were carried out in triplicate wells, so that N=1 is composed

of 3 triplicate wells per sample. The error bars all represent 1 standard deviation of all

individual well readings needed to generate the average.

Modified PAR assay

The chromophores were intact after the addition of 4 M GnHCl usually employed in the

PAR assay. This resulted in significant protein absorbance around 515 nm (which

interfered with the absorbance of the PAR-Zn complex at 500 nm).40 Additionally, the

affinity of the intact protein for zinc is very comparable to that of PAR, leading to

competition for zinc.40 Due to the protein’s resistance to denaturation and competition with

PAR over zinc, the protein was instead boiled with 4 M urea for around an hour, before

the addition of 60 μM PAR, followed by absorption measurements at 500 nm. Zinc

standards ranging from 1-10 μM zinc were made using 100 μM ZnSO4 stocks. The

standards were similarly diluted with FRET buffer, and boiled with urea before addition

of PAR.

ICP-MS (initial metal loading level)

ICP-MS (inductively coupled mass spectrometry) was carried out with 270 μl of not less

than 1 μM protein samples diluted in FRET buffer and incubating overnight with 230 μl

nitric acid (46%) to make a 0.5 ml sample. Standards were similarly prepared with 46%

nitric acid, and small volumes of 1:100 or 1:10000 dilutions of AA multi-element mix.

FRET buffer was also used for dilution.

19

Size Exclusion Chromatography

Protein at a concentration of around 30 μM was concentrated to around 50 μM using micro

spin concentrators. Around 120 μl of the protein was incubated with 100 μM zinc for 3

hours before injecting into a 200 30/100 GL Superdex™ column equilibrated with MOPS

buffer containing 100 mM MOPS, 100 mM KCl, 0.5 mM TCEP, and 100 μM zinc at pH

7.4. Gel filtration molecular weight standard was diluted in a 1:1 ratio with water to make

around 120 μl, and ran under identical conditions.

PDC treatment

About 1 ml of around 20 μM protein was dialyzed overnight in a buffer containing 0.2 M

phosphate, 0.075 M pyridine-2,6-dicarboxylic acid (dipicolinic acid),41 0.1 M NaCl and 1

mM TCEP, at pH 7.4. The protein was then dialyzed overnight in FRET buffer.

EDTA treatment

Around 10 mM of EDTA was added to around 20 μM protein and incubated overnight.

This was then later dialyzed for not less than 6 hours in FRET buffer.

PMB treatment

To 300-500 μl of 30-60 μM protein, 10 molar equivalents of PMB (p-

Hydroxymercuribenzoic acid) were added followed by 15 minutes of incubation on ice.

This was followed by 2 mM EDTA to chelate the released metal, in addition to 10 more

equivalents of PMB. The mixture was again incubated for 15 minutes at room temperature.

The mercury thiolate bonds between the PMB and cysteines were reversed by adding 10

mM DTT and incubating for 10 minutes at room temperature. The protein was dialyzed

against 3 changes of TCEP containing MOPS buffer at 4 °C for 2 days.42

20

PCR mutagenesis

Phusion® PCR was done using between 100 pg-1 ng of template plasmid, in addition to 2-

2.6 μl of 100 μM pre-mixed forward and reverse primer, with 5 μl of 5 mM dNTP mixture,

50 μl Thermo Fisher Phusion® buffer, and 2.5 μl Phusion® polymerase. This produced 250

μl which was divided into 5 x 50 μl reaction mixtures. The PCR protocol is as follows:

Initial denaturation at 98 °C for 30 seconds, denaturation at 98 °C for 10 seconds, annealing

at 60-72 °C for 30 seconds, extension at 72 °C for 3 minutes and 30 seconds, for 25 cycles,

final extension at 72 °C for 10 minutes. The PCR product was ran on a 0.5% agarose gel

to confirm for PCR success by checking for bands around 7 kb. A ligation reaction was set

up by mixing 5 μl of successful PCR product with 2 μl of 5X rapid ligation buffer, 2.5 μl

H2O, and 0.5 μl T4 DNA ligase. The mixture was then incubated at room temperature for

5 min. The circularized PCR product was either stored at -20 °C or immediately used to

transform NEB turbo® competent cells for later plasmid extraction and sequencing to

confirm successful mutagenesis (See section 5.2 for primers).

21

3. Results and Discussion

3.1 Protein purification

3.1.1 Initial characterization

The molecular weight of the purified ZapCY2 was confirmed using ESI-MS

(electrospray ionization mass spectrometry), where the protein’s molecular weight in its

apo form (including the strep-tag) is 62711.82 Da. Additionally, the protein was run on a

10% SDS-PAGE gel. The resultant bands consistently displayed a higher mobility than

expected, possibly as a result of incomplete denaturation (Figure 6). The purified protein

samples also contained some dimers. This was confirmed by size exclusion

chromatography, followed by electronic absorption spectroscopy on the dimer containing

fraction collected. These features were present in all mutants.

Monomer

kDa

113

66.2

45

35

ZapCY2 Mutant C2D Mutant WF-H

Dimer

Figure 6. 10% SDS-PAGE gel of multiple preps of purified ZapCY2 and mutants. Bands displayed

a downward band shift from the expected position of 62.7 kDa, possibly as a result of incomplete

denaturation leading to high mobility. Dimer formation was also apparent.

22

The sensor’s electronic absorption spectrum from 250-600 nm was used to determine the

concentration by measuring the absorbance values at 280 nm and 515 nm, representing the

absorbance of the tryptophan residues and the YFP/acceptor chromophore, respectively

(Figure 7).

The fluorescence emission of the sensor at 475 and 525 nm was recorded upon excitation

at 434 nm, and in the absence of metal (Figure 8). Emission at 475 nm represented the

CFP/donor and decreased with zinc addition, while emission at 525 nm represented the

YFP/acceptor and increased with zinc addition.

The fluorescence emission intensity of both CFP (475 nm) and YFP (525 nm) displayed a

linear relationship with protein concentration (Figure 9), and concentrations lower than

around 0.2 μM resulted in signals too similar to the background signals for both

wavelengths. Due to the poor response below this point, these protein concentration were

never used for fluorescence measurements. The apparent emission at 525 nm, representing

Figure 7. Absorbance spectrum of a

~ 6 μM sample of ZapCY2. Peaks are

present at 280, 434, and 515 nm

representing the absorption of the

tryptophans, CFP, and YFP

respectively. This absorbance

spectrum was seen with all sensor

variants.

Trp (280 nm) YFP (515 nm)

CFP (434 nm)

Figure 8. The fluorescence emission

of ZapCY2 from 465-545 nm upon

excitation at 434 nm reveals two

emission peaks at 475 and 525 nm

coinciding with CFP and YFP,

respectively. Zinc addition resulted

in a decrease in emission at 475 nm,

and an increase at 525 nm. This

emission spectrum and subsequent

change with zinc addition was seen

with all sensor variants.

23

the YFP/acceptor upon excitation at 434 nm (required to excite the CFP/donor), indicates

that a degree of energy transfer (FRET) was occurring in the absence of metal.

ICP-MS in conjunction with PAR assays were used to determine the degree of metal

loading of fully purified protein (see methods). Fully purified protein samples varied in

metal content from 0.097-0.81 equivalents for zinc, and 0.019-0.063 equivalents for nickel.

Despite identical purification procedures, the highest metal loading levels were associated

with ZapCY2, while the lower metal loading levels were usually detected in the mutant

protein samples. This difference is thought to be reflective of the relative metal binding

affinities of the sensor variants (see below).

As explained in the methods section, during protein purification, the protein is subjected

to three dialysis steps. The first step is carried out after retrieving the protein from the

strepactin® affinity column and before further purifying it using an anion exchange column

(to reduce the NaCl concentration and remove desthiobiotin used for elution from the

strepactin® column). The second dialysis step is to ensure the protein is not oxidized, where

a large amount of DTT, a reducing agent, is included in the buffer. Finally, the third dialysis

step introduces TCEP to make the buffer more suited to prolonged storage of the protein,

due to the greater stability of TCEP compared to DTT.

Figure 9. The linear relationship

between ZapCY2 concentration and

the fluorescence intensities of CFP

and YFP in ZapCY2 at 475 and 525

nm, respectively, in the absence of

metal. The excitation wavelength

used was 434 nm. The emission at

525 nm reflects background FRET

occurring in the absence of metal.

24

The effectiveness of the protein purification protocol was inspected by measuring the

starting FRET ratio in the absence of added metal (Rmin) at different stages of the protein

treatment, as well as upon the addition of zinc. As the protein purification process

progressed, the starting FRET ratio (Rmin) increased, possibly because of metal

contamination during treatment. However, the relative increase in FRET ratio (R-Rmin)

with the addition of zinc is much larger towards the end of purification (Figure 10).

It can be said that the purification procedure gradually improved both the increase in

FRET ratio for a given amount of metal added, as well as the dynamic range of the sensor,

although it resulted in an increase in the starting FRET ratio “Rmin”.

Figure 10. The FRET ratio “R” (top) and the increase in FRET, or “R-Rmin” (bottom) of ZapCY2 during

the protein purification process. The FRET ratio in the absence of metal (Rmin), and the increase in FRET

with metal addition both increased after the final FPLC purification. N=2

25

3.1.2 Metal removal

Several metal removal strategies were attempted after finding that the starting

FRET ratio “Rmin” increases throughout the progression of protein purification, possibly

due to accumulation of metal contamination of the protein. These strategies aimed to

achieve the lowest starting Rmin, which is indicative of an apo or unloaded sensor. The

tested methods included treatment with the metal chelators EDTA and PDC, as well as

PMB, which binds to thiol groups, releasing any coordinated metals. These treatments

were followed by dialysis to remove the small molecules from the protein solution.

EDTA and PDC treatments both caused an increase in the FRET ratio in the absence of

added metal (Rmin) when compared to the untreated protein, possibly because of additional

metal introduction during the dialysis step required to remove the EDTA and PDC from

the protein (Figure 11).

Increasing protein concentration without metal addition led to a decrease in FRET ratio for

treated as well as untreated protein. It is speculated that this is the result of a decrease in

the background metal, protein ratio.

Figure 11. FRET ratio of fully purified ZapCY2, without metal addition, after treatment with EDTA,

PDC, or PMB to remove bound metal during purification. The EDTA, PDC, and PMB treated proteins

were dialyzed overnight to remove the small molecules. The untreated protein was not dialyzed. EDTA

and PDC treatment led to an increase in Rmin while PMB treatment led to no significant change. N=2

26

Due to the increase in the Rmin value (Figure 11), it was decided that the protein should not

be subjected to EDTA or PDC treatment, particularly that residual EDTA and PDC might

interfere with later attempts to quantify the response to metal addition. The decrease in

Rmin seen with PMB treated protein was not deemed significant. The treated protein did

not have a different dynamic range when compared to the untreated protein. Additionally,

the PMB treated protein displayed an altered response profile to zinc in terms of affinity

compared to untreated protein (Data not shown), which might very likely be the result of

residual PMB forming adducts with the cysteines. More importantly, the PMB treatment

appeared to result in protein degradation as apparent from ESI-MS (see section 5.3 for

spectra). For these reasons, PMB treatment was not adopted as a metal removal strategy.

3.1.3 Protection from oxidation

Due to the presence of 9 cysteines in ZapCY2 and WF-H, and 7 in C2D and C2H,

the sensor is highly prone to oxidation. The sensors were accordingly kept in buffer

containing 1 mM TCEP during storage and for all assays. To determine the impact of

oxidation on the FRET response, the FRET ratio in the absence of metal (Rmin), was

measured in the presence and absence of 1 mM TCEP (Figure 13).

It was confirmed that the absence of TCEP led to an increased FRET ratio compared to

when TCEP was present, suggesting protein oxidation. This effect was observed with all

protein variants, although it was more apparent with ZapCY2 and C2D. It should also be

Figure 12. (Left) Dipicolinic acid,

PDC, used for chelating metals, and

(right) 4-(hydroxymercuri)benzoic

acid, PMB, that binds thiol group to

release bound metals.

27

noted that each protein variant has a “signature” Rmin value, even in the presence of

reducing agent, that is reflective of its affinity for background zinc, and initial loading with

trace metal (Figure 13).

It is suspected that oxidation causes a conformational change in the linker that leads to a

permanent decreased distance between chromophores, resulting in a higher FRET ratio in

the absence of metal. Oxidation was also found to greatly decrease the dynamic range for

all sensors. Figure 14 displays the effect on C2D, where the increase in the FRET ratio

with zinc addition was reduced by more than half (Data is not shown for the other sensors).

Figure 13. FRET ratio of purified ZapCY2, C2D, C2H, and WF-H in the absence and presence of 1 mM

reducing agent, TCEP, without any additional metal. N=2

Figure 14. FRET response of C2D with added zinc in the presence N=8 and absence N=2 of TCEP. The

absence of TCEP leads to a much smaller increase in FRET ratio for the same amount of added zinc.

Lines of best fit and error bars were omitted for clarity.

28

3.2 Response to zinc

3.2.1 ZapCY2

The initial step in characterizing the sensor was to measure its dissociation constant

for zinc, Kd. This value was determined to be necessary for future comparisons of the

sensor’s performance with that of the designed mutants.

With the Kd previously predicted by Palmer et. al32 to lie in the picomolar range, the sensor

was titrated with chelator-buffered zinc, using EGTA as a competing ligand, to achieve

picomolar concentrations of free zinc (see methods).

The determined Kd of 540 pM is comparable to the value of 811 pM reported by Palmer

et. al32 (see table 2). The maximum R-Rmin value, or Rmax-Rmin was not explicitly reported

by Palmer et. al32, but it can be deduced to be around 0.6 from the binding curve presented

in the paper. The dynamic range or Rmax/Rmin reported by Palmer et. al was 1.2532. Here,

the average Rmin is 1.78 ± 0.05, and with the calculated Rmax-Rmin of 0.52 ± 0.01, the

dynamic range comes to: (1.78+0.52)/1.78= 1.3 ± 0.01 (Figure 15).

The Hill coefficient calculated by Palmer et. al32 was n=0.44, indicating negative

cooperativity. However, fitting the data reported here revealed a value of 1.2 ± 0.1.

Figure 15. EGTA

buffered zinc titration of

ZapCY2 fit using the

Hill equation. N=8.

Kd: 0.54 ± 0.04 nM n: 1.2 ± 0.1 Rmax-Rmin: 0.52 ± 0.01 R2: 0.88 N: 8 Dynamic range:

1.30 ± 0.01

29

3.2.2 Mutant sensors

The first mutant, C2D, with the two cysteines substituted by aspartic acids

displayed a 10-fold decreased affinity for zinc (Figure 16). The sensor had a lower Rmin of

1.49 ± 0.05 compared to ZapCY2. This could be the result of a lower level of initial loading

due to the lower affinity. The sensor also has a higher dynamic range of 1.60 ± 0.03.

The C2H sensor has an approximately 15-fold lower affinity than the original sensor, and

an average Rmin of 1. 74 ± 0.03 (Figure 17). This value is very comparable with the C2D

sensor, with the Rmax-Rmin being slightly higher.

Kd: 7.6 ± 0.5 nM n: 1.7 ± 0.2 Rmax-Rmin: 0.92 ± 0.02 R2: 0.90 N: 4 Dynamic range: 1.53 ± 0.01

Kd: 6.0 ± 0.5 nM n: 1.2 ± 0.08 Rmax-Rmin: 0.90 ± 0.03 R2: 0.84 N: 9 Dynamic range: 1.60 ± 0.03

Figure 16. EGTA

buffered zinc titration of

the mutant sensor C2D fit

using the Hill equation.

The sensor displayed an

increased dynamic range

and a 10-fold decreased

affinity for zinc compared

to ZapCY2. N=9

Figure 17. EGTA buffered

zinc titration of the mutant

sensor C2H fit using the

Hill equation. The sensor

displayed an increased

dynamic range like C2D,

and a 15-fold decreased

affinity for zinc compared

to ZapCY2. N=4

30

All mutants were subjected to titrations with zinc concentrations ranging from around 0.5

pM to 800 nM, however only concentrations above 30 pM are shown in figures 15, 16, and

17. Zinc concentrations between 0.5 – 30 pM did not illicit a response that is different from

the response seen up until the increase at 1 nM with C2D and C2H (Figure 16 and 17), and

the increase at 0.4 nM with ZapCY2 (Figure 15). WF-H however displayed an unusual

primary saturation event towards 1 nM (Figure 18).

The WF-H mutant had an average Rmin of 1.57 ± 0.04 nM, with a dynamic range

comparable to C2D and C2H. The mutation of the two aromatic residues could have

achieved the intended disruption of π-stacking that might have provided the linker region

with added rigidity. This increased flexibility of the linker in the WF-H mutant might have

resulted in better “resolving” of the binding process, to show two distinct binding events.

Figure 18. EGTA buffered zinc titration of the mutant sensor WF-H fit using the Hill equation (left) or

two binding site model (right). The sensor displayed an increased dynamic range similar to C2H and

C2D (larger than ZapCY2), and around a 30-fold decreased affinity for zinc compared to ZapCY2.

A notable distinct primary saturation event displayed was not present with the other sensor variants. N=9

Kd: 15 ± 9 nM Rmax-Rmin: 1.04 ± 0.09 R2: 0.78 N: 9 Dynamic range: 1.66 ± 0.06 n: 0.36 ± 0.03

1st Kd: 0.012 ± 0.006 nM Initial Rmax-Rmin: 0.26 ± 0.01 2nd Kd: 12 ± 1 nM Final Rmax-Rmin: 0.72 ± 0.02 R2: 0.85 N: 9 Dynamic range: 1.46 ± 0.02

31

Looking at the response of C2D, C2H, and WF-H to zinc, it can be stated that substituting

the cysteines in the linker resulted in a modest reduction of zinc affinity. Since WF-H had

the weakest affinity for zinc, which was 30-fold less than ZapCY2, it is possible that

substituting the two cysteines in the WF-H mutant with a histidine or aspartic acid might

further reduce the affinity for zinc.

3.3 Response to nickel

3.3.1 ZapCY2

In accordance with previous literature on the interaction of zinc fingers with nickel,

the Kd of ZapCY2 with nickel was expected to be in the low micromolar range.43-44

Consequently, nickel titrations were carried out without the use of a competitor.

Titrations revealed two unusual but reproducible phenomena. First, the FRET ratio

continued to increase even when more than the theoretical amount of nickel necessary for

saturation was added. The theoretical amount refers to 2 μM nickel, since 1 μM protein

was consistently used for titrations. Second, the FRET ratio initially decreased as

compared with the ratio prior to metal addition, or Rmin. The FRET ratio resumed to

increase only after the addition of nickel concentrations higher than 2 μM (Figure 19).

Kd (nM) n Rmin Rmax-Rmin Rmax/Rmin N

ZapCY2 0.54 ± 0.04 1.2 ± 0.1 1.78 ± 0.05 0.52 ± 0.01 1.30 ± 0.01 8

C2D 6.0 ± 0.5 1.2 ± 0.08 1.49 ± 0.05 0.90 ± 0.03 1.60 ± 0.03

9

C2H 7.6 ± 0.5 1.7 ± 0.2 1.74 ± 0.03 0.92 ± 0.02 1.53 ± 0.01

4

WF-H 15 ± 9 0.36 ± 0.03 1.57 ± 0.04 1.04 ± 0.09 1.66 ± 0.06 9

Table 4. Summary of the response to zinc of ZapCY2, and the three variants C2D, C2H, and WF-H.

Fitting to the Hill equation and calculating the associated error (reflecting one standard deviation) was

done using Prism®. Error for dynamic range is calculated through propagation of the error associated

with Rmin and Rmax

32

As a result, the FRET ratio resulting from the addition of 1.75-2 equivalents of nickel is

almost equal to the Rmin. This dip in response is also reflected in the YFP/acceptor

fluorescence intensity. Although no explanation can be found for the drop in FRET

response after the addition of less than 2 equivalents of nickel, the subsequent increase in

FRET ratio upon further nickel addition might be the result of non-specific binding. The

non-specific nickel binding at unknown sites in the sensor might be resulting in a structural

arrangement that illicits an increase in the FRET response.

It was decided that reporting the Kd is not appropriate for fitting the nickel titration data,

particularly given that the point of saturation is not well defined as it is with zinc. However,

a line of best fit that represents fitting the nickel response to the Hill equation is presented.

Figure 19. The FRET response upon addition of micromolar nickel amounts to ZapCY2 (left). Close-

up on 1-3 µM added nickel (right). A FRET response lower than Rmin is seen up until the addition of

around 2 equivalents of nickel. The FRET ratio also increases beyond 2 equivalents of added nickel

without apparent saturation. N=5

33

3.3.2 Mutant sensors

The response profile of C2D with nickel was almost identical to ZapCY2, except

for a slight decrease in the dynamic range (Figure 20). The decrease in FRET ratio was

also apparent, with the lowest point occurring at around 1.75 instead of 1 equivalent of

added nickel.

The C2H mutant exhibited a slight increase in the dynamic range for nickel compared to

C2D and ZapCY2. The sensor again presents a decrease in FRET upon addition of low

micromolar amount of nickel, with the greatest decrease occurring at 1 equivalent of nickel

similar to ZapCY2 (Figure 21).

Figure 20. The FRET response upon nickel addition to the mutant sensor C2D (left). A close-up of the

1-3 μM added nickel range (right) reveals a dip in FRET below the Rmin as seen with ZapCY2. The

greatest drop in FRET ratio occurs at around 1.75 instead of 1 equivalent of nickel. N=3

Figure 21. The FRET response upon nickel addition to the mutant sensor C2H (left). Close-up of the 1-3

μM added nickel range (right). The sensor displayed a dynamic range for nickel that is very comparable

to ZapCY2 and C2D. A dip in FRET below the Rmin was also seen as with ZapCY2. N=3

34

The WF-H mutant exhibited an identical response to nickel as the three previous variants,

except for a slightly larger response at 20 µM added nickel (Figure 22).

Looking at the combined response to zinc and nickel addition to the sensor variants, it

can be stated that while the mutations targeting the cysteines have reduced the affinity for

zinc, the variants did not have an altered response profile to nickel in terms of the

affinity, the increase in FRET ratio for a given nickel concentration, nor the

unexplainable initial dip in FRET ratio.

Figure 22. The FRET response upon nickel addition to the mutant sensor WF-H (left). Close-up of the

1-3 μM added nickel range (right). The sensor displayed a dynamic range for nickel that is very

comparable to ZapCY2, C2D and C2H. A dip in FRET below the Rmin was also seen as with all the other

variants. N=3

35

3.4 High metal concentrations

In an effort to find the metal concentrations at which the sensor’s FRET response

ceases to change, a titration was performed by using concentrations at the millimolar level,

which represents 5000-fold more than the sensor concentration utilized (1 μM). The FRET

response was found to increase, albeit less steeply than at lower concentrations (Figure

23). Zinc concentrations higher than 5 mM resulted in a continuous increase in FRET ratio,

until the apparent precipitation of the protein in the solution resulted in a sharp drop in

FRET ratio (Data not shown).

Figure 23. Titration of ZapCY2 with millimolar zinc and nickel. These metal concentrations produced

a steady increase in the FRET ratio without apparent saturation. Increase in FRET ratio (top) N=2,

YFP/acceptor fluorescence (bottom left), CFP/donor fluorescence (bottom right). N=1

36

Zinc addition led to a well-defined increase in YFP fluorescence and decrease in CFP

fluorescence, which produced an over-all increase in FRET ratio. The FRET response with

nickel addition however was accompanied with an expected decrease in CFP fluorescence

but almost no increase of YFP fluorescence intensity. The explanation as to why that

occurred is unknown.

Similar to the original sensor, the addition of close to 5000-fold more metal than protein

still does not induce saturation in all of C2D, C2H, and WF-H, giving an identical response

to ZapCY2 (Data not shown).

The most likely explanation for the increase in FRET ratios upon the addition of very high

metal concentrations is sensor oligomerization. Excess zinc ions cause protein

precipitation,45 and divalent cations are known to result in the “salting-out” effect. With

the “salting-out” effect, the interactions between solvent molecules and cations are

stronger than those between solvent and protein molecules. This results in hydrophobic

protein-protein interactions becoming the dominant attractive force, leading to protein

oligomerization and eventually complete aggregation.46-47 As previously mentioned,

protein precipitation was visible when zinc concentrations exceeded 5 mM

(See section 3.4).

This phenomenon would not pose a problem in terms of sensor use in vivo because these

metal concentrations greatly surpass those present in living systems. Nonetheless, it does

make in vitro analysis difficult in terms of confidently calculating Kd’s because the

threshold metal concentration at which such event starts has not been empirically defined.

37

3.5 Size exclusion chromatography on ZapCY2

To test if oligomerization is indeed induced in the presence of metal, purified

ZapCY2 was analyzed by size exclusion chromatography. This experiment was performed

with and without 100 μM zinc in the running buffer. Recall that ZapCY2 has a molecular

weight of 62.7 kDa.

670 158 44 17 1.35 kDa

670 158 44 17 1.35 kDa

No added zinc

100 μM zinc

Figure 25. Size exclusion chromatography on purified ZapCY2 in the presence of 100 μM

zinc revealed monomer, dimer, and trimer populations.

Figure 24. Size exclusion chromatography of purified ZapCY2 revealed a monomer and

small dimer peak as observed by SDS-PAGE.

Elu

ent

abso

rban

ce (

AU

) E

luen

t ab

sorb

ance

(A

U)

Fraction number

38

Size exclusion chromatography in the presence of zinc did indeed confirm oligomerization

of the sensor, because larger peaks corresponding to dimers and trimers were present.

These results also reflect the stability of these complexes.

3.6 Response to magnesium

The ZapCY2 sensor displayed a modest response to magnesium addition, with an

increase of around 0.05 FRET units upon the addition of 5 mM MgCl2 (Figure 26).

Compare this with an average increase of around 7 FRET ratio units for zinc, and 5.5 units

for nickel (Figure 23).

Figure 26. Magnesium addition to ZapCY2 yielded a FRET increase that is around 100-fold lower than

the change observed upon addition of 5 mM zinc and nickel (0.05 FRET units compared to 7 and 5.5

units for zinc and nickel respectively). N=2

39

4. Conclusion and Future Directions

A total of 3 modifications have been made to ZapCY2 to produce 3 independent

sensor variants: C2D, C2H, and WF-H. The dissociation constant for zinc binding to the 4

variants was measured to be 0.54 ± 0.04 nM for ZapCY2, and 6.0 ± 0.5, 7.6 ± 0.5, 15 ± 9

nM respectively for the mutants, with only WF-H exhibiting a response that might be

characteristic of a two binding-site model. It can be stated that eliminating the cysteines

have in fact slightly reduced the zinc affinity by 10 to 15-fold, and eliminating Trp240 and

Phe282 to replace them with histidines reduced the affinity by around 30-fold.

The response to nickel was almost identical across all 4 variants, showing no improvement

in the affinity. More importantly, nickel addition resulted in FRET ratios below the Rmin

(FRET ratio in the absence of metal) for all variants up until the addition of around 2

equivalents of nickel. The greatest drop in FRET ratio coincided with around 1 equivalent

of added nickel, and was also reflected in the YFP/acceptor’s fluorescence intensity.

The increase in the FRET response upon addition of more than 2 equivalent of nickel might

be the product of non-specific binding to other sites in the sensor. This possibility was

confirmed by several experiments. The FRET response continued to increase and did not

saturate upon addition of millimolar amounts of zinc or nickel (nearly 5000-fold). Size

exclusion chromatography on ZapCY2 in the presence of 100 μM zinc revealed the

formation of dimer and trimer complexes.

The initial decrease in FRET ratio, followed by the hard to define non-specific response

did not allow for confident Kd calculation for nickel. Future work must aim to quantify the

impact of non-specific metal binding for greater confidence in the in vitro characterization.

40

Due to the apparent difficulty of improving ZapCY2’s response and affinity towards

nickel, alternative sensor systems could be explored. One such system is CLY9-2His

designed by Merkx et. al.31 This system uses a very similar FRET pair: eCFP, and eYFP

but in this case the linker is a flexible sequence of 9 GGSGGS repeats. Each chromophore

is additionally fixed to a hexahistidine tag. The two tags can sandwich a metal ion to result

in improved FRET. Although originally designed for zinc, this sensor has displayed a

response that is 20-fold weaker for nickel compared to zinc, giving a Kd of 0.88 ± 0.07 μM

for nickel, compared to 47 ± 4 nM for zinc. Although this affinity is deemed modest, there

is however a smaller difference in affinity between zinc and nickel compared to the current

system. This might make the modification process more likely to have a noticeable impact

on the affinity for nickel compared to the current system, where the difference in affinity

is not less than 2000-fold (compare Kd’s of around 500 pM for zinc and not less than 1 μM

for nickel).

41

5. Appendix

5.1 Calculating half-life of binding

Kd=koff/kon and half life= ln2/koff

For zinc, Kd is around 8 x 10-12 M. kon is assumed to be the diffusion limit of 108 s-1 M-1.

This gives a koff of 8 x 10-4 s-1, which gives a half-life of 14 min, so that 10 X the half life

is 140 minutes.

For nickel, the Kd is speculated to be around 1 x 10-6 M, making the half life around 7

milliseconds, so that 10 X the half life is 70 milliseconds.

5.2 PCR mutagenesis primers

MUTANT PRIMER NAME PRIMER SEQUENCE 5’=>3’

C2D C243D Fwd phosCAAATGGAAAGAAGATCCTGAGTCTTGTA

Rev phosTGTTTTAAGTCATTGTTTTTATGCATGCG

C280D Fwd phosATTGGGAGGACGATGATTTCCTTG

Rev phosTATGAGCTAATGGTTCCATAGGGTGTTT

C2H C243H

Fwd phosACAAATGGAAAGAACATCCTGAGTCTTGTAGC

Rev phosGTTTTAAGTCATTGTTTTTATGCATGCGGG

C280H Fwd phosAATTGGGAGGACCATGATTTCCTTGG

Rev phosATGAGCTAATGGTTCCATAGGGTGTTTG

WF-H W240H Fwd phosGACTTAAAACACAAACATAAAGAATGTCCTGAGTC

Rev phosATTGTTTTTATGCATGCGGGCG

F282H Fwd phosGGAGGACTGTGATCATCTTGGAGATGATAC

Rev phosCAATTATGAGCTAATGGTTCCATAGGGTG

Table 5. List of Phusion PCR mutagenesis primers

42

5.3 ESI-MS spectrum of PMB treated ZapCY2

Untreated ZapCY2

PMB treated ZapCY2

*Expected mass is 62711.82 Da

43

5.4 ZapCY2 sequence

ORIGIN

1 ATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGAGCTGGAC 60 M V S K G E E L F T G V V P I L V E L D

61 GGCGACGTAAACGGCCACAGGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATGCCACCTAC 120

G D V N G H R F S V S G E G E G D A T Y

121 GGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCCTGGCCCACC 180

G K L T L K F I C T T G K L P V P W P T

181 CTCGTGACCACCCTGACCTGGGGCGTGCAGTGCTTCAGCCGCTACCCCGACCACATGAAG 240

L V T T L T W G V Q C F S R Y P D H M K

241 CAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGCGTACCATCTTC 300

Q H D F F K S A M P E G Y V Q E R T I F

301 TTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAGGGCGACACCCTG 360

F K D D G N Y K T R A E V K F E G D T L

361 GTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAACATCCTGGGGCAC 420

V N R I E L K G I D F K E D G N I L G H

421 AAGCTGGAGTACAACTACATCAGCCACAACGTCTATATCACCGCCGACAAGCAGAAGAAC 480

K L E Y N Y I S H N V Y I T A D K Q K N

481 GGCATCAAGGCCCACTTCAAGATCCGCCACAACATCGAGGACGGCAGCGTGCAGCTCGCC 540

G I K A H F K I R H N I E D G S V Q L A

541 GACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCTGCTGCCCGACAACCAC 600

D H Y Q Q N T P I G D G P V L L P D N H

601 TACCTGAGCACCCAGTCCGCCCTGAGCAAAGACCCCAACGAGAAGCGCGATCACATGGTC 660

Y L S T Q S A L S K D P N E K R D H M V

661 CTGCTGGAGTTCGTGACCGCCGCCCGCATGCATAAAAACAATGACTTAAAACACAAATGG 720

L L E F V T A A R M H K N N D L K H K W

721 AAAGAATGTCCTGAGTCTTGTAGCTCACTATTTGACCTACAAAGACATCTTTTGAAGGAT 780

K E C P E S C S S L F D L Q R H L L K D

781 CATGTCTCTCAAGATTTCAAACACCCTATGGAACCATTAGCTCATAATTGGGAGGACTGT 840

H V S Q D F K H P M E P L A H N W E D C

841 GATTTCCTTGGAGATGATACATGTTCCATAGTGAACCATATTAATTGTCAACATGGTATC 900

D F L G D D T C S I V N H I N C Q H G I

901 GAGCTCATGGTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGAG 960

E L M V S K G E E L F T G V V P I L V E

961 CTGGACGGCGACGTAAACGGCCACAAGTTCAGCGTGTCCGGCGAGGGCGAGGGCGATGCC 1020

L D G D V N G H K F S V S G E G E G D A

1021 ACCTACGGCAAGCTGACCCTGAAGTTCATCTGCACCACCGGCAAGCTGCCCGTGCCCTGG 1080

T Y G K L T L K F I C T T G K L P V P W

1081 CCCACCCTCGTGACCACCTTCGGCTACGGCCTGATGTGCTTCGCCCGCTACCCCGACCAC 1140

P T L V T T F G Y G L M C F A R Y P D H

1141 ATGAAGCAGCACGACTTCTTCAAGTCCGCCATGCCCGAAGGCTACGTCCAGGAGCGCACC 1200

M K Q H D F F K S A M P E G Y V Q E R T

1201 ATCTTCTTCAAGGACGACGGCAACTACAAGACCCGCGCCGAGGTGAAGTTCGAGGGCGAC 1260

I F F K D D G N Y K T R A E V K F E G D

1261 ACCCTGGTGAACCGCATCGAGCTGAAGGGCATCGACTTCAAGGAGGACGGCAACATCCTG 1320

T L V N R I E L K G I D F K E D G N I L

1321 GGGCACAAGCTGGAGTACAACTACAACAGCCACAACGTCTATATCATGGCCGACAAGCAG 1380

G H K L E Y N Y N S H N V Y I M A D K Q

1381 AAGAACGGCATCAAGGTGAACTTCAAGATCCGCCACAACATCGAGGACGGCAGCGTGCAG 1440

K N G I K V N F K I R H N I E D G S V Q

1441 CTCGCCGACCACTACCAGCAGAACACCCCCATCGGCGACGGCCCCGTGCTGCTGCCCGAC 1500

L A D H Y Q Q N T P I G D G P V L L P D

1501 AACCACTACCTGAGCTACCAGTCCGCCCTGAGCAAAGACCCCAACGAGAAGCGCGATCAC 1560

N H Y L S Y Q S A L S K D P N E K R D H

1561 ATGGTCCTGCTGGAGTTCGTGACCGCCGCCGGGATCACTCTCGGCATGGACGAGCTCTAC 1620

M V L L E F V T A A G I T L G M D E L Y

1621 AAGTAA 1626

K *

eCFP and YFP are bolded and the linker region is underlined

*Does not include the N-terminal strep-tag

44

6. References

1.Li, Y.; Zamble, D. B., Nickel homeostasis and nickel regulation: An overview. Chemical

Reviews 2009, 109 (10), 4617-43.

2.Rowinska-Zyrek, M.; Zakrzewska-Czerwinska, J.; Zawilak-Pawlik, A.; Kozlowski, H.,

Ni2+ chemistry in pathogens - a possible target for eradication. Dalton Transactions 2014,

43 (24), 8976-8989.

3.Mulrooney, S. B.; Hausinger, R. P., Nickel uptake and utilization by microorganisms.

FEMS Microbiology Rreviews 2003, 27 (2-3), 239-61.

4.Vignais, P. M.; Billoud, B., Occurrence, classification, and biological function of

hydrogenases: An overview. Chemical Reviews 2007, 107 (10), 4206-72.

5.Can, M.; Armstrong, F. A.; Ragsdale, S. W., Structure, Function, and Mechanism of the

Nickel Metalloenzymes, CO Dehydrogenase, and Acetyl-CoA Synthase. Chemical

Reviews 2014, 114 (8), 4149-4174.

6.Shima, S.; Thauer, R. K., Methyl-coenzyme M reductase and the anaerobic oxidation of

methane in methanotrophic Archaea. Current Opinion in Microbiology 2005, 8 (6),643-8.

7.Zerner, B., Recent advances in the chemistry of an old enzyme, urease. Bioorganic

Chemistry 1991, 19 (1), 116-131.

8.Cabiscol, E.; Tamarit, J.; Ros, J., Oxidative stress in bacteria and protein damage by

reactive oxygen species. International Microbiology : The Official Journal of the Spanish

Society for Microbiology 2000, 3 (1), 3-8.

9.De Pina, K.; Desjardin, V.; Mandrand-Berthelot, M. A.; Giordano, G.; Wu, L. F.,

Isolation and characterization of the nikR gene encoding a nickel-responsive regulator in

Escherichia coli. Journal of Bacteriology 1999, 181 (2), 670-4.

10.Abraham, L. O.; Li, Y.; Zamble, D. B., The metal- and DNA-binding activities of

Helicobacter pylori NikR. Journal of Inorganic Biochemistry 2006, 100 (5-6), 1005-14.

11.Macomber, L.; Hausinger, R. P., Mechanisms of nickel toxicity in microorganisms.

Metallomics : Integrated Biometal Science 2011, 3 (11), 1153-62.

12.Imlay, J. A., Pathways of oxidative damage. Annual Review of Microbiology 2003, 57,

395-418.

13.Imlay, J. A., Cellular defenses against superoxide and hydrogen peroxide. Annual

Review of Biochemistry 2008, 77, 755-76.

14.Maret, W., Analyzing free zinc(II) ion concentrations in cell biology with fluorescent

chelating molecules. Metallomics : Integrated Biometal Science 2015, 7 (2), 202-11.

15.Chivers, P. T.; Sauer, R. T., NikR Repressor. Chemistry & Biology 2002, 9 (10), 1141-

1148.

16.Wang, S. C.; Dias, A. V.; Bloom, S. L.; Zamble, D. B., Selectivity of Metal Binding

and Metal-Induced Stability of Escherichia coli NikR. Biochemistry 2004, 43 (31), 10018-

10028.

17.Iwig, J. S.; Leitch, S.; Herbst, R. W.; Maroney, M. J.; Chivers, P. T., Ni(II) and Co(II)

sensing by Escherichia coli RcnR. Journal of the American Chemical Society 2008, 130

(24), 7592-606.

18.Dodani, S. C.; He, Q.; Chang, C. J., A Turn-On Fluorescent Sensor for Detecting Nickel

in Living Cells. Journal of the American Chemical Society 2009, 131 (50), 18020-1.

45

19.Jiang, L.-J.; Luo, Q.-H.; Wang, Z.-L.; Liu, D.-J.; Zhang, Z.; Hu, H.-W., A

dioxotetraamine fluorenyl ligand and its nickel(II) complex — crystal structure and

fluorescent sensing properties in aqueous solution. Polyhedron 2001, 20 (22–23), 2807-

2812.

20.Salins, L. L.; Goldsmith, E. S.; Ensor, M. C.; Daunert, S., A fluorescence-based sensing

system for the environmental monitoring of nickel using the nickel binding protein from

Escherichia coli. Analytical and Bioanalytical Chemistry 2002, 372 (1), 174-180.

21.Donadio, G.; Di Martino, R.; Oliva, R.; Petraccone, L.; Del Vecchio, P.; Di Luccia, B.;

Ricca, E.; Isticato, R.; Di Donato, A.; Notomista, E., A new peptide-based fluorescent

probe selective for zinc(ii) and copper(ii). Journal of Materials Chemistry B 2016, 4 (43),

6979-6988.

22.Pearce, D. A.; Walkup, G. K.; Imperiali, B., Peptidyl chemosensors incorporating a

FRET mechanism for detection of Ni(II). Bioorganic & Medicinal Chemistry Letters 1998,

8 (15), 1963-8.

23.Wang, B.; Hu, Y.; Su, Z., Synthesis and photophysical behaviors of a blue fluorescent

copolymer as chemosensor for protons and Ni2+ ion in aqueous solution. Reactive and

Functional Polymers 2008, 68 (7), 1137-1143.

24.Wang, B.-Y.; Liu, X.-Y.; Hu, Y.-L.; Su, Z.-X., Synthesis and photophysical behavior

of a water-soluble coumarin-bearing polymer for proton and Ni2+ ion sensing. Polymer

International 2009, 58 (6), 703-709.

25.Lei, Y.; Li, H.; Gao, W.; Liu, M.; Chen, J.; Ding, J.; Huang, X.; Wu, H., Highly

sensitive conjugated polymer fluorescent sensors based on benzochalcogendiazole for

nickel ions in real-time detection. Journal of Materials Chemistry C 2014, 2 (35), 7402-

7410.

26.Bajar, B. T.; Wang, E. S.; Zhang, S.; Lin, M. Z.; Chu, J., A Guide to Fluorescent Protein

FRET Pairs. Sensors (Basel, Switzerland) 2016, 16 (9).

27.Broussard, J. A.; Rappaz, B.; Webb, D. J.; Brown, C. M., Fluorescence resonance

energy transfer microscopy as demonstrated by measuring the activation of the

serine/threonine kinase Akt. Nature Protocols 2013, 8 (2), 265-281.

28.Carter, K. P.; Young, A. M.; Palmer, A. E., Fluorescent Sensors for Measuring Metal

Ions in Living Systems. Chemical Reviews 2014, 114 (8), 4564-4601.

29.Park, J. G.; Palmer, A. E., Properties of genetically-encoded indicators used for

cytosolic and organellar Ca(2+) measurements. Cold Spring Harbor Protocols 2015, 2015

(1), pdb.top066043-pdb.top066043.

30.Vinkenborg, J. L.; Koay, M. S.; Merkx, M., Fluorescent imaging of transition metal

homeostasis using genetically encoded sensors. Current Opinion in Chemical Biology

2010, 14 (2), 231-237.

31.Evers, T. H.; Appelhof, M. A.; Meijer, E. W.; Merkx, M., His-tags as Zn(II) binding

motifs in a protein-based fluorescent sensor. Protein Engineering, Design & Selection :

PEDS 2008, 21 (8), 529-36.

32.Qin, Y.; Dittmer, P. J.; Park, J. G.; Jansen, K. B.; Palmer, A. E., Measuring steady-state

and dynamic endoplasmic reticulum and Golgi Zn(2+) with genetically encoded sensors.

Proceedings of the National Academy of Sciences of the United States of America 2011,

108 (18), 7351-7356.

33.Qiao, W.; Mooney, M.; Bird, A. J.; Winge, D. R.; Eide, D. J., Zinc binding to a

regulatory zinc-sensing domain monitored in vivo by using FRET. Proceedings of the

National Academy of Sciences of the United States of America 2006, 103 (23), 8674-8679.

46

34.Wang, Z.; Feng, L. S.; Matskevich, V.; Venkataraman, K.; Parasuram, P.; Laity, J. H.,

Solution structure of a Zap1 zinc-responsive domain provides insights into

metalloregulatory transcriptional repression in Saccharomyces cerevisiae. Journal of

Molecular Biology 2006, 357 (4), 1167-83.

35.Boer, J. L.; Mulrooney, S. B.; Hausinger, R. P., Nickel-dependent metalloenzymes.

Archives of Biochemistry and Biophysics 2014, 544, 142-52.

36.Cubitt, A. B.; Heim, R.; Adams, S. R.; Boyd, A. E.; Gross, L. A.; Tsien, R. Y.,

Understanding, improving and using green fluorescent proteins. Trends in Biochemical

Sciences 1995, 20 (11), 448-55.

37.Rizzo, M. A.; Springer, G. H.; Granada, B.; Piston, D. W., An improved cyan

fluorescent protein variant useful for FRET. Nature Biotechnology 2004, 22 (4), 445-449.

38.Griesbeck, O.; Baird, G. S.; Campbell, R. E.; Zacharias, D. A.; Tsien, R. Y., Reducing

the environmental sensitivity of yellow fluorescent protein. Mechanism and applications.

The Journal of Biological Chemistry 2001, 276 (31), 29188-94.

39.Fahrni, C. J.; O'Halloran, T. V., Aqueous Coordination Chemistry of Quinoline-Based

Fluorescence Probes for the Biological Chemistry of Zinc. Journal of the American

Chemical Society 1999, 121 (49), 11448-11458.

40.Kocyla, A.; Pomorski, A.; Krezel, A., Molar absorption coefficients and stability

constants of metal complexes of 4-(2-pyridylazo)resorcinol (PAR): Revisiting common

chelating probe for the study of metalloproteins. Journal of Inorganic Biochemistry 2015,

152, 82-92.

41.Maret, W., Cobalt(II)-substituted class III alcohol and sorbitol dehydrogenases from

human liver. Biochemistry 1989, 28 (26), 9944-9949.

42.Atanassova, A.; Zamble, D. B., Escherichia coli HypA Is a Zinc Metalloprotein with a

Weak Affinity for Nickel. Journal of Bacteriology 2005, 187 (14), 4689-4697.

43.Krizek, B. A.; Berg, J. M., Complexes of zinc finger peptides with nickel(2+) and

iron(2+). Inorganic Chemistry 1992, 31 (13), 2984-2986.

44.Bal, W.; Schwerdtle, T.; Hartwig, A., Mechanism of nickel assault on the zinc finger

of DNA repair protein XPA. Chemical Research in Toxicology 2003, 16 (2), 242-8.

45.Cohn, E. J.; Gurd, F. R. N.; Surgenor, D. M.; Barnes, B. A.; Brown, R. K.; Derouaux,

G.; Gillespie, J. M.; Kahnt, F. W.; Lever, W. F.; Liu, C. H.; Mittelman, D.; Mouton, R. F.;

Schmid, K.; Uroma, E., A System for the Separation of the Components of Human Blood:

Quantitative Procedures for the Separation of the Protein Components of Human

Plasma1a,b,c. Journal of the American Chemical Society 1950, 72 (1), 465-474.

46.Arakawa, T.; Timasheff, S. N., Mechanism of protein salting in and salting out by

divalent cation salts: Balance between hydration and salt binding. Biochemistry 1984, 23

(25), 5912-5923.

47.Grover, P. K.; Ryall, R. L., Critical Appraisal of Salting-Out and Its Implications for

Chemical and Biological Sciences. Chemical Reviews 2005, 105 (1), 1-10.