In vivo protein transduction to the CNS.

Proteins and peptides are useful research and therapeutic tools, however applications are limited because delivery to the desired location is not easily achievable. There are two hurdles in protein/peptide delivery to the brain: the blood-brain barrier and intracellular penetration. Penetration to both brain and the intracellular space can be achieved by adjusting hydrophilicity, and small molecule pharmacological agents have been successfully developed using this approach. But with proteins and peptides, it is difficult to modify the hydrophilicity without influencing biological functions. Trans-acting factor protein from the human immunodeficiency virus contains a highly conserved cationic peptide sequence necessary for transduction across the cell membrane. While trans-acting factor peptide has been used for in vitro protein transduction, its in vivo application is very limited because it is rapidly degraded by proteolysis. Polyethylenimine is a chemically synthesized small molecule cationization agent; the charge density is greater than a peptide-based cationic cluster such as trans-acting factor, and it is resistant to proteolysis in vivo. We first tested intracellular protein transduction following direct brain injection in mice using polyethylenimine-conjugated green fluorescence protein and beta-galactosidase (molecular weights 29 and 540 kDa, respectively). Polyethylenimine-conjugates penetrated to the intracellular space immediately surrounding the injection site within one hour. We further tested polyethylenimine-mediated protein transduction following intranasal administration, which bypasses the blood-brain barrier. Polyethylenimine-conjugates in pH 7.5 solution did not reach the brain, probably because the polyethylenimine-conjugates penetrated into the intracellular space where first exposed to the tissue, i.e. at the nasal mucosae. We temporarily reduced the electrostatic interaction between cationized polyethylenimine-conjugates and cellular surfaces by adjusting the pH to 4.5; solution rapidly reached the brain and penetrated to the intracellular space. This study suggests that polyethylenimine is a useful protein transduction agent in the brain in vivo, and adjusting cationic charge interaction can determine the extent of brain penetration.

Preparation of potent cytotoxic ribonucleases by cationization: enhanced cellular uptake and decreased interaction with ribonuclease inhibitor by chemical modification of carboxyl groups.

Carboxyl groups of bovine RNase A were amidated with ethylenediamine (to convert negative charges of carboxylate anions to positive ones), 2-aminoethanol (to eliminate negative charges), and taurine (to keep negative charges), respectively, by a carbodiimide reaction. Human RNase 1 was also modified with ethylenediamine. Surprisingly, the modified RNases were all cytotoxic toward 3T3-SV-40 cells despite their decreased ribonucleolytic activity. However, their enzymatic activity was not completely eliminated by the presence of excess cytosolic RNase inhibitor (RI). As for native RNase A and RNase 1 which were not cytotoxic, they were completely inactivated by RI. More interestingly, within the cytotoxic RNase derivatives, cytotoxicity correlated well with the net positive charge. RNase 1 and RNase A modified with ethylenediamine were more cytotoxic than naturally occurring cytotoxic bovine seminal RNase. An experiment using the fluorescence-labeled RNase derivatives indicated that the more cationic RNases were more efficiently adsorbed to the cells. Thus, it is suggested that the modification of carboxyl groups could change complementarity of RNase to RI and as a result endow RNase cytotoxicity and that cationization enhances the efficiency of cellular uptake of RNase so as to strengthen its cytotoxicity. The finding that an extracellular human enzyme such as RNase 1 could be effectively internalized into the cell by cationization suggests that cationization is a simple strategy for efficient delivery of a protein into cells and may open the way of the development of new therapeutics.

Three-dimensional structure of human RNase 1 delta N7 at 1.9 A resolution.

Human pancreatic ribonuclease 1 (RNase 1) is considered to be the human counterpart of bovine pancreatic RNase A. Truncation of seven amino-acid residues from the amino-terminal sequence resulted in RNase 1 Delta N7, which has a reduced ribonucleolytic activity and a lower affinity for the human placental RNase inhibitor (PRI). This RNase 1 variant has been cloned, heterologously overexpressed, purified and crystallized. Its crystal structure has been determined and refined using data to 1.9 A resolution. The molecule displays the alpha + beta folding topology typical of members of the RNase A superfamily. The main distinct features found in RNase 1 Delta N7 are basically located in three loops affecting the fitting of the enzyme to the active site of subtilisin and the shape of the B2 subsite. These changes, taken with the lack of the catalytically active residue Lys7, may explain the reduced affinity of RNase 1 Delta N7 for PRI and the low ribonucleolytic activity of the protein when compared with the native enzyme.

Stabilization of human RNase 1 by introduction of a disulfide bond between residues 4 and 118.

In order to stabilize human RNase 1 by introduction of an intramolecular cross-link, a mutant protein (4-118CL RNase 1), in which Arg4 and Val118 are replaced with cysteine residues and linked by a disulfide bond, was designed and expressed in Escherichia coli as inclusion bodies. The 4-118CL RNase 1 that refolded under redox conditions was a monomer without free SH groups and retained 11% of the activity of the wild-type recombinant RNase 1, indicating that the mutant enzyme was correctly folded with the formation of an additional disulfide bond between Cys4 and Cys118. From guanidium chloride denaturation experiments based on the assumption of a two-state transition for unfolding, it was demonstrated that the introduction of the present cross-link increased the thermodynamic stability of RNase 1 by 2.0 kcal/mol. This value was lower than that, 5.4 kcal/mol, theoretically calculated from the reduction of chain entropy of the unfolded state due to the introduction of the cross-link. These results suggest that the present cross-link also destabilized the folded state of RNase 1 by 3.4 kcal/mol. Along with the increase in the thermodynamic stability, the stability of RNase 1 against trypsin digestion was also significantly increased by the introduction of this cross-link. It is likely, although not proven, that stabilized human RNases are favorable for clinical use, because human RNase-based immunotoxins should have long half-lives as to proteolytic degradation after endocytosis.

Convenient and efficient in vitro folding of disulfide-containing globular protein from crude bacterial inclusion bodies.

We investigated how the folding yield of disulfide-containing globular proteins having positive net charges from crude bacterial inclusion bodies was affected by additives in the folding buffer. In screening folding conditions for human ribonucleases and its derivative, we found that addition of salt (about 0.4 M) to a folding buffer increased the folding yield. This suggested that electrostatic interaction between polyanionic impurities such as nucleic acids and cationic unfolded protein led to the formation of aggregates under the low-salt conditions. Since inclusion bodies were found to contain nucleic acids regardless of the electrostatic nature of the expressed protein, the electrostatic interaction between phosphate moieties of nucleic acids and basic amino acid residues of a denatured protein may be large enough to cause aggregation, and therefore the addition of salt in a folding buffer may generally be useful for promotion of protein folding from crude inclusion bodies. We further systematically investigated additives such as glycerol, guanidium chloride, and urea that are known to act as chemical chaperons, and found that these additives, together with salt, synergistically improved folding yield. This study, suggesting that the addition of salt into the folding buffer is one of the crucial points to be considered, may pave the way for a systematic investigation of the folding conditions of disulfide-containing foreign proteins from crude bacterial inclusion bodies.

Inhibition of cell growth by a fused protein of human ribonuclease 1 and human basic fibroblast growth factor.

Pancreatic-type RNases are considered to have cytotoxic potential due to their ability to degrade RNA molecules when they enter the cytosol. However, most of these RNases show little cytotoxicity because cells have no active uptake mechanism for these RNases and because the ubiquitous cytoplasmic RNase inhibitor is considered to play a protective role against the endocytotic leak of RNases from the outside of cells. To study the cytotoxic potential of RNase toward malignant cells targeting growth factor receptors, the C-terminus of human RNase 1 was fused to the N-terminus of human basic fibroblast growth factor (bFGF). This RNase-FGF fused protein effectively inhibited the growth of mouse melanoma cell line B16/BL6 with high levels of cell surface FGF receptor. This effect appeared to result from prolongation of the overall cell cycle rather than the killing of cells or specific arrest in a particular phase of the cell cycle. Thus, human RNase 1 fused to a ligand of cell surface molecules, such as the FGF receptor, is shown to be an effective candidate for a selective cell targeting agent with low toxic effects on normal cell types.

Tissue-specific expression of pancreatic-type RNases and RNase inhibitor in humans.

The tissue-specific expression of five human pancreatic-type RNases and RNase inhibitor was analyzed by Northern hybridization against poly(A)+ RNA prepared from 16 normal tissues. The widespread expression of RNase 1 was observed in almost all of the tissues. RNase 4 and angiogenin showed a similar distribution of expression abundantly present in the liver. This suggested the identity of the cell types producing these two molecules. However, no relativity appeared to be present between the vascularization of the tissues and the angiogenin expression. A narrow range of expression of the eosinophil-derived neurotoxin gene was observed. This localization seems related to the phagocytic cells in the tissues. The undetectable level of the eosinophil cationic protein mRNA in normal tissues suggests that the differentiation of eosinophils, triggered by inflammation and/or atopy, is required. The expression of RNase inhibitor was found to be ubiquitous. The regulatory function of inhibitor against RNases in the cell should be considered in studying the physiological significance of the pancreatic-type RNase family.

Molecular Targeting for Epidermal Growth Factor Receptor Expressed on Breast Cancer Cells by Human Fusion Protein.

Recombinant human ribonuclease 1 (RNase 1) was chemically linked to recombinant human epidermal growth factor (EGF). The cytotoxicity of this conjugate was assayed using MTT assay. The EGF-RNase conjugate showed dose-dependent cytotoxicity against breast and squamous cell carcinomas overexpressing the EGF receptor (EGFR). The cytotoxicity of the conjugate correlated positively with the level of EGFR expression by each cell line. These results suggest that the EGF-RNase conjugate is a more effective anticancer agent with less immunogenicity and toxicity than conventional chimeric breast cancer toxins.

Development of photosynthesis and respiration measurement system using mass spectrometry.

A closed gas-exchange system was developed to measure gross photosynthesis and respiration discriminately and simultaneously. The system developed in this study included a high performance mass spectral analyzer for gas measurements. The gas-exchange system consisted of a 3L assimilation leaf chamber, a 1L flexible metallic bag, gas supply apparatus, and a lighting system. The lights were turned on and gas measurements were started after the initial 12CO2 concentration level was increased to 500 ppm. The 13CO2 gas was added to the chamber 10 min after the start of the light period. The lights were turned off 15 min after the addition of 13CO2. The 12CO2 and 13CO2 concentrations in the chamber during the light and dark periods were measured for pothos and maize leaves. The 13CO2 absorption rate by the leaves was larger than that for 12CO2 during the light period. It was assumed that the 13CO2 absorbed by photosynthesis was not evoluted by respiration during the first 25 min from the start of the light period. Based on this assumption, gross photosynthetic rate and respiration rate were estimated by calculating the difference in uptake rates of 12CO2 and 13CO2 during the light period.

Discriminating and continuous measurement of photosynthesis and respiration by monitoring 13CO2 and 12CO2 as tracers.

The simultaneous and discriminative measurement of the photosynthesis and the respiration of the plant was attained by simultaneous monitoring of 13CO2 and 12CO2 by artificial control of 13CO2 abundance of ambient air. The principle of the measurement is based on the following physiological processes. 6CO2 + 12H2O --> C6H12O6 + 6O2 + 6H2O, 6(13C)O2 + 12H2O --> (13C6)H12O6 + 6O2 + 6H2O, 6CO2 + 12H2(18O) --> C6H12O6 + 6(18O)18O + 6H20. Assuming that respiratory consumption of the new born carbon substrate fixed by photosynthesis is negligible during the measurement, the photosynthetic CO2 consumption VPCO2 and the respiratory CO2 production VRCO2 are measured according to the estimation (1) or (2), (1) for closed method, VPCO2 = k(V0 - V t)¿ F13CO2 + (F12CO2/F13CO2)F13CO2 ¿, VRCO2 = k(V0 - V t)¿ F12CO2 - (F12CO2/F13CO2)F13CO2 ¿, (2) for open method, VPCO2 = kVE ¿ (FI13CO2 - FE13CO2) + (F12CO2/F13CO2)(FI13CO2 - FE13CO2) ¿, VRCO2 = kVE ¿ (FI12CO2 - FE12CO2) - (F12CO2/F13CO2)(FI13CO2 - FE13CO2) ¿ where V0 is initial volume of growth chamber including attached flexible bag, FICO2 is the inlet or initial gas concentration of CO2 and FECO2 is the ambient gas concentration of CO2 in the chamber, V and VE are the sampling rate of mass spectrometer and the ventilation rate of the growth chamber respectively, k is the STPD conversion factor = ¿273(PB-PH2O)/760(273+tE)¿, tE(degrees C) is the ambient gas temperature. In the closed method, the gas container of the growth chamber is circulated, resulting FECO2 is varied according to the balance of consumption and production of CO2, while in the open method VE is controlled to keep FECO2 at a constant value. Both (1) and (2) methods were examined and evaluated on the measurements of komatsuna and maize.

Recombinant human pancreatic ribonuclease produced in E. coli: importance of the amino-terminal sequence.

Human pancreatic ribonuclease 1 (hRNase 1) in the mature form has been produced in E. coli using T7 expression system. The recombinant hRNase 1 protein was solubilized from the inclusion bodies, refolded in glutathione redox system, and purified through chromatographic procedures by utilizing cation-exchange and reversed-phase columns. The ribonucleolytic activity of recombinant hRNase 1 was examined on yeast RNA and cytidylyl-3',5'-adenosine revealing the distinctive ribonucleolytic activity. The activity was perfectly inhibited by human placental RNase inhibitor. Truncation of 7 amino acid residues in the amino-terminal sequence resulted in much reduction in ribonucleolytic activity and in affinity to human placental RNase inhibitor with the disintegration of secondary structures of the protein observed by circular dichroism spectra. The present study has revealed the important contribution of the amino-terminal sequence of hRNase 1 to the characteristics of the protein.

Molecular cloning and expression of human ribonuclease 4 cDNA.

A cDNA coding for human ribonuclease 4 was isolated from a pancreas cDNA library and sequenced. This cDNA (996 bp) includes an entire open reading frame encoding mature protein (119 aa) following signal peptide (28 aa). Expression of mature protein in Escherichia coli showed an apparent molecular mass of about 16 kDa, which was slightly lower than the mature form of human RNase 1, in SDS-PAGE.

Nucleotide sequence encoding human pancreatic ribonuclease.

A cDNA coding for human pancreatic ribonuclease was isolated from a pancreas cDNA library and sequenced. This cDNA (1620 bp) includes an entire open reading frame encoding mature protein (128 aa) following a signal peptide (28 aa) as well as 5'- and 3'-untranslated regions.