Betraying Paracelsus, Ignoring Newton: A Flaw in the Nanotoxicology Paradigm

Justin Teeguarden, PHD, DABT

[Paul Hinderliter, Joel Pounds, Brian Thrall,, Galya Orr, Katrina Waters, Tom Weber, Barbara Tarasevich, Bobbie-Jo Webb Robertson]

Funded by the Environmental Biomarkers Initiative, DOE

Nano, Nanomaterial, Nanotechnology

Nano is 10-9. Nanotechnology is the understanding and control of matter at dimensions of roughly 1 to 100 nanometers, where unique phenomena enable novel applications.

Nanomaterials are 1-100 nm in at least one dimension

National Nanotechnology Initiative, 2002

SARS Virus Avain Flu Virus 100 nmSayes, Tox Sci 2006

TiO2 Agglomerates

Nanotoxicology is in its Infancy

Do the unique material properties of nanomaterials equate to unique biological properties as well?

Which of these properties are responsible for the biological effects?

Are the effects limited by the traditional defintion of nanomaterial (1-100 nm)?

Who should be studying the biocompatibility or toxicity of these new materials?

What studies should be done?

Where do we obtain high quality materials for testing?

Nanotoxicology is in its Infancy

236 January ‘08

What We Know

Nanomaterial Human Health Risks and Risk Assessment

Unknown

Insufficient Data

In Vitro Dosimetry

The State of The Science (Ignorance?)

Res

pons

e

?

Major Challenges for Dose-Response Assessment

What is dose?Size, shape, particle number, agglomeration state, surface chemistry, reactivity, surface area? Dose rate is also important.Do researches have all the tools necessary to measure these characteristics?

What is dose for in vitro studies?67% of published NM toxicity studies use in vitro systems.Mass concentrations are typically reported.Cannot be compared to doses for in vivo studies.How do you measure NM characteristics in liquid systems?

How do you extrapolate animal study doses to human dosesThis extrapolation is required for all risk assessments based on animal studies

Cell Culture is a Standard Tool for Mechanistic Work and Hazard

ScreeningThe paradigm for chemicals is being used without consideration of the unique kinetic differences between chemicals and particles in solution:

Response is considered a function of the nominal media mass or surface area concentration (cm2/ml, µg/ml)

Expose cells to selected concentrations of suspended

particles

Report dose-response on a nominal media

concentration basis

Dosimetry for Particles is Important!

Response is proportional to concentration at the target site!

So What’s the Problem?

In solution, particles settle and diffuse at rates which depend on:

SizeDensityShapeAgglomeration state

Surface area changes with sizeNumber concentrations change with sizeWhile nominal media concentrations stay the same across density and size, DELIVERY RATES DO NOT!

Nominal Media Concentration is A Poor Metric of Dose

1 nm Particle Size 1000 nm

Mas

s C

on

cen

trat

ion

(u

g/m

l)

SA

Co

nce

ntr

atio

n (

cm2 /

ml)

Decreases as the square of the radius

1 nm Particle Size 1000 nmM

ass

Co

nce

ntr

atio

n (

ug

/ml)

# co

nce

ntr

atio

n (

#/m

l)

Decreases as the cube of the radius

Nominal Media Concentration is A Poor Metric of Dose

1 Particle Size 1000(nm)

Dif

fusi

vity

(C

m2 /

s)

Set

tlin

g R

ate

(cm

/s)

1 Density 20(g/cm3)

# co

nce

ntr

atio

n (

#/m

l)

SA

Co

nce

ntr

atio

n (

cm2 /

ml)

Mass Concentration is Constant

Nominal Media Concentration is A Poor Metric of Dose

1 Density 20(g/cm3)

Set

tlin

g R

ate

(cm

/s)

Dif

fusi

vity

(C

m2/

s)

Mass Concentration is Constant

Dose-Related Parameters are Not Constant Across Particle Types

How Misleading can Nominal Media Concentrations Be?

Different Particles = Different Delivery Rates to Cells

Slo

wer

F

ast

er

T = 23 Hours Post Mixing

From First Principles: TiO2 and Gold Nanoparticle Dosimetry

Particle Diameter (nanometers)

1 10 100 1000

Su

rfac

e A

rea

Co

nce

ntr

atio

n (

cm2 /m

l)

0.0001

0.001

0.01

0.1

1

10

100

1000

10000

TiO2AuTiO2AuTiO2Au

] NMC

] Gravitational Settling Adjusted NMC

] Gravitational and Diffusional Transport Adjusted NMC

Particle Diameter (nm)

1 10 100 1000M

ass

Co

nc

entr

atio

n ( g

/ml)

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

October 3, 2007 California EPA

Nominal Media Concentration Obscures Underlying Dose-Response Behavior

Delivery Adjusted Surface Area Concentration (cm2/ml)

0.1 1 10 100

"Res

po

ns

e"

100

1000

10000

100000

1000000

1 nm 10 nm 100 nm 1000 nm

Nominal Surface Area Concentration (cm2/ml)

0.1 1 10 100 1000

"Re

sp

on

se

"

1

10

100

1000

10000

100000

1000000

1 nm 10 nm 100 nm 1000 nm

Material Particle Diameter (nm) EC50’s

Nominal Media Mass Concentration (μg/ml)

Surface Area Concentration

(cm2/ml)

1Delivery Adjusted Surface Area Concentration

(cm2/ml)

Gravity Diffusion & Gravity

CdO 1000 0.75 0.005 0.005 0.005

Ag 100 24 1.37 0.017 0.020

Ag 15 50 19.0 0.005 0.272

MoO3 150 250 21.3 0.232 0.262

MoO3 30 210 89.4 0.039 0.663

Material

Particle Size(nm)

Compartative Potency is affected by the Choice of Dose Metric

October 3, 2007 California EPA

Delivered Dose Reveals Surface Area Relationship

Is Any of this Real, or is it Just a Theory?

Ceria Oxide Nano Particles 25-300 nm

Cellular Uptake is Diffusion Driven for Small Nanoparticles and Gravity Driven for Larger particles

Which Particle is Most Toxic?

Settling Impacts CNT Toxicity?

Supernatant Pellet

Rapid Settling

Slow Settling

Material Property

Size (nm) Affect on Nanoparticle Transport1

<1000 >1000 Diffusion Gravitational Settling

Size -/+ + ↓ with ↑ Diameter ↑ with Square of Diameter

Shape -/+ + Uncertain Spheres most efficient

Density -/+ + - ↑ with Density

Surface Chemistry + + Agglomeration1 Agglomeration

Zeta Potential2 + + Agglomeration Agglomeration

Concentration + + Agglomeration Agglomeration

Media Property

Density -/+ + - ↓ with ↑ Media Density

Viscosity -/+ + ↓ with ↑ Viscosity ↓ with ↑ Media Viscosity

26

Add LD50 Table

27

The Evidence this is Nonsense

28

Delivery of These Particles is Driven by Diffusion

Three Cell Types, Three Laboratories, Several Endpoints

One Particle Size Dose Not Test The Hypothesis that Size and Density Affect Delivery

29

Response is proportional to Concentration!

Wait, no its proportional to mass!

Lison et al. 2008

Changing mass &concentration

Changing mass

Changing concentration

30Lison et al. 2008

Response is proportional to Concentration!

Wait, no its proportional to mass!

Changing mass &concentration

Changing mass Changing concentration

31

Cellular Dose is Related to Mass and Concentration

Lison et al. 2008

32

Response Could be Related to Mass and Concentration if Agglomerates are Settling Out

But we really need to know better what is going on in these systems!

Conclusions

In vitro, untested assumptions and a flawed paradigm are the norm Biologically and kinetically relevant measures of dose are not being used, imperiling interpretation of many studies..

What should the future be?Research to improve our understanding of kinetics and dosimetry in vitro is essential, but not often appreciatedDirect measures of particle behavior in solutionDirect measures of cellular particle doseComputational models of delivery

These studies will support interpretation as well as extrapolation (NRC Vision for Toxicology)

What is the Future of Nanomaterial Dosimetry at PNNL?

Establish a Paradigm for Rapid Development of Biokinetic Models From In Vitro Kinetics and Cellular Uptake Data

Complete development of predictive models of in vitro kinetics and cellular uptake (Macrophages, and other elements of the RES system)

Correlate material properties and serum protein binding to rates of uptake in tissues of the RES system

Scale models of cellular uptake to full tissues/in vivo and integrate within our PBPK models

Test and revise predictive models of in vivo dosimetry

This is best accomplished as one element of an integrated program in nanomaterial biocompatibility.

PBPK Model Development and In Vivo Rodent Kinetics Completes our Dosimetry Program

18 nm gold +, - and PEG28 day blood, tissue kinetics in rats

36

Particle Settling Rate Equation:

Where

x

nB

x

nA

t

n

2

2

aN

RTA

6

9

2 2agB

Partial differential equations representing Navier-Stokes settling and Fickian diffusion are combined to form a single partial differential equation describing macroscopic particle transport. The solution to this equation is the basis for the computational model of nanoparticle solution dynamics and dosimetry in vitro (NanoDose). The model is written in MatLab. Parameters: n, number-density of particles at place x and time t; t, time; x, vertical distance from bottom of well; R, gas constant; T, temperature; N, Avogadro’s number; m, viscosity of liquid, a particle radius; g, gravitational acceleration; d, effective particle density (density of particle – liquid density). From Mason and Weaver (1924).

Computational In Vitro Dosimetry

Galya’s CellularUptake Data