CORDIS - EU research results

Development of novel MRI and PET probes for the detection of apoptosis

Final Report Summary - APOPTOPROBES (Development of novel MRI and PET probes for the detection of apoptosis)


The membranes of mammalian cells consist of a bilayer of a mixture of different phospholipids. The composition of the phospholipid mixture varies between the inner and outer layers and, in healthy mammalian cells, anionic species (principally phosphatidylserine, PS) are arranged largely on the inner layer. This is not the case In some abnormal cells and a considerable amount of anionic lipids are displayed on the outer membrane surface. This phenomenon is especially prevalent in cells undergoing the early/intermediate stages of apoptosis (programmed cell death), cells of the tumour vasculature, bacteria and viruses. The overall aim of the project is to develop new magnetic resonance imaging (MRI) probes that can be used to detect apoptotic cells (i.e. cells undergoing programmed cell death) by targeting phosphatidylserine. Such molecular probes could eventually be used to monitor the extent and speed of the onset of apoptosis in tumours following a treatment. This could therefore be a good indicator of the treatment outcome. Figure 1 represents a simplification of the concept.

During the first year, three bifunctional molecules, two polyaminocarboxylates based molecules and a molecule based on cyclodecapeptide, were designed and their synthesis started. Details of the ligands are reported Figure 2. The second year of research was focused on finishing up the synthetic route to obtain ligands to be tested by biological assays.

Results and discussion:

DTPA based ligands, (L1):
The free ligand was obtained after the first year and a molecule containing one gadolinium and four zinc ions were obtained during the second year (figure 3). This molecule is fully characterised and it is currently undergoing cell testing (apoptotic vs non apoptotic cells) to validate the efficacy of four recognition units as compared to two recognition units. If the assays are conclusive the data will be written up for publication. The molecule L1(Zn)4Gd was presented as a poster during the Eurobic 16 conference last summer.

Pyridine based ligands, (L2):
The recognition part was optimised and obtained after the first year. Eight steps out of ten were optimised for the synthesis route of the imaging part during the second year (Figure 4). The two last steps should be completed by another member in the group as it is a ligand not previously tested with gadolinium.

Cyclodecapeptide ligands, (L3):
After obtaining the cyclic peptide, four zinc ions were added to be complexed by the DPA motifs (Figure 5). This molecule was characterized by NMR and UV-Vis (See figure 6).
Anionic species are involved in various chemical and biological processes where phosphate groups are residue commonly involved in half of biological protein interaction reactions. We studied the selectivity of the receptor L3(Zn)4 on different anionic phosphate species by using indicator displacement assays (IDAs) are a simple and effective method of determining binding affinity between receptors and analytes. Pyrocatechol violet (PV) was chosen as the indicator for the IDAs. The receptor binds to PV changing the colour from yellow to blue-green, a process which was readily monitored by UV-Vis spectroscopy. Several ions were tested on the complex receptor: PV = 1 : 4; Adenosine triphosphate (ATP), Adenosine monophosphate (AMP), Adenosine diphosphate (ADP), pyrophosphate, phosphate, O-phospho-serine, O-phospho-threonine, phosphor-6 Glucose, inositol hexasulfate and phylic acid. This assay shows selectivity between pyrophosphate and phosphate ions and selectivity between AMP/ADP and ATP (Figure 7). Titrations between the PV/receptor complex and these ions were carried out. They confirmed a high selectivity for pyrophosphate as only 0.6 equivalents of this ion are added to the complex PV:L3(Zn)4 to displace PV entirely (Figure 8). We believe that the formation of intramolecular hydrogen bonds stabilize the pyrophosphate in between two DPA-Zn units (see Figure 9). The titrations also showed selectivity between ATP/ADP and AMP which can be explained by the number of four recognition units on L3(Zn)4 which have tendency to recognise multi-phosphate (see Figure 10). However, the high selectivity for pyrophosphate over ATP was expected and can be attributed in part to the difference in the total anionic charge density involved in the binding. O-P oxygen atoms have a relatively smaller total negative charge in ATP to complex Zn2+ compared to those of PPi. This results in the reduction of the binding affinity. Another factor is the rigidity of the peptide scaffold, the presence of two β-turns gives little flexibility, which can interfere with the larger ATP and ADP binding.
Inositol phospholipids have emerged as key players in a wide variety of cellular functions. Phosphatidylinositol 4,5-bisphosphate [PI(4,5)P2] is by far the most abundant of all phosphoinositides and it has attracted much attention in recent years due to its important role in numerous cellular signaling events and regulations, which in turn impact several human diseases. This particular lipid is recognized in the cell by specific lipid binding domains. In order to determine whether the receptors can successfully compete with protein domains to bind PPis, competitive Enzyme Linked Immunosorbent Assays (ELISA) were employed. ELISA are routinely used to quantify the binding of protein domains to their targets. Five PPis where tested with our receptor, Pi(3,4,5)P3, Pi(4,5)P2, Pi(3,4)P2, Pi3P and Pi4P. Unfortunately for Pi3P, the protein domain FYVE contains a zinc finger domain, so the results were not conclusive. The ELISA results concerning L3(Zn)4 showed a high affinity for Pi(4,5)P2, 13.32 ± 1.94 µM (calculated using the method outlined by Orosz and Ovadi) compared to Pi(3,4)P2; 243 µM ± 1.5 µM resulting in a difference of ionization state of the phosphate ions. The receptor was tested with the tri-phospholipid Pi(3,4,5)P2 and presented no affinity or recognition for it. To finish, we investigated the affinity between our receptor and a monophosphate Pi4P. In comparison, the affinity of Pi4P and Pi(4,5)P2 is lower with our receptor which can be explained by the presence of the four recognition motif which will have a stronger affinity for bi-phosphate compared to monophosphate. However, the affinity is strong (74.95 ± 3.89 µM). These results are extremely promising and are currently being written for publication.

During the two years of my fellowship, I developed three different ligands. The first – L1 – was a polyaminocarboxylate based ligand with four recognition units is currently being tested as a contrast agent for apoptotic cells to validate the importance of increasing the number of recognition units. The second ligand – L3 – with its four recognition units has shown high affinity and selectivity for Pi(4,5)P2 which is the most abundant in cell membrane and also responsible for signalling. These primary results are really promising for future work in the field of protein-lipid inhibition. L2, the third ligand, is a new polyaminocarboxylate-based ligand containing two recognition units and requires more time to complete its synthesis. Only two more steps need to be optimized before complexation and testing.