Illuminating Risk: New Horizons in Toxics Assessment

Houston’s landscape of chemical infrastructure seemed pretty normal to me, growing up. The vast arrays of storage tanks and refineries sprawling east of Downtown were exotic — a point of fascination on the rare occasion that my family drove toward Kemah or Louisiana — but not unexpected. As a young adult, I actually used to take dates and close friends on late-night drives back and forth across the arc of the bridge over the Ship Channel, awed by the endless hive of blazing sodium lights mirrored by the water, the dancing flames blooming here and there from atop the refinery stacks.

I’ve learned over the years that most cities don’t look much like Houston, however, and that this infrastructure is often tied to health and environmental impacts that may or may not be obvious.  I can’t see refining equipment now without also seeing its potential to change the invisible chemical landscapes in which we all live and operate, as well as the unevenly distributed impacts those changes can have on everything from physical well-being to ecosystem integrity to housing and property values.

But measuring changes in our chemical environment — a first step in understanding the likely risks they pose — is actually an extraordinarily complex process. For one thing, scale matters: one big spill from a tank is (usually!) much easier to measure and track than, say, the tiny trickles of gas leaking from a hundred pipeline welding defects along a miles-long easement — each all but undetectable on their own, but maybe adding up to a big problem.

Knowing what to look for also poses a challenge. Maybe we have a sensor to measure how much of Chemical A is coming out of a factory drainpipe, but do we even know to try to measure the trace chemicals created as byproducts when Chemical A is made in big batches for sale? What about when Chemical A gets moved into a new tank — and reacts, unexpectedly, with the tank’s inside coating, to form another new substance with its own properties?

And how often is our monitoring station taking its measurements? Do the pesticides and fertilizers building up in an aquifer get flushed out into a downstream river all at once during rainstorms, while the water quality samples collected by workers on a sunny day show no problems?

Capturing the real picture of complex environmental exposures over time is a daunting task; to do it better than we currently do, we’ll need to apply new approaches and technologies. Fortunately, there are people already at work on bringing to life some of the next-generation methods that may carry this field forward. Whether it’s by sending roving monitoring stations into Houston’s most polluted neighborhoods, building devices containing cell cultures to show directly how pollution affects the cells of the lung, or other creative ways to illuminate this landscape, a lot of people are working on new methods to measure the real scope and impacts of complex pollution.

As part of my Environmental Health course this semester, I got the chance to learn about another cool set of technologies that may bring big changes to the chemical risk assessment paradigm in coming years — and the regulatory landscape along with it. But to understand what’s so special about it, let’s talk a little bit first about how risk assessment is currently done.


The two primary pieces of the chemical risk assessment puzzle are effects (what a chemical can potentially do to a human or an ecosystem) and exposure (how much of that chemical is likely to actually make contact with the people or organisms in question).

Putting these pieces together lets an assessor characterize and quantify “risk”. For example, if a chemical’s potential to cause health effects is very small, then a tiny or moderate degree of exposure may not matter. But if a substance is super toxic, a tiny exposure may be a serious threat to health.  If the likely exposure faced by people or animals is less than some critical threshold dose or “concentration of concern”, then a substance isn’t deemed a major risk. Above that threshold, however, it is.

The problem with this framework is that chemicals don’t exist in isolation. Maybe Chemical A, coming out of that factory water outlet pipe into a small creek, isn’t a threat to the local ecosystem, since it’s below that certain concentration level deemed “safe”. But what happens when Chemical A is being released in a slurry that also contains Chemicals B, C, D, E, F, X, and Z? Studying mixtures is a whole different ballgame.

Some compounds may behave in very similar ways to one another — meaning that getting a “safe” dose of Chemical A, along with just a little bit of B, C, and D, may actually cause the same health effects as getting dosed with a whole lot of Chemical A. A risk assessor may believe that small amounts of each chemical, considered on their own, don’t pose a likely threat to a community or ecosystem. But taken all together, these exposures could add up to do real damage.

What’s more, some chemical mixtures can actually interact in ways that make the total impact of an exposure worse than exposure to just the individual components. For example, the dispersant mixtures used to break up the 2010 Gulf oil spill likely made the spill far more toxic to some marine life (and perhaps to people) than just the oil alone. A more familiar example is drug interactions: drinking too much alcohol along with Tylenol is well known to boost your chance of liver damage, and even grapefruit juice can interact with some kinds of medicines to change how these drugs behave in the body. Some chemicals that aren’t harmful on their own may only become toxic when mixed with the right combination of other stuff.

What this all means is that the specific kind and amount of pollution coming out of your refinery may not be a problem for the local community — but taken altogether with what’s coming out of every other building in the region, the public health picture could seriously change for the people nearby. Under the current risk assessment paradigm, that picture may not be captured.


In a number of classes and field trips this semester, I’ve had the chance to talk to researchers, regulatory agency workers, and industry employees who have suggested that big changes to this risk paradigm might be on the horizon. Rather than just focusing on untangling what each kind of pollution in a given environment does on its own, future assessments might begin to incorporate studies of what a relevant mixture, all together, does directly to a specific set of body systems.  A great example of what this might look like can be found in ongoing research around endocrine disruptors — classes of chemicals that can interfere with how the body handles hormones.

Hormones are the body’s messaging system. Hormones produced in response to genes, to stress, or even to activities like eating, can trigger or control all sorts of changes: fetal growth in the womb, fight-or-flight reactions, puberty, weight gain, and much, much more. When the body naturally produces molecules like testosterone, estrogen, or adrenaline, cells “read” these chemical signals: the hormone molecules bounce around until they hit a special receptor site on or in a target cell, shaped just right to receive them. This binding of the hormone to a cell’s receptor causes some kind of a change in cell’s activity, often triggering a gene to be translated into a protein or some other process to ramp up.

Endocrine disruptors are chemicals that can interact with the body’s hormone systems in a few different ways. Hormone “mimics” can send bogus messages to cells in the body by binding to the target cell sites that are supposed to receive a hormone — basically tricking the cell into acting like it’s just been hit by testosterone, estrogen, or some other real signal molecule. Alternatively, these disruptors can prevent the “real” messages from being read by blocking the receptor site altogether, keeping the hormone signals from doing what they’re supposed to do. Either type of disruption can be bad news for a cell, or for an organism as a whole.

Endocrine disruptors are a great example of how lots of below-threshold exposure to chemicals that act in similar ways can add up to produce major impacts – the “something from ‘nothing'” effect, as it’s been called. These kinds of chemicals may be different enough from one another that a traditional risk-assessment approach would consider them separately, and not find a likely risk at low levels – but when they are all present together, a surprisingly large impact on health may be observed. In other cases, chemicals that wouldn’t do anything to the body on their own may be changed by reactions in mixtures into chemicals that do have endocrine effects, or may somehow work together with other chemicals in ways they wouldn’t on their own.

Dr. Chris Kassotis, a post-doc in one of Duke’s toxicology labs, visited our Environmental Health class this semester to talk about his previous work toward understanding the health impacts of the chemical mixtures used in hydraulic fracking of oil and gas wells. Through a variety of routes peripheral to the actual fracking process, these chemicals sometimes find their way into drinking water in the areas near and downstream of drilling sites. Some of these often-proprietary chemicals are known or thought to be endocrine disruptors – the question is, at what concentration do they matter? And how do they function as part of a complicated mix?

Dr. Kassotis talked to us about studying these mixtures not in terms of the impacts from each of their 100+ components, but in terms of their collective impact on the way the body processes hormones. Using samples of actual fracking wastewater diluted down to levels that can be found in the environment in some places, Kassotis and others can directly measure hormone impacts using a kind of biological test that is becoming more and more common: they create a set of engineered cells in which hormone target receptors are linked to a gene that controls the same bioluminescent reaction that makes fireflies glow. These cells are then exposed to the fracking chemical mixture; when something in the frackwater mixture triggers one of the hormone receptors being tested (no matter which specific chemical it might be) scientists can measure the light produced, and use this to quantify the endocrine-disrupting action of the mixture as a whole.

Kassotis pointed out to me later that other labs are applying this same kind of firefly reporter assay (so called because the light-emitting reaction “reports” on the action of interest) technique to other kinds of complex environmental mixtures. He’s currently working as part of Dr. Heather Stapleton‘s lab here at Duke, which is looking at how chemicals in normal house dust (complete with traces of everything from food to beauty products to cleaning products) might impact other body hormone systems; their lab has been running similar tests using extracts from this dust as the source of potential endocrine disruption.

(Other labs use this same reporter gene to show different types of genetic activity, not just in animal cells but in plants; a group in the Netherlands at Wageningen University has used it to show circadian rhythms gene action, for example — which I’m including below mostly because it’s really cool to see.)

Will these kinds of new tests and new technologies geared toward assessing chemical impacts and exposure from a different angle ever translate into regulation? That’s still a long way off, according to some of the EPA and industry folks I’ve gotten to speak to this year — but several of them have told me such effects-focused testing may well be the future of the field. The ability to directly measure how complex chemical mixtures impact critical functions of cells and other body systems, simply by dropping those mixtures into a petri dish, opens the door for a totally new framework for measuring — and making rules based on — what the real impact of the chemical landscape may be on human and environmental health.

The future of the  field is hazy — but, maybe, it’s getting brighter.

Images: LyondellBasell (Houston refinery), USDA (risk assessment framework diagram), NIEHS (endocrine disruptor diagram), University of Missouri School of Medicine (luciferase assay diagrams, courtesy of Chris Kassotis)

Video of plant timelapse: Laboratory of Plant Physiology at Wageningen University