Microinjection of Macromolecules

Oberholzer, Thomas, see, Fischer, Aline, 6, 37 see also Microinjection of Macromolecules in Giant Vesicles Prepared by Electroformation, 6, 285. 

Oberholzer, T. and Fischer, A., Microinjection of macromolecules in giant vesicles prepared by electroformation, in Giant Vesicles, P.L. Luisi and P. Walde, Eds., John Wiley Sons, New York, 2000. 

Computer-Automated Capillary Microinjection of Macromolecules into Living Cells... 

Microinjection of Macromolecules in Giant Vesicles Prepared by Electroformation... 

The nuclear envelope acts as a permeability barrier for a variety of macromolecules, and DNA is no exception. In 1980, Mario Capecchi demonstrated that when a pBR322-derived plasmid expressing thymidine kinase (TK) was microinjected into the nuclei of TK-deficient mouse fibroblasts, between 50% and 100% of the injected cells showed TK activity at 24 hours post-injection (Capecchi, 1980). By contrast, in over 1000 cytoplasmically injected cells, no gene expression was detected in any cell during the same time frame. Similar results were obtained in another microinjection study using rat TK-deficient cells and a plasmid expressing chloramphenicol acetyl transferase (CAT) driven by a herpes TK promoter. When 1000-2000 copies of the plasmid were injected into the cytoplasm, less than 3% of the activity was seen as compared to cells injected in the nucleus with the same number of plasmids (Graessman et al., 1989). Zabner et al. 

Diffusion of macromolecules in the cytoplasm is slower than diffusion in water this effect is more pronounced with larger molecules (Figure 4.27) [29, 30, 124, 125]. The functional dependence of diffusion coefficient on molecular size is similar to that observed for diffusion of proteins in concentrated polymer solutions or gels [126] (Equations 4-28 and 4-29 are frequently used to analyze diffusion in actin solutions and cytoplasm [127]). Globular proteins—lactalbu-min, ovalbumin, and serum albumin—diffuse approximately five times slower in the cytoplasm of cultured neurons than in water [9]. When size-fractionated dextrans were microinjected into neuron processes, the reduction in diffusion coefficient was greater for larger molecules (Figure 4.27). The filamentous cytoskeleton appears to create this size-dependent reduction in the diffusion... 

Low transfection rates are a further problem. Reasons are, among others, insufficient lysosomal release of the therapeutic DNA from its carrier and/or its rapid enzymatic degradation within the lyso-some or in the cytosol. The DNA that was integrated in the Kposomes must be released to become effective (see Fig. 5). Studies on microinjections of Kpid/DNA complexes directly into the nucleus clearly show reduced transfection rates [54]. As the nuclear pores have mean diameters of only 25 to 50 nm and passage for macromolecules is further controlled by the nuclear pore complex, passive diffusion into the karyosol is limited to particle sizes below 45 kDa [51, 55]. Conventional recombinant plasmids normally have a molecular weight ranging from 50 to 100 kDa, and therefore require active transport into the nucleus [55]. 

Therefore, the experiments were designed in a way that the enzymatic reaction could only begin in the GV. There were basically two different strategies for performing these experiments (1) multiple loading of the selected GV with different substances and (2) loading of the macromolecules into the selected GV by microinjection and addition of the substrate molecules from the external medium. Because nucleotide triphosphates are the substrate molecules for enzymes such as polymerases, the GV membranes had to be made semipermeable. Their inherent permeability toward nucleotides was too low for enzymatic reactions to be carried out in this way. 

In order to perform a simple competition study it is necessary to first determine the minimal concentration of competitor needed to saturate its own transport. To calculate intracellular concentrations of microinjected macromolecules we have taken the accessible volume of the nucleus to be 40 nl and the cytosol to be 500 nl (Gurdon and Wickens, 1983). In practice, saturation concentrations are operationally defined as the minimal unlabeled substrate concentration necessary to reduce significantly the rate of labeled tracer substrate transport. If too high a concentration of unlabeled competitor is used, then nonspecific inhibitory effects on tracer substrate transport can be expected. 

Microinjections into such spherical GVs are possible and, in most cases, these punctured vesicles remain stable over a long period and do not lose their injected material. This work shows that macromolecules such as nucleic acids and proteins can be entrapped by microinjection and that the detection limit with the nucleic acid dye YO-PRO-1 is in the range 10-50fg. With RNAs (mixture of tRNAs or midi variant RNA [13]) this limit is about 10 times higher (A. Fischer and T. Oberholzer, unpublished observation). A limitation of the present microinjection technique is the fact that the vesicles, in which a liquid has been injected, become less stable because of osmotic effects. Therefore, the injections of highly concentrated salt solutions have to be avoided. Such an injection would immediately lead to shrinkage (maybe it does grow, and then the membrane bursts) of the vesicle in such a way that it cannot be observed by this microscopy technique. 

---