3.1 Experimental design
3.1.1 Copper isotopes
Chelation and transchelation experiments with Cu2+ and adamanzanes were performed both with 64Cu
(detected by its γ-radiation), and with stable copper, where changes in the UV-visible spectra were measured.
Experiments with stable isotopes have the advantage of being simple and inexpensive. Perhaps more importantly, it is possible to follow chelation in real time with a spectrophotometer. The disadvantages of spectrophotometric measurements are that the concentration range is small - for the absorbance to be linear related to concentration, and that chelation to protein bound adamanzanes cannot be measured due to protein absorbance. Given that the absorption peaks of copper-adamanzane complexes in the UV and visual area differ in magnitude by a factor 20, the UV peak was chosen for chelation experiments to enable
concentrations in the submillimolar range; the peak in the visual field was used as control. In competition experiments, where a high concentration is desirable, the visual peak was chosen for measurement.
The 64Cu experiments have almost no limitations in concentration range, but some kind of separation step is required to distinguish between bound and unbound 64Cu species. In such experiments, only the chelation result after a separation can be measured. Furthermore, the duration of the assays is limited to a few days by the half-life of 64Cu (12.7 h).
In the stable isotope experiments, where chelation speed was important, near stoichiometric ratios of copper and adamanzane were used. Conversely, in the 64Cu experiments, where a high LE is the primary objective, a large surplus of adamanzanes was used.
3.1.2 Irradiation setup
Samples that were irradiated by UV light, were irradiated either in a HP8435 diode array spectrophotometer (figure 3.1), or in a closed UV lamp from a LKB / Pharmacia 8300 UVicord II UV Monitor (figure 3.2). In the spectrophotometer, the sample cuvette could be thermostated. Thermostating in the closed UV lamp where done by placing the entire lamp in a heating cupboard. The closed UV lamp consisted of a metal box with a UV fluorescent tube in one side, two cuvette holding clips in the middle and a detection system in the other side, which was not used. The fluorescent tube and the sample cuvette were originally separated by a slit that was removed in later experiments for more intense irradiation.
Figure 3.1: Irradiation and measurement in spectrophotometer. The light from the UV lamp is not red but is drawn that way for visualization. Likewise, the spectrum is a UV spectrum, not a visible spectrum (unless the samples were irradiated by visible light as well) and the intensity is greatly enhanced in the figure.
Figure 3.2: Irradiation in closed UV lamp. The actual cuvettes were closed in the top by a stopper, and separated by 1 cm. Although the figure shows a light beam irradiating only a part of the sample, the entire sample was irradiated in the actual setup. The slit was an opening in a metal plate, impenetrable by UV light.
Filter UV lamp
Detector system, not to scale Sample
Filter
Sample UV lamp
Filter
Sample UV lamp
Slit
3.2 Synthesis of adamanzanes
3.2.1 Caution
Mechanical handling of perchlorate salts especially of organic molecules can constitute an explosion risk.
However, we have never experienced any explosions with the presented compounds.
3.2.2 Ligand synthesis
[24.31]adz · 3HBr, [35]adz · 4HBr and pure [35]adz were prepared as previously described [29,208,209].
3.2.3 Functionalization of [24.31]adz
Modification of previously described method [106]:
N,N’-(CH2COOCH 2CH3)2[24.31]adz•HBr: 7.28 g (H3[24.31]adz)Br3 (16 mmol), 5.09 g anhydrous Na2CO3
(48 mmol) and 8 g molecular sieve (0.3 nm) is dissolved in 160 mL dry acetonitrile. The solution is refluxed over night while being protected from moisture by a drying tube filled with CaCl2. After cooling to room temperature, 3.56 mL ethyl bromoacetate is added drop wise. This solution is refluxed for 24 hours while being protected from moisture by a drying tube filled with CaCl2. After cooling to room temperature the reaction mixture is filtered and the filtrate dried by rotatory evaporation.
Yield: 96 %, 1H NMR (D2O), δ(ppm): 4.20 (quartet, 4H), 3.64 (singlet, 4H), 3.14 (triplet, 4H), 3.04 (multiplet, 8H), 2.90 (triplet, 8H), 1.89 (multiplet, 2H), 1.27 (triplet, 6H). 13C NMR (D2O), δ(ppm): 174.21 (C=O), 61.69 (CH2-O), 55.75, 54.15, 53.08, 50.43 (CH2-N), 20.41 (CH2-CH2-CH2), 13.40 (CH2-CH2).
N,N’-(CH2CO2H)2[24.31]adz•HBr: 3.29 g N,N’-(CH2COOCH 2CH3)2[24.31]adz•HBr (7.1 mmol) is dissolved in 118 mL 6 M HCl (710 mmol). The solution is refluxed for 24 hours and after cooling to room temperature filtered. The filtrate is dried by rotatory evaporation. Yield: 99 %,
electrospray ionization
mass spectrometry (ESI MS) m/z: 329.15 (M + H+), expected: 329.22.3.3 Chelation of stable copper, by heat and by light
3.3.1 Synthesis of [Cu(adz)]2+ in mg scale
[Cu([24.31]adz)Br]ClO4 was prepared as previously described [211].
In short: CuSO4 and adamanzane were mixed in an ammonium/ammonia buffer and refluxed. Addition of HBr and HClO4 yielded blue crystals, which were washed in ice cold water and ethanol before drying.
Solutions of [Cu([24.31]adz)]2+ were prepared by mixing:
x mL 20.0 mM [Cu([24.31]adz)Br]ClO4 with
0.2088⋅x mL 95.8 mM AgClO4 and removal of AgBr by centrifugation.
[Cu([35]adz)Br]Br was prepared as previously described [29].
Solutions of [Cu([35]adz)](ClO4)2 were prepared by mixing:
x mL 20.0 mM [Cu([35]adz)Br]Br with
0.4175⋅x mL 95.8 mM AgClO4 and removal of AgBr by centrifugation.
3.3.2 Real time measurement of Cu2+ chelation by adamanzanes
The chelation rate was measured spectrophotometrically by exploiting the difference in UV-visible absorption between Cu2+ + [24.31]adz and [Cu([24.31]adz)]2+. [Cu([24.31]adz)]2+ has absorption maxima at 285 nm, 600 nm and 1050 nm, when water is coordinated as the fifth ligand (figure 3.3). The same principle can be used for [35]adz with maxima at 291 nm, 658 nm and 911 nm, and for other adamanzanes.
0 1500 3000 4500 6000
200 250 300 350 400
nm
M-1 cm-1 A
B
0 100 200 300
400 500 600 700 800 900 1000 1100
nm
M-1 cm-1
Figure 3.3: UV-visible molar absorptivity of: A: [Cu([24.31]adz)H2O]2+ and B: [Cu([35]adz)H2O]2+.
UV filters, including 50 mM solutions of anions
0 0.5 1 1.5 2
175 200 225 250 275 300 325 350
nm
Absorption
Nitrate Cuvette side Object glas Water Phosphate Perchlorate Chloride Sulphate Bromide
Figure 3.4: Spectra of UV filters and other anions employed in chelation of Cu2+ to adamanzanes in this thesis. All anion filter solutions were of sodium salts.
Cu(ClO4)2, a solution of the perchlorate salt of the ligand and either salt solutions or distilled water was mixed in a 1.000 cm 100 µL cuvette to a final volume of 125 µL. The ligands used in this way were [35]adz, [24.31]adz, N,N’-(CH2CO2H)2[24.31]adz and N,N’-(CH2COC2CH3)2[24.31]adz. Chelation was induced and the UV-spectrum measured by a HP8435 diode array spectrophotometer. Temperature dependence was
measured using a thermostated cuvette holder. Aqueous salt solutions in a 0.200 cm quartz cuvette placed between the light source and the sample were used as filters to block high energy UV light (figure 3.4). The data was analysed using UV Visible ChemStation software. The shutter of the spectrophotometer is open for approximately 1 second more than the measurement time, so if a sample is measured for 1 second every 4 seconds, it is irradiated half of the time.
A Shimadzu UV-3600 UV-Vis-NIR Spectrophotometer was employed for measuring UV spectra with minimal UV irradiation of the samples.
3.4 Kinetic stability
3.4.1 Pilot studies
Solutions of 5.0 mM [Cu([24.31]adz)]2+, [Cu([(2.3)2.21]adz)]2+ or [Cu([35]adz)]2+ were mixed with 0.10 M NanEDTA, pH 8 or 10. Then incubated at 60 °C and the visible absorption spectra measured over the following days.
Afterwards, 5.0 mM [Cu([24.31]adz)](ClO4)2 and 5.0 mM [Cu([35]adz)](ClO4)2 were incubated with 2.0 mM EDTA in 1x PBS (phosphate buffered saline) at 60 °C and pH 5.5.
10x PBS: 80.0 g NaCl, 2.0 g KCl, 14.4 g Na2HPO4 · 2H2O and 2.4 g KH2PO4 in 1.0 L distilled water.
1x PBS: 10x PBS diluted 10 fold in distilled water.
3.4.2 5th ligand stability
[Cu([35]adz)](ClO4)2 and a complexane (
N-(hydroxyethyl)-ethylenediaminetriacetic acid
= HEDTA or nitrilotriacetic acid =
NTA) were mixed to final concentrations of: [Cu([35]adz)]2+: 4.0 mMHEDTA/NTA: {0; 0.2; 0.4; 1.0; 2.0; 4.0; 10; 20; 40; 100}mM
The [Cu([35]adz)]2+ solution and HEDTA/NTA solutions were adjusted to pH 7.5 at room temperature before mixing to final concentrations and measured immediately afterwards by a spectrophotometer with the
deuterium lamp turned off to prevent any transchelation induced by UV light.
The resulting spectra were adjusted to enforce an isosbestic point at 787 nm and the changes in absorption at 675 nm and 910 nm were calculated at each complexane concentration. The values were then plotted as
( )
⎟⎟⎠⎜⎜ ⎞
⎝
⎛
Δ Δ
− Δ Δ
max max
Abs Abs 1
Abs
Log Abs as a function of the complexane concentration, a Hill plot. The values lower than -1, were omitted as the uncertainty for those in a Hill plot are too big. The until then estimated ΔAbsmax was then fitted to ensure a slope of exactly 1, as there is no cooperativity in the 5th ligand binding.
3.4.3 Transchelation to competing chelators
[Cu([24.31]adz)]2+ or [Cu([24.31]adz)]2+ was mixed with HEDTA or NTA and pH adjusted to 5.5 at 25 °C before the volume was adjusted with water to 1.0 mL and a final concentration of 5.0 mM Cu-adz and 5.0 mM complexane. The solutions were incubated at 50 °C and samples of 120 µL were taken out for
measurement and discarded afterwards. When the 7th sample had been measured, the remaining 160 µL were used the measure the pH at 25 °C.
3.5 Binding of 1b to Cytochrome c. and to β-lactoglobulin
N
N N
N
O
O
OH
OH N C N
NH Cl EDC
+
N,N'-(CH2CO2H)2[24.31]adz
O O
N OH
N C NH
NH Cl (E)
N O
O
NHS +
O O N O
HN C NH
N O NH
O + N
C NH
NH Cl (E)
N O
O
H2N OH hydroxylamine
N
N N
N
O
O
HN
Cc
O H2N
Cc
O O N O
O
+ +
O O N HO
N
N N
N
O
O
HN
Cc
OH +
O
O N N
N N
O N O O
O N
N
N N
N
O
O
HN
Cc
O O
O N
+ H2N
OH + H2O
Figure 3.5: Coupling of N,N’-(CH2CO2H)2[24.31]adz to Cc with EDC and NHS. The first 3 reaction lines take place simultaneously. The reactions stop when hydroxylamine is added (the last line).
3.5.1 Cytochrome c.
4 mg EDC, 6 mg NHS and 5 mL 20 mM HEPES pH 7.3 were mixed.
4.0 mL was removed, added 8.0 mg N,N’-(CH2CO2H)2[24.31]adz and left for 15 min at room temperature.
Meantime 48 g Cytochrome c (Cc) was dissolved in 4.0 mL 20 mM HEPES pH 7.3.
After the 15 min the two solutions were mixed and left to react for 2.5 h (figure 3.5).
Then 24 µL 50 % hydroxylamine was added to stop the reaction.
15 min later a spatula of sodium dithionite was added to reduce the Cc.
The different reaction products were separated on a UNO-S12 cation exchange column prepared with 20 mM sodium phosphate buffer pH 7.0 and eluded with the same buffer including 0.25 M NaCl in fractions of 5.0 mL.
Fractions of interest were desalinated by ultrafiltration and the concentration measured by the absorption at 409 nm.
3.5.2 β-lactoglobulin
71.5 g β-lactoglobulin was coupled to N,N’-(CH2CO2H)2[24.31]adz in similar manner to the Cc coupling but no sodium dithionite was added.
3.6 Radiochemistry
64Cu for all experiments was produced locally by the 64Ni (p,n) 64Cu process, using a 13 MeV (energy degraded) proton beam from a GE PETtrace cyclotron with beamline. To minimize foreign metal
contamination, the 64Ni target was plated onto high density graphite backings. Irradiations were performed at 20 uA for 1-2 hours and yielded 1-1.5 GBq 64Cu EOB. Target dissolution and the separation of carrier-free
64Cu was performed according to the method by Hou et al. [95].
64CuCl2 was mixed with Cu(ClO4)2 to a 6 MBq/mL solution of 3.0 µM (64)Cu(ClO4)2. 400 µL 30 µM CC with or without [24.31]adz conjugated was mixed with 200µL (64)Cu(ClO4)2. The samples were then incubated for 4 h at 45 °C where the first 2 hours were either in a 0.200 cm quartz cuvette in a UV light source (LKB UVICORD type 4701 A) or in darkness. The last 2 hours were always in darkness. The UV light was filtered through a 0.200 cm quartz cuvette containing a 50 mM aqueous NaBr solution (cut off at 220 nm) to remove high energy UV light.
The Cc (Cytochrome c) solutions were then filtered in a Centricon tube (10 kd). The filtrates were removed and the activity measured. The Centricon tubes were then refilled with 2 mL 5.0 mM EDTA (pH 7.5) and centrifuged for 30 min. removing most of the EDTA. The Centricon tubes were then refilled with 1 mL EDTA and centrifuged for 30 min 6 times. The last time the tubes were refilled, they were left for 14 h before centrifugation, which was continued to near dryness. The protein was resuspended in 1 mL water.
Activity in filter, protein and washing fractions was measured by a 3x3 “ NaI scintillation well crystal coupled to conventional HV, amplifier and single channel scaler equipment. Counting was done in an energy window covering the range from 300 keV and up to at least 1200 keV (to include the possible sum peak at 1022 keV). Measurements were deadtime and background corrected, and errors from reduction in activity due to decay was eliminated by compared to 64Cu counting standards taken from the same batch. The activity of the samples was calculated as percentage of the batch standard. If a sample was added e.g. 5 times the activity of the standard and the sample afterwards was split up in subsamples, the sum of all the subsample percentages should equal 500 % if there was no loss. A large deviation between added and measured activity
in percentage of the standard would render the sample measurements invalid. Activity was then expressed as percentage of the total administered activity to each labeling experiment.
With the sample size used (less than 1.8 mL total volume in 2 mL glass vials) no sample geometry effects were found necessary to correct for.
3.7 Mass spectrometry
Aqueous solutions of samples were mixed with methanol and acetic acid to a final concentration of 25 % and 0.1 % respectively. All spectra were recorded at room temperature on an ESI ion trap mass spectrometer (Esquire 4000, Bruker) and analysed by HYSTAR version 3.2 software.