Evidence for intracellular aggregation of hypericin and the impact on its photocytotoxicity in PAM 212 murine keratinocytes

We have assessed photoinduced toxicity of hypericin in PAM 212 murine keratinocytes and the relationship between concentration, incubation time and light fluence to evaluate the effect of intracellular aggregation at high concentrations. Confocal microscopy was used to establish the subcellular localization of hypericin at 5 and 50 microM and incubation times of 1 and 3 h. From fluorescence uptake time course studies, intracellular hypericin was demonstrated to exist predominantly in the monomeric form for up to 26 h incubation at 5 microM. However, there was a pronounced aggregation effect at 50 microM, with intracellular hypericin fluorescence levels initially showing an increase followed by a decrease with incubation time. This effect was subsequently shown to exert an effect on the phototoxicity of hypericin. On irradiation, the photocytotoxicity for 1 and 7 h incubation with 50 microM hypericin was comparable, whereas using 5 microM the photocytotoxicity showed good correlation with the intracellular fluorescence measurements at 1 and 7 h incubation.     

Subcellular distributions and excited-state processes of hypericin in neurons

The photodynamic drug, hypericin, is studied in fetal rat neurons using fluorescence microscopy. Hypericin has an extremely high affinity for the cell membrane and is found to a smaller extent in the nucleus. Fluorescent excitation of hypericin is shown to cause irreversible damage to the cell membranes of living neurons. Fixed cells were used to make ultrafast time-resolved measurements to avoid the deleterious effects of long-term exposure to intense light and room temperatures. To our knowledge, these are the first ultrafast time-resolved measurements of the fluorescence lifetime of hypericin in a subcellular environment. Nonexponential fluorescence decay is observed in hypericin in the neurons. This nonexponential decay is discussed in terms of other examples where nonexponential decay is induced in hypericin upon its binding to biomolecules. The nonradiative processes giving rise to the nonexponential hypericin decay are attributed to excited-state electron transfer, excited-state proton transfer or both.     

Hypericin as a Marker for Determination of Myocardial Viability in a Rat Model of Myocardial Infarction

The aim of this study was to investigate the necrosis‐avid agent hypericin as a potential indicator for determination of myocardial infarction (MI). Male Sprague‐Dawley rats (n = 30) weighing 350 ± 20 g were subjected to acute reperfused MI. Animals were divided into four groups (n = 6), in which hypericin was intravenously injected at 0, 1, 2 and 5 mg kg^?1 respectively. One day after injection, rats were euthanized with their hearts excised for qualitative and quantitative studies by means of microscopic fluorescence examination to decide the dosage of hypericin. Another group was injected with hypericin at the decided dose and evaluated by fluorescence macroscopy in colocalization with triphenyltetrazoliumchloride (TTC) and histomorphology. Infarct‐to‐normal contrast ratio and relative infarct size were quantified. Hypericin‐induced red fluorescence was significantly brighter in necrotic than in viable myocardium as proven by a six times higher mean fluorescence density. Mean MI area was 35.66 ± 22.88% by hypericin fluorescence and 32.73 ± 21.98% by TTC staining (R² = 0.9803). Global MI‐volume was 34.56 ± 21.07% by hypericin and 35.11 ± 20.47% by TTC staining (R² = 0.9933). The results confirm that hypericin specifically labeled necrosis, and enhanced the imaging contrast between the infarcted and normal myocardium, suggesting its potential applications for the assessment of myocardial viability.     

Selective Examination of Plasma Membrane–Associated Photosensitizers Using Total Internal Reflection Fluorescence Spectroscopy: Correlation between Photobleaching and Photodynamic Efficacy of Protoporphyrin IX

Fluorescence spectra, fluorescence decay kinetics, photobleaching kinetics and photodynamic efficacy of protoporphyrin IX (PP) were investigated in endothelial cells in vitro after different incubation times. Fluorescence spectra and photobleaching kinetics were determined during total internal reflection (TIR) illumination or epiillumination. Because penetration depth of the excitation light during TIR illumination was limited to about 100 nm, plasma membrane-associated PP was almost selectively examined. Spectra obtained by TIR fluorescence spectroscopy (FS) showed a very low background, where-as spectra obtained by epi-illumination exhibited considerable background by autofluorescence and scattered light. For photobleaching kinetics during TIR illumination after 1 h or 24 h incubation, a biexponential fluorescence decrease was observed with a rapidly and a slowly bleaching portion. After 1 h incubation, the rapidly bleaching portion was the predominant fraction, whereas after 24 h incubation comparable relative amounts of the rapidly and slowly bleaching portion were determined. The rapidly and slowly bleaching portion were assigned to PP monomers and aggregated species in close vicinity to the plasma membrane. Fluorescence decay measurements after epi-illumination support the decrease of PP monomers within the whole cell with increasing incubation time. In contrast to TIR illumination, photobleaching of PP during epi-illumination was characterized by slow monoexponential fluorescence decrease after 1 h or 24 h incubation. Photodynamic efficacy of PP using epi-illumination was found to depend strongly on incubation time. Considerable cell inactivation was determined for short incubation times (1 h or 3 h), whereas photodynamic efficacy was diminished for longer incubation times. Reduced photodynamic efficacy after long incubation times was assigned to the lower amount of photodynamically active monomers determined close to the plasma membrane as well as within the whole cell. In conclusion, TIRFS measurements are suggested to be an appropriate tool for the examination of the plasma membrane-associated photosensitizer fraction in living cells.     

PHOTOREACTION OF 5-METHOXYPSORALEN WITH THYMIDINE and THE THYMINE MOIETY OF ISOLATED and Saccharomyces cerevisiae DNA. CHARACTERIZATION and MEASUREMENT OF THE TWO cis-syn FURAN-SIDE MONOCYCLOADDUCTS

The photoreaction of the furan-side moiety of 5-methoxypsoralen (5-MOP) with thymidine used as a DNA model compound was investigated in the dry state. Under these conditions, two main fluorescent photoadducts were formed and isolated by HPLC. The two modified nucleosides were characterized as the two cis-syn diastereoisomers of furan-side monoadducts of 5-MOP to thymidine on the basis of spectroscopic measurements including UV, fluorescence, ¹H-NMR and circular dichroism analysis. The identification and quantification of the latter photoproducts within naked DNA exposed to photoexcited 5-MOP were achieved by enzymatic digestion completed by HPLC separation and fluorescence detection. Similarly, the two cis-syn furan-side monoadducts were found to be formed in the DNA of Saccharomyces cerevisiae cells after incubation with 5-MOP and subsequent exposure to 365 nm at an incident dose of 38.4 kJ m^?2. Under these conditions, the rate of induction of two diastereoisomeric photoadducts was as low as one modification per 10⁶ and 2 × 10⁵ bases, respectively.     

Hypericin-Induced Phototoxicity in Cultured Fibroblasts and Swine Erythrocytes

Hypericin is a naturally occurring photosensitizer, whose presence in plants has been responsible for cutaneous phototoxicity in grazing animals. The photosensitizing properties of this agent have recently been exploited in models for anti-tumor and anti-viral activity. The cytotoxicity of hypericin and light was assessed in 3T3 mouse fibroblasts using the MTT [3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyl tetrazolium bromide)] assay and the lactate dehydrogenase (LDH) leakage assay. Membrane damage was assessed in swine erythrocytes using hemolysis, potassium (K⁺) leakage and formation of lipid hydroperoxides. Concentration- and light-dependent decreases in fibroblast viability were seen starting at hypericin concentrations of 1.25 μM and light power flux levels of 24 J/cm² using a visible light source and at 0.417 μM hypericin and a similar light dose using a solar simulator, No LDH leakage was observed at hypericin concentrations up to 30 μM and visible light up to 144 J/cm². Light-and/or concentration-dependent increases in hemolysis, K⁺ leakage and formation of lipid hydroperoxides in red blood cell (RBC) membranes were observed, but at concentrations and light doses much greater than those required to induce cytotoxicity in fibroblasts. Lipid peroxidation and hemolysis occurred at 15 μM hypericin and 24 J/cm² (visible light source). Potassium ion leakage occurred at concentrations and light levels as low as 5 μM and 12 J/cm² or 15 μM and 4.8 J/cm² (visible light source) but was still a less sensitive indicator than fibroblast cytotoxicity. Evidence for both type I and type II reactions was shown in RBC membranes by TLC analysis of cholesterol products. In the absence of light, hypericin appears to be relatively nontoxic in the models tested.     

Photodynamic Mechanisms induced by a Combination of Hypericin and a Chlorin Based‐Photosensitizer in Head and Neck Squamous Cell Carcinoma Cells

The aim of this study was to elucidate photodynamic therapy (PDT) effects mediated by hypericin and a liposomal meso‐tetrahydroxyphenyl chlorin (mTHPC) derivative, with focus on their 1:1 mixture, on head and neck squamous cell carcinoma cell lines. Absorption, excitation and photobleaching were monitored using fluorescence spectrometry, showing the same spectral patterns for the mixture as measured for single photosensitizers. In the mixture mTHPC showed a prolonged photo‐stability. Singlet oxygen yield for light‐activated mTHPC was Φ_Δ = 0.66, for hypericin Φ_Δ = 0.25 and for the mixture Φ_Δ = ~0.4. A linear increase of singlet oxygen yield for mTHPC and the mixture was found, whereas hypericin achieved saturation after 35 min. Reactive oxygen species fluorescence was only visible after hypericin and mixture‐induced PDT. Cell viability was also more affected with these two treatment options under the selected conditions. Examination of death pathways showed that hypericin‐mediated cell death was apoptotic, with mTHPC necrotic and the 1:1 mixture showed features of both. Changes in gene expression after PDT indicated strong up‐regulation of selected heat‐shock proteins. The application of photosensitizer mixtures with the features of reduced dark toxicity and combined apoptotic and necrotic cell death may be beneficial in clinical PDT. This will be the focus of our future investigations.     

RUTHENIUM RED INHIBITION OF OXYGEN EVOLUTION AND SPECIFIC RELEASE OF THE EXTRINSIC 16 kDa POLYPEPTIDE IN A PHOTOSYSTEM II PREPARATION

The inhibitory effect of the dye ruthenium red was studied in photosystem II-enriched submembrane fractions. A number of distinct types of interaction were found, which differed in their concentration range and required incubation time. Ruthenium red instantaneously quenches the initial chlorophyll a fluorescence level (F₀) and the maximum fluorescence level (F_m) by enhancing radiationless deactivation in the chlorophyll light harvesting complex. Associated with this quenching of fluorescence is an instantaneous decrease in the quantum yield of oxygen evolution. Ruthenium red also inhibited the light saturated rate of oxygen evolution and the variable fluorescence, monitored 80 µs after a saturating excitation-flash. These inhibitions increased with incubation time and became greater than 50% within 5 min. Although ruthenium red was known to affect Ca²⁺ or Cl^? sites specifically, the inhibitory action was more pronounced than simple Ca²⁺ or Cl^? depletion. Incubation with ruthenium red for 5 min blocks the Z P680⁺ → Z⁺ P680 charge transfer reaction. Upon mixing with the photosystem II preparation, ruthenium red induced specific release of the extrinsic 16 kDa polypeptide associated with water-splitting without release of Mn. It is proposed that the inhibitor produces an ionic imbalance which alters the configuration of the donor side of photosystem II.     

Synthesis and In Vitro Photodynamic Activities of an Integrin‐Targeting cRGD‐Conjugated Zinc(II) Phthalocyanine

A 1,4‐disubstituted zinc(II) phthalocyanine conjugated with a cyclic Arg‐Gly‐Asp‐D ‐Phe‐Lys (cRGDfK) moiety through a triazole linker was prepared and characterized by UV/Vis spectroscopy and high‐resolution ESI‐MS. The conjugate showed a relatively weak fluorescence emission in N,N‐dimethylformamide (Φ_F=0.08), but it was a very efficient singlet oxygen generator (Φ_Δ=0.80) as a result of the di‐α‐substituted structure. Owing to the presence of the cyclic peptide sequence cRGDfK, which is a well‐known α_vβ₃‐integrin antagonist, this conjugate exhibited significantly higher cellular uptake toward the α_vβ₃⁺ U87‐MG cells compared with the α_vβ₃^? MCF‐7 cells, as determined by flow cytometry and fluorescence microscopy. The photocytotoxicity of this compound against these two cell lines, however, was comparable owing to the similar efficiency of intracellular reactive oxygen species generation. Confocal microscopic studies also revealed that this conjugate localized preferentially in the lysosomes, but not in the nucleus, endoplasmic reticulum, and mitochondria of the U87‐MG cells.     

On the homo- and heteroassociation of hypericin

Summary Hypericin exhibits rather complicated homo- and heteroassociation behavior. Whereas in common polar solvents hypericin dissolves monomolecularly up to concentrations of 10^–3 mol/l, the presence of water in these solvents leads to homoassociation. As derived by spectroscopic measurements, these homoassociates exhibit a stacking pattern similar to the one observed for the crystalline material. Tetrahydrofuran seems to be an exception, as it is the only solvent which results in 1,6-dioxo tautomer formation. Heteroassociation of hypericin involves two distinct types of behavior. In the majority of cases, hypericin forms homoassociates which then heteroassociate with the co-solvate to yield stabilized solutions of these homoassociates. Only with human serum albumin a specific heteroassociate is formed. By means of competition experiments it could be established that hypericin is binding to the active site of the IIIA subdomain of the protein.Dedicated to Prof. Dr.Karl Schlöpl on the occasion of his 70th birthday     

A Remarkable Photoreaction of 3‐O‐Benzylhypericin

Irradiation of 3‐O‐benzylhypericin dissolved together with at least 1 equiv. of `proton sponge' in benzene was shown, by means of UV/VIS, ¹H‐NMR, and mass spectrometric measurements, to produce the rearranged blepharismin‐analogous 11‐phenyl‐11H‐benz[4,10]anthra[2,1,9,8‐nopqa]pleiadene system. The latter compound yielded hypericin when treated with light and oxygen in acetone solution, followed by aqueous NH₄Cl. This reaction is considered to proceed via the oxybleopharismin‐analog 3,4‐benzal acetal of hypericin. The implications of the novel phototransformation for synthesis and biosynthesis of the natural products, as well as the structural peculiarities of the photoproducts, are discussed.     

Dose-dependent induction of apoptosis or necrosis in human cells by organotin compounds

In the present study, the mode of cell death induced by highly toxic trialkylated tin compounds has been evaluated. Treatment of undifferentiated HL-60 cells with submicromolar to micromolar concentrations of tri-n-butyltin (TBT) led only to a slight decrease in cell viability measured with trypan blue exclusion. Nevertheless, cell membrane blebbing was observed by means of light microscopy and condensation of nuclear chromatin and formation of apoptotic bodies was demonstrated in Hoechst 33342 stained cells. The nuclear chromatin condensation was associated with an extensive DNA fragmentation. Visualized by agarose gel electrophoresis, genomic DNA appeared as a characteristic ladder-like pattern of DNA fragments which is the biochemical hallmark of apoptosis. The typical internucleosomal DNA digestion was concentration-dependent and began within 2 to 3 h of incubation. During the incubation period a persistent and steady elevation of intracellular free calcium concentration ([Ca²⁺]_i) could be detected. Furthermore, the chromatin condensation and DNA fragmentation could be blocked by supplementation of the incubation medium with zinc pointing to an activation of a zinc-sensitive and calcium-dependent endogenous endonuclease. Higher concentrations of tributyltin (≥ 5 μmol/L TBT) led within hours to a cell killing with degenerative changes indicative of necrosis, demonstrated by plasma membrane disruption which was accompanied by random DNA breakdown. Furthermore, these concentrations also provoked a persistent elevation in [Ca²⁺]_i which reached, even after 10 min, higher levels in comparison with apoptosis-inducing concentrations. The loss in membrane integrity observed at these concentrations of TBT could be due to an activation of calcium-dependent phospholipases. Here it is shown that activation of cytosolic phospholipase A₂ (cPLA₂) leads to liberation of arachidonic acid (AA) out of the phospholipid membrane. The results presented here demonstrate that organometals are able to induce different cell death pathways depending on the applied concentration: low concentrations led to apoptosis whereas high concentrations stimulate necrosis. We suggest that there exists a direct correlation between the intracellular free calcium concentration and the mode of cell death. Received: 1 August 1997 / Revised: 8 October 1997 / Accepted: 10 October 1997     

Dose-dependent induction of apoptosis or necrosis in human cells by organotin compounds

In the present study, the mode of cell death induced by highly toxic trialkylated tin compounds has been evaluated. Treatment of undifferentiated HL-60 cells with submicromolar to micromolar concentrations of tri-n-butyltin (TBT) led only to a slight decrease in cell viability measured with trypan blue exclusion. Nevertheless, cell membrane blebbing was observed by means of light microscopy and condensation of nuclear chromatin and formation of apoptotic bodies was demonstrated in Hoechst 33342 stained cells. The nuclear chromatin condensation was associated with an extensive DNA fragmentation. Visualized by agarose gel electrophoresis, genomic DNA appeared as a characteristic ladder-like pattern of DNA fragments which is the biochemical hallmark of apoptosis. The typical internucleosomal DNA digestion was concentration-dependent and began within 2 to 3 h of incubation. During the incubation period a persistent and steady elevation of intracellular free calcium concentration ([Ca²⁺]_i) could be detected. Furthermore, the chromatin condensation and DNA fragmentation could be blocked by supplementation of the incubation medium with zinc pointing to an activation of a zinc-sensitive and calcium-dependent endogenous endonuclease. Higher concentrations of tributyltin (≥ 5 μmol/L TBT) led within hours to a cell killing with degenerative changes indicative of necrosis, demonstrated by plasma membrane disruption which was accompanied by random DNA breakdown. Furthermore, these concentrations also provoked a persistent elevation in [Ca²⁺]_i which reached, even after 10 min, higher levels in comparison with apoptosis-inducing concentrations. The loss in membrane integrity observed at these concentrations of TBT could be due to an activation of calcium-dependent phospholipases. Here it is shown that activation of cytosolic phospholipase A₂ (cPLA₂) leads to liberation of arachidonic acid (AA) out of the phospholipid membrane. The results presented here demonstrate that organometals are able to induce different cell death pathways depending on the applied concentration: low concentrations led to apoptosis whereas high concentrations stimulate necrosis. We suggest that there exists a direct correlation between the intracellular free calcium concentration and the mode of cell death. Received: 1 August 1997 / Revised: 8 October 1997 / Accepted: 10 October 1997     

INTERACTION OF THE PEPTIDE HORMONE ADRENOCORTICOTROPIC ACTH(l-24), WITH A MEMBRANE MODEL SYSTEM: A FLUORESCENCE STUDY*

The peptide hormone adrenocorticotropin and a related peptide were studied in solution and in interaction with a model system of membranes (small unilamellar vesicles of dipalmitoylphosphatidylcholine and 17% dimyristoylphosphatidylglycerol) via fluorescence spectroscopy. In aqueous solution, intramolecular distances between the fluorescent residues R(Tyr²-Trp⁹) = 9.2 Å and R(Trp⁹-Tyr²³) 18 Å were obtained, in agreement with molecular models. Interaction of the peptide with the negatively charged membrane is evident from the alteration of the Trp photophysical parameters (quantum yield, fluorescence spectra and anisotropy), with a partition constant between the lipidic and aqueous phase of Kp =1–2 times 10³. The existence of two populations of Trp in the membrane, which are distinctly accessed by acrylamide, was concluded from the tryptophan fluorescence quenching study; the two fractions are located near the membrane interface as inferred from its fluorescence quenching by the 5-doxylstearate and 16-doxylstearate lipophilic quenchers. This result is further supported by energy transfer experiments to the 3-(9-anthroyloxyl)stearic acid and 12-(9-anthroyloxyl)stearic acid probes.     

CELLULAR UPTAKE OF HYDROXYETHYLVINYLDEUTEROPORPHYRIN(HVD) AND PHOTOINACTIVATION OF CULTIVATED HUMAN LEUKEMIA (REH6) CELLS

Abstract— The in vitro incorporation of purified hydroxyethylvinyldeuteroporphyrin (HVD) into cells (Reh6) derived from an acute lymphocytic human leukemia is investigated using quantitative extraction techniques and fluorescence spectroscopy. A fast incorporation step (< 2 min) is characterized by its dependence on the porphyrin concentration in the incubation medium which suggests a saturation process. It is followed by a slower uptake, the rate of which linearly depends on the porphyrin concentration. No preferential uptake of aggregated form of HVD, which is shown to dimerize with an equilibrium constant of 9.7 × 10⁵ M ⁻¹, can be evidenced. As inferred from fluorescence spectra of cell suspensions and those of HVD dissolved in aqueous and micellar solutions as references, the porphyrin is mainly located in membrane structures and to a lower extent in cytoplasm. Cell photoinactivation does not depend on the incubation time but is only related to the intracellular porphyrin concentration.     

Fluorescence imaging of Foscan and Foslip in the plasma membrane and in whole cells

A fluorescence microscope equipped with a condenser for total internal reflection (TIR) illumination was combined with a pulsed laser diode and a time-gated image intensifying camera for fluorescence lifetime measurements of single cells. In particular, fluorescence patterns, decay kinetics, and lifetime images of the lipophilic photosensitizers Foscan and Foslip were studied in whole cells as well as in close vicinity to their plasma membranes. Fluorescence lifetimes of both photosensitizers in cultivated HeLa cells decreased from about 8 ns at an incubation time of 3 h to about 5 ns at an incubation time of 24 h. This seems to result from an increase in aggregation (or self-quenching) of the photosensitizers during incubation. Selective measurements within or in close proximity to the plasma membrane indicate that Foscan and Foslip are taken up by the cells in a similar way, but may be located in different cellular sites after an incubation time of 24 h. A combination of TIR and fluorescence lifetime imaging microscopy (FLIM), described for the first time, appears to be promising for understanding some key mechanisms of photodynamic therapy (PDT).     

ω,ω′-Appended nucleo-base derivatives of hypericin

ω,ω′-Disubstituted hypericin derivatives with the nucleo-bases thymine, cytosine, and adenine in these positions were prepared starting from tri-O-methyl-ω-bromoemodin. The most promising derivative proved to be that with a thymine moiety. It displayed the best solubility of the three products together with a potency to produce singlet oxygen and/or reactive oxygen species comparable to the parent compound hypericin. In addition, although no specific interaction with DNA or poly(2′-deoxyadenylic acid) could be detected, it proved to be significantly better accumulating in the nucleus of prostatic cancer LNCaP cells than hypericin making it a promising candidate for a second-generation photodynamic hypericin agent.     

New fluorescent probes for the measurement of cell membrane viscosity

BACKGROUND: Molecular rotors are fluorescent molecules that exhibit viscosity-dependent fluorescence quantum yield, potentially allowing direct measurements of cell membrane viscosity in cultured cells. Commercially available rotors, however, stain not only the cell membrane, but also bind to tubulin and migrate into the cytoplasm. We synthesized molecules related to 9-(dicyanovinyl)-julolidine (DCVJ), which featured hydrocarbon chains of different length to increase membrane compatibility.RESULTS: Longer hydrocarbon chains attached to the fluorescent rotor reduce the migration of the dye into the cytoplasm and internal compartments of the cell. The amplitude of the fluorescence response to fluid shear stress, known to decrease membrane viscosity, is significantly higher than the response obtained from DCVJ. Notably a farnesyl chain showed a more than 20-fold amplitude over DCVJ and allowed detection of membrane viscosity changes at markedly lower shear stresses.CONCLUSIONS: The modification of molecular rotors towards increased cell membrane association provides a new research tool for membrane viscosity measurements. The use of these rotors complements established methods such as fluorescence recovery after photobleaching with its limited spatial and temporal resolution and fluorescence anisotropy, which has low sensitivity and may be subject to other effects such as deformation.     

Concerning the Diastereomerization of Stilbenoid Hypericin Derivatives

 Experiments on a newly prepared (E)-configured ω-benzal-hypericin derivative using TLC and ¹H NMR together with quantum chemical calculations revealed that in stilbenoid hypericin derivatives photodiastereomerization between the (E)- and (Z)-diastereomers occurs in principle. However, due to its low diastereomerization quantum yield and photo and thermal equilibria, which reside mostly on the side of the (E)-diastereomer, this photoreaction is only of marginal importance to the photochemistry of stilbenoid hypericin derivatives. Thus, photodiastereomerization does not appreciably interfere with the photoreactions important for photodynamic therapy. This was demonstrated by comparing the sensitized bilirubin photodestruction of hypericin and the ω-benzal-hypericin derivative.     

Comparison of subcellular responses for the evaluation and prediction of the chemotherapeutic response to cisplatin in lung adenocarcinoma using Raman spectroscopy

Confocal Raman Micro-spectroscopy (CRM) is employed to examine the chemical and physiological effects of anticancer agents, using cisplatin and A549 adenocarcinoma cells as a model compound and test system respectively. Spectral responses of the membrane and cytoplasm of the cell are analysed independently and the results are compared to previously reported spectroscopic studies of the nucleus. Moreover, Raman spectra from the proteins extracted from the control and exposed samples are acquired and analysed to confirm the origin of the molecular changes of the cell membrane and cytoplasm of the A549 cells. Multivariate data analysis techniques including Principal Component Analysis (PCA) and Partial Least Squares Regression (PLSR) along with PLS-Jackknifing are used to analyse the data measured from the cell membrane and cytoplasm of the A549 cells and results are correlated with parallel measurements from the cytotoxicity assay MTT. A PLSR model is used to differentiate between the chemical effect of the chemotherapeutic agent and the physiological response of the A549 cells and to identify regions of the spectrum that are associated with these processes respectively. The PLSR model is also employed to predict, on the basis of the Raman spectra, the effective dose as well as the level of physiological response, using spectra data from the cytoplasmic and cell membrane regions. The effectiveness of the models based on spectral datasets from the cell membrane and cytoplasm is compared to similar models constructed using spectral data from the nuclear region as well as one combining spectral data from all regions. In all cases, higher prediction accuracy is found for regression against the cisplatin dose, and for both regression against the dose and the physiological response, nuclear data yield higher precision.