Synthesis of biologically active molecules by imprinting polymerisation

Highly cross-linked molecularly imprinted polymers (MIPs) are synthetic materials with properties mimicking those of natural receptors. Here we describe an ability of МІР nanoparticles to manifest biological activity. Molecularly imprinted polymers were synthesised by co-polymerisation of urocanic acid, N,N'-bisacryloyl piperazine in the presence of herbicide binding Dl protein, ground and separated from the template by washing and ultrafiltration. It was demonstrated that МІР nanoparticles retained affinity to the template. Moreover, imprinted polymers were able to activate chloroplast photosystem II in in vitro experiments. This provides the first example of the use of imprinted polymers for the attenuation of a biological system and opens new possibilities for their application in pharmacology, biotechnology and medicine.

Introduction. Since being first described in 1972, molecularly imprinted polymers (MIPs) have attrac ted broad interest from scientists engaged in the development of chromatographic adsorbents, sensors, catalysts, enzymes and receptor mimics [1][2][3][4]. The particular features which make them the target of intense investigation are striking resemblance of the МІР binding properties to those of natural receptors [4][5][6] and their inherent stability, low cost and ease of preparation [7,8]. However, MIPs suffer in comparison with natural antibodies and receptors in clinical and pharmaceutical applications due to their relatively large particle size and the lack of true water compatibility.
We report here the preparation of water-soluble cross-linked nano-sized МІР particles through the extensive polymer grinding and ultrafiltration. The imprinted polymer was prepared in the presence of thylakoid membrane Dl protein. Dl protein repre sents a good model system because it is stable in vitro, its role in photosynthetic process, where it mediates the electron flow passing through the Pho tosystem II, is well known and its activity can be easy monitored using Hill reaction [9,10]. In Dl structure one can envisage two possible interaction sites for МІР: internal (herbicide binding site) and external, exposed to cytoplasm.
Our intention was to investigate the ability of Dl-specific МІР upon binding to the Dl to alter the photosynthetic process either through the displa-cement of the electron acceptor Q B from herbicide binding site, or through the interaction with Dl external site. Evidence for such a biological activity of synthetic polymers can open a new way for the application of МІР nanoparticles in pharmacology, analytical chemistry and biotechnology as well as in the fundamental research of molecular recognition phenomena. Furthermore, the analysis of the in fluence of extensive grinding on the recognition characteristics of polymer was of intend in order to establish the minimal particle size necessary for maintaining molecular «memory» and specificity.
Materials and Methods. All chemicals were obtai ned from commercial sources and used as received.

Chromatography of the polymers on affinity co lumn with immobilised Dl protein (Dl-column).
To prepare Dl-Sepharose, Dl protein (0.1 g) was added to activated CH-Sepharose 4B (5 g) («Pharmacia Fine Chemicals*, Sweden) in phosphate buffer (10 mM, pH 7.5) and left overnight at 40 °С. The immobi lization rate was 6 mg/g of sorbent. Chromatography column was packed with 3g of Dl-Sepharose. The column was equilibrated with eluent (25 mM sodiumphosphate buffer, pH 7.5, flow rate 0.22 ml/min) at room temperature. An aliquot (100/Л) of the polymer solution (fraction 5-10 kDa) was injected into the column and the retention times of the peaks were measured. All eluted peaks were analyzed addi tionally by GFC.
Influence of the polymers on the activity of thylakoid membranes. The thylakoid activity was measured as previously described [10]. in the microplate wells. The samples were illuminated for 10 min with a 100 W lamp and the absorbance was measured at 530 nm using a Dynatech reader (Ger many). The activity of thylakoid membranes was calculated from the amount of reduced DPIP, concen tration of chlorophyll and the time of illumination.
Results and Discussion. We have polymerised the functional monomer urocanic acid, containing carboxylic and imidazole residues, essential for the protein recognition and water-soluble cross-linker N,N'-bisacryloyl piperazine [11] in the presence of Dl protein. Judging from our experience, UA is a mo nomer superior to methacrylic acid, traditionally used in molecular imprinting, being able to form a stronger complex with a template. It can also be co-poly merised with BAP giving polymer with 99 % yield. The synthesised polymer was ground, washed from the template and centrifuged through microfilters to yield three fractions: < 5 kDa, 5-10 kDa and > 10 kDa. These fractions were analysed by GFC giving peaks with average molecular weight 0.8 kDa and 0.2 kDa (fraction < 5 kDa), 6.7 kDa (fraction 5-10 kDa) and 18 kDa (fraction > 10 kDa). Taking into account the fact that the molecular weight of Dl protein is 32 kDa, we did not expect the template to be present in the polymer fractions. Minor quantities of the protein residues (presumably short peptides) however were found in the fraction with molecular weight < 5 kDa, as was determined using BCA-assay [12].
The affinities of the imprinted and blank polymer (prepared in the absence of a template) for templates were analysed by liquid chromatography using a Sepharose column derivatised with Dl protein (Dlcolumn). The elution profiles of the МІР and blank polymers on the Dl-column are presented in Fig. 1.
The peaks eluted first correspond to residues of non-polymerised monomers. The capacity factors, which reflect polymer affinity, were much higher for the МІР than for the blank polymer (/C MIP =1.11 ± ± 0.06 and KBlank = 0.56 ± 0.05; separation factor a = All measurements were done in triplicate = 1.98). For a non-derivatised column the capacity factor for МІР (fraction 5--10 kDa) was the same as for a blank polymer (A: MIP = 0.60 ± 0.06). This pro vided a clear indication of the role of the imprinting process and proved that the affinity of the МІР to Dl-column results from the interaction of the polymer with the immobilised template. From these results we can assume that at least a part of the polymer specific structure is preserved upon grinding. The biological activity of the polymers was analy sed in vitro using the Hill reaction [9]. Upon illumination thylakoids reduced the dichlorphenolindophenol (DPIP)-electron acceptor, changing its ab sorbance spectra. The activity of thylakoids can be calculated from the amount of DPIP reduced. We had anticipated that the imprinted polymer upon binding to Dl protein would influence this reaction. In our experiments the clear increase (up to 45 % for 30/jg/ml of polymer concentration) in activity of thylakoid membranes was observed (Fig. 2) which suggests that the target of МІР action is not the herbicide-binding site but rather the external site of the Dl protein (Fig. 3).
The ability of polymers to enhance the stability and activity of thylakoids and enzymes has previously been demonstrated [13][14][15], though this effect has been ascribed to the inhibition of proteolysis or non-specific stabilisation of the protein structure by restricting unfolding [16]. In our experiments, how ever, we were able to select the molecular target in the photosystem complex for the interaction with MIPs and enhance the biological effect using imprin ting with Dl protein. A good indication of the selectivity of the interaction is the negligible influence of the blank polymer and monomers (UA and BAP) on the Hill reaction ( Fig. 3 and 4). Additionally the blank polymer and Dl protein were not able to change the activity of thylakoids. We also found that polymer prepared in the presence of BSA had no biological activity. We thus conclude that the acti vation of the thylakoids by Dl-specific МІР is indu ced by the selective interaction with Dl protein so as to provide partial stabilisation of the photosystem complex structure, and preventing its interaction with proteases or free radicals. The influence of МІР particle size on thylakoid activity was also inves tigated (Fig. 4).
The lower activation achieved for the polymer fraction with a particle size smaller than 5 kDa can be explained by the partial destruction of the binding sites. The reason for slightly lower activity of > 10 kDa fraction in comparison with 5-10 kDa fraction could be relative difficulty in permeation of this fraction through the thylakoid membranes. It can be concluded that in order to possess affinity for Dl protein MIPs should contain at least 25-30 monomer units with pre-determined primary monomer se quence.
Conclusions. To our knowledge, this is the first report of а МІР capable of mediating biological activity. These water-soluble MIPs have an affinity to the original template and superior activity to ran domly synthesised polymers. They can be prepared easier than specifically designed discrete organic structures. Although the yield of the polymer is low, МІР nanoparticles can be used in the area of funda mental research of molecular recognition phenomena. It is anticipated that imprinted polymers can be practically useful in analytical chemistry, biotech nology, in the pharmaceutical and food industries.