An indium-111-labelled membrane-targeted peptide for cell tracking with radionuclide imaging

Cell labelling agents that enable longitudinal in vivo tracking of administered cells will support the clinical development of cell-based therapies. Radionuclide imaging with gamma and positron-emitting radioisotopes can provide quantitative and longitudinal mapping of cells in vivo. To make this widely accessible and adaptable to a range of cell types, new, versatile and simple methods for directly radiolabelling cells are required. We have developed [111In]In-DTPA-CTP, the first example of a radiolabelled peptide that binds to the extracellular membrane of cells, for tracking cell distribution in vivo using Single Photon Emission Computed Tomography (SPECT). [111In]In-DTPA-CTP consists of (i) myristoyl groups for insertion into the phospholipid bilayer, (ii) positively charged lysine residues for electrostatic association with negatively charged phospholipid groups at the cell surface and (iii) a diethylenetriamine pentaacetate derivative that coordinates the γ-emitting radiometal, [111In]In3+. [111In]In-DTPA-CTP binds to 5T33 murine myeloma cells, enabling qualitative SPECT tracking of myeloma cells’ accumulation in lungs immediately after intravenous administration. This is the first report of a radiolabelled cell-membrane binding peptide for use in cell tracking.

[ 111 In]In-DTPA-CTP binds to 5T33 murine myeloma cells, enabling qualitative SPECT tracking of myeloma cells' accumulation in lungs immediately after intravenous administration. This is the first report of a radiolabelled cell-membrane binding peptide for use in cell tracking.

Background
Cell-based therapies are potentially transformative in oncological, 1 cardiac, 2 neurological, 3 transplant 4 and regenerative therapies. 5 Chimeric Antigen Receptor (CAR) T-cells for treatment of leukaemia and lymphoma have recently entered the clinic, with durable responses in some patients. 1,6 The ability to track the migration and distribution of therapeutic cells in a living organism will (i) aid development of cell-based therapeutics, (ii) allow clinical assessment of migration and safety of small populations of cells in patients before administering larger doses, and (iii) help predict efficacy of cellular therapies and understand outcomes. Whole body in vivo, non-invasive cell tracking methods can give real-time information about the distribution of administered cells. 7 Additionally, in vivo imaging of tumor cells labelled with contrast agents gives detailed information about their metastatic potential in animal models. 8 In direct cell labelling approaches, cells take up contrast agent prior to being administered to subjects or patients. Magnetic resonance imaging (MRI) (with nanoparticles 9 ) and optical imaging (with fluorophores 10 ) have enabled whole body imaging of cell distribution. However, MRI requires large amounts of contrast agent, and optical methods are hampered by low tissue penetration of light for excitation and emission of fluorophores. Neither method is quantitative. Radionuclide imaging methods (Single Photon Emission Computed Tomography (SPECT) and Positron Emission Tomography (PET)) are both quantitative and more sensitive than magnetic resonance and optical imaging. 11 There are several strategies for direct radiolabelling of cells for PET and SPECT imaging. They include radiolabelled receptortargeted compounds that are taken up by cell surface receptors (e.g. [ 18 F]FDG targeting GLUT, 12 radiolabelled antibodies 13 ), radiolabelled compounds with a pendant reactive group for covalent linkage to cell membrane proteins/glycoproteins (e.g. via endogenous primary amines 14 or thiols 15 ), radiolabelled nanoparticles that internalise in cells, 16 and lipophilic, neutral radiometallic complexes (ionophores) that diffuse across cell membranes and dissociate, releasing their radioactive cargo inside the cell. [17][18][19][20] Ionophores including [ 111 In]In-(oxine) 3 19 and [ 99m Tc]Tc-HMPAO 20,21 have been in clinical use since the 1980s to radiolabel leukocytes, for administration to patients to image inflammation and infection with g-scintigraphy or SPECT. Lipophilic zirconium-89-and copper-64-labelled ionophores for PET cell tracking have also been developed. 17,18,22,23 Lastly, in recent years, elegant methods for indirect radiolabelling of cells have been developed: in this approach, cells are genetically modified with a reporter gene that leads to expression of a cell-surface receptor. [24][25][26][27][28] The cells can then be administered to a subject or patient. A complementary radiotracer with high specificity and affinity for the receptor is subsequently used to track this introduced cell population. There are many advantages to this approach. First, the cell population can be tracked over an indefinite period of time, simply by administration of the radiotracer followed by imaging. In direct cell labelling, longitudinal cell tracking is limited by the half-life of the selected radioisotope. Second, provided that reporter gene expression is stable, this approach allows imaging and quantification of expansion or proliferation of the cell population in vivo. In instances where populations of cells require genetic manipulation to elicit a therapeutic response (e.g. CAR T-cells 24,[26][27][28] ), introduction of a reporter gene to enable radionuclide imaging is viable. However, in other instances (e.g. leukocyte imaging), the practical complexity of genetic manipulation of cells (including GMP manufacturing and safety management) is undesirable.
Herein we describe a new approach for direct radiolabelling of cells, 25 utilising a synthetic ''cytotopic'' peptide (CTP) that Smith et al. have pioneered for attachment of functional cargo to the surface of cell membranes. [29][30][31][32][33][34][35][36] CTP consists of two longchain C 13 myristoyl groups for insertion into cellular lipid bilayers, and multiple lysine residues for favourable electrostatic interactions with negatively charged phospholipid groups of the cell surface membrane. 31,35,37 We have prepared a radiolabelled chelator bioconjugate of CTP: to the best of our knowledge, this is the first report of a radiolabelled peptidic bioconjugate designed to associate with cell membranes, for cell tracking with radionuclide imaging. Such a construct could have potential advantages over existing cell radiolabelling platforms: (i) it localises the radionuclide to the cell membrane, increasing the distance between the radionuclide and the cell nucleus and thus potentially decreasing toxicity or impaired cellular function associated with intracellular emission of secondary electrons (e.g. Auger electrons); 38,39 (ii) in contrast to methods that covalently append radiolabelled motifs to cell surface proteins, 14,15 our approach does not involve chemical modification of cell surface proteins, which could be detrimental to cell function and migration; (iii) it is a versatile approach that can be used for radiolabelling many cell types unlike receptor-targeted approaches 13 that are restricted to only some cell types and cannot be applied when the target receptor is not expressed; and (iv) radiosynthesis does not require a large excess of chelator (as is required for preparations of [ 111 In]In-(oxine) 3 , in which traces of oxine can be toxic to cells).
The cell membrane binding properties of this class of cytotopic peptide have been well characterised: both hydrophobic and electrostatic interactions contribute to stable cell membrane binding. 35,37 Cytotopic bioconjugates of the CTP type ( Fig. 1(a)) have been used in vitro and in vivo to anchor proteins that prevent or modulate complement activation, 31,32,36 thrombosis 30,33 and hormone signalling 29 to the cell membrane. They have low systemic toxicity. A CTP-protein bioconjugate of a complement inhibitor is in clinical trials, where it is administered to kidneys prior to transplantation, to decrease transplant rejection and perfusion-related complications. 32 In an ex vivo kidney model, a fluorescent CTP bioconjugate has demonstrated uptake and stable adherence to kidney vasculature in the presence of normal blood flow. 30 CTP is thus a promising agent for delivery of radioactive cargo to the surface of cells for in vivo cell tracking.

Preparation of [ 111 In]In-DTPA-CTP
We elected to radiolabel CTP with the g-emitting isotope, indium-111, as this allows direct preclinical comparison with the clinically used cell tracking agent, [ 111 In]In-(oxine) 3 . To this end, precursor peptide, with the sequence K(a,e-bis-myristoyl)-SSKSPSKKDDKKPGD-C(S-(2-pyridyldithio))-OH, was reduced with tris(2-carboxyethyl)phosphine to cleave the peptide disulfide bond. The reduced peptide was then conjugated at the thiolate of its C-terminal cysteine to a derivative of the diethylenetriaminepentaacetate (DTPA) chelator that contains a pendant maleimide. DTPA was chosen because it is known to complex radiopharmaceutical concentrations of [ 111 In]In 3+ rapidly and stably at room temperature, 40  (in aqueous 0.2 M ammonium acetate, pH 6) furnished [ 111 In]In-DTPA-CTP ( Fig. 1(a)) in radiochemical yields of 496% at specific activities of 16 MBq mg À1 without need for postlabelling purification ( Fig. 1(b)).

In vitro cell binding, viability and retention
The in vivo migration of 5T33 murine myeloma cells in immunocompromised mouse models is well characterised, 18,45 and so these cells were selected for radiolabelling and cell tracking studies with [ 111 In]In-DTPA-CTP.
In preliminary studies, 2 Â 10 6 5T33 cells were incubated with [ 111 In]In-DTPA-CTP or [ 111 In]In 3+ (to quantify any uptake of indium-111 that is not CTP-mediated) in RPMI media supplemented with foetal bovine serum, for up to 6 h to establish an optimal cell radiolabelling protocol ( Fig. 2(a)). Cell uptake of [ 111 In]In 3+ was negligible. Cell radiolabelling yields for [ 111 In]In-DTPA-CTP increased over 0-3 h, to 3.9 AE 0.1%AR (percentage added radioactivity) with no significant gains in yield after 3 h. Thus, an incubation period of 3 h was chosen and this growth medium was used for all subsequent [ 111 In]In-DTPA-CTP 5T33 cell radiolabelling experiments.
[ 111 In]In-(oxine) 3 -labelled 5T33 cells localised to the lungs and then liver 0-2 h PI, consistent with prior studies (Fig. 3(c)). Mice injected with cell-free [ 111 In]In-DTPA-CTP exhibited indium-111 activity in the blood pool and liver 0-2 h PI (Fig. 3(b)). Mice injected with [ 111 In]In-DTPA-CTP-labelled cells exhibited indium-111 activity in the lungs, liver and blood pool (Fig. 3(a)). For the latter group of mice, SPECT/CT data indicated that although some indium-111 radioactivity accumulated in the lungs 0-30 min PI, this decreased over the course of the next 90 min. In contrast to mice administered [ 111 In]In-(oxine) 3 -labelled 5T33 cells, there was significantly greater indium-111 localisation in the liver relative to the lungs.

Paper RSC Chemical Biology
At 1 day PI, SPECT/CT scans were acquired and all animals were culled, with organs harvested for ex vivo radioactivity counting (Fig. 4). There were high concentrations of indium-111 in the liver and spleen of mice administered [ 111 In]In-(oxine) 3 -labelled 5T33 cells or [ 111 In]In-DTPA-CTP-labelled cells, with no statistically significant differences in these organs across these two groups.
We suggest that some of the initial lung indium-111 uptake observed for animals administered In contrast, [ 111 In]In-(oxine) 3 -labelled cells migrate from the lungs to the liver and spleen over a longer time frame, with lower blood pool and liver localisation over 0-2 h PI (Fig. 3) and high liver and spleen localisation by 1 day PI (Fig. 4). This is consistent with data from prior studies. 18,45 Although animals administered [ 111 In]In-DTPA-CTP-labelled 5T33 cells demonstrate high indium-111 radioactivity accumulation in the liver and spleen 1 day PI, this is likely to be a result of accumulation of [ 111 In]In-DTPA-CTP that has dissociated from 5T33 cells.

Protein binding
To better understand the behaviour of [ 111 In]In-DTPA-CTP in biological media, a solution of [ 111 In]In-DTPA-CTP was incubated in murine serum and analysed by both size exclusion HPLC and C 8 reverse phase HPLC. C 8 chromatograms demonstrated that over the course of 24 h, greater than 70% of [ 111 In]In-DTPA-CTP remained intact (Fig. 5(a)), with respect to the stability of the [ 111 In]In-DTPA complex. However, size exclusion chromatograms indicated that [ 111 In]In-DTPA-CTP interacted non-specifically with many serum proteins ( Fig. 5(b)-i,ii). In contrast, size exclusion chromatograms of solutions of [ 111 In]In 3+ in serum indicated that [ 111 In]In 3+ bound to one specific serum component (Fig. 5(b)-iii).  It is also possible that the intracellular location of 111 In in [ 111 In]In-(oxine) 3 -labelled cells, 47 rather than the cell surface location for [ 111 In]In-DTPA-CTP-labelled cells, leads to a corresponding higher degree of radiation damage to nuclear DNA from emitted short-range Auger electrons. 38,39 The majority of PET or SPECT cell tracking agents rely on cellular internalisation of the radionuclide. There are two examples in which the radiolabelled motifs are covalently appended to endogenous cell surface molecules: (i) a ''desferrioxamine'' (DFO) chelator that complexes [ 89 Zr]Zr 4+ has been covalently attached via a reactive isothiocyanate to cell surface primary amines (for PET cell tracking); 14 (ii) a fluorescent motif labelled with 124 I has also been covalently attached to chemically reduced thiol groups on cell surfaces (also for PET cell tracking). 15 In both instances, radiolabelling of cells did not decrease cell viability, similar to the case of [ 111 In]In-DTPA-CTP-labelled cells. However, it is notable that neither 89 Zr nor 124 I emit Auger electrons.

Discussion and concluding remarks
Although [ 111 In]In-DTPA-CTP-labelled cells show prolonged cell viability, the relatively rapid dissociation of [ 111 In]In-DTPA-CTP from cell membranes precludes the use of this agent for quantitative radionuclide cell tracking applications. A CTPprotein bioconjugate, APT070, that prevents organ transplant rejection, is entering clinical trials. 32 Perfusion of APT070 ''paints'' endothelial and epithelial cells of transplant organs with a cell-surface protein that inhibits in vivo complement activation pathways that lead to transplant rejection. APT070 (with a molecular weight of B24 kDa 34 ) is significantly larger than In-DTPA-CTP (2886 Da). It is likely that the increased size of the APT070 cell-surface cargo protects the CTP tail from membrane dissociation for sufficient periods of time posttransplantation, enabling amelioration of transplant rejection.
This first cell tracking study on a radiolabelled cell membrane binding peptide has shown that [ 111 In]In-DTPA-CTP-labelled myeloma cells can be qualitatively tracked to the lungs using SPECT imaging immediately after intravenous administration. Although [ 111 In]In-DTPA-CTP dissociates from cell membranes within 1-2 h post-administration of cells and thus falls short of cell adhesion stability requirements for quantitative long-term cell tracking with SPECT, this proof-of-principle study demonstrates the simplicity and feasibility of using synthetic cell membrane binding/penetrating peptides for radiolabelling cells. Numerous advances in peptide research have enabled efficient cellular attachment and delivery of biomolecules, fluorescent motifs and nanoparticles either via cell membrane localisation or cell membrane penetration. [48][49][50][51] Such peptides will be very efficient vectors for stable delivery of radioactive cargo to cells for quantitative whole body cell tracking with SPECT or PET imaging.

CTP-DTPA synthesis
CTP (K(a,e-bis-myristoyl)-SSKSPSKKDDKKPGD-C(S-(2-pyridyldithio))-OH) (prepared by solid phase peptide synthesis as previously described 31 ) (1 mg) was reacted with tris(2-carboxyethyl)phosphine (2.5 mM) dissolved in dimethyl sulfoxide/water (15%/85%, 1 mL) for 30 min. The solution was applied to a Waters Sep-Pak C18 Plus Short cartridge and the cartridge was washed with a solution (2.5 mL) of acetonitrile/water (20%/80%) containing trifluoroacetic acid (0.1%). Reduced CTP was eluted from the cartridge with acetonitrile containing trifluoroacetic acid (0.1%), diluted in water (to a volume of 5 mL) and lyophilised. Reduced CTP was dissolved in aqueous ammonium acetate (20 mM) containing dimethylsulfoxide (10%) (1 mL total volume) and added to a solution containing an excess of DTPA-maleimide (50 mL, 20 g mL À1 ) (CheMatech, France). After reaction at room temperature for 30 min, the conjugated peptide was applied to a Sep-Pak C18 Plus Short cartridge (Waters) and the cartridge was washed with a solution (2.5 mL) of acetonitrile/ water (35%/65%) containing trifluoroacetic acid (0.1%). Reduced CTP was eluted from the cartridge with acetonitrile containing trifluoroacetic acid (0.1%), diluted in water (to a volume of 5 mL) and lyophilised. binding to chelators in our high resolution mass spectrometric analyses. 52 DTPA-CTP (300 mg) was dissolved in a solution (1 mL) of ammonium acetate (0.2 mM) containing dimethylsulfoxide (10%) and stored at À20 1C for further use. min. An aliquot of the reaction solution was applied to an analytical reverse phase C 8 HPLC column to determine radiochemical yield. HPLC chromatograms were acquired using an Agilent 1200 LC system, with an Agilent Eclipse XDB-C8 5 mM, 4.6 Â 250 mm column coupled to a LabLogic Flow-Count detector with a sodium iodide probe (B-FC-3200). Gradient conditions: 1 mL min À1 flow rate, with 100% A at 0 min, with the concentration of B increasing at a rate of 5% min À1 .
[ 111 In]In-(oxine) 3 radiosynthesis [ 111 In]In-(oxine) 3 was prepared using previously described methods. 18 An aqueous solution of [ 111 In]In 3+ (50 MBq) diluted in water (500 mL) was added to a solution of 8-hydroxyquinoline in chloroform (50 mL, 10 mg mL À1 ). The biphasic mixture was vigorously mixed using a vortex for 5 min. A further aliquot of chloroform (450 mL) was added and the solution mixed using a vortex for 10 min. The chloroform phase was collected and chloroform was evaporated by heating the solution at 50 1C. The residue, containing [ 111 In]In-(oxine) 3 was dissolved in PBS containing dimethyl sulfoxide (2%).
Cell culture 5T33 myeloma cells stably expressing GFP, originating from C57Cl/ KaLwRij strain, 53 were cultured in RPMI media supplemented with 10% FBS, 1% L-glutamine and 1% penicilin/streptomycin. Cells were sub-cultured every two to three days at a ratio of 1 : 10.
Comparative uptake, viability and efflux studies 5T33 myeloma cells (10 5 3 cell labelling was undertaken in serumfree growth media. After cell labelling, cells were washed twice with PBS and 300 mL of fresh growth media was added. Uptake efficiency was quantified at this time-point by measuring pellet and supernatant in the g-counter. At 2, 5 and 20 h after addition of fresh media, cells were washed twice and activity in both the supernatant (original supernatant and washes) and pellet was measured using a g-counter. Efflux was quantified as the activity of the supernatant divided by the total activity (supernatant + pellet). Cell viability was assessed prior to radiolabelling and at each time point by trypan blue exclusion. 54 Serum stability studies  (2 MBq) were incubated in 200 mL mouse serum for 24 h and subsequently analysed using both C 8 and size exclusion HPLC. For C 8 analysis, serum proteins were precipitated by addition of 200 mL acetonitrile to each sample, followed by centrifugation and separation of supernatant from pellet. The supernatant was applied to an analytical reverse phase C 8 HPLC column using conditions described above. For size exclusion, the sample was centrifuged to remove any particulate matter, and the supernatant was applied to an analytical size exclusion column. Analytical size exclusion radio-HPLC traces were acquired using an Agilent 1200 Series HPLC system and a Phenomenex Biosep 2000 (300 Â 7.8 mm) size exclusion column, coupled to a LabLogic Flow-Count detector with a sodium iodide probe (B-FC-3200). Isocratic mobile phase conditions: 1 mL min À1 flow rate of phosphate buffered saline solution. For one animal in each of these groups, a SPECT/CT scan was acquired on a dedicated small animal SPECT system, NanoSPECT/CT Silver Upgrade (Mediso Ltd, Budapest, Hungary), calibrated for indium-111. The whole body SPECT scan time was 30 min Â 4, (conducted sequentially) with a frame time of 40 s (using a 4-head scanner with 4 Â 9 [1.4 mm] pinhole collimators in helical scanning mode) followed by a helical CT (45 kVP X-ray source, 1000 ms exposure time in 180 projections over 7.5 min). After this, animals were allowed to recover. SPECT/CT scans were acquired again, 1 day PI, after which all animals were culled (by increasing the dose of anaesthesia followed by cervical dislocation to confirm death) and organs/tissues harvested, weighed and radioactivity counted using a gamma counter. SPECT/CT images were reconstructed in a 256 Â 256 matrix using HiSPECT (ScivisGmbH), a reconstruction software package and visualised and quantified using VivoQuant v.3.0 software (InVicro LLC., Boston, USA). All animal experiments on mice complied with the Animals (Scientific Procedures) Act (UK 1986) and Home Office (UK) guidelines. No human participants were involved in this study.

Author contributions
J. P. contributed to the study design, prepared the radiotracers, undertook in vitro and in vivo experiments, analysed data and co-wrote the manuscript, J. E. B, T. T. P., A. V. and K. S. contributed to the study design and experiments, P. C. contributed to the study design and undertook preliminary in vitro experiments, G. E. D. M and P. B. contributed to the study design, R. A. G. S. contributed peptide and contributed to the study design, M. T. M. led study design, supervised the project and led compilation of the manuscript.

Conflicts of interest
There are no conflicts to declare.