PHOTOINACTIVATION OF CHINESE HAMSTER CELLS BY HEMATOPORPHYRIN DERIVATIVE AND RED LIGHT

Abstract— The use of hematoporphyrin derivative (HpD) has previously been demonstrated to be beneficial in clinical cancer therapy. This paper describes cell culture studies used to examine HpD phototherapy in Chinese hamster ovary cells (line CHO). Survival curves have been obtained for both direct HpD toxicity and HpD induced photoinactivation. Examination of HpD induced photoinactivation as a function of stage in the cell growth cycle has also been performed, as has the quantitative measurement of HpD uptake in cells (using ³H-HpD) as a function of cellular incubation time, serum concentration in the incubation medium, and cell cycle position. In the absence of light, no toxicity was observed for HpD incubation levels of up to 400 μg/m/ when incubations times were 3 h or less. Exposure of cells to light alone (> 590 nm, 4.0 mW/cm²) for 9 min was also found to be completely nontoxic. Survival curves obtained for exponentially growing cells labeled with various concentrations of HpD and subsequently illuminated with red light exhibited a threshold or shoulder region at short exposure times followed by exponential killing at longer exposure times. The cell cycle response curves for HpD induced photoinactivation of synchronized CHO cells was nearly flat, indicating no variation in sensitivity for cells treated at time periods from 6 to 15 h after mitosis. Additon of serum to the incubation medium resulted in improved plating efficiency and reproducible survival curves but decreased cellular uptake of HpD.     

ACTION SPECTRUM (620–640 nm) FOR HEMATOPORPHYRIN DERIVATIVE INDUCED CELL KILLING

Abstract— Monochromatic red light generated by a tunable dye laser is currently being utilized for the treatment of solid tumors with hematoporphyrin derivative (HpD) photoradiation therapy (PRT). Experiments were performed using mammalian cells to determine the most efficient wavelength of red light (620 to 640 nm range) for HpD induced cellular photoinactivation. Decrease in the clonogenic potential of Chinese hamster ovary (CHO) cells was examined following both short (I h) and extended (12 h) HpD incubation times. Maximal photosensitization was observed with wavelengths ranging from 630 to 632.5 nm and the action spectra for cell killing matched the absorption spectra for HpD bound to cells. Similar observations were obtained following both short and extended HpD-cell incubation times. The potential relevance of these results as they relate to clinical HpD PRT are discussed.     

BINDING OF HEMATOPORPHYRIN DERIVATIVE TO BRAIN TUMOR CELLS—A FLUORESCENCE SPECTROSCOPIC STUDY

The binding of hematoporphyrin derivative (HpD) to brain tumor cells and their photosensitivity was studied as a function of HpD concentration, time of incubation and growth phase of cells. Upon binding to cells, HpD showed three fluorescence bands at 616, 636 and 678 nm. In plateau phase cells a fluorescence band at 636 nm was predominant, which was further enhanced by increasing HpD concentration and/or increasing incubation time. In exponential phase cells the maximum fluorescence was exhibited at 616 nm. After 1 h incubation of exponential phase cells with increasing HpD concentration an overall intensity enhancement occurred with no change in the distribution of bands, whereas longer incubation time caused an increase in relative intensity of the 636 nm band similar to that observed in plateau phase cells. After 1 h incubation with HpD plateau phase cells were more photosensitive than exponential phase cells, although cell bound HpD was much less in the former case. Incubation of cells for 24 h drastically enhanced the photosensitivity irrespective of the growth phase. Our results suggest a relationship between the fluorescence emission band of HpD at 636 nm and photosensitivity of cells.     

LOCALIZATION OF MONO-L-ASPARTYL CHLORIN e6 (NPe6) IN MOUSE TISSUES

Abstract It is known that HpD is retained longer by malignant tissue than normal tissue and is therefore a useful material for photodynamic therapy (PDT). Currently, vigorous research is being conducted throughout the world to discover a new material which can have greater cancer cell affinity than hematoporphyrin derivative (HpD) and will be used effectively for PDT. Investigation has been conducted to determine the spectral characteristics and cancer cell affinity of NPe6, a recently developed material.
Structurally, a double bond on the D-ring of the porphyrin ring of mono-L-aspartyl chlorin e6 (NPe6) has been reduced, thereby changing its spectral properties from that of HpD. This difference accounts for the stronger absorption bands in wavelengths longer than those of HpD. Furthermore, NPe6 in tumor showed stronger absorption at 660 nm than HpD. Absorption by hemoglobin (Hb) in the blood occurs at wavelengths in the range 500-600 nm, thereby lowering light transmittance. A compound which has a strong absorption band at wavelengths longer than 600 nm and consequently is not affected by Hb will naturally be activated by light at a greater depth in tissue than compounds which do not share this characteristic. The localization of NPe6 in sarcoma and various internal organs was examined with an endoscopic spectrophotometer using an excimer dye laser. After 72 h i.v. NPe6 injection, the results indicate that NPe6 has 10 times greater uptake in malignant tissue cells than in normal organs. Based on the above observations, it was concluded that NPe6 could be effective for PDT if toxicity is low and that this compound has a high malignant tissue affinity.     

PHOTODYNAMIC EFFECTS OF PROFLAVINE ON BACTERIOPHAGE φ times 174 AND ITS ISOLATED DNA

Abstract. Proflavine-mediated photoinactivation of φ times 174 phage and its isolated DNA was studied under identical irradiation conditions. The inactivations followed single-hit kinetics and a linear relationship was obtained in reciprocal plots of the inactivation rates vs the proflavine concentrations for both phage and isolated DNA. The phage photoinactivation rate was increased with an increase in the amount of proflavine bound to the phage DNA in a strong binding range (0.01-0.04 proflavine/ nucleotide) as the total proflavine concentration was increased or the ionic strength decreased. Further, a phage-specific factor was also found to affect the inactivation rate. The photodynamic treatment induced mutations in three phage strains from "amber" to "wild type" at a mutation rate per lethal hit of 0.3 times 10^-5 to 2.6 times 10^-5. In contrast to phage infectivity, the φ times 174 DNA infectivity was measurable only at a high multiplicity of infection, and its photoinactivation occurred only at high proflavine concentrations. The photoinactivation rate was enhanced either with a decrease in the multiplicity of infection or with the use of spheroplasts of recA mutants strains. The results are discussed in terms of the nature of and possible repair mechanisms of photodynamically induced lesions in φ times 174 phage DNA.     

PHOTOINACTIVATION OF ENZYMES BY LINEAR AND ANGULAR FUROCOUMARINS

Abstract Furocoumarins with linear (psoralen, 8-methylpsoralen, 8-methoxypsoralen and 3-carbethox-ypsoralen) and angular molecular structures (angelicin and 4,5'-dimethylangelicin) were found to inactivate enzymes to different extents through UV-A irradiation. Moreover, enzymes with different structures (glutamate dehydrogenase, lysozyme, 6-phosphogluconate dehydrogenase, enolase, thermoly-sine and ribonuclease) are inactivated to different extents by the same furocoumarin. UV-A irradiation produces both covalent incorporation of the furocoumarins into the protein molecule and photodegra-dation of amino acid residues; the latter phenomenon seems to be mainly responsible for the photoinactivation process. A close correlation was found between the capacity of the furocoumarins to photoinactivate enzymes and their capacity to modify free amino acids.
A study of the effects of quenchers of various forms of activated oxygen on the photoinactivation of glutamate dehydrogenase, used as a model enzyme, and psoralen and 8-methylpsoralen as a reference for furocoumarins, showed that singlet oxygen is the species most involved in the photoinactivation process.     

LOCAL NEUROTOXICITY OF HEMATOPORPHYRIN DERIVATIVE FOLLOWING INTRACEREBRAL INJECTION

The present study reports on toxicity of hematoporphyrin derivative (HpD) for normal brain tissue in vivo without the addition of light. Hematoporphyrin derivative was injected by slow infusion in rat brains. Histological examination was carried out for intervals after HpD administration, ranging from 0 h to 15 days. Ultrastructural changes were examinated with transmission electron microscopy. The extent of the necrosis was determined for different HpD concentrations and compared with control animals infused with 0.9% saline. Leukocytic infiltration was observed at day 5. Transmission electron microscopy showed that nuclei of neurons were completely disintegrated 4 h after HpD administration. Furthermore disruption of myelin sheaths was observed. The extent of the necrosis decreased with lower HpD doses. Injection of 2 μg HpD in a volume of 4 μL (0.5 mg/mL) resulted in a virtually equal extension of the tissue damage, as compared to the mechanical damage in the control animals caused by the infusion procedure.     

LASER FLASH PHOTOLYSIS AND PHOTOINACTIVATION OF SUBTILISIN BPN'

Abstract— Laser flash photolysis of subtilisin BPN'at 265 nm has shown that photoionization of tryptophanyl (Trp) and tyrosinyl (Tyr) residues are the principal initial photochemical reactions. The initial products are the corresponding oxidized radicals. Trp and Tyr, and hydrated electrons (e_aq) which react with the enzyme at: k (e_aq+ subt. BPN') = 2.1 × 10¹⁰ M⁻¹ s⁻¹. The photoionization quantum yield was 0.032 ± 0.005 at 265 nm, which was enhanced 3.5-fold by simultaneous excitation at 265 and 530 nm. The photoionization yields were unchanged by 3 M bromide ion and 8 M urea. which did affect the enzyme fluorescence excited at 265 and 295 nm. A similar lack of correlation between the effects of perturbants on the photionization yields and fluorescence yields was found for subtilisin Carlsherg. The results indicate that the monophotonic and biphotonic ionization of the Trp residues does not involve the thermally-equilibrated. lowest excited singlet state and that singlet energy transfer from Tyr to Trp does not contribute to Trp photoionization. The photoinactivation quantum yield was 0.014 for 265 nm laser excitation. which was not changed by simultaneous 530 nm excitation. The corresponding quantum yield was 0.009 for low intensity 254 nm radiation, indicative of a biphotonic contribution to photoinactivation. The results are explained by postulating that photolysis of Trp-113 leads to disruption of hydrogen bonding to Asn-117 and a shift in the primary chain sequence associated with the aromatic substrate binding sites. The photoionization quantum yields in subtilisin BPN'and subtilisin Carlsberg agree with a model based on the assumption that exposed Trp and Tyr residues contribute independently at intrinsic photoionization efficiencies characteristic of the chromophores.     

EFFECT OF PHOTODYNAMIC THERAPY ON ANTITUMOR IMMUNE DEFENSES: COMPARISON OF THE PHOTOSENSITIZERS HEMATOPORPHYRIN DERIVATIVE AND CHLORO-ALUMINUM SULFONATED PHTHALOCYANINE

The effects of the two photosensitizers chloroaluminum sulfonated phthalocyanine (ClAlSPc) and hematoporphyrin derivative (HpD) on the functional activities of macrophages and natural killer (NK) cells, two immunocyte populations implicated in the control of tumor development and spread, have been investigated. Murine peritoneal macrophages treated in vivo with ClAlSPc or HpD at 10 mg/kg body weight showed no impairment of Fc-mediated phagocytic capacity and only minor disturbances of in vitro tumoricidal/tumoristatic function. The NK cell activity of splenocytes obtained from photosensitizer-treated mice, assayed 24 or 48 h after i.v. injection of ClAlSPc or HpD at 10 mg/kg was unaffected compared to controls. However significant inhibition of NK activity was observed when splenocytes obtained from mice with or without subcutaneous Colo 26 tumors, treated with ClAlSPc plus laser therapy (675 nm) were used as effector cells. The results show that impairment of some anti-tumor activity can be observed in phthalocyanine treated or phthalocyanine + laser-treated animals but this relatively minor impairment may augur well for the use of systemic phthalocyanine administration in photodynamic therapy.     

SINGLET OXYGEN GENERATION BY PORPHYRINS AND THE KINETICS OF 9,10-DIMETHYLANTHRACENE PHOTOSENSITIZATION IN LIPOSOMES

Abstract— Two new sensitizers are introduced for a potential use in photodynamic therapy: Zn²⁺- and MG²⁺-tetrabenzoporphyrin (ZnTBP and MgTBP). A comparative study of the quantum yields of singlet oxygen generation (Φ_Δ) of hematoporphyrin derivative (HpD), Photofrin II (PF-II), Zn²⁺-phthalocyanine tetrahydroxyl [ZnPC(OH)₄] and the newly introduced sensitizers ZnTBP and MgTBP in liposomes, as well as the kinetics of a photochemical reaction sensitized by them, was made by employing the fluorescent membrane probe 9,10-dimethylanthracene (DMA). We followed the photosensitization of DMA in real time by monitoring its fluorescence decrease at 457 nm and found that DMA's photosensitization is oxygen mediated. The kinetic traces of the photosensitization reactions were fitted to an analytical function, and the Φ_Δ values were evaluated. At 10 μ M sensitizer in an aqueous suspension of 2 mg/mL egg phosphatidylcholine (EPC), HpD was found to have the largest value of Φ_Δ (0.215), followed by PF-II (0.191), ZnTBP (0.023), MgTBP (0.019) and ZnPC(OH)₄ (0.005). As a test of the method, Φ_Δ for methylene blue in ethanol was measured and found to be 0.45 as compared to 0.52 reported in the literature. Due to difference in the sensitizers' absorbances at the laser's wavelength, the reaction photosensitized by ZnTBP was the fastest with a time constant of 6.7 min, followed by MgTBP (8.7), PF-II (11.9), HpD (17.1) and ZnPC(OH)₄ (31.2), all at equal sensitizers' concentrations and laser intensities. The binding constants of the sensitizers to EPC liposomes are also reported.     

FLUORESCENCE SPECTRAL CHANGES OF HEMATOPORPHYRIN DERIVATIVE UPON BINDING TO LIPID VESICLES, Staphylococcus aureus AND Escherichia coli CELLS

Abstract— The binding of hematoporphyrin derivated (Hpd) to lipid vesicles and bacterial membranes was determined by fluorescence spectroscopy. The fluorescence measurements of Hpd in aqueous solutions showed two bands at 613 and 677 nm. In lipid environments of lecithin vesicles the fluorescence spectrum was shifted to 631 and 692 nm, respectively. Hpd was rapidly bound to the cell membrane of Staphylococcus aureus while much less binding occurred in the presence of Escherichia coli. At the same time, spheroplasts of both bacteria were shown to bind Hpd to a similar extent. These results are well correlated with the photoinactivation of the gram positive bacteria with Hpd while the gram negative cells were shown to be resistant. The pH dependence of both Hpd binding to S. aureus as well as the photodynamic inhibitory effect of the same bacteria are similar. It is concluded that the segregation of Hpd to the cell membrane is a prerequisite for its photodynamic effect.     

SUBCELLULAR LOCALIZATION OF HEMATOPORPHYRIN DERIVATIVE IN BLADDER TUMOR CELLS IN CULTURE

Mitochondria have been implicated as a primary subcellular site of porphyrin localization and photodestruction. However, other organelles including the cell membrane, lysosomes and nucleus have been shown to be damaged by hematoporphyrin derivative (HpD) photosensitized destruction as well. In this study we attempted to follow the translocation of the fluorescent components of HpD in human bladder tumor cells (MGH-U1) in culture to determine whether specific subcellular localization occurs over time. Following a 30 min exposure to HpD the cellular fluorescence was examined immediately and 1, 2, 4, and 24 h after HpD removal using fluorescence microscopy and an interactive laser cytometer. The in vitro translocation of dye appeared to be fairly rapid with fluorescence present at the cell membrane and later (1-2 h) within a perinuclear area of the cytoplasm. To determine whether HpD had become concentrated into a specific subcellular organelle, these fluorescence distribution patterns were compared with fluorescent marker dyes specific for mitochondria, endoplasmic reticulum and other membranous organelles. The HpD fluorescence did not appear to be as discrete as the dyes specific for mitochondria or endoplasmic reticulum but appeared similar to the diffuse cytomembrane stain. Finally, the interaction between the fluorescent components of HpD and the cellular constituents was evaluated using a "fluorescence redistribution after photobleaching" technique. The results indicated that the mean lateral diffusion for HpD in MGH-U1 cells was 1.05 x 10(-8) cm2/s, a rate closer to that of lipid diffusion (10(-8)) than that of protein diffusion (10(-10)).(ABSTRACT TRUNCATED AT 250 WORDS)     

FEMTOSECOND STUDIES OF HEMATOPORPHYRIN DERIVATIVE IN SOLUTION: EFFECT OF pH AND SOLVENT ON THE EXCITED STATE DYNAMICS

Abstract We report direct femtosecond measurements of the excited state dynamics of hematoporphyrin derivative (HpD) in solution. The dynamics are found to be very sensitive to the solvent and pH of aqueous solutions. The decay of the excited singlet states is much faster in acidic and pH 7 buffer aqueous solutions (<230 ps) than in basic aqueous solutions or organic solvents (> 10 ns). The dynamical results show strong correlation with static fluorescence measurements: weaker fluorescence in acidic and pH 7 buffer solutions corresponding to shorter-lived excited states. A new fast decay component with a time constant around 5 ps is identified both in acidic aqueous solutions and in organic solvents such as acetone and attributed to internal conversion from the second to the first excited singlet state of aggregates or certain oligomers in HpD, in accord with the observation that the fast decay component is larger at a higher concentration. Oxygen is found to have no effect on the dynamics on the time scale investigated, 1 ns, indicating that oxygen quenching of the singlet excited states is insignificant on this time scale. The sensitive solvent and pH dependence of the excited state dynamics has important clinical implications in the use of HpD as a photosensitizing agent.     

INACTIVATION OF WILD-TYPE AND rad MUTANT Caenorhabditis elegans BY 8-METHOXYPSORALEN AND NEAR ULTRAVIOLET RADIATION

Survival of wild-type and four radiation-sensitive (rad) mutants of the nematode Caenorhabditis elegans was determined after near-UV irradiation in the presence of 8-methoxypsoralen (8-MOP). Three sets of inactivation profiles were generated for each strain by irradiating synchronous populations of either early embryos, late embryos or first-stage larvae (L1s). Late embryos were consistently the most sensitive. Curiously, none of the four rad mutants were even moderately hypersensitive. Split-dose experiments indicated that DNA-DNA crosslinks were primarily responsible for lethality. Crosslink induction and repair were determined using two different assays. In both cases, little if any repair was observed in wild-type. This lack of repair thus explains why the rad mutants were not hypersensitive to 8-MOP photoinactivation. Since early embryos undergo extensive cell cycling, their resistance to 8-MOP photoinactivation suggests that replication is highly refractory to both monoadducts and crosslinks, as has been demonstrated previously for UV radiation-induced photoproducts (Hartman et al., 1991, Mutat. Res., 255, pp. 163-173).     

COMPARISON OF PHOTODYNAMIC ACTION BY ROSE BENGAL IN GRAM-POSITIVE AND GRAM-NEGATIVE BACTERIA

Abstract— The photodynamic inactivation by illuminated Rose Bengal of a number of bacterial species was compared. The gram-positive species, Bacillus subtilis, Staphylococcus aureus, Streptococcus faecalis and Streptococcus salivarius, were inactivated about 200x more quickly (99% inactivation) than a Salmonella typhimurium wildtype strain. The Salmonella inactivation curve exhibited an initial lag time during which bacteria were not significantly inactivated. The lag time for inactivation of a derivative of the wildtype Salmonella strain that is deficient in a large portion of its cell wall lipopolysaccharide coat was approximately half of the lag time for the wildtype strain but the subsequent rate of inactivation was approximately the same for the two strains. Dark preincubation of both Salmonella strains with Rose Bengal before illumination shortened the lag time, but did not increase the final rate of inactivation. Dark preincubation prior to illumination did not measurably change the inactivation curve of the gram-positive species. The lag time observed in the inactivation curves for Salmonella bacteria may reflect the time required for penetration of the Rose Bengal anion through the outer portion of the gram-negative cell wall to a critical location within the cell for effective photosensitization.     

ACUTE SKIN RESPONSE IN ALBINO MICE FOLLOWING PORPHYRIN PHOTOSENSITIZATION UNDER OXIC AND ANOXIC CONDITIONS

Abstract Acute normal skin toxicity induced by porphyrin photosensitization has been examined using albino mice. Oxic and anoxic (clamped) skin was exposed to red light (630 nm) 24 h following administration of hematoporphyrin derivative (HpD) or Photofrin II (the active component of HpD). Experiments were also performed to determine the effect of sodium pentobarbital anesthesia on HpD and Photofrin II photosensitization of normal skin. Results from this study demonstrated that comparable levels of acute skin damage were induced by HpD and Photofrin II under oxic conditions but neither porphyrin produced any apparent phototoxicity under anoxic conditions. In addition, the level of skin damage induced by porphyrin photosensitization was not affected by sodium pentobarbital anesthesia.     

COMPARISON OF THE EFFECTS OF ULTRAVIOLET RADIATION ON THE INTRACELLULAR CATALASE OF AEROBIC AND ANAEROBIC YEAST

Abstract— Suspensioris of aerobic and anaerobic yeast were subjected to ultraviolet radiation (principally 254 mµ ) under closely comparable experimental conditions, and changes in the level and in the temperature dependence of their catalase activity were determined. Qualita tively, the effects of U.V. on the enzyme of the anaerobic cells were similar to those on that of the aerobic cells. The effect of U.V. on the anaerobic catalase differed from that on the aerobic enzyme in the following respects: I, a considerably greater dose of U.V. was necessary in order to attain the maximum activity and the minimum activation energy of the enzyme-substrate system; 2, a far greater dose was required before appreciable photoinactivation of the maxi mally active enzyme occurred; 3, photoinactivation proceeded at less than one-half the rate; 4, the u.v.-induced increase in the catalase activity of the suspension was virtually complete before appreciable reduction in activation energy occured. The first three of these differences were interpreted in terms of a model, which pictures the anaerohic catalase as being tightly bound to an intracellular chromophore group.     

THE EFFECT OF PORPHYRIN PHOTOSENSITIZERS ON THE TOXICITY OF l,3-BIS(2-CHLOROETHYL)-1- NITROSOUREA IN C57BL/6J MICE

Abstract The most widely used agents for photodynamic therapy are the porphyrin photosensitizers. It has been shown that hematoporphyrin derivative (HpD) can cause murine marrow hypercellularity and splenic hypertrophy. We have examined the effect on survival and marrow cellularity of high dose l,3-bis(2-chloroethyl)-l-nitrosourea (BCNU) after HpD or dihematoporphyrin ether (DHE) pretreatment in C57BL/6J mice.
The lethal toxicity of the LD_S0+ 10% dose of BCNU (60 mg kg⁻¹) was significantly reduced by pretreatment with HpD when the HpD was administered at least 3 days prior to the BCNU. HpD administered 1 or 2 days prior to BCNU or after BCNU had no effect. The percent death rate was reduced from 80 to 0% when HpD was administered 7 and 5 days prior to BCNU.
No alteration of the lethal toxicity rate of BCNU at doses of 80 mg kg⁻¹ were identified with DHE pretreatment although some increase in median survival was noted in two groups. Some reduction in lethal toxicity was noted when 60 mg kg⁻¹ BCNU was used and the pretreatment dose of DHE was 10 or 25 mg kg⁻¹ given twice 3 days apart. Furthermore, a significant reduction of BCNU induced marrow cell depletion was found when low doses of DHE were used as pretreatment. High doses of DHE resulted in marrow depletion. Both HpD and DHE altered the toxicity of BCNU.
Should porphyrin photosensitizers, which alone have little toxicity, prove to protect against nitrosourea toxicity then an important dose limiting factor (myelotoxicity) could be altered if not reduction in the tumouricidal activity occurs.     

ABSORPTION SPECTRUM OF HEMATOPORPHYRIN DERIVATIVE in vivo IN A MURINE TUMOR MODEL

Abstract-Time-resolved reflectance was used to measure the absorption spectrum of hematoporphyrin derivative (HpD) in vivo in a murine tumor model. Reflectance measurements were performed in the 600–640 nm range on mice bearing the L1210 leukemia. Then the animals were administered 25 mg/kg body weight of HpD intraperito-neally. One hour later the reflectance measurements were repeated. Fitting of the data using the diffusion theory allowed assessment of the absorption coefficient before and after the administration. As a difference between the latter and the former data, the in vivo absorption spectrum of HpD was evaluated. Maximum absorption was measured at 620–625 nm. Similar spectral behavior was obtained for HpD in solution in the presence of low-density lipoproteins.     

农药久效磷分子印迹聚合物合成及其亲合性评估

有机磷农药药效高、易于被水、酶和微生物所降解，残留毒性小，但由于不合理利用等因素，粮食、蔬菜、水果和环境中的有机磷农药残留严重超标，其危害引起了各国政府、学术界、民众的高度重视．目前有色谱法、酶法和免疫法等，这些方法的应用都受到客观因素的限制．因此研制类似于酶和抗体的高选择性、低造价、稳定性好的吸附材料用于食品和环境中有机磷农药残留的富集、分离纯化和分析尤为重要．