Tailored Mobility in a Zeolite Imidazolate Framework (ZIF) Antibody Conjugate

Abstract Zeolitic imidazolate framework (ZIF) hybrid fluorescent nanoparticles and ZIF antibody conjugates have been synthesized, characterized, and employed in lateral‐flow immunoassay (LFIA). The bright fluorescence of the conjugates and the possibility to tailor their mobility gives a huge potential for diagnostic assays. An enzyme‐linked immunosorbent assay (ELISA) with horseradish peroxidase (HRP) as label, proved the integrity, stability, and dispersibility of the antibody conjugates, LC‐MS/MS provided evidence that a covalent link was established between these metal‐organic frameworks and lysine residues in IgG antibodies.


Reagents for HPLC
All HPLC solvents used were of high purity HPLC grade. Water was acquired from a Milli-Q® Direct Water Purification System [18.2 MΩ•cm (25°C)]. Methanol for HPLC was purchased from CHEMSOLUTE as HPLC grade and TFA was purchased from Sigma Aldrich as Reagent-Plus 99 %.
Water was obtained from an ultrapure water purification system (Elga).

Reagents for buffers
All reagents for buffer preparation, and casein, were purchased from Merck.
Mouse anti-CBZ IgG (AK2512/01 ), 1.75 mg/mL, in PBS buffer, ordering code BAM-mab 01 (CBZ) was obtained from BAM. [1] 1.7 Solvent for NMR d6-DMSO (CAS: 2206-27-1, 99.9%) in 0.75 ml ampules was purchased from Sigma-Aldrich 2 Preparation of buffers 2.1 Preparation of the phosphate buffer for HPLC (pH 8.8) 1.75 g of Na 2 HPO 4 (9.83 mmol) and 16.29 mg of NaH 2 PO 4 (0.12 mmol) were weighed using a 5-figure balance, transferred to a 1 L HPLC glass solvent bottle and added 1 L of water obtained from a Milli-Q® Direct Water Purification System [18.2 MΩ•cm (25°C)], measuring the added volume with a 1 L graduated glass cylinder. After complete dissolution, the buffer was filtered through a 0.45 μm nylon membrane (47 mm diameter) using a 1 L solvent vacuum microfiltration apparatus to remove any undissolved crystal. The pH was determined to be 8.60 using a pH meter, and the buffer was ready to use.

Preparation of the buffers for immunoanalysis
2.2.1 Borate buffer 9.64 g (48 mmol) of di-sodium tetraborate decahydrate and 8.29 g (142 mmol) of sodium chloride were weighed and transferred to a 1 L HPLC glass solvent bottle. Using a 1L graduated glass cylinder 1 L of ultrapure water, obtained from a Milli-Q® Direct Water Purification System [18.2 MΩ•cm (25°C)], was added to the bottle. After full dissolution 1 mL of a sodium azide solution (10% w/v in water) was added and mixed. The final pH of the buffer was measured and determined to be pH 9.05.

Phosphate buffered saline (PBS)
1.56 g (10 mmol) of NaH 2 PO 4 ·2H 2 O, 12.46 g of Na 2 HPO 4 ·2H 2 O (70 mmol) and 8.47 g of NaCl (142 mmol) were weighed, transferred to a 1 L HPLC glass solvent bottle, 1 L of ultrapure water was added and the pH determined to be pH 7.60. 1.21 g of C 4 H 11 NO 3 (TRIS, 10 mmol) and 8.77 g of NaCl (150 mmol) were weighed and, as described in SI-2.2.2, used to prepare 1 L of buffer, pH 8.50.

Citrate buffer
47.10 g of sodium dihydrogen citrate (220 mmol) were weighed and, as described in SI-2.2.2, used to prepare 1 L of buffer, pH 4.00. For reasons of stability, the buffer was kept in the fridge at 4°C.

Wash buffer
6.124 g of KH 2 PO 4 (45 mmol), 65.32 g of K 2 HPO 4 (375 mmol), 0.225 g of C 6 H 7 KO 2 (potassium sorbate, 1.5 mmol) were weighed and, as described in SI-2.2.2, used to prepare 1 L of buffer. 30 ml of a 0.05% (v/v) Tween™ 20 solution using a 100 mL graduated glass cylinder was added. The final pH of the buffer was measured and determined to be pH 7.60.

TMB solution
102.93 mg of tetrabutylammonium borohydride (8 mmol) and 480.7 mg of 3,3′,5,5′tetramethylbenzidine (TMB, 40 mmol) were weighed using a 5-figure analytical balance, transferred to a 200 mL new glass bottle and 50 mL of Milli-Q water added. For stability, the buffer was kept in the fridge at 4°C.

FTIR:
Fourier-transform infrared spectra were recorded on a Nicolet FT-IR NEXUS instrument with a Golden Gate-Diamond-ATR Unit and a DTGS KBr detector. For each measurement 32 sample scans and 32 background scans were recorded. The spectral range was from 200 cm -1 to 4000 cm -1 with a resolution of 4 cm -1 .

TEM:
Transmission electron microscopy (TEM) images were obtained in a Talos F200S Microscope (Thermo Fisher Scientific) by using a 200 kV microscopy technique in which a beam of electrons is transmitted through a specimen to form an image. A Ceta 16M camera (TEM mode) and a HAADF detector (STEM mode) were used to capture the images.
Additionally, a side-entry retractable Energy Dispersive X-ray Spectroscopy (EDS) detector were also used in the STEM mode to collect detailed elemental information.
The specimens were prepared by dropping sample solutions onto a 3 mm copper grid (Lacey Carbon, 400 mesh) and leaving them to air-dry at room temperature. S 8

NMR:
Nuclear magnetic resonance measurements were performed on a 400 MHz NMR spectrometer (Mercury-400 BB) at 29.0 °C, with a relaxation delay of 2.000 sec, a pulse of 45.0°, an acquisition time of 2.274 sec and 16 repetitions. Solvent was deuterated DMSO.

Bragg-Brentano diffractometry:
Powder X-ray diffractometry (PXRD) patterns were collected with Cu Kα radiation (λ = 1.50406 nm) on a D8 Advanced diffractometer (Bruker AXS, Germany) equipped with a LYNXEYE detector. Samples were measured in reflection geometry in a 2Ɵ range from 4° to 50° with a step size of 0.009° with spinning setup.

Fluorescence spectrometry:
Fluorescence spectra were obtained on an all-in-one microplate reader Synergy H1 with dispenser (BioTek). The software used was Gen5. The spectra was acquired from 450 to 600 nm, with a step of 2 nm, measurement time of 10 ms for the particle characterization, and of 50 ms for the fluorescence ELISA, with delay between measurements of 100 ms. The excitation wavelength was 335 nm. The samples were handled in non-binding or high-binding polypropylene 96-well black microtiter plates from Greiner Bio-One.

HPLC:
High performance liquid chromatography analysis of the Z8P samples were performed using a modular Agilent 1100 Series HPLC system composed of a high pressure binary pump, auto-sampler with injector programming capabilities, Peltier type column oven with 6 µL heat exchanger and a Diode Array Detector with a micro flow cell (1.6uL, 6mm pathlength) to reduce peak dispersion when using short columns as in this case. The flow-path was connected using 0.12 mm ID stainless steel tubing to minimize peak dispersion. ChemStation software was used to process the HPLC data.
3.6.1 HPLC method HPLC column was Excel C18-PFP with 2 µm particle size (4.6 mm internal diameter × 50 mm length) manufactured by ACE. The C18-PFP stationary phase was prepared by covalently bonding a pentafluorophenyl octadecylsilane to pure silica. The HPLC column is manufactured to be stable up to pH 10 and 1000 bar back pressure. The chosen wavelength for 2-methylimidazole was 210 nm (8 nm bandwidth), for imidazolate-2-carboxaldehyde 280 nm (8 nm bandwidth) and for 1-aminopyrene 338 nm (8 nm bandwidth), all having a reference of 550 nm (100 nm bandwidth, Fig. S11).

ICP-OES:
For inductively coupled plasma optical emission spectrometry an Agilent 5110 VDV ICP-OES instrument was employed (Agilent Technologies, USA). The emission line of zinc at wavelength 202.548 nm at radial mode was used. Detailed operating conditions are listed in Table  S8.

Absorbance spectrometry:
Absorbance, with measurements in microtiter plates usually designated as optical density (OD), was measured on a SpectraMax Plus384 microplate reader (Molecular Devices, Ismaning, Germany) at 450 nm, referenced to 620 nm. All calibration standard solutions were determined in triplicate. SoftMax®Pro Software (v.5.3, Molecular Devices) was used for data acquisition
Positive and negative electrospray ionization (ESI+/-) was carried out at 320 °C using capillary voltages of 5500 V and 4500 V, respectively. Mass spectrometric detection was performed in full-scan mode (m/z 50-600) at a scan rate of 4 s -1 using data-dependent MS/MS acquisition (DDA Top-4). For data analysis, raw data files were converted to mzML format and imported into an R workspace. Peak pre-processing was performed using package xcms [2] . For MS/MS fragment substructure annotation, competitive fragmentation modelling was used [3] (CFM-ID).

DLS measurements:
The dynamic light scattering measurements were performed on a Zetasizer Nano ZS, Malvern Instruments, UK) equipped with a He-Ne laser (λ = 632.8 nm). Scattering data were recorded in backscattering mode at a scattering angle of 2θ = 173°, which corresponded to a scattering vector of q = 4π n/λ sinθ (0.02636 nm −1 ).

Nanoparticle tracking:
Nanoparticle tracking analysis was performed on a Malvern Panalytical NanoSight LM14.  16.08 mg (0.075 mmol) of 1-aminopyrene and 47.28 mg (0.493 mmol) of Imidazolate-2carboxaldehyde were accurately weighed and transferred into a sealable glass flask. 5 mL of methanol was added using a 10 mL pipette, the glass flask was sealed with an aluminium cap and afterwards heated at 80°C for 30 min with stirring. After cooling the solution to ambient temperature, 40.4 mg (0.493 mmol) of 2-methylimidazole was added to the solution as well as 250 µL of a nearly saturated solution of MeONa in methanol (0.25 g/mL) using a 1 mL pipette. Obtaining a well dissolved solution of MeONa requires heavy stirring or ensuring that the solution is being stirred while pipetting.
After the addition of MeONa, the solution was stirred for 1 minute. Finally, 5 ml of a Zn(NO 3 ) 2 •6•H 2 O solution was added using a 10 ml micropipette. This Zn(NO 3 ) 2 solution was prepared by weighting 146 mg of Zn(NO 3 ) 2 •6•H 2 O (0.494 mmol) and transferring it to a glass vial, then adding 10 ml of methanol with a 10 ml pipette and mixing by shaking gently.
Almost immediately after the addition of the Zn salt solution, a yellow solid appeared dispersed in the solution. Stirring was stopped. The mixture was left undisturbed for 4 hours. A yellow solid sedimented to the bottom of the vial.
The sedimented mixture was redispersed in the solvent with a short burst of vigorous agitation and aliquots were quantitatively transferred into a 2 mL Eppendorf tube. The suspended solid was then sedimented via centrifugation (15000 rpm, 3 mins), the supernatant removed with a pipette and discarded. The process was repeated until all the reaction volume had been centrifuged. All the solid residue remained at the bottom of the Eppendorf.
Fresh methanol was added to fill up the Eppendorf tube, the solid residue was then redispersed by agitation, the suspensed solid was then sedimented via centrifugation (15000 rpm, 3 mins) and the coloured supernatant methanolic solution was removed. This procedure was repeated 5 additional times, until the methanolic solution became colourless. The solid residue was finally dried overnight at 5 mbar and 75 o C and let to cool down under vacuum.
For the three synthesises performed the final masses of Z8P obtained were: S1= 31.24 mg; S2= 30.02 mg; S3= 29.54 mg. When compared with the initial 119.8 mg of total mass weighted for synthesis (16. WARNING: When handling Z8P solid to weight for HPLC analysis it must be kept away from humidity as much as possible.

Fig. S1
Chemical structures of the constituents to form Z8P, and their proposed abbreviations.

Preparation of the nanoparticulated ZIF-90
94.56 mg (0.984 mmol) of Imidazolate-2-carboxaldehyde was weighed, transferred to a sealable glass flask with an aluminium cap. 5 mL of methanol was added, using a 10 mL pipette. The flask was sealed and then heated to 80°C until total dissolution (5 min) was achieved.
After the addition of MeONa, the solution was stirred for 1 minute. Finally, 5 ml of a Zn(NO 3 ) 2 •6•H 2 O solution was added using a 10 ml micropipette. This Zn(NO 3 ) 2 solution was prepared by weighting 146 mg of Zn(NO 3 ) 2 •6•H 2 O (0.494 mmol) and transferring it to a glass vial, then adding 10 ml of methanol with a 10 ml pipette and mixing by shaking gently.
Almost immediately after the addition of the Zn salt solution, a white solid precipitated. The suspension was left undisturbed for 4 hours without stirring. The white solid sedimented to the bottom of the vial.
The mixture was redispersed with a short vigorous agitation and aliquots were transferred into a 2 mL Eppendorf tube. The suspended solid was then sedimented via centrifugation (15000 rpm, 3 mins), the supernatant removed with a pipette and discarded. The process was repeated until all the reaction volume had been centrifuged; all the solid residue remained at the bottom of the Eppendorf.
Then fresh methanol was added to fill the Eppendorf tube, the solid residue was then redispersed by vigorous shaking, the suspended solid was then sedimented via centrifugation (15000 rpm, 3 mins) and the methanolic supernatant removed. This procedure was repeated 5 additional times. The solid residue was finally dried overnight at 5 mbar and 75 o C, let to cool down under vacuum.

WARNING:
When handling ZIF-90 solid to weight for HPLC it must be kept away from humidity as much as possible.

Preparation of ZIF-8
This synthetic procedure was the same as in SI-4.2, with the following difference: 82.20 mg (0.984 mmol) of 2-methylimidazole (2-mim) was used instead of imidazolate-2-carboxaldehyde. No heating is required to dissolve 2-mim. It dissolves almost immediately.

FTIR characterization of the synthesised ZIFs
We performed FTIR measurements, in order to verify that the aldehyde functionality is incorporated in the Z8P. We used ZIF-8 as a negative control and ZIF-90 as positive control for comparison. For the measurements of each sample, a small amount of the solid powder was deposited on the diamond accessory sample holder.
All the characteristic peaks for each of the samples are consistent with the published spectra [4][5] . The band corresponding to the aldehyde functionality on the Imidazolate-2carboxaldehyde moiety (broad band at 1670 nm [6] ) can only be observed on the ZIF-90 and the modified ZIF-8, Z8P and is not present in the unmodified ZIF-8 (Fig. S2).  Size and shape of the nanoparticles were analysed via TEM with respect to reported ZIFs [7] The TEM analysis were complemented with EDS. [8] To prepare the Z8P sample, a dispersion of the Z8P ZIF in methanol (1 mg/mL) was prepared by weighing 4.97 mg of Z8P in a glass vial and adding 5 ml of methanol with a 10 ml pipette. After dispersion by ultrasonicating for 10 min, 10 μL aliquot of the methanol solution of the Z8P dispersion was deposited on a copper ultrathin C film on Lacey Carbon support film with a 400 mesh that was carefully deposited on top of a small glass crystallizer. The excess of solvent was left to evaporate open to the atmosphere, and the sample was then ready to be analysed.
To prepare the ZIF-90 sample, a dispersion of the ZIF-90 in methanol (1mg/mL) was prepared by weighting 5.24 mg of ZIF-90 glass vial and adding 5 ml of methanol with a 10 ml pipette; the rest of the process mirrors the TEM sample preparation described above for Z8P.

Fig. S3
TEM pictures for ZIF-90. The rhombic dodecahedral shape for ZIF particles [9] can be seen, with some polydispersity in size, between 30 to 350 nm and averaging 150 nm. These sizes are determined using only TEM results and are not absolute values.
The ZIF-antibody conjugates were also analysed by TEM measurements, complemented with S-TEM and EDS (Fig. S29) in Section SI-9. To prepare the sample, 100 μL of a borate buffer dispersion of Z8P-IgG (0.75 mg/mL, SI-6.1) was transferred to a 2 mL Eppendorf tube with a 200 µL pipette and centrifuged at 15000 rpm for 3 min. The supernatant was carefully removed and the sedimented solid was then redispersed in fresh methanol (200 μL). A 10 μL aliquot was deposited on a copper ultrathin C film on Lacey Carbon support film with a 400 mesh. This film was carefully deposited on top of a small glass crystallizer. The excess of solvent was left to evaporate, and the sample was then ready to be analysed by TEM.

Synthesis of the imine formed by Schiff condensation of 1-AP and i2ca
To study the imine formation, 200 mg of 1-AP (0.8755 mmol) and 86.06 mg of i2ca (0.875 mmol) were transferred into an aluminium-capped sealable glass flask. 10 mL methanol was added. The solution was then heated at 80 °C for 30 min with slow stirring (300 rpm) on a magnetic stirrer. The imine precipitated as an intense yellow-mustard solid. After the reaction was completed, the suspension was allowed to cool down to room temperature while stirring. The suspension was transferred quantitatively to a 15 mL Falcon tube and the solid sedimented via centrifugation (4000 rpm, 5 min). The supernatant was carefully removed and discarded. At this stage there maybe residual starting materials.
The wash workup is as follows. 5 mL of a fresh methanol was added to the sedimented residue and redispersed by strong agitation; the suspension was the centrifuged to sediment the solid. After removing the supernatant, the open Falcon tube was subjected to vacuum for 3 hours at 5 mbar and 60°C in a vacuum oven. Some impurities can be observed on the NMR measurements ( Fig S6) which may have resulted from the preparation method to obtain imines Caution: Perform the wash procedure with methanol only once to prevent the imine from dissolving in methanol, hence reducing the final yield of the imine.

Characterisation of 1-AP, i2ca and the imine formed by Schiff condensation of 1-AP and i2ca
4.6.1 Characterization by 1H NMR 1 H NMR was used to confirm the structures of 1-AP and i2ca and to characterise the structure of the imine and assess its purity.
All samples were prepared using the same protocol: a small amount of each analyte (12.54 mg for i2ca, 13.54 mg for 1-AP and 11, 54 mg for the imine) was transferred into a 5 mL glass vial, then the contents of a 0.75 mL d 6 -DMSO ampule was quantitatively transferred into the glass vial, and after achieving complete dissolution, the solution was transferred using a Pasteur pipette into an NMR tube. In the 1 H NMR spectrum for 1-AP matches the one reported on the literature [10] (See Fig.  S5), and it has a strong singlet that can be seen at 6.311 ppm, corresponding to the 2 protons of the amino group of aminopyrene.
In the 1 H NMR spectrum of the imine (Fig. S6) this peak at 6.311 ppm is almost completely gone (a very small peak can be seen), thus indicating the significant reduction of the amino group. When comparing the spectra of i2ca and the imine, the peak corresponding to the proton on the aldehyde functionality on i2ca can be seen at 9.647 ppm as a singlet. This singlet is also almost completely gone in the imine, instead a new singlet appears at 8.760 ppm, corresponding to the proton on the carbon of the C=N bond of the imine.

Fig. S8
Scheme representing the chemical shift on 1 H NMR that the aldehyde proton on the imidazole moiety suffers during the Schiff base condensation. The replacement of the aldehyde for an imine functionality reduces the electronegativity of the carbon, shifting the peak to higher fields We can also see that the imine is not completely pure (Fig. S7), since there is around 1% of i2ca and around 7.5% of 1-AP present on the mixture. This is most likely indicative of the imine not fully washed with the methanol, as well as maybe suffering small amounts of hydrolysis with the water molecules present on the system.

Characterization of imine by PXRD
Other evidence of the formation of the imine is supported by comparing the PXRD analysis of 1-AP, i2ca and the Imine (Fig. S9). The new imine has a different crystallographic structure to that of the precursors.
To prepare the samples for PXRD analysis, each powder was softly ground with an agar pestle on an agate mortar, the homogeneous fine powder was then placed on a Bragg-Brentano PVC sample holder with an opening of 0.8 cm in diameter, the cavity was filled with the powder and the top surface of the sample was compacted flat and flash with the top surface of the holder with the help of a glass slide.

Additional analysis and characterisation of ZIF and ZIF conjugates
We complement the studies and characterization for our ZIFs and ZIF-IgGs with more relevant analytical techniques: fluorescence of the Z8P-IgG conjugate particles (SI-5.1), organic composition of the ZIF-90 and Z8P particles (SI-5.2), metallic composition of the Z8P particles (SI-5.3), crystal structure of the ZIFs and ZIF conjugates (SI-5.4).

Fluorescence characterization the conjugated Z8P particles versus DMP
An experiment was designed using the analytical conditions developed in previous work. [11] to assess the changes in fluorescence quenching of the conjugate, Z8P-IgG, when exposed to a model contaminant, dimethyl phthalate, DMP.
In each well of a black non-binding polypropylene 96-well black microtiter plate, 10 μL of a Z8P-IgG dispersion in borate buffer (0.75 mg/mL, SI-6.1), was pipetted. The respective aliquots of a DMP stock solution listed in Table S2 are added to each well and then brought to a final volume of 210 μL with borate buffer. Each well had a final concentration of 47 μg/mL of Z8P-IgG. The excitation wavelength was 335 nm, emission wavelength was observed at 445 nm. The gain was set to 100%.  The fact that the fluorescence emission for the Z8P-IgG vs DMP fits a modified Stern-Volmer model that static and dynamic quenching are happening [12][13] . This could indicate that the sensing is happening on two different areas of the particle: on the surface, the exposed pyrene cores suffer collisional quenching, this being the dominant contribution, and in the bulk, the phthalates can diffuse through and be trapped with the pyrenes. However, when considering that the absorption spectra for the DMP [14] and for the imidazolate species forming the modified ZIF ( Figure S11) showed they do not absorb at the bands for excitation and emission of the imine within the Z8P ZIF, it must be assumed that the quenching is not due to quenching effects by absorption of the emitted radiation by the imine. Instead, we believe a π-π stacking that leads to the formation of an exciton state that no longer absorbs radiation. The formation of a π-π stacking exciplex between the imine and the DMP is driven by aggregation often due to hydrophobic effects-the dye molecules stack together to minimize contact with water. Planar aromatic molecules that are matched for association through hydrophobic forces can enhance static quenching. The formation of an exciplex via π-π stacking of the DMP and the imine needs further studies to be confirmed.
We also postulate that an "inner-filter quenching" [15] is happening on the Z8P ZIF nanoparticles. The fluorescence of the aminopyrene moieties in the Z8P particles are quenched as a result of the tight packing of the dye inside the ZIF, where it leads to an inner filter effect. Part of the filter effect is also caused by absorbing moieties in the antibodies surrounding the particles. This inner filter effect has also been reported for other nanocarrier systems such as liposomes [16] . More evidence supporting this inner filter effect will be shown later, when the breakdown of the Z8P in Z8P-iGg, and liberation of the dye, leads to an increase of the fluorescence emission ( Figure S22).
The effect the antibodies exert on the Z8P particles surface is more difficult for us say, since it would be likely that the antibodies can absorb part of the incident excitation radiation, as well as quenching the fluorophores on the surface via excimer formation (some amino acids have aromatic moieties).

HPLC procedure to obtain the chemical composition of ligands in Z8P
To quantify the concentration of the organic ligands in Z8P, we selected phosphate buffer as the aqueous solvent (solvent A) for HPLC because it is the only buffer that is completely transparent at 210 nm, which was the wavelength required to detect the presence of i2ca by HPLC.
The pH was chosen considering the pK a values of the imidazole species (7.46 for 2-mim and for i2ca) to make them less-dissociated and therefore more lipophilic, increasing their retention time on reverse phase HPLC. It was possible to achieve baseline separation for both species at the selected pH of 8.8. (Fig. S12).
Caution: This Excel C18-PFP column has been manufactured to be stable up to pH 10. However, to avoid unnecessary damage to the column and the HPLC system, both should be flushed after use for at least 20 minutes at a low flowrate with water/methanol to remove damaging residues from buffer which could crystallise in the HPLC equipment and column and also to avoid subjecting the silica-based column to damaging alkaline conditions.

S 23
The HPLC method is explained in SI-3.6.1.

Preparation of Standard solutions
Standards stock solutions for the quantification of imidazolate-2-carboxaldehyde (i2ca) and 1-aminopyrene (1-AP) by HPLC were prepared as follows.
Weigh 25 mg of the commercially available starting materials (97% purity for i2ca and 95% for 1-AP) and transfer each to a separate glass flask. Dissolve each in 8 mL of methanol. Full dissolution is only achieved by heating each of them to 60 °C in the sealed glass flask. Once dissolved, allow the solution to cool down to room temperature. Quantitatively transfer the content of the flask to a 10 mL graduated flask. Use 1 mL methanol to clean the glass flask and transfer this rinse solution to the 10 mL graduated flask. Make the solution up to 10 mL volume with methanol. One additional standard solution for each species was independently prepared to confirm that the working standard can be reproducibly prepared.
The stock standard (Stock 1) which will be later used to prepare the set of calibration standard solutions and the duplicate independent standard (Stock 2) were prepared as follows: For i2ca: Stock 1: 24.20 mg (0.249 mmol) in 10 mL methanol. This stock was used in the preparation of the linearity standard solutions. Stock 2: 25.87 mg (0.261 mmol) in 10 mL methanol was independently prepared, and used to prepare Std 5bis in Table S3.  Table S3.  Table S3 from a 2.5mg/mL (stock 1), then adding the corresponding methanol aliquot to accurately make a total volume of 1.000 µL.
The concentration of each working standard solution was calculated from the known concentration of Stock 1 or Stock 2. The peak area obtained by HPLC for each working standard solution was plotted against the known concentration of the calibration standard solution used. The concentrations range chosen demonstrates good linearity for all the analytes (Fig. S13).
Once linearity has been demonstrated as part of the validation of the HPLC method, all subsequent analyses can be performed by using just one working standard solution prepared at the expected concentration ensuring that the concentration of the working standard and sample solutions are within the demonstrated linearity range. It is desirable to independently weigh 2 working standards to demonstrate that they have been accurately prepared. It is best to prepare the working standard directly and not by dilution of a stock solution. It is recommended to freshly prepare the working standard every time.

Preparation of Z8P acid digestion sample solutions
This protocol was done for each Z8P batch in triplicate. Approximately 4 mg of Z8P were weighed, transferred to a glass vial and dispersed in 1.5 ml of MeOH. Then 150 µL of concentrated trifluoroacetic acid (TFA) measured with a 500 µL glass syringe with metallic needle was added.
Note: after adding the TFA to the methanolic dispersion of the ZIF, the solution will turn to an intense yellow-mustard.
After complete dissolution of the ZIF, an additional 1.3 ml of MeOH is added, then 1.8 mL of a 1 M NaOH solution in water was added using a 2.5 mL automatic pipette, to neutralize the solution. We made sure, using a laser pointer, that there were no colloids present in the solution. We also made sure that the pH of the solution remained slightly acidic (around pH 3-5) using pH paper to avoid any reprecipitation. After cooling, the solution was transferred to a 5 mL volumetric flask and the volume was adjusted with a Pasteur pipette using deionized water as solvent.
Note: after 2-3 min of time the colour of the solution will turn from an intense yellowmustard to almost completely transparent and colourless.
1 mL volume was transferred to a 2 mL HPLC glass vial using a 1 mL pipette. Each sample was injected on to the HPLC column in duplicate.
The results of the HPLC quantification are shown on Table S6, using the peak areas obtained with the HPLC method previously described to calculate their concentration in relation to the peaks areas obtained from the working standard solutions of known concentration. We decided to calculate the final composition of the Z8P ZIF in relation to Zn 2+ ions. We believe that this assist in a clearer understanding of the relative composition.

Quantification of Zn 2+ content on Z8P by ICP-OES
ICP-OES (inductively coupled plasma atomic emission spectroscopy) was used in order to obtain the quantitative content of Zn 2+ in ZIF samples. ZIF must undergo a wet digestion procedure before the samples can be analysed by ICP-OES.

Suitable disintegration procedure for Z8P analyses by ICP-OES
The disintegration procedure is performed as follows. The samples were divided into four synthetic batches. The solution for each batch was independently prepared in duplicate and each solution was measured five times by ICP-OES.
1. Around 1 to 3 mg of each sample of Z8P was weighted 2. The sample was dissolved in 2 mL of HNO 3 30 %v/v. Concentrated HNO 3 , previously purified by sub-boiling in a PFA device. 3. The disintegrated ZIF sample in concentrated HNO 3 was diluted with ultrapure deionized water to a volume of 10 mL and mixed well.
Note: When weighting the samples for ICP quantification, the weighted amounts were very small (1,115 mg and 1,118mg for sample 1; 2.863 mg and 2,865 for sample 2; 0.76 mg and 0.78 mg sample 3) and they were dissolved it to 10 ml as final volume. This dilution magnifies any error that is carried from the whole process.
ICP determination of a Zn 2+ in the samples was performed against an external standard solution of a Zn 2 which is directly traceable to the NIST SRM 3169 (Certipur© Certified Reference Material, Merck, Germany).

ICP-OES parameter for analysis of disintegrated Z8P
The operating conditions for the ICP-OES (Agilent 5110 VDV) are listed on Table S8. The quantitation of Zn 2+ in samples are summarized in Table S9.

ICP Determination of Zn 2+ ; Results
The results of the ICP determination of the Zn content are as shown on Table S9. It reinforces the results of the composition obtained via HPLC on Fig. S14.

Powder X-Ray diffractograms for the ZIFs, both conjugated and unconjugated
Both conjugated and unconjugated ZIFs were analysed by X-ray powder diffraction with Bragg-Brentano symmetry to determine the crystal structure of the obtained ZIFs. The parameters for all diffractograms are: 2θ 5-50°; step size 0.02°; integration time 6 sec.
These measurements demonstrate that the ZIFs conserve its characteristic SOD (sodalite) crystalline structure even after the conjugation process has taken place.

Preparation of PXRD sample of non-conjugated ZIFs
To prepare the non-conjugated ZIF samples, each ZIF powder was softly ground in an agate mortar with an agate pestle, then placed on a Bragg-Brentano PVC sample holder with an opening of 0.8 cm in diameter. The cavity of the sample holder was filled with the powder and with the help of a glass slide, the top surface of the sample was compacted flat and flash with the top surface of the holder.

Preparation of PXRD sample of conjugated ZIFs
For the conjugated ZIFs, 400 μL of a freshly prepared ZIF solution (SI-6.1) from each conjugate (a total mass of 0.3 mg for each measurement), was transferred to an Eppendorf tube, centrifuged and all excess supernatant to 50 μL carefully removed with a pipette. This remaining 50 μL suspension was placed on a zero-background silicon monocrystalline (SIOV) sample holder using a 100 µL pipette. The measurement was started immediately, with the sample still wet.  [17] .
In this section we present the two-step protocol to covalently bind the free amino functionality of the lysine residue (-(CH 2 ) 4 -NH 2 of the antibody tested to the free aldehyde functionality part of the imidazole 2-carboxaldehyde, included on the ZIFs of interest (ZIF-90 and the fluorescently modified ZIF-8/ZIF-90 hybrid, Z8P).
In SI-9 we discuss how we demonstrate that this binding has been successfully achieved. S 37 4) Add 50 μL of a 2 mg/mL solution of sheep anti-mouse IgG (see SI-1.6) to this dispersion without any previous purification step using a 100 µL pipette. 5) Add a magnetic stirrer bar, seal the glass vial, and stir the mixture for 15 min at room temperature.

6) Add 50 μL of a commercially prepared aqueous solution of Na[BH 3 (CN)] (5 M) in 1 M
NaOH using a 200 µL pipette and allow to react for an additional 4 hours at room temperature.
Warning: Exercise extreme care when handling Cyanoborohydride. Always wear gloves, glasses and work in a properly enclosed fume hood.

Protocol for washing the freshly prepared ZIF-90-IgG or Z8P-IgG
The next steps were designed to remove any anti-mouse IgG not covalently bonded to ZIF-90 or Z8P. The washing protocol of this conjugation reaction is as follows: 1) Transfer the 1.5 mL suspension to a 2 mL Eppendorf tube with the help of a 1 mL pipette and centrifuge it at 15000 rpm for 5 min. 2) Carefully remove the supernatant taking care not to disturb the solid residue and transfer it to a common container labelled as "Wash 1". This wash is reserved for later use, in case the ELISA experiment fails suggesting unsuccessful binding between the IgG and the ZIF. "Wash 1" will help to investigate if the free IgG in the wash was active or inactive. "Wash 1" is disposed of, after successful completion of ELISA experiments Warning: The very first supernatant of "Wash 1" contain cyanide derivatives. Extreme care must be exercised when dealing with and disposing of "Wash 1". GHS05, GHS06, GHS08, GHS09 3) Add 1 mL of borate buffer to the Eppendorf tube to wash the solid residue. 4) Sonicate the suspension for 2 min to ensure full dispersion of the solid particles 5) Centrifuge the dispersion at 15000 rpm for 5 min 6) Remove the supernatant carefully avoiding contact with the solid residue and transfer to a separately container labelled as "Wash 2". Wash-2. is also reserved for later use, in case the ELISA experiment fails suggesting unsuccessful binding between the IgG and the ZIF. "Wash 2" will further help to investigate if the free IgG in the wash was active or inactive. "Wash 2" is disposed of, after successful completion of ELISA experiments 7) Add 1 mL of TRIS buffer (see SI-2.2.3) to the residue 8) Redisperse the solid residue by vigorous agitation, and centrifuge at 15000 rpm for 5 min 9) Carefully remove this TRIS buffer supernatant and discard it S 38 10) Repeat steps 7-9 three more times to wash out any IgG not properly conjugated to the ZIF 11) Perform steps 7-9 two additional times, this time with borate buffer 12) Disperse the properly washed solid residue containing the ZIF conjugate in 2 mL of borate buffer. This is the ZIF-IgG stock solution (0.75 mg/ml) used in all the ELISA, LFIA, LC/MS and PXRD characterization.
Addendum: The amount of washing steps was developed so that only the covalently bound antibody remains on the ZIF particles.

Studies on conjugated antibody activity
In this section we study the residual activity of the antibodies bound to the nanoparticulate ZIFs. We show the viability of the ZIFs as fluorescent bio labels on a series of non-competitive ELISA experiments.

Fig. S18 Schematic representation of the general non-competitive indirect ELISA procedure developed to study the antibody activity of the conjugate. The electrostatically charged surface of the high-binding microtiter plates (MTP) was coated with the mouse IgG and incubated on a plate shaker. After washing, the remaining binding sites on the MTP were blocked with casein to avoid any passive adsorption of the conjugates to the surface of the MTP ("non-specific binding"), which would lead to false positive results. After another incubation, the plates were washed again and dilutions of an anti-mouse IgG antibody (or a modified system including anti-mouse IgG) at various concentrations were added and the MTP incubated for 18 hours on a plate shaker. After an additional washing step, a solution with the HRP-labelled detection antibody (anti-sheep IgG) was added and incubated for 1 hour on a microplate shaker. After a final washing step, the substrate 3,3',5,5'-tetramethylbenzidine (TMB) was added to each well and incubated for 15 min on a microplate shaker, then 100 µL of 1 M sulphuric acid was added to each well and the optical density (OD) was measured.
The response that should be obtained, should the system response correctly, would be a characteristic sigmoidal calibration curves, obtained by fitting a four-parameter logistic function to the data points.
The basic model to assess antibody activity by a non-competitive ELISA methodology is depicted in Fig. S18. It is a layered sandwich system which has a base formed by:

Detector: HRP labelled donkey anti-sheep IgG
This sandwich system is used to characterize the activity response of the primary, analyteselective antibody. This system will be later modified by switching the primary antibody for our synthesised ZIF-IgGs.
The system will be quantified via HRP (Horse Radish Peroxidase) redox reaction driven TMB colour development, in which HRP triggers the conversion of TMB (transparent) to its oxidized coloured product (first intense blue, then bright yellow after the addition of 1M sulfuric acid) [12]; the coloration is proportional to the amount of HRP present on the system which, in turn, is proportional to the amount of primary present on the system. The calibrating solutions were prepared by making serial dilutions from a stock solution of the antibody species to be assessed (anti-mouse IgG or ZIF-IgG conjugate). The rest of the species taking part on the experiment should be maintained at constant concentrations. In order to compare accurately the amounts of soluble antibodies, each well was processed under the identical conditions, and 4 replicates of each measurement was performed.

Protocol of the non-competitive indirect ELISA for the anti-sheep IgG primary
In this version of the protocol we assess the activity of our base ELISA system [mouse IgG (target) / sheep anti-mouse IgG (from sheep, primary, selective antibody) / anti-sheep IgG-HRP (from donkey, enzyme-labelled detector antibody)].
The standard curve was prepared by making serial dilutions of one known concentration of the primary (sheep anti-mouse IgG) across a range of concentrations (0.25-750 ng/ml), while keeping every other parameter of the system constant. Each well was treated under the same conditions, and each measurement was done in quadruplicate.
For details on the antibodies see SI-1.6.
The protocol is performed at follows: Characteristic sigmoidal calibration curves were obtained by fitting a four-parameter logistic function in Origin to the data points obtained from OD measurements. The sigmoidal shape of Standard calibration curves is expected when the binding occurs via intact antibody binding sites and not, e.g., due to non-specific binding. The calibration standard solutions were prepared by making serial dilutions from a stock solution of the antibody species to be assessed (anti-mouse IgG), while keeping the rest of the solution constituents at constant concentrations. In order to compare the amounts of soluble antibodies, each well was processed under the same conditions, and 4 replicates of each measurement were performed in.

Preparation of the dilution series for assessing antibody activity of anti-mouse IgG
We prepare 11 new solutions from these four stock solutions (MS1, MS2 and MS3) to varying concentrations for sheep anti-mouse IgG.
The procedure for each dilution series stock (number #) is performed as follows: 1) Transfer 1 mL of borate buffer into a new Eppendorf tube. 2) Using the suitable micropipette, remove from the Eppendorf tube a volume of borate buffer equivalent to the corresponding cell on the column named "aliquot volume from MS stock solution" as listed on Table S11. 3) Using the same micropipette, transfer the aliquot volume corresponding to the exact same cell as on step 3 from Table S11, with a new tip on the micropipette. Wash the tip by aspirating/dispensing the solution several times 4) Seal the Eppendorf tube and homogenise the solution using a mechanical shaker WARNING; In order to keep the antibodies stable during the whole process all solutions are kept on ice.   (Table S13). The logistic function (4-parameter function) has been calculated which fits perfectly the binding events in immunoanalytical methods. Rodbard discovered around 1985 that these functions perfectly fit immunoassay data. The parameters are here named A1, A2, Xo and p. x is the analyte concentration, A1 is the maximum signal (upper asymptote of the curve), A2 is the signal background (lower asymptote of the curve, lowest point on "y-axis", Xo is the x coordinate of the inflection point ("test mid point") of the curve, in concentration units and p is a parameter that reflects the slope of the curve. HRP is Horseradish peroxidase.

Protocol of the non-competitive indirect ELISA for the ZIF-IgG conjugates
In this version of the protocol we assess the residual activity of the antibodies bound to both our ZIF-IgG conjugates by a modified ELISA system [mouse IgG (target) / ZIF-IgG (Z8P-IgG and Z90-IgG, primary, selective antibody) / anti-sheep IgG−HRP (from donkey, enzyme-labelled detector antibody)].
The standard curve was prepared by making serial dilutions of one known concentration of the analyte (ZIF−IgG conjugates, both Z90-IgG and Z8P-IgG) across a range of concentrations (0-150 ng/ml), while keeping every other parameter of the system constant. Each well was treated under the same conditions, and each measurement was done in quadruplicate.
The protocol described in SI-7.2 was performed with the following modification: Step 9 in SI-7.

Preparation of 4 stock solutions of ZIF-IgG
Prepare in total 4 stock solutions, each to a total volume of 1 mL. Stock solution MS1 was prepared from the initial stock solution of Z8P-IgG (0.75 mg/mL) and Z90-IgG (0.75 mg/mL); Stock solution MS2 was prepared by dilution from MS1; Stock solution MS3 was prepared by dilution from MS2; Stock solution MS4 was prepared by dilution from MS3.

Preparation of the dilution series for assessing antibody activity of ZIF-IgG
We prepare 23 new solutions from these four stock solutions (MS1, MS2, MS3 and MS4) to varying concentrations for ZIF-IgG (both Z90-IgG and Z8P-IgG).
The procedure for each dilution series stock (number #) is performed as follows: 1) Transfer 1 mL of borate buffer into a new Eppendorf tube. 2) Using the suitable micropipette, remove from the Eppendorf tube a volume of borate buffer equivalent to the corresponding cell on the column named "aliquot volume from MS stock solution" as listed on Table S14. 3) Using the same micropipette, transfer the aliquot volume corresponding to the exact same cell as on step 3 from Table S14, with a new tip on the micropipette. Wash the tip by aspirating/dispensing the solution several times 4) Seal the Eppendorf tube and homogenise the solution using a mechanical shaker S 46 WARNING; In order to keep the antibodies stable during the whole process all solutions are kept on ice.

Conclusion:
A similar response to the one we obtained on our previous ELISA system (SI-7.2) can be observe, proving that the antibodies bound to the ZIF conjugates retain their activity.

Protocol of the non-competitive ELISA for the antibody conjugates Z8P-IgG / non-competitive fluorescent ELISA tandem
To study the viability of our fluorescently modified ZIF-8/ZIF-90 hybrid as a fluorescent bio label in immunoassays, we design a new ELISA by modifying our previous methodology, comparing the response for Z8P-IgG between non-competitive ELISA and non-competitive fluorescent ELISA.
In both these ELISAs the standard curve was prepared by making serial dilutions of one known concentration of the primary, Z8P-IgG, across a range of concentrations (0-1500 ng/ml), while keeping every other parameter of the system constant. Each well was treated under the same conditions, and each measurement was done in quadruplicate. Both ELISA protocols took place using the exact same concentration for all species. Two measurements took place in parallel: -Non-competitive ELISA: On a high-binding transparent microtiter plate (SI-7.4.1), where OD was measured as a final step -Non-competitive fluorescent ELISA: On a high-binding black microtiter plate (SI 7.4.2), were fluorescence was measured as a final step Note: the significant increase on the concentrations for our target and our primary is due to the fact that the fluorescence sensitivity is not high enough to work on the previously used range of concentrations (SI-7.3), thus we are forced to increase them to reasonably detect and quantify fluorescence.
The protocols area as follows: 7.4.1 Protocol of the non-competitive ELISA for the antibody conjugates Z8P-IgG The protocol described in SI-7.2 was followed with the following modifications:

Preparation of Capture antibody solution
1) Transfer 20 mL of borate buffer (see SI-2.2.1) to a long glass vial with a plastic cap using a 10 mL pipette 2) Transfer 5.0 µL of mouse IgG using a 5 µL pipette. Wash the tip by aspirating/dispensing the solution several times 3) Close the glass vial and homogenize the mixture d using a mechanical shaker Prepare in total 4 stock solutions, each to a total volume of 1 mL. Stock solution MS1 was prepared from the initial stock solution of Z8P-IgG (0.75 mg/mL); Stock solution MS2 was prepared by dilution from MS1; Stock solution MS3 was prepared by dilution from MS2; MS4 was prepared by dilution from MS3. MS1 was prepared as follows: 1) Transfer 1 mL of borate buffer inside a 1.5 mL Eppendorf tube 2) Remove 100 µL of the buffer using a 100 µL pipette 3) Transfer 100 µL of the original stock solution of the Z8P-IgG with a new tip on the micropipette. Wash the tip by aspirating/dispensing the solution several times 4) Seal the Eppendorf tube and homogenise the solution using a mechanical shaker MS2 was prepared as follows: 1) Transfer 1 mL of borate buffer inside a new 1.5 mL Eppendorf tube 2) Remove 100 µL of borate buffer with a 100 µL pipette 3) Transfer 100 µL of the MS1 stock solution with a new tip on the micropipette. Wash the tip by aspirating/dispensing the solution several times 4) Seal the Eppendorf tube and homogenise the solution using a mechanical shaker MS3 was prepared as described for MS2 but using MS2 as the stock solution MS4 was prepared as described for MS2 but using MS3 as the stock solution The concentration for the MS solutions will be: MS1 = 75 µg/mL; MS2 = 7.5 µg/mL; MS3 = 0.75 µg/mL; MS4 = 0.075 µg/mL.

Preparation of the dilution series for Z8P-IgG
We prepare 23 new solutions from these four stock solutions (MS1, MS2, MS3 and MS4) to varying concentrations for Z8P IgG.
The procedure for each dilution series stock (number #) is performed as follows: 1) Transfer 1 mL of borate buffer into a new Eppendorf tube. 2) Using the suitable micropipette, remove from the Eppendorf tube a volume of borate buffer equivalent to the corresponding cell on the column named "aliquot volume from MS stock solution" as listed on Table S16. 3) Using the same micropipette, transfer the aliquot volume corresponding to the exact same cell as on step 3 from Table S16, with a new tip on the micropipette. Wash the tip by aspirating/dispensing the solution several times 4) Seal the Eppendorf tube and homogenise the solution using a mechanical shaker WARNING; In order to keep the antibodies stable during the whole process all solutions are kept on ice.

Protocol of the non-competitive ELISA for the antibody conjugates Z8P-IgG via non-competitive fluorescence immunoassay
We developed a non-competitive indirect immunoassay for the assessment of the activity of the Z8P-IgG conjugate and its fluorescence capabilities.
The protocol described in SI-7.2 was followed with the following modifications: liberate the fluorophores. 9) The plate was incubated in a plate shaker for 30 min 10) Fluorescence was then measured again By determining the variation of the crystallite sizes over a specific window of time (5 days) keeping the conjugates in a borate dispersion (0.75 mg ZIF-IgG/ml of Borate Buffer) at 4°C and with stirring, it was possible to demonstrate that the crystallinity structure of the ZIFs after the conjugation reaction has not been heavily affected. The ZIF particle stability is dependent on the concentration of the dispersion in water.

PXRD measurements of the Z90-IgG and Z8P-IgG conjugates
To study how the crystallinity of the ZIFs (both Z8P and ZIF-90) could be degrading over time because of the action of the antibodies (antibody conjugated particles need to be immersed in aqueous buffer, which is known to slowly degrade ZIF-particles if not otherwise protected), we obtained X-ray diffractograms of a sample of the conjugates over time.
For each diffractogram, 400 µL aliquot (a total mass of 0.3 mg for each measurement) of the stock solution for each ZIF-IgG conjugate was freshly added to an Eppendorf tube, centrifuged (15000 rpm, 3 min) to reduce its volume to 50 µL by removing excess supernatant. The 50 µL dispersion was placed on a zero-background silicon monocrystalline (SIOV) sample holder. The measurement was started with the sample still wet. The following measurement conditions were used. 2θ 5-50°; step size 0.02°; integration time 6 sec.

S 55
To asses that the crystallinity of the system was not severely damaged over time we measured the crystallite size for each system, applying the Scherrer equation to the first three reflections, (110), (200) and (211), for all diffractograms.
The Scherrer equation is a formula that relates the size of sub-micrometre crystallites in a solid to the broadening of a peak in a diffraction pattern and was used in the determination of size of crystals in the powder. [18] If the Scherrer value remains stable over time, this is an indication that there is no significant degradation of the ZIF particles in the conjugates. = Equation S 1 . Scherrer equation for calculating the crystallite size, where "τ" is the mean size of the ordered (crystalline) domains, "K" is a dimensionless shape factor [we took 1], "λ" is the X-ray wavelength (1.50406 nm), "β" is the line broadening at half the maximum intensity (FWHM) in radians, "θ" the Bragg angle. We calculated all the FWHM for each diffraction peak after background subtraction and applying a Gaussian fitting model with Origin.  We plot the crystallite sizes obtained of each diffraction peak versus decomposition time.

Fig. S24
Scherrer crystallite sizes for the first three reflections for the studied ZIF−IgG conjugates, Z8P−IgG and Z90−IgG. No marked changes in crystallite size were observed in the conjugated ZIFs under storage at 4°C for 120 hours using this crystallographic method. The pronounced variation on the (110) reflection for Z8P−IgG does not seem to be representative for the system. A significant crystallinity degradation over time could not be observed in the studied time frame, since both systems have a similar range of crystallite sizes for all diffractions and over time. The pronounced variation on the diffraction [1,0,0] for Z8P-IgG does not seem to be representative of the system, and was probably caused by the sample not being completely immobilized on the sample holder (was deposited on a water dispersion and started measurement immediately), thus generating the disparity on the first 30 mins of measurement (corresponding to the first 8-9°, approximately).

Dynamic laser scattering (DLS) of the Z90-IgG conjugates
To study how the ZIF nanoparticles (both ZIF-90 and Z90-IgG)) could be degrading over time on dispersion, DLS measurements were performed

Dispersant phase preparation
To measure on DLS it is critical that the dispersion has no dust particles present. To eliminate all non-desired particles from the dispersion, the dispersant (borate buffer) is double filtered before use. This is done my making the dispersant go through a double filter setup, using two different filters:

DLS sample preparation and results
A dispersion on double filtered borate buffer of Z90-IgG with a concentration for approximately 0.1 mg/ml is prepared. Before the measurement on DLS is critical to sonicate the sample for at least 5 min, to break all the micro aggregates. This, however, must be performed carefully, since ultrasonicating can lead to fragmentations and exfoliation of the particles [19] .
The systems behaviour on borate buffer shows two important points: 1-On the measured concentration, the particles tend to stay as microaggregates of around 1 µm, even though TEM shows the particles are 150-300 nm in diameter; this microaggregates cannot be broken down by ultrasonication alone, being necessary the introduction of stabilising agents, such as DMSO or other surfactants. This is likely a hydrophobicity issue, since ZIF-8 and -90 tends to stick together in very polar solvents due to the un-polarity of aged ZIF, which leads to forming colloidosomes. 2-Over time the microaggregates tend to increase in size and in size dispersity, hinting a tendency for the system to precipitate overtime. This phenomenon can be easily reverted by sonicating the sample for 3-5 minutes, unlike the microaggregates.   Table S20. The coloured area around each line represents the standard deviation for the particle sizes as calculated by DLS (also in Table S20).

Nanoparticle tracking measurements for Z90-IgG.
To measure on NTA it is critical that the dispersion has no dust particles present. The big advantage over scattering methods NTA offers is that this is suitable for multi-disperse samples, as well as allowing small particles to be more visible than large particles of the same sample.
To eliminate all non-desired particles from the dispersion, the dispersant (borate buffer) is double filtered before use. This is done my making the dispersant go through a double filter setup, using two different filters:

NTA sample preparation and results
A dispersion on double filtered borate buffer of Z90-IgG with a concentration of approximately 0.01 mg/ml is prepared. Before the measurement on NTA, it is critical to sonicate the sample for at least 5 min, to break all the micro aggregates.
In this technique is possible to see the nanoparticles individually on dispersion. A similar size dispersion can be seen (150-300 nm in diameter) as observed in TEM.

Fig. S27
Particle size distribution and standard deviation measured by NTA for the Z90-IgG dispersion in borate buffer.

Fig. S28
Stationary image of the nanoparticle Z90-IgG dispersion seen through NTA. The rings around the particles are diffraction spheres, due to the crystallographic nature of the conjugate.

Studies on the ZIF-antibody bond
To study the nature of the bond between our ZIFs (ZIF-90 and Z8P) and the antibody (sheet anti-mouse IgG) we design a high-resolution mass spectroscopy methodology combined with TEM pictures.

TEM images of the ZIF-IgG conjugates
The protocol to characterize the particles via TEM is described in SI-4.4.2.
TEM images are consistent with the ZIF structure and particle size being retained upon formation of both ZIF conjugates. The antibodies appear as a rugged layer surrounding the ZIF particles.

Fig. S29
TEM pictures for the ZIFs. A rhombic dodecahedra morphology can be observed for the Z8P-ZIF particles (pictures a-c), with some polydispersity in size, between 50 and 150 nm and averaging 100 nm. The pictures were complemented with EDS analysis (picture c), confirming that the signal for Zn was contained within the particles. For Z8P-IgG (pictures d-g) the particles were covered by a rugged layer of antibodies. The relative size of the whole system was in the same range as the one for the ZIF particles without the antibodies. The relative size of each entity was around 10 nm, with was consistent with the relative size of the antibodies themselves. S-TEM pictures were taken to detect the ZIF particles (picture e) and they were located within the yellow circles. EDS analysis complements the images (picture g). Most of the Zn signals were inside the rugged particles, which was consistent with the system description. There was also a not insignificant number of signals for zinc outside of the ruggedness, suggesting that after the conjugation procedure some of the ZIF have been damaged.

Study of the nature of the bond on the ZIF-IgG conjugates via HRMS
In order to confirm covalent conjugation through the imidazole carboxaldehyde moiety of the ZIF particles to a free amino group of the antibody, conjugates were digested (SI-9.2.1). Digestion results in the cleavage of most amide bonds in the protein and the potential release of L-lysine-bound Imidazolate-2-carboxaldehyde. The hydrolysate was then screened with liquid chromatography high-resolution mass spectrometry (LC-HRMS) (Fig. S30). After developing the methodology to detect the fragment (SI-9.2), it revealed a potential lysine-imidazole derivative at a retention time of 558 s (Fig. S31b). The compound was present in both ZIF-IgG conjugates and absent in all control experiments (individually digested IgG and ZIF without conjugation), thus supporting the notion of it being formed exclusively in ZIF−IgG conjugates.

Digestion of the conjugate
The protocol for digesting the samples was as follows: To a distinct amount of sample (See Table S21 for more information) 1 mL of an aqueous solution of HCl 6N was added, the air was purged with the use of a Schlenk line with vacuum and Argon, and then heated to 107 °C for 24 h. After the digestion, the samples were cooled down and transferred to a 1.5 mL Eppendorf tube, and the solvent was removed using a SpeedVac.

High-resolution mass spectrometry
First, the efficiency of this hydrolysis was tested using pure antibody instead of the conjugate. Extracted ion chromatograms (EICs) of protonated and deprotonated molecular peaks ([M+H] + , [M-H] -) acquired in both positive and negative electrospray ionisation modes suggest the presence of 17 of the 20 proteinogenic amino acids in the sample (Fig. S32), indicating efficient protein breakdown. Next, ZIF-IgG conjugates were hydrolysed and subjected to LC-HRMS. Chromatograms of hydrolysed ZIF-IgG conjugates contained approximately 140 chromatographically deconvoluted compounds ("features") in the mass range of 50 to 600 Da (Fig.  S31a). This shows extensive hydrolytic breakdown of conjugates into smaller molecules. In these chromatograms, ion traces of possible amino acid imidazole derivatives were checked using a target list of theoretical precursor ions (Table S22).
This approach revealed a potential lysine imidazole derivative at a retention time of 558 s (Fig. S31b). The compound was present in both ZIF/IgG conjugates but in none of the control measurements (IgG alone, ZIFs alone), supporting the notion that it was formed exclusively in ZIF/IgG conjugates.
Structural evidence for the compound could be obtained from both isotope pattern and MS/MS spectrum. First, isotopologue intensities matched calculated ones within instrumentlimited error margins (<2% ; Table S23). Second, all major MS/MS fragments could be assigned substructures that explained the putative molecular structure very consistently (Fig. S31c).
Taken together, while only co-elution of a chemical standard or NMR could give definite structural evidence, covalent bonding of an Imidazolate-carboxaldehyde moiety to an amino acid residue of the antibody was well supported by the present results.
S 66 S 67

Studies on the mobility of the ZIF-antibody conjugate
The mobility of the ZIF conjugates was important for their application in bioanalytical assays. The feasibility of our ZIF-IgG system as a mobile conjugate on a solid matrix for lateral flow assays is next tested. PVDF matrix is chosen from the candidates available; tests on nitrocellulose were performed but proved to be an unsuitable matrix given the strong interactions between the nitrocellulose and the conjugates.
In order to facilitate the proof of concept for the Z8P platform as a bio label, we create a new conjugate system, Z8P-IgG HRP, by changing the antibody conjugated to the ZIFs from sheep-anti-mouse IgG to HRP labelled donkey anti-sheep IgG For details on the antibodies see SI-1.6. was added without any previous purification step using a 100 µL pipette to this dispersion 5) A magnetic stirrer bar was added, the glass vial closed, and the mixture stirred for 15 min at room temperature 6) 50 μL of a commercially prepared aqueous solution of Na[BH 3 (CN)] (5 M) in 1 M NaOH was added using a 200 µL pipette and allowed to react for an additional 4 hours at room temperature.

Conjugation reaction of Z8P particles to IgG−HRP
Warning: Exercise extreme care when handling Cyanoborohydride. Always wear gloves, glasses and work in a properly enclosed fume hood.

Protocol for washing the freshly prepared Z8P-IgG HRP
Next step was designed to remove any anti-sheep IgG HRP not covalently bonded to Z8P. The washing protocol of this conjugation reaction is as follows: 1) Transfer the 1.5 mL suspension to a 2 mL Eppendorf tube with the help of a 1 mL pipette and centrifuge it at 15000 rpm for 5 min. 2) Carefully remove the supernatant taking care not to disturb the solid residue and transfer it to a common container labelled as "Wash 1". This wash is reserved for later use, in case the ELISA experiment fails suggesting unsuccessful binding between the S 73 IgG and the ZIF. "Wash 1" will help to investigate if the free IgG in the wash was active or inactive. "Wash 1" is disposed of, after successful completion Elisa experiments Warning: The very first supernatant of "Wash 1" contains cyanide derivatives. Extreme care must be exercised when dealing and disposing with "Wash 1". GHS05, GHS06, GHS08, GHS09 3) Add 1 mL of borate buffer to the Eppendorf tube to wash the solid residue. 4) Sonicate the suspension for 2 min to ensure full dispersion of the solid particles 5) Centrifuge the dispersion at 15000 rpm for 5 min 6) Remove carefully the supernatant avoiding contact with the solid residue and transfer it to a separately container labelled as ""Wash 2" 7) Add 1 mL of TRIS buffer (see SI-2.2.3) to the residue 8) Redisperse the solid residue by vigorous agitation, and centrifuge at 15000 rpm for 5 min 9) Carefully remove this TRIS buffer supernatant and discard it 10) Repeat steps 7-9 three more times to wash out any IgG not properly conjugated to the ZIF 11) Perform steps 7-9 two additional times, this time with borate buffer 12) Disperse the properly washed solid residue containing the ZIF conjugate in 2 mL of borate buffer. This is the ZIF-IgG HRP stock solution (0.75 mg/ml) used in all the ELISA analysis

Conclusion:
The amount of washing steps was developed so that only the covalently bound antibody remains on the ZIF particles.

Mobility studies of ZIF conjugates on a PVDF membrane
The mobility was studied on the commercial membrane TransBlot Turbo Midi-size PVDF Membrane (Bio-Rad, Cat. # 10026933).

Protocol for the PVDF mobility tests of Z8P-IgG
1) The activation of the PVDF strip is done by submerging the membrane in pure ethanol for 10 sec 2) Remove the excess ethanol by rinsing it in deionized water for 3 sec 3) Allow the strip to dry at open atmosphere 4) Place 1 µL of a sheep anti-mouse IgG (2 mg/ml) solution on the immobilization point using a 5 µL micropipette operating in reversed mode 5) Allow the PVDF strip with the immobilized sheep IgG dry 6) Submerge the strip into a casein solution in PBS (1mg/mL, see SI-7.2.2) for ten seconds to block all active sites on the strip 7) Allow the strip to dry at room temperature. S 74 8) Dispense 10-30 µL of the Z8P-IgG HRP dispersion in borate buffer (0.75 mg/mL) on the elution starting point using a reverse micropipette. 9) Allow to dry 10) Attach an adsorption pad on the opposite side of the strip in relation to the elution starting point. 11) Elute the Z8P-IgG HRP using borate buffer until the deposited Z8P-IgG HRP reaches the adsorption pad (yellow colouring from the Z8P) 12) Allow dry.
The Z8P-IgG HRP can be seen by the naked eye on the capture zone and on the adsorption pad (yellow colouring), indicating antibody binding.
13) TMB solution is used to reveal the location of the HRP-labelled antibodies. This will trigger an intense blue coloration within minutes wherever IgG HRP labelled are present, with by extension shows were the ZIF-IgG HRP conjugates are present. 14) Cut out the relevant area of the strip and digest it in 200 µL of 1 M H 2 SO 4 for 15 min using a mechanical shaker and a desk centrifuge to aid the process. 15) Pipetted the digestion solution into a 96-well black microtiter plate. 16) Measure the fluorescence of the solution measured Note: TMB and PVDF samples were also prepared as negative controls.

Fig. S34
Experimental set-up to demonstrate the aptitude of Z8P-IgG for the bioanalytical application lateral flow immunoassay (LFIA). a) Assay. The protocol begins by activating the PVDF membrane through submersion in pure ethanol followed by rinsing in deionized water. After air drying, a solution containing sheep IgG as target is placed in a central part of the strip, the later capture zone. The PVDF strip with the immobilized sheep IgG is then submerged in a casein solution to block all remaining active sites on the strip, and afterwards allowed to dry. A varying volume of the Z8P-IgG-HRP dispersion is pipetted on the elution starting point and allowed to dry. Then the absorption pad is attached to the strip. Subsequently, the Z8P-IgG-HRP is eluted by adding enough of a mobile phase (borate buffer) to the strip that the phase's front-line flows across the capture zone into the absorption pad. This is followed by drying the strip in air. b) Revelation of mobility of the composite and retention of binding specificity of the antibody: The capturing of Z8P-IgG by the target (sheep antibody) in the capture zone is visible by the colour of the ZIF particles. For stronger evidence that the labelled antibody is binding selectively to the immobilized target, the HRP enzyme linked to Z8P-IgG is revealed via its reaction with the added substrate TMB. c) Proof of intact ZIF label. The area to which the antibody has bound is cut out and digested in H2SO4 to decompose the ZIF and liberate the pyrene. The supernatant is transferred to a black microwell and fluorescence of the solution measured. The emission spectra show that fluorophore is released in amounts related to the volume of conjugate employed (PVDF and TMB samples that were not run through the assay serve as negative control).