The acute effects of daily nicotine intake on heart rate--a toxicokinetic and toxicodynamic modelling study.

Joint physiologically-based toxicokinetic and toxicodynamic (PBTK/TD) modelling was applied to simulate concentration-time profiles of nicotine, a well-known stimulant, in the human body following single and repeated dosing. Both kinetic and dynamic models were first calibrated by using in vivo literature data for the Caucasian population. The models were then used to estimate the blood and liver concentrations of nicotine in terms of the Area Under Curve (AUC) and the peak concentration (Cmax) for selected exposure scenarios based on inhalation (cigarette smoking), oral intake (nicotine lozenges) and dermal absorption (nicotine patches). The model simulations indicated that whereas frequent cigarette smoking gives rise to high AUC and Cmax in blood, the use of nicotine-rich dermal patches leads to high AUC and Cmax in the liver. Venous blood concentrations were used to estimate one of the most common acute effects, mean heart rate, both at rest and during exercise. These estimations showed that cigarette smoking causes a high peak heart rate, whereas dermal absorption causes a high mean heart rate over 48h. This study illustrates the potential of using PBTK/TD modelling in the safety assessment of nicotine-containing products.

The effect of EIF dynamics on the cryopreservation process of a size distributed cell population.

Typical mathematical modeling of cryopreservation of cell suspensions assumes a thermodynamic equilibrium between the ice and liquid water in the extracellular solution. This work investigates the validity of this assumption by introducing a population balance approach for dynamic extracellular ice formation (EIF) in the absence of any cryo-protectant agent (CPA). The population balance model reflects nucleation and diffusion-limited growth in the suspending solution whose driving forces are evaluated in the relevant phase diagram. This population balance description of the extracellular compartment has been coupled to a model recently proposed in the literature [Fadda et al., AIChE Journal, 56, 2173-2185, (2010)], which is capable of quantitatively describing and predicting internal ice formation (IIF) inside the cells. The cells are characterized by a size distribution (i.e. through another population balance), thus overcoming the classic view of a population of identically sized cells. From the comparison of the system behavior in terms of the dynamics of the cell size distribution it can be concluded that the assumption of a thermodynamic equilibrium in the extracellular compartment is not always justified. Depending on the cooling rate, the dynamics of EIF needs to be considered.

Enhancing the X-ray contrast of polymeric biochromatography particles for three-dimensional imaging.

A major limitation of the three-dimensional imaging of polymeric biochromatography particle packings using X-ray computed tomography is that the particles have a low density and a high porosity, making them almost undistinguishable from the surrounding liquid phase. Additionally, the employed media are typically composed of materials with low atomic numbers, which exhibit low X-ray absorption. We report an improvement of packed column reconstruction using micro X-ray computed tomography. A simple, inexpensive, and fast method to increase the contrast factor of highly porous polymer-based chromatographic particles was developed by applying a modified pore-blocking method. This approach relies on the selective filling of the porous chromatographic particles with a hydrophilic phase while a hydrophobic phase occupies the void spaces between the particles. The hydrophilic phase contains a dissolved X-ray absorbing radiocontrast agent. No chemical modifications of the chromatographic beads or columns were necessary. The developed method can be applied in-situ in a previously packed column and can be used for media with different organic backbones. We show the applicability of this method by carrying out the first 3D-reconstruction of packed micro columns with an inner diameter of 760 µm. The micro column contained agarose- and methacrylate-based particles commonly used in preparative biochromatography with mean diameters of 40 and 65 µm, respectively. Based on the obtained high-resolution 3D-reconstructions, we exemplarily computed packing properties such as global extraparticle porosity and radial porosity profiles, and visualized the presence of void spaces using 3D image analysis.

The Differential Anti-HIV Effect of a New Humic Substance-Derived Preparation in Diverse Cells of the Immune System.

The anti-HIV activity of a new humic substance-derived preparation has been studied in individual pools of immune cells (CD4+ T lymphocytes, macrophages, dendritic cells). Near-complete inhibition of the HIV infection (by more than 90%) was achieved by treating each of the abovementioned cell types with non-toxic concentrations of the preparation. The inhibitory effect demonstrates the possibility of preventing the depletion of a significant portion of functionally important immune cells. A comparative study of infection inhibition in individual cell pools has allowed us to reveal the differences in the preparation's effectiveness in each of the cell populations. A R5-tropic HIV-1 infection in macrophages exhibited maximum sensitivity to the preparation: 90% and 50% inhibition of the infection were observed in the presence of concentrations as low as 1.4 and 0.35 µg/ml, respectively. A 15- and 19-fold higher concentration was required to achieve the same extent of inhibition in dendritic cells infected with the same strain. The effectiveness of the drug in CD4 + T lymphocytes is quite comparable to its effectiveness in macrophages. The drug is universally effective for both the T- and M-tropic variants of HIV-1.

Towards a Full Automation of the ELISpot Assay for Safe and Parallelized Immunomonitoring.

The ELISpot assay, as a sensitive and specific method, enables the detection of cytokines for immunological purposes and in vaccine development. Here we describe the successful transfer of the manual procedure to a commercially available automated liquid handling platform, based on the work described in Neubauer et al. (Cytotechnology 69:57-73, 2017). Different kinds of technical issues (dead volume reduction, instrumental handling limitations, liquid class improvement) have been solved and biological effects (reagents concentration, selectivity tests, dispensing way, etc.) have been controlled during the implementation process. At the end a maximum of 6% mean delta difference and a lower mean dispersion than the manual assay were reached as well as a turnaround time of four to six times higher than the manual process.

Towards standardized automated immunomonitoring: an automated ELISpot assay for safe and parallelized functionality analysis of immune cells.

The ELISpot assay is used for the detection of T cell responses in clinical trials and vaccine evaluations. Standardization and reproducibility are necessary to compare the results worldwide, inter- and intra-assay variability being critical factors. To assure operator safety as well as high-quality experiment performance, the ELISpot assay was implemented on an automated liquid handling platform, a Tecan Freedom EVO. After validation of the liquid handling, automated loading of plates with cells and reagents was investigated. With step by step implementation of the manual procedure and liquid dispensing optimization on the robot platform, a fully automated ELISpot assay was accomplished with plates remaining in the system from the plate blocking step to spot development. The mean delta difference amounted to a maximum of 6%, and the mean dispersion was smaller than in the manual assay. Taken together, we achieved with this system not only a lower personnel attendance but also higher throughput and a more precise and parallelized analysis. This platform has the potential to guarantee validated, safe, fast, reproducible and cost-efficient immunological and toxicological assays in the future.

Is there a difference in the efficacy of peripartum antiretroviral regimens in reducing mother-to-child transmission of HIV in Africa?

BACKGROUND: Peripartum antiretroviral regimens have been shown to prevent mother-to-child transmission of HIV (MTCT) in randomized clinical trials; however, direct comparison of published results is impossible given methodological and population differences. OBJECTIVE: To directly compare the efficacy of different antiretroviral regimens in reducing the risk of 6-week MTCT rate in African breastfeeding populations. METHODS: Pooled analysis including all mother-infant pairs from any relevant trial: West African ZDV-placebo trials, Petra ZDV+3TC [two regimens A (pre/intra/post-partum) and B (intra/post-partum), placebo from Uganda and Tanzania], SAINT (NVP and Petra arm B), HIVNET012 (NVP, ultra short ZDV pp) and the Vitamin A trial (as placebo arm in South Africa). Peripartum HIV infection was any positive RNA or DNA polymerase chain reaction test < day 60. The MTCT risk was estimated at 6 weeks for each treatment arm and compared with placebo or single-dose NVP using logistic regression adjusting for maternal CD4 cell count, breastfeeding and birthweight. RESULTS: Overall, 4125 singleton live-births were included; 3629 (88%) were assessed for HIV status at 6 weeks of age. In comparison with placebo, zidovudine + lamivudine (ZDV+3TC) arm A [adjusted odds ratio (AOR), 0.23; P < 0.0001], ZDV+3TC arm B (AOR, 0.49; P < 0.001), antenatal ZDV short (AOR, 0.55; P = 0.006) and nevirapine (NVP) (AOR, 0.60; P = 0.0007) significantly reduced MTCT. In comparison with NVP, only the longest regimen of ZDV+3TC (AOR, 0.39, P < 0.0005) was significantly more effective. CONCLUSION: These results are in line with current World Health Organisation guidelines suggesting equivalence of choice between single-dose NVP and short-course ZDV, and confirm the greater efficacy of ZDV+3TC than with any single antiretroviral drug.

In vitro-to-in vivo correlation of the skin penetration, liver clearance and hepatotoxicity of caffeine.

This work illustrates the use of Physiologically-Based Toxicokinetic (PBTK) modelling for the healthy Caucasian population in in vitro-to-in vivo correlation of kinetic measures of caffeine skin penetration and liver clearance (based on literature experiments), as well as dose metrics of caffeine-induced measured HepaRG toxicity. We applied a simple correlation factor to quantify the in vitro and in vivo differences in the amount of caffeine permeated through the skin and concentration-time profiles of caffeine in the liver. We developed a multi-scale computational approach by linking the PBTK model with a Virtual Cell-Based Assay to relate an external oral and dermal dose with the measured in vitro HepaRG cell viability. The results revealed higher in vivo skin permeation profiles than those determined in vitro using identical exposure conditions. Liver clearance of caffeine derived from in vitro metabolism rates was found to be much slower than the optimised in vivo clearance with respect to caffeine plasma concentrations. Finally, HepaRG cell viability was shown to remain almost unchanged for external caffeine doses of 5-400 mg for both oral and dermal absorption routes. We modelled single exposure to caffeine only.

Determination of HIV-1 subtypes in injecting drug users in Bangkok, Thailand, using peptide-binding enzyme immunoassay and heteroduplex mobility assay: evidence of increasing infection with HIV-1 subtype E.

OBJECTIVES: To evaluate the sensitivity and specificity of peptide-binding enzyme immunoassay (PEIA) and heteroduplex mobility assay (HMA) for the determination of HIV-1 subtypes B and E; to determine the proportions of infections due to subtypes B and E over time; and to generate data on DNA sequences of the C2-V3 region of the env genes. METHODS: HIV-1 subtyping was conducted by PEIA and HMA on blood specimens obtained from 97 injecting drug users (IDU) infected with HIV between 1988 and 1993. Genetic sequencing was performed on 84 specimens. RESULTS: Both laboratory methods were highly sensitive and specific for the determination of HIV-1 subtypes B and E. The two tests were complementary; samples which could not be typed by HMA were correctly typed by PEIA and vice versa. While subtype B accounted for 80.4% (78 out of 97) of infections overall, the proportion of new infections due to subtype E increased from 2.6% (one out of 38) in 1988-1989 to 25.6% (11 out of 43) in 1990-1991, and to 43.8% (seven out of 16) in 1992-1993 (chi 2 for linear trend, P < 0.001). CONCLUSIONS: HMA and PEIA are practical, sensitive and specific laboratory methods for the determination of HIV-1 subtypes in Thailand, and may be useful in other geographic areas to define the molecular epidemiology of the global HIV-1 pandemic. Data suggest that the proportion subtype E infections have increased among Bangkok IDU from 1988 through 1993.

HIV-1 strains from India are highly divergent from prototypic African and US/European strains, but are linked to a South African isolate.

OBJECTIVE: To gain molecular insights into different HIV-1 strains present in two different states of India, nucleotide sequences derived from the env region of four HIV-1 strains were analysed. DESIGN: HIV-1 was isolated from high-risk patients from the states of Maharashtra (city of Bombay) and Goa. The molecular analysis of the env region encompassed all variable domains of the external glycoprotein, gp120. METHODS: Genomic DNA from cultured cells infected with each of the four Indian HIV-1 strains independently was amplified by polymerase chain reaction (PCR). PCR fragments were cloned and sequenced and a phylogenetic tree constructed. RESULTS: All four Indian HIV-1 sequences were closely related to each other. The closest related sequence to them was from a South African isolate, HIV-1NOF, with a homology of 85-87%. In the phylogenetic tree, the Indian and the South African HIV-1 sequences cluster together and constitute a subtype different from the North American/European, Central African, Uganda/Rwanda and Northern Thailand subtypes. Interestingly, the viruses of this subtype are characterized by an additional potential N-glycosylation site C-terminal to the CD4-binding domain. CONCLUSION: The low variation between the HIV-1 sequences from randomly chosen individuals from high-risk cohorts in two Indian states suggests a rapid and recent spread of HIV and, possibly, introduction of the virus by the same route, most probably heterosexual transmission. The rapid spread of HIV-1 variants in India, which form a subgroup of their own together with a South African strain, necessitate consideration of these strains in vaccine development.

Selection of HIV-1 genotypes by cultivation in different primary cells.

OBJECTIVES: To determine the representation of particular HIV-1 genotypes during cultivation in different primary cell-culture systems compared with the spectrum of the quasispecies in vivo. METHODS: Primary isolates of HIV-1 were recovered by isolation in cultures of lymphocytes, mixed mononuclear cells (MNC), and monocytes/macrophages. Nucleotide sequence determination of the C2-V3 region of gp120 of HIV was performed on 10-20 independently isolated clones derived by polymerase chain reaction from the culture systems, the uncultured peripheral blood MNC (PBMC) as well as plasma. RESULTS: Several predominant HIV genotypes were found in the uncultured PBMC from each of the patients. The most frequent genotypes in PBMC were also the most frequent types in plasma. In addition, lymphocytes, macrophages or mixed MNC cultures allowed the outgrowth of variants that were underrepresented in uncultured PBMC. We showed that the virus cultivation systems used in this study selected differently for the genetic variants. Whereas some genotypes were present in all three culture systems, although at different frequencies, others were exclusively found in a specific culture system. CONCLUSIONS: These results demonstrate that monocyte/macrophage and mixed MNC culture systems complement the standard lymphocyte culture in terms of the spectrum of genotypically different virus variants obtained in vitro.

HIV-1 and HIV-2 infections in a high-risk population in Bombay, India: evidence for the spread of HIV-2 and presence of a divergent HIV-1 subtype.

A high-risk population (patients of a sexually transmitted disease clinic and the GT hospital in Bombay) was tested for antibodies against HIV-1 and HIV-2. Among 405 serum samples, 226 had previously been classified HIV-positive in India using different locally available enzyme-linked immunosorbent assay (ELISA) tests. The serology of 179 samples was unknown. All 405 samples were tested at the Georg-Speyer-Haus (GSH) with the Pasteur HIV-1/2-Combi-ELISA. Positive samples were further analyzed with HIV-1 and HIV-2 Western blot kits from Dupont and Pasteur, respectively. A very high seroprevalence of HIV was found in this population. Among the 179 unscreened samples, 69 (38.5%) were positive in the ELISAs as well as the Western blots for HIV-1 or HIV-2. Among the prescreened samples, only 174 (77%) were confirmed HIV-positive. Altogether, 243 of 405 sera were HIV-positive. Of these, 184 (76%) were reactive with HIV-1, 10 (4%) were reactive with HIV-2, and 49 (20%) had dual reactivity to HIV-1 and HIV-2. Previous data from the Indian Council of Medical Research had already suggested a possible high prevalence of HIV-1 in India. Our results confirm this view. The finding of a substantial spread of HIV-2 infection was, however, totally unexpected in India, but confirms our previous study which had already demonstrated the existence of HIV-2 in this country. Asia can thus no longer be considered free of HIV-2, and testing for HIV-2 appears mandatory, at least in India.(ABSTRACT TRUNCATED AT 250 WORDS)

Restricted HIV type 1 replication under serum-free culture conditions in human monocyte-derived macrophages.

Monocytes (MOs) and macrophages (MACs) are well-known targets for HIV-1 infection. Even though the virus load is contributed mainly to lymphocytes during the asymptomatic phase of infection, the expression of HIV-1 in MO/MACs seems to be important for the course of the disease. To establish a model for restricted HIV-1 expression in MACs in vitro, we cultured MO-derived MACs under different culture conditions and analyzed their susceptibility to HIV-1 infection as well as their capacity for virus replication in vitro. MACs cultured under serum-free conditions with M-CSF (M-MACs) remain viable and functionally active as assessed by the analysis of cytokine production. In addition, the levels of CD4, CD14, CCR5, and HLA-DR expression are comparable to those of serum-derived MACs (SER-MACs). However, serum-free MACs were less susceptible to HIV-1 infection, with only 9.5+/-4.5% (mean+/-SEM) of all cells being p24 antigen positive on day 22 as compared with 51+/-9% under serum conditions (p < 0.005). Reverse transcriptase (RT) activity in the culture supernatant of M-MACs was always about 100-fold lower than that of SER-MACs even when comparable amounts of cells were infected. The addition of serum to serum-free cultures increased the percentage of HIV-1 p24 antigen-positive cells (21+/-8% positive cells on day 22) and increased the RT activity, indicating that serum factors could be important for HIV-1 replication in MACs. Therefore we also switched SER-MACs to serum-free culture conditions and found a sharp decrease in RT activity. However, the RT level could always be rescued by the addition of serum, even after a long serum-free culture period. This effect was dependent on the serum concentration added, with as little as 0.1% serum being effective in reestablishing viral production as measured by RT activity. In conclusion, we show that serum has an important role in the replication of HIV-1 in MACs. Our results suggest that besides the role of CD4 and CCR5 other microenvironmental factors, e.g., growth factors, cytokines, or hormones, which are not provided by the target cell itself, are involved in the regulation of MAC infection and of replication by HIV-1.

Standard conditions of virus isolation reveal biological variability of HIV type 1 in different regions of the world. WHO Network for HIV Isolation and Characterization.

HIV-1 isolates were obtained from four countries within the framework of the WHO Network for HIV Isolation and Characterization. The use of standard HIV isolation procedures allowed us to compare the biological properties of 126 HIV-1 isolates spanning five genetic subtypes. In primary isolation cultures, viruses from Uganda and Brazil appeared early and replicated without delay, whereas the replication of Thai viruses was delayed by several weeks. Regardless of genetic subtype or country of origin, blood samples collected more than 2 years after seroconversion yielded virus that replicated efficiently in the primary isolation cultures. None of the isolates obtained from Thailand or Rwanda replicated in cell lines, whereas 5 of the 13 Brazilian isolates and 7 of the 11 Ugandan isolates replicated and induced syncytia in MT-2 cells. As expected for virus isolates obtained early in HIV-1 infection (within 2 years of seroconversion), all viruses from Brazil, Rwanda, and Thailand showed a slow/low replicative pattern. For the Ugandan samples, the time from seroconversion was known precisely for a few of the samples and only in one case was less than 2 years. This may explain why the five viruses that were able to replicate in all cell lines, and thus classified as rapid/high, were of Ugandan origin. Viruses able to induce syncytia in MT-2 cells, also induced syncytia in PBMC. However, 8 slow/low viruses (out of 27) gave discordant results, inducing syncytia in PBMC but not in MT-2 cells. Furthermore, using syncytium induction as a marker, changes in virus populations during early in vitro passage in PBMC could be observed.(ABSTRACT TRUNCATED AT 250 WORDS)

Resistance mutations selected in vivo under therapy with anti-HIV drug HBY 097 differ from resistance pattern selected in vitro.

The quinoxaline derivative HBY 097, an orally active nonnucleoside inhibitor of HIV-1 reverse transcriptase (NNRTI), showed an efficient suppression of viral load in a dose-escalating phase I study with mean trough concentrations increasing from 137-1299 ug/l [Rübsamen-Waigmann et al., Lancet 349:1517]. Half-maximal inhibitory concentrations (IC50) for viruses grown from the patients at entry of the study were 0.1-3 nM, except for one patient who had a virus with reduced susceptibility to HBY 097 at entry (IC50: 160 nM). During therapy, only two patients developed a virus with a moderately increased IC50 (2.2 and 15 nM). This reduced susceptibility was associated with the known NNRTI-resistance mutation K ==> N at position 103, in contrast to resistance selection in vitro, which had yielded predominant mutations at positions 179 and 190. The Tyr mutation at position 181, inducing high resistance for other NNRTIs, was never observed. The resistant virus at study entry (IC50 = 160 nM) had a mutation at position 103 as well, combined with an AZT resistance mutation (K ==> R) at position 70, suggesting that nucleoside-resistance mutations may help increasing resistance to HBY 097. This is in line with our in vitro selection studies, where resistance mutations at the 'nucleoside sites' 74 and 75 increased the resistance phenotype of NNRTI mutations. Our findings highlight the crucial importance of IC50 determinations from cultured virus for determination of phenotypic resistance development during therapy and demonstrate that in vivo resistance development cannot be predicted from in vitro selection.

Body distribution of azidothymidine bound to hexyl-cyanoacrylate nanoparticles after i.v. injection to rats.

Cells of the reticuloendothelial system (RES e.g. macrophages) play an important role in the immunopathogenesis of AIDS. The objective of the present study was to investigate the possibility of specifically targeting antiviral drugs such as azidothymidine (AZT) to macrophages using nanoparticles as colloidal drug carriers. In a first series of experiments the body distribution of 14C-labelled AZT bound to nanoparticles and a similarly prepared control solution with unbound AZT were studied in rats after intravenous injection. In a second series of experiments polysorbate 80-coated nanoparticles and a solution of AZT in saline were tested. 14C-labelled AZT was bound to nanoparticles using the surfactant bis(2-ethylhexyl) sulphosuccinate sodium (DOSS). The radioactivity in several organs, including those containing large numbers of macrophages, was measured after intravenous injection of the AZT-nanoparticles and the AZT-control solutions. AZT concentrations were up to 18 times higher in organs belonging to the RES if the drug was bound to nanoparticles compared with unbound AZT. These results demonstrate that nanoparticles are a potential drug targeting system for anti-AIDS drugs. The increase in drug concentration at the sites containing abundant macrophages may allow a reduction in dosage to reduce systemic toxicity.

Preparation of avidin-labeled protein nanoparticles as carriers for biotinylated peptide nucleic acid.

The possibility of preparing protein nanoparticles followed by covalent linkage of avidin was investigated. Free sulfhydryl groups were introduced onto the surface of protein nanoparticles either by aldehyde quenching with cysteine or reaction of free amino groups with 2-iminothiolane. The number of primary amino groups and sulfhydryl groups on the surface of the resulting particles was quantified with site-specific reagents. Avidin was attached to the surface of the thiolated nanoparticles via a bifunctional spacer which reacted in a first step with amino groups of avidin and in a second step with the sulfhydryl groups introduced onto the surface of the nanoparticles. Biotinylated peptide nucleic acid (PNA) as a model compound for biotinylated drugs was effectively coupled to the nanoparticles by complex formation with the covalently attached avidin. Since the formation of the interaction between biotin and avidin is very rapid and stable a highly effective drug carrier system for biotinylated compounds such as PNAs was achieved.

Preparation of avidin-labelled gelatin nanoparticles as carriers for biotinylated peptide nucleic acid (PNA).

The possibility of preparing uniform nanoparticles consisting of proteins such as gelatin followed by covalent linkage of avidin was investigated. Gelatin nanoparticles were prepared by two step desolvation. Functional groups at the surface of the particulate system were quantified with site-specific reagents. The surface of the nanoparticles was thiolated and avidin was covalently attached to the nanoparticles via a bifunctional spacer at high levels. Biotinylated peptide nucleic acid (PNA) was effectively complexed by the avidin-conjugated nanoparticles. Avidin-conjugated protein nanoparticles should prove as potential carrier system for biotinylated drug derivatives in antisense therapy.

Optimization of the preparation process for human serum albumin (HSA) nanoparticles.

Nanoparticles prepared by desolvation and subsequent crosslinking of human serum albumin (HSA) represent promising carriers for drug delivery. Particle size is a crucial parameter, in particular for the in vivo behaviour of nanoparticles after intravenous injection. The objective of the present study is the development of a desolvation procedure for the preparation of HSA-based nanoparticles under the aspect of a controllable particle size between 100 and 300 nm in combination with a narrow size distribution. A pump-controlled preparation method was established which enabled particle preparation under defined conditions. Several factors of the preparation process, such as the rate of addition of the desolvating agent, the pH value and the ionic composition of the HSA solution, the protein concentration, and the conditions of particle purification were evaluated. The pH value of the HSA solution prior to the desolvation procedure was identified as the major factor determining particle size. Varying this parameter, (mean) particle diameters could be adjusted between 150 and 280 nm, higher pH values leading to smaller nanoparticles. Washing the particles by differential centrifugation led to significantly narrower size distributions. The reproducibility of the particle size and particle size distribution under the proposed preparation conditions was demonstrated by sedimentation velocity analysis in the analytical ultracentrifuge and the cellular uptake of those nanoparticles was studied by confocal microscope imaging and FACS analysis. The stability of the resulting nanoparticles was evaluated by pH and buffer titration experiments. Only pH values distinctly outside the isoelectric pH range of HSA and low salt concentrations were able to prevent nanoparticle agglomeration.

Application of physiologically-based toxicokinetic modelling in oral-to-dermal extrapolation of threshold doses of cosmetic ingredients.

The application of physiologically based toxicokinetic (PBTK) modelling in route-to-route (RtR) extrapolation of three cosmetic ingredients: coumarin, hydroquinone and caffeine is shown in this study. In particular, the oral no-observed-adverse-effect-level (NOAEL) doses of these chemicals are extrapolated to their corresponding dermal values by comparing the internal concentrations resulting from oral and dermal exposure scenarios. The PBTK model structure has been constructed to give a good simulation performance of biochemical processes within the human body. The model parameters are calibrated based on oral and dermal experimental data for the Caucasian population available in the literature. Particular attention is given to modelling the absorption stage (skin and gastrointestinal tract) in the form of several sub-compartments. This gives better model prediction results when compared to those of a PBTK model with a simpler structure of the absorption barrier. In addition, the role of quantitative structure-property relationships (QSPRs) in predicting skin penetration is evaluated for the three substances with a view to incorporating QSPR-predicted penetration parameters in the PBTK model when experimental values are lacking. Finally, PBTK modelling is used, first to extrapolate oral NOAEL doses derived from rat studies to humans, and then to simulate internal systemic/liver concentrations - Area Under Curve (AUC) and peak concentration - resulting from specified dermal and oral exposure conditions. Based on these simulations, AUC-based dermal thresholds for the three case study compounds are derived and compared with the experimentally obtained oral threshold (NOAEL) values.