Synthesis and biological studies of 5-aminolevulinic acid-containing dendrimers for photodynamic therapy.

Using a convergent growth approach, a series of novel 5-aminolevulinic acid (ALA)-containing dendrimers have been synthesized. In these molecules, ALA residues are attached to the periphery by ester linkages, with amide bonds connecting the dendrons. Three first-generation dendrimers, bearing either 6 or 9 ALA residues, were synthesized by attachment of a tris(Boc-protected ALA)-containing wedge (1) to a di- or tripodent aromatic, or tripodent aliphatic core. Two second generation 18-ALA-containing dendrimers were also synthesized using a 3,3'-iminodipropionic acid spacer unit between wedge 1 and the aromatic core. These compounds differed only in the distance between the core and the linker unit. The Boc-protected dendrimers were deprotected using trifluoroacetic acid and isolated as their TFA salts. The potential of these ALA ester dendrimers as macromolecular prodrugs for photodynamic therapy has been demonstrated in the tumorigenic keratinocyte PAM 212 cell line.

A quantitative assessment of protoporphyrin IX metabolism and phototoxicity in human skin following dose-controlled delivery of the prodrugs 5-aminolaevulinic acid and 5-aminolaevulinic acid-n-pentylester.

BACKGROUND: Topical 5-aminolaevulinic acid (ALA) is widely used in photodynamic therapy (PDT) to generate protoporphyrin IX (PpIX) in the skin. However, other prodrugs may be more effective. OBJECTIVES: The pharmacokinetics of ALA- and ALA-n-pentylester-induced PpIX, together with the phototoxicity after PDT, were compared in human skin in vivo, using iontophoresis as a quantitative drug delivery system. METHODS: A series of six increasing doses of equimolar prodrug solutions was iontophoresed into normal skin of the upper inner arms of 20 healthy subjects. The kinetics of PpIX metabolism in skin (n = 4) and the response to light exposure, performed at 4.5 h (n = 6) and 6 h (n = 10) after application, were assessed by skin surface PpIX fluorescence and postirradiation erythema. RESULTS: ALA and ALA-n-pentylester showed a linear correlation between logarithm of dose and PpIX fluorescence (P < 0.005), and logarithm of dose and skin phototoxicity with irradiation at 4.5 h (P < 0.001 and P < 0.005, respectively) and 6 h (P < 0.05 and P < 0.0001, respectively) after iontophoresis. Higher phototoxicity was observed with ALA-n-pentylester than with ALA when sites were irradiated at 6 h, as indicated by the significantly lower theoretical threshold dose for erythema (P < 0.05) and the shift of the PpIX fluorescence/phototoxicity curve towards greater skin erythema at equal PpIX fluorescence levels. Depth of PpIX fluorescence in skin, as determined by fluorescence microscopy, was similar for both prodrugs, but a more homogeneous distribution of PpIX was seen with the more lipophilic ALA-n-pentylester. CONCLUSIONS: The observed greater phototoxicity of ALA-n-pentylester relative to ALA may be attributable to a more favourable PpIX localization in tissue and/or greater intrinsic toxicity.

Comparison of the pharmacokinetics and phototoxicity of protoporphyrin IX metabolized from 5-aminolevulinic acid and two derivatives in human skin in vivo.

Our novel approach was to compare the pharmacokinetics of 5-aminolevulinic acid (ALA), ALA-n-butyl and ALA-n-hexylester induced protoporphyrin IX (PpIX), together with the phototoxicity after photodynamic therapy (PDT) in human skin in vivo, using iontophoresis as a dose-control system. A series of four increasing doses of each compound was iontophoresed into healthy skin of 10 volunteers. The kinetics of PpIX metabolism (n = 4) and the response to PDT (n = 6) performed 5 h after iontophoresis, were assessed by surface PpIX fluorescence and post-irradiation erythema. Whilst ALA-induced PpIX peaked at 7.5 h, highest PpIX fluorescence induced by ALA-n-hexylester was observed at 3-6 h and no clear peak was seen with ALA-n-butylester. With ALA-n-hexylester, more PpIX was formed after 3 (P < 0.05) and 4.5 h, than with ALA or ALA-n-butylester. All compounds showed a linear correlation between logarithm of dose and PpIX fluorescence/phototoxicity at 5 h, with R-values ranging from 0.87 to 1. In addition, the ALA-n-hexylester showed the tendency to cause greater erythema than ALA and ALA-n-butylester. Fluorescence microscopy (n = 2) showed similar PpIX distributions and penetration depths for the three drugs, although both ALA esters led to a more homogeneous PpIX localization. Hence, ALA-n-hexylester appears to have slightly more favorable characteristics for PDT than ALA or ALA-n-butylester.

The active site structure of ba3 oxidase from Thermus thermophilus studied by resonance raman spectroscopy.

The ba3 cytochrome oxidase from Thermus thermophilus was studied by resonance Raman spectroscopy. The component spectra of both heme groups were determined by using different excitation wavelengths. In the ferric state the heme a3 group reveals resonance Raman marker bands characteristic for two high spin species with the heme iron in an in-plane and an out-of-plane configuration that reflects a coordination equilibrium. This equilibrium obviously results from protonation of one of the axial ligands that is ascribed to a hydroxide. Coordination by its protonated form, a water molecule, may be too weak to keep the heme iron in the porphyrin plane. The corresponding Fe-OH2 stretching mode was attributed to a weak H/D-sensitive band at 464 cm(-1). The coordination equilibrium not only depends on the pH but is also affected by the buffer, the salt concentration, and the binding of the natural redox partner cytochrome c552. These changes of the coordination equilibrium are attributed to the perturbation of the hydrogen bonding network at the catalytic center that is connected to the protein surface via a relay of hydrogen bonds. Environmental changes at the catalytic site are sensitively reflected by the formyl stretching of heme a3. The unique structural properties of the ba3 oxidase may be related to the unusual proton pump efficiency and heme a3 redox potential.

Resonance Raman spectroscopic study of the caa3 oxidase from Thermus thermophilus.

The terminal caa3 oxidase of Thermus thermophilus has been studied by resonance Raman spectroscopy. Using different excitation wavelengths in the Soret band region, it was possible to disentangle the resonance Raman spectra of the fully oxidized and fully reduced state in terms of the component spectra of the individual hemes a, a3, and c. For the heme a and a3 groups, the spectra reveal only minor differences compared to those of beef heart cytochrome c oxidase attributable to subtle modifications of the heme environment. These differences are not more pronounced than those between the oxidases from beef heart and Paracoccus denitrificans confirming the view that this oxidase of Th. thermophilus is a typical member of the aa3 oxidase superfamily. The heme c component spectra display far-reaching similarities with those of c-type cytochromes which serve as mobile electron carriers in the respiratory chain. These results imply that caa3 oxidase represents an integrated version of the noncovalent redox complex between cytochrome c and cytochrome c oxidase in higher organisms. On the other hand, the structural changes of cytochrome c in the noncovalent complex have no counterpart in the heme c component of the caa3 oxidase indicating a specific cytochrome c binding site for the mitochondrial enzyme.

Chromophore-anion interactions in halorhodopsin from Natronobacterium pharaonis probed by time-resolved resonance Raman spectroscopy.

Halorhodopsin of Natronobacterium pharaonis which acts as a light-driven chloride pump is studied by time-resolved resonance Raman spectroscopy. In single-beam experiments, resonance Raman spectra were obtained of the parent state HR578 and the first thermal intermediate HR520. The parent state is structural heterogeneous including ca. 80% all-trans and 20% 13-cis isomers. The resonance Raman spectra indicate that the all-trans conformer exhibits essentially the same chromophoric structure as in the parent states of bacteriorhodopsin or halorhodopsin from Halobacterium salinarium. Special emphasis of the resonance Raman spectroscopic analysis was laid on the C=C and C=N stretching region in order to probe the interactions between the protonated Schiff base and various bound anions (chloride, bromide, iodide). These investigations were paralleled by spectroscopic studies of retinal Schiff base model complexes in different solvents in an attempt to determine the various parameters which control the C=C and C=N stretching frequencies. From these data, it was concluded that in the parent state the anion is not involved in hydrogen bonding interactions with the Schiff base proton but is presumably bound to a nearby (positively charged) amino acid residue. On the other hand, the anion still exerts an appreciable effect on the chromophore structure which is, for instance, reflected by the variation of the isomer composition in the presence of different anions and in the anion-depleted form. In contrast to the parent state, the intermediate HR520 reveals frequency shifts of the C=N stretching in the presence of different anions. These findings indicate a closer proximity of the bound anion to the Schiff base proton which is sufficient for hydrogen bonding interactions. These changes of the anion-chromophore interaction upon transition from HR578 to HR520 may be related to the coupling of the chromophore movement with the anion translocation.

Resonance Raman spectroscopy of the integral quinol oxidase complex of Sulfolobus acidocaldarius.

The integral quinol oxidase complex of Sulfolobus acidocaldarius (DSM 639) was investigated by resonance Raman spectroscopy. The complex includes four heme a groups which constitute two functional entities, a587 and aa3, containing two low-spin hemes and a low-spin as well as a high-spin heme, respectively. RR spectra were obtained from the fully oxidized and fully reduced states of the complex using different excitation wavelengths in the Soret band region in order to disentangle the contributions from the four heme groups. For the oxidized state, this approach allowed for the identification of two spectrally different types of heme a which were assigned to the bishistidine ligated hemes a of aa3 and a587 (type II) and to the additional heme a of a587 which is ligated by a histidine and methionine (type I). The spectra of both heme a types differ substantially from that of beef heart cytochrome c oxidase. In particular, the formyl stretching modes of types II and I are upshifted by 8 and 15 cm-1, respectively, implying a largely hydrophobic environment of the formyl groups in the quinol oxidase of Sulfolobus. Furthermore, the RR spectra of the oxidized state reveal the characteristic marker bands of a five-coordinated and a six-coordinated high-spin state, indicating that heme a3 exists in a coordination equilibrium, which is in sharp contrast to the purely six-coordinated high-spin configuration of heme a3 in any (quinol or cytochrome) oxidases studied so far. Also the formyl stretching mode of heme a3 appears to be unusual as its frequency is substantially lower than in beef heart oxidase. In the fully reduced state, no heterogeneity of heme a3 is observed and also the spectra of the various hemes a are nearly indistinguishable. Moreover, the formyl stretching vibrations of all hemes a and a3 apparently coincide to one prominent peak at 1658 cm-1 characteristic for a non-hydrogen-bonded carbonyl group. This finding is unique compared to other aa3 oxidases in which the formyl stretchings give rise to widely separated bands at approximately 1610 and approximately 1665 cm-1 for heme a and a3, respectively. In both the oxidized and the reduced states, the spectra of the aa3 entity in the integral complex differ significantly from those of the isolated aa3 entity studied previously [Heibel, G., Anzenbacher, P., Hildebrandt, P., & Schäfer, G. (1993a) Biochemistry 32, 10878-10884], indicating substantial interactions between the various subunits of the integral complex.