Measurement of Nanoparticle Uptake by Alveolar Macrophages - Safety of Nanomaterials along Their Lifecycle - page 153

showed that the phagocytized mass for AlOOH and CeO 2 amounted to 150 pg/cell
when the cells had a nonlimited access to particles. This is a plausible result as it is
in the range formerly known from in vitro experiments in which AM engulfed up to
240 pg/cell (Rehn et al. 1999) and even in the range of in vivo results, in which AM
were loaded with up to 90 pg (Pauluhn 2009). Thus, the proposed concept of measur-
ing the uptake of agglomerated particles via digital image processing appears sound.
Nevertheless, results have to be discussed with respect to possible sources of error.
The first is the assumption that >99% of the suspended particle mass agglomerates
and sediments onto the bottom of the measurement chamber (Hinderliter et al. 2010).
At least for AlOOH and CeO 2 NP this could be confirmed as no NP were detected in
the supernatant by NanoSight tracking analysis. Also an inhomogeneous distribution
of particles at the bottom may strongly influence the outcome, but this obstacle has
been made unlikely by the mode of particle application and was further on controlled
by using different sites to measure. A small fraction of particles may also bind to the
vessel wall. However, given the dimension of the Petri dish also this mistake appears
negligible.
Another source of error emanates from data processing itself. We estimate that
up to 1% of the mass equivalent particles may have been removed from evaluation
by smoothing digital images. However, the correct setting of parameters for the
automatic analysis or compensating the influence of particle overlap is certainly
more important. The correct identification of particles, macrophages, and cell con-
tours, therefore, was a special issue since it strongly influences all calculations.
For all these parameters, a constant accuracy of 95%-99% was reached if the
outcome was compared to the detection result of an experienced user. As outlined
earlier, a review of the optical information by the user is mandatory for a sound
interpretation of single results. However, despite some uncertainties with respect
to absolute values, the method allows a rapid comparison of the uptake kinetics for
many different particles or cell types and, by this, enables a better interpretation
of biologic effects.
Statistic evaluation was needed at several steps of evaluation and helped to better
analyze the data. The developed sedimentation model enables us to characterize the
sedimentation process and to precompute the quantity of particles at every point in
time inside the ROI as it considered the collected sedimentation parameters from
the image data (Schippritt et al. 2010). The sedimentation model not only eliminated
indispensable statistic variance as it eliminated noise from advection- or diffusion-
based dislocation of particles but also was helpful to better calculate the in vitro dose
for macrophage experiments before or after an experiment. For the determination of
phagocytized particles the difference between control ROIs (without macrophage)
and active ROIs (with macrophage) was used. In this context, it was helpful that the
characteristics of particle quantity for the control ROI (formula 7.3) and the active
ROI (cf. formula 7.5) could be described by kinematic models.
What are the future perspectives of light microscopy and image analysis meth-
ods for describing particle uptake by cells in vitro? As previously outlined, the
approach described here is largely limited to agglomerated NP, that is, to study
their sedimentation and uptake kinetics. However, single NP close to or inside cells
may be viewed by light microscopy as well. Recently, a method based on dark field