In vitro assays for hazard identification of nanoparticles

Samantha K. Kloet

Research output: Thesisinternal PhD, WU

Abstract

The production of nanoparticles (NPs) has increased in the last decades and the number of products in which NPs are being incorporated is still growing. The rapid increase of nanotechnology has several benefits for society, yet there is an increasing concern that exposure to NPs may result in significant adverse health effects. Since NPs are incorporated in a variety of consumer products, it is likely that the general population will be exposed to NPs. It would be desirable that the safety and risk assessment of NPs could be largely based on studies using in vitro models instead of in vivo models as this would reduce the use of test animals, costs and time required to test the large numbers of NPs. The aim of the present thesis was to investigate the potential of in vitro testing strategies to detect hazards of NPs, focusing on toxicokinetic as well as toxicodynamic endpoints. Toxicokinetic studies focused on translocation of NPs in in vitro models of the placental barrier, while toxicodynamic studies were directed at two endpoints that represent potential hazards of NPs that have not yet been well characterized including: developmental toxicity and immunotoxicity.

In the present thesis different types of NPs were used. Polystyrene nanoparticles (PS-NPs) were selected because of their commercial availability, with high quality and a wide variety of available physicochemical properties like surface charge, and fluorescent labeling enabling easy detection in toxicokinetic (translocation) studies. Several metal (oxide) NPs were selected as well, of which some are possible constituents of food additives like TiO2, Fe2O3, SiO2 and Ag. Other metal oxide NPs that were selected were Mn­2O3, CuO, Cr2O3, CoO and NiO to which we may be exposed via products like paints, catalysts, construction materials, coatings and batteries.

Placental translocation of NPs was studied as an important toxicokinetic aspect, since part of the toxicodynamic studies of the present thesis were directed at developmental toxicity testing of NPs. In order to obtain insight in toxicity and translocation of NPs across the placental barrier, cytotoxicity and translocation was studied for one positively and two negatively charged PS-NPs of 50 nm in an in vitro model of the placenta. In this study it appeared that in spite of similar size, surface charge and type of proteins in the protein corona, the differently charged NPs displayed a remarkable difference in cytotoxicity, with only the PS-NPs with an original positive charge inducing cytotoxicity. Translocation of PS-NPs appeared not to be related to PS-NP charge alone. A remarkable difference in translocation was found between the two 50 nm negatively charged PS-NPs that were obtained from different manufacturers. Since none of the characterized parameters, including size, surface charge and protein corona revealed remarkable differences between the two negatively charged NPs, the difference may originate from the chemical groups on the surface of the NPs generating the negative charge. The general conclusion from this study was that the in vitro BeWo b30 model can be used as a fast method to get an initial qualitative impression about the capacity of NPs to translocate across the placental barrier and to set priorities for further in vivo studies on translocation of NPs to the fetus.

The same PS-NPs as tested for placental translocation were investigated whether they are able to cause in vitro developmental toxicity in the ES-D3 cell differentiation assay of the embryonic stem cell test (EST) focusing also on the effect that charge may have. The study showed that the two negatively charged PS-NPs did not show any effect in the ES-D3 cell differentiation assay up to the highest concentration tested while the positively charged PS-NP showed a concentration-dependent inhibition of ES-D3 cell differentiation. However, effect concentrations in the ES-D3 cell differentiation assay were close to cytotoxic concentrations, which indicated that the inhibition of the ES-D3 cell differentiation may be due to cytotoxic effects of the positively charged PS-NPs. This indicated that the inhibition of the ES-D3 cell differentiation by the positively charged PS-NPs may be caused by non-specific effects. Although the experiments on placental translocation of the present thesis showed that positively charged PS-NPs are more toxic than negatively charged PS-NPs, it appeared that this may not be generalizable to other NPs. This  follows from the fact that in SiO2, Ag and TiO2 NPs that were reported in other studies to inhibit ES-D3 cell differentiation were negatively charged, while the negatively charged PS-NPs of the present study did not affect ES-D3 cell differentiation. Although the limited data available indicate that charge, size and coating of NPs may be important characteristics that determine the developmental toxicity potential of NPs, more (systematic) studies are needed to assess how physicochemical characteristics of NPs relate to their developmental toxicity. This information may help to prioritize NPs for in vitro and in vivo developmental toxicity testing.

In addition, toxic effects of a series of metal (oxide) NPs were tested in macrophage RAW264.7 cells in order to obtain insight in effect of these NPs on cells of the innate immune response. In these macrophage RAW264.7 cells the effects of the metal (oxide) NPs were characterized on cell viability, TNF-α production and mitochondria-related parameters like production of reactive oxygen species (ROS), mitochondrial permeability transition pore (MPTP) opening, and intracellular ATP levels. Altogether, results obtained showed no or limited effects of the NP formulations of metal (oxide) food additives on cell viability, ROS production, MPTP opening, ATP levels and TNF-α production in RAW264.7 macrophages. Effects were only observed at high concentrations that may not be physiologically relevant, indicating that related adverse effects upon exposure to the respective NPs in vivo may be limited.

Taken together, the present thesis provided further evidence of the influence of physicochemical properties of NPs in driving toxicity in in vitro models. However, the determination of the fate and toxicity of NPs using in vitro or in vivo models is a challenge that needs further evaluation. A combination of several factors likely play a role in determining the outcome of exposure including factors like NP core material and presence and type of coating agents resulting in various physicochemical properties (size, charge, etc.). This appears to hamper conclusive evaluation of the role of physicochemical characteristics of NPs in their potential hazards and risks so far. The results obtained do show however that in vitro assays can detect differences in potential hazards posed by NPs. Therefore it is concluded that the results of the work presented in this thesis will contribute to the further development and use of non-animal based testing strategies for safety testing of NPs providing insight into selected potential hazards of the tested NPs.

 

Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Wageningen University
Supervisors/Advisors
  • Rietjens, Ivonne, Promotor
  • Louisse, Jochem, Promotor
  • van den Brink, Nico, Promotor
Award date7 Dec 2016
Place of PublicationWageningen
Publisher
Print ISBNs9789462579415
DOIs
Publication statusPublished - 2016

Fingerprint

Nanoparticles
Polystyrenes
Metal Nanoparticles
In Vitro Techniques
Cell Differentiation
Oxides
Food Additives
Poisons

Keywords

  • nanotechnology
  • particles
  • in vitro
  • models
  • hazards
  • toxicity
  • toxicokinetics

Cite this

Kloet, Samantha K.. / In vitro assays for hazard identification of nanoparticles. Wageningen : Wageningen University, 2016. 213 p.
@phdthesis{7df47a2aa123448187a55b596b01b683,
title = "In vitro assays for hazard identification of nanoparticles",
abstract = "The production of nanoparticles (NPs) has increased in the last decades and the number of products in which NPs are being incorporated is still growing. The rapid increase of nanotechnology has several benefits for society, yet there is an increasing concern that exposure to NPs may result in significant adverse health effects. Since NPs are incorporated in a variety of consumer products, it is likely that the general population will be exposed to NPs. It would be desirable that the safety and risk assessment of NPs could be largely based on studies using in vitro models instead of in vivo models as this would reduce the use of test animals, costs and time required to test the large numbers of NPs. The aim of the present thesis was to investigate the potential of in vitro testing strategies to detect hazards of NPs, focusing on toxicokinetic as well as toxicodynamic endpoints. Toxicokinetic studies focused on translocation of NPs in in vitro models of the placental barrier, while toxicodynamic studies were directed at two endpoints that represent potential hazards of NPs that have not yet been well characterized including: developmental toxicity and immunotoxicity. In the present thesis different types of NPs were used. Polystyrene nanoparticles (PS-NPs) were selected because of their commercial availability, with high quality and a wide variety of available physicochemical properties like surface charge, and fluorescent labeling enabling easy detection in toxicokinetic (translocation) studies. Several metal (oxide) NPs were selected as well, of which some are possible constituents of food additives like TiO2, Fe2O3, SiO2 and Ag. Other metal oxide NPs that were selected were Mn­2O3, CuO, Cr2O3, CoO and NiO to which we may be exposed via products like paints, catalysts, construction materials, coatings and batteries. Placental translocation of NPs was studied as an important toxicokinetic aspect, since part of the toxicodynamic studies of the present thesis were directed at developmental toxicity testing of NPs. In order to obtain insight in toxicity and translocation of NPs across the placental barrier, cytotoxicity and translocation was studied for one positively and two negatively charged PS-NPs of 50 nm in an in vitro model of the placenta. In this study it appeared that in spite of similar size, surface charge and type of proteins in the protein corona, the differently charged NPs displayed a remarkable difference in cytotoxicity, with only the PS-NPs with an original positive charge inducing cytotoxicity. Translocation of PS-NPs appeared not to be related to PS-NP charge alone. A remarkable difference in translocation was found between the two 50 nm negatively charged PS-NPs that were obtained from different manufacturers. Since none of the characterized parameters, including size, surface charge and protein corona revealed remarkable differences between the two negatively charged NPs, the difference may originate from the chemical groups on the surface of the NPs generating the negative charge. The general conclusion from this study was that the in vitro BeWo b30 model can be used as a fast method to get an initial qualitative impression about the capacity of NPs to translocate across the placental barrier and to set priorities for further in vivo studies on translocation of NPs to the fetus. The same PS-NPs as tested for placental translocation were investigated whether they are able to cause in vitro developmental toxicity in the ES-D3 cell differentiation assay of the embryonic stem cell test (EST) focusing also on the effect that charge may have. The study showed that the two negatively charged PS-NPs did not show any effect in the ES-D3 cell differentiation assay up to the highest concentration tested while the positively charged PS-NP showed a concentration-dependent inhibition of ES-D3 cell differentiation. However, effect concentrations in the ES-D3 cell differentiation assay were close to cytotoxic concentrations, which indicated that the inhibition of the ES-D3 cell differentiation may be due to cytotoxic effects of the positively charged PS-NPs. This indicated that the inhibition of the ES-D3 cell differentiation by the positively charged PS-NPs may be caused by non-specific effects. Although the experiments on placental translocation of the present thesis showed that positively charged PS-NPs are more toxic than negatively charged PS-NPs, it appeared that this may not be generalizable to other NPs. This  follows from the fact that in SiO2, Ag and TiO2 NPs that were reported in other studies to inhibit ES-D3 cell differentiation were negatively charged, while the negatively charged PS-NPs of the present study did not affect ES-D3 cell differentiation. Although the limited data available indicate that charge, size and coating of NPs may be important characteristics that determine the developmental toxicity potential of NPs, more (systematic) studies are needed to assess how physicochemical characteristics of NPs relate to their developmental toxicity. This information may help to prioritize NPs for in vitro and in vivo developmental toxicity testing. In addition, toxic effects of a series of metal (oxide) NPs were tested in macrophage RAW264.7 cells in order to obtain insight in effect of these NPs on cells of the innate immune response. In these macrophage RAW264.7 cells the effects of the metal (oxide) NPs were characterized on cell viability, TNF-α production and mitochondria-related parameters like production of reactive oxygen species (ROS), mitochondrial permeability transition pore (MPTP) opening, and intracellular ATP levels. Altogether, results obtained showed no or limited effects of the NP formulations of metal (oxide) food additives on cell viability, ROS production, MPTP opening, ATP levels and TNF-α production in RAW264.7 macrophages. Effects were only observed at high concentrations that may not be physiologically relevant, indicating that related adverse effects upon exposure to the respective NPs in vivo may be limited. Taken together, the present thesis provided further evidence of the influence of physicochemical properties of NPs in driving toxicity in in vitro models. However, the determination of the fate and toxicity of NPs using in vitro or in vivo models is a challenge that needs further evaluation. A combination of several factors likely play a role in determining the outcome of exposure including factors like NP core material and presence and type of coating agents resulting in various physicochemical properties (size, charge, etc.). This appears to hamper conclusive evaluation of the role of physicochemical characteristics of NPs in their potential hazards and risks so far. The results obtained do show however that in vitro assays can detect differences in potential hazards posed by NPs. Therefore it is concluded that the results of the work presented in this thesis will contribute to the further development and use of non-animal based testing strategies for safety testing of NPs providing insight into selected potential hazards of the tested NPs.  ",
keywords = "nanotechnology, particles, in vitro, models, hazards, toxicity, toxicokinetics, nanotechnologie, deeltjes, in vitro, modellen, gevaren, toxiciteit, toxicokinetiek",
author = "Kloet, {Samantha K.}",
note = "WU thesis 6520 Author's name on overleaf title page: Samantha Kristina Kloet Includes bibliographic references. - With summary in English",
year = "2016",
doi = "10.18174/390912",
language = "English",
isbn = "9789462579415",
publisher = "Wageningen University",
school = "Wageningen University",

}

Kloet, SK 2016, 'In vitro assays for hazard identification of nanoparticles', Doctor of Philosophy, Wageningen University, Wageningen. https://doi.org/10.18174/390912

In vitro assays for hazard identification of nanoparticles. / Kloet, Samantha K.

Wageningen : Wageningen University, 2016. 213 p.

Research output: Thesisinternal PhD, WU

TY - THES

T1 - In vitro assays for hazard identification of nanoparticles

AU - Kloet, Samantha K.

N1 - WU thesis 6520 Author's name on overleaf title page: Samantha Kristina Kloet Includes bibliographic references. - With summary in English

PY - 2016

Y1 - 2016

N2 - The production of nanoparticles (NPs) has increased in the last decades and the number of products in which NPs are being incorporated is still growing. The rapid increase of nanotechnology has several benefits for society, yet there is an increasing concern that exposure to NPs may result in significant adverse health effects. Since NPs are incorporated in a variety of consumer products, it is likely that the general population will be exposed to NPs. It would be desirable that the safety and risk assessment of NPs could be largely based on studies using in vitro models instead of in vivo models as this would reduce the use of test animals, costs and time required to test the large numbers of NPs. The aim of the present thesis was to investigate the potential of in vitro testing strategies to detect hazards of NPs, focusing on toxicokinetic as well as toxicodynamic endpoints. Toxicokinetic studies focused on translocation of NPs in in vitro models of the placental barrier, while toxicodynamic studies were directed at two endpoints that represent potential hazards of NPs that have not yet been well characterized including: developmental toxicity and immunotoxicity. In the present thesis different types of NPs were used. Polystyrene nanoparticles (PS-NPs) were selected because of their commercial availability, with high quality and a wide variety of available physicochemical properties like surface charge, and fluorescent labeling enabling easy detection in toxicokinetic (translocation) studies. Several metal (oxide) NPs were selected as well, of which some are possible constituents of food additives like TiO2, Fe2O3, SiO2 and Ag. Other metal oxide NPs that were selected were Mn­2O3, CuO, Cr2O3, CoO and NiO to which we may be exposed via products like paints, catalysts, construction materials, coatings and batteries. Placental translocation of NPs was studied as an important toxicokinetic aspect, since part of the toxicodynamic studies of the present thesis were directed at developmental toxicity testing of NPs. In order to obtain insight in toxicity and translocation of NPs across the placental barrier, cytotoxicity and translocation was studied for one positively and two negatively charged PS-NPs of 50 nm in an in vitro model of the placenta. In this study it appeared that in spite of similar size, surface charge and type of proteins in the protein corona, the differently charged NPs displayed a remarkable difference in cytotoxicity, with only the PS-NPs with an original positive charge inducing cytotoxicity. Translocation of PS-NPs appeared not to be related to PS-NP charge alone. A remarkable difference in translocation was found between the two 50 nm negatively charged PS-NPs that were obtained from different manufacturers. Since none of the characterized parameters, including size, surface charge and protein corona revealed remarkable differences between the two negatively charged NPs, the difference may originate from the chemical groups on the surface of the NPs generating the negative charge. The general conclusion from this study was that the in vitro BeWo b30 model can be used as a fast method to get an initial qualitative impression about the capacity of NPs to translocate across the placental barrier and to set priorities for further in vivo studies on translocation of NPs to the fetus. The same PS-NPs as tested for placental translocation were investigated whether they are able to cause in vitro developmental toxicity in the ES-D3 cell differentiation assay of the embryonic stem cell test (EST) focusing also on the effect that charge may have. The study showed that the two negatively charged PS-NPs did not show any effect in the ES-D3 cell differentiation assay up to the highest concentration tested while the positively charged PS-NP showed a concentration-dependent inhibition of ES-D3 cell differentiation. However, effect concentrations in the ES-D3 cell differentiation assay were close to cytotoxic concentrations, which indicated that the inhibition of the ES-D3 cell differentiation may be due to cytotoxic effects of the positively charged PS-NPs. This indicated that the inhibition of the ES-D3 cell differentiation by the positively charged PS-NPs may be caused by non-specific effects. Although the experiments on placental translocation of the present thesis showed that positively charged PS-NPs are more toxic than negatively charged PS-NPs, it appeared that this may not be generalizable to other NPs. This  follows from the fact that in SiO2, Ag and TiO2 NPs that were reported in other studies to inhibit ES-D3 cell differentiation were negatively charged, while the negatively charged PS-NPs of the present study did not affect ES-D3 cell differentiation. Although the limited data available indicate that charge, size and coating of NPs may be important characteristics that determine the developmental toxicity potential of NPs, more (systematic) studies are needed to assess how physicochemical characteristics of NPs relate to their developmental toxicity. This information may help to prioritize NPs for in vitro and in vivo developmental toxicity testing. In addition, toxic effects of a series of metal (oxide) NPs were tested in macrophage RAW264.7 cells in order to obtain insight in effect of these NPs on cells of the innate immune response. In these macrophage RAW264.7 cells the effects of the metal (oxide) NPs were characterized on cell viability, TNF-α production and mitochondria-related parameters like production of reactive oxygen species (ROS), mitochondrial permeability transition pore (MPTP) opening, and intracellular ATP levels. Altogether, results obtained showed no or limited effects of the NP formulations of metal (oxide) food additives on cell viability, ROS production, MPTP opening, ATP levels and TNF-α production in RAW264.7 macrophages. Effects were only observed at high concentrations that may not be physiologically relevant, indicating that related adverse effects upon exposure to the respective NPs in vivo may be limited. Taken together, the present thesis provided further evidence of the influence of physicochemical properties of NPs in driving toxicity in in vitro models. However, the determination of the fate and toxicity of NPs using in vitro or in vivo models is a challenge that needs further evaluation. A combination of several factors likely play a role in determining the outcome of exposure including factors like NP core material and presence and type of coating agents resulting in various physicochemical properties (size, charge, etc.). This appears to hamper conclusive evaluation of the role of physicochemical characteristics of NPs in their potential hazards and risks so far. The results obtained do show however that in vitro assays can detect differences in potential hazards posed by NPs. Therefore it is concluded that the results of the work presented in this thesis will contribute to the further development and use of non-animal based testing strategies for safety testing of NPs providing insight into selected potential hazards of the tested NPs.  

AB - The production of nanoparticles (NPs) has increased in the last decades and the number of products in which NPs are being incorporated is still growing. The rapid increase of nanotechnology has several benefits for society, yet there is an increasing concern that exposure to NPs may result in significant adverse health effects. Since NPs are incorporated in a variety of consumer products, it is likely that the general population will be exposed to NPs. It would be desirable that the safety and risk assessment of NPs could be largely based on studies using in vitro models instead of in vivo models as this would reduce the use of test animals, costs and time required to test the large numbers of NPs. The aim of the present thesis was to investigate the potential of in vitro testing strategies to detect hazards of NPs, focusing on toxicokinetic as well as toxicodynamic endpoints. Toxicokinetic studies focused on translocation of NPs in in vitro models of the placental barrier, while toxicodynamic studies were directed at two endpoints that represent potential hazards of NPs that have not yet been well characterized including: developmental toxicity and immunotoxicity. In the present thesis different types of NPs were used. Polystyrene nanoparticles (PS-NPs) were selected because of their commercial availability, with high quality and a wide variety of available physicochemical properties like surface charge, and fluorescent labeling enabling easy detection in toxicokinetic (translocation) studies. Several metal (oxide) NPs were selected as well, of which some are possible constituents of food additives like TiO2, Fe2O3, SiO2 and Ag. Other metal oxide NPs that were selected were Mn­2O3, CuO, Cr2O3, CoO and NiO to which we may be exposed via products like paints, catalysts, construction materials, coatings and batteries. Placental translocation of NPs was studied as an important toxicokinetic aspect, since part of the toxicodynamic studies of the present thesis were directed at developmental toxicity testing of NPs. In order to obtain insight in toxicity and translocation of NPs across the placental barrier, cytotoxicity and translocation was studied for one positively and two negatively charged PS-NPs of 50 nm in an in vitro model of the placenta. In this study it appeared that in spite of similar size, surface charge and type of proteins in the protein corona, the differently charged NPs displayed a remarkable difference in cytotoxicity, with only the PS-NPs with an original positive charge inducing cytotoxicity. Translocation of PS-NPs appeared not to be related to PS-NP charge alone. A remarkable difference in translocation was found between the two 50 nm negatively charged PS-NPs that were obtained from different manufacturers. Since none of the characterized parameters, including size, surface charge and protein corona revealed remarkable differences between the two negatively charged NPs, the difference may originate from the chemical groups on the surface of the NPs generating the negative charge. The general conclusion from this study was that the in vitro BeWo b30 model can be used as a fast method to get an initial qualitative impression about the capacity of NPs to translocate across the placental barrier and to set priorities for further in vivo studies on translocation of NPs to the fetus. The same PS-NPs as tested for placental translocation were investigated whether they are able to cause in vitro developmental toxicity in the ES-D3 cell differentiation assay of the embryonic stem cell test (EST) focusing also on the effect that charge may have. The study showed that the two negatively charged PS-NPs did not show any effect in the ES-D3 cell differentiation assay up to the highest concentration tested while the positively charged PS-NP showed a concentration-dependent inhibition of ES-D3 cell differentiation. However, effect concentrations in the ES-D3 cell differentiation assay were close to cytotoxic concentrations, which indicated that the inhibition of the ES-D3 cell differentiation may be due to cytotoxic effects of the positively charged PS-NPs. This indicated that the inhibition of the ES-D3 cell differentiation by the positively charged PS-NPs may be caused by non-specific effects. Although the experiments on placental translocation of the present thesis showed that positively charged PS-NPs are more toxic than negatively charged PS-NPs, it appeared that this may not be generalizable to other NPs. This  follows from the fact that in SiO2, Ag and TiO2 NPs that were reported in other studies to inhibit ES-D3 cell differentiation were negatively charged, while the negatively charged PS-NPs of the present study did not affect ES-D3 cell differentiation. Although the limited data available indicate that charge, size and coating of NPs may be important characteristics that determine the developmental toxicity potential of NPs, more (systematic) studies are needed to assess how physicochemical characteristics of NPs relate to their developmental toxicity. This information may help to prioritize NPs for in vitro and in vivo developmental toxicity testing. In addition, toxic effects of a series of metal (oxide) NPs were tested in macrophage RAW264.7 cells in order to obtain insight in effect of these NPs on cells of the innate immune response. In these macrophage RAW264.7 cells the effects of the metal (oxide) NPs were characterized on cell viability, TNF-α production and mitochondria-related parameters like production of reactive oxygen species (ROS), mitochondrial permeability transition pore (MPTP) opening, and intracellular ATP levels. Altogether, results obtained showed no or limited effects of the NP formulations of metal (oxide) food additives on cell viability, ROS production, MPTP opening, ATP levels and TNF-α production in RAW264.7 macrophages. Effects were only observed at high concentrations that may not be physiologically relevant, indicating that related adverse effects upon exposure to the respective NPs in vivo may be limited. Taken together, the present thesis provided further evidence of the influence of physicochemical properties of NPs in driving toxicity in in vitro models. However, the determination of the fate and toxicity of NPs using in vitro or in vivo models is a challenge that needs further evaluation. A combination of several factors likely play a role in determining the outcome of exposure including factors like NP core material and presence and type of coating agents resulting in various physicochemical properties (size, charge, etc.). This appears to hamper conclusive evaluation of the role of physicochemical characteristics of NPs in their potential hazards and risks so far. The results obtained do show however that in vitro assays can detect differences in potential hazards posed by NPs. Therefore it is concluded that the results of the work presented in this thesis will contribute to the further development and use of non-animal based testing strategies for safety testing of NPs providing insight into selected potential hazards of the tested NPs.  

KW - nanotechnology

KW - particles

KW - in vitro

KW - models

KW - hazards

KW - toxicity

KW - toxicokinetics

KW - nanotechnologie

KW - deeltjes

KW - in vitro

KW - modellen

KW - gevaren

KW - toxiciteit

KW - toxicokinetiek

U2 - 10.18174/390912

DO - 10.18174/390912

M3 - internal PhD, WU

SN - 9789462579415

PB - Wageningen University

CY - Wageningen

ER -

Kloet SK. In vitro assays for hazard identification of nanoparticles. Wageningen: Wageningen University, 2016. 213 p. https://doi.org/10.18174/390912