Catherine ​Metayer, MD, PhD

Goal: Characterizing In Utero Chemical Exposures, Immune Status, and Childhood Leukemia Risk

Team Members: Dr. Catherine Metayer (UC Berkeley), Dr. Todd Whitehead (UC Berkeley), Dr. Xiaomei Ma (Yale University), Dr. Joseph Wiemels (University of Southern California), and Dr. Scott Kogan (UC San Francisco).

The Immune System and Childhood Leukemia

Leukemia is a malignancy of immune cells (called “leukocytes”), in the blood. As such, abnormal immune-system development plays an important role in the initiation of childhood leukemia. Several studies have indicated that early-life infections have the potential to influence the development of the immune system and, by extension, the risk of childhood leukemia. One leading theory – referred to as the “delayed infection hypothesis” – suggests that when a child is NOT exposed to common infections early in life, the uneducated immune system is more likely to react poorly to infections later in life, potentially leading to the development of leukemia. Epidemiological studies using indirect measures of early-life immune challenges – such as being born via vaginal delivery, living with older siblings, having social contacts in daycare settings, or being vaccinated – have consistently reported a subsequent reduction in childhood leukemia risk, supporting the role of early-life immune-system education in protecting children from leukemia-causing infections. The two most common molecular subtypes of childhood leukemia are initiated in utero and studies of biomarkers in newborns suggest that abnormal immune-system development before birth may also play a role in the initiation of childhood leukemia.

How CIRCLE Measures Immune Development

Project 1 leverages two population-based studies – the California Childhood Leukemia Study and the California Mother-Child Birth Cohort – to support CIRCLE with biospecimens and data from birth certificates and participant interviews. To assess immune status at birth, CIRCLE measures a panel of nine cytokines — interleukins IL1β, IL4, IL6, IL8, IL10, IL12p70, granulocyte-macrophage colony-stimulating factor (GM-CSF), tumor necrosis factor alpha (TNFα), and vascular endothelial growth factor (VEGF) — in blood samples collected from newborn babies. CIRCLE investigators have reported that infants who went on to develop childhood leukemia were born with profoundly depressed levels of IL10, a critical immunosuppressive cytokine that regulates immune responses to infection after birth in healthy individuals. Finding these abnormal IL-10 levels in the blood of newborns who went on to develop childhood leukemia provides more evidence of the role of immune development in the etiology of childhood leukemia. More recent CIRCLE research has implicated a wider array of cytokines (in addition to IL-10) as potential biomarkers of immune dysregulation in archived neonatal biospecimens collected from childhood leukemia patients. At the same time, CIRCLE investigators are using a mouse model to reveal connections between in utero chemical exposures and disruptions in cytokine levels at birth. These controlled experiments provide important insight into the relationships between environmental exposures, neonatal cytokine levels, and childhood leukemia risk.

What are Cytokines?

A cytokine is a small protein which is released by cells. Once released, cytokines influence the behavior of other cells through receptors, a process which can be thought of as an intercellular interaction or as intercellular communication. The cytokines include interleukins, lymphokines and cell signal molecules, such as tumor necrosis factor and the interferons. An important group of cytokines are those which modulate the balance between humoral and cell-based immune responses through the regulation of the maturation, growth, and responsiveness of particular immune cell populations. It is this group of cytokines — the ones which trigger inflammation and respond to infections — that are the focus of CIRCLE’s Immune Development Project.

Goal: Identifying In Utero Exposures that Are Risk Factors for Childhood Leukemia

Team members: Dr. Steve Rappaport a leading proponent of the exposome concept (UC Berkeley); Dr. Lauren Petrick (Icahn School of Medicine at Mount Sinai), and Dr. Sandrine Dudoit (UC Berkeley).

What is the Exposome?

The exposome is the lifetime of exposures (coming from both inside and outside of the body) to which a person comes in contact. The exposome concept (originally proposed by Christopher Wild) was made practical by Dr. Rappaport and his colleague Martyn Smith who set out to measure the collection of chemicals that are found circulating in the blood – the blood exposome. The blood exposome provides a record of a person’s exposures to a broad range of environmental agents including radiation, stress, infections, drugs, diet, and pollutants. Additionally, internal processes, such as inflammation, lipid peroxidation, oxidative stress, and gut flora also shape the blood exposome.

An Untargeted Approach to Measuring Exposure to Chemicals

There is an allegory about a man searching for his keys under a lamppost which explains why many well-intentioned scientific efforts fail. In the context of an epidemiological study, the “keys” are important risk factors for disease and the “light from the lamppost” is a validated analytical assay. Inherently, scientists tend to focus on the assays and chemicals that they know well in their efforts to identify important risk factors for disease. However, the allegory teaches us that we should also be looking “across the street in the dark”, that is, in the universe of previously unidentified exposures that may be most relevant to disease. To move away from the lamppost, the project uses an ‘untargeted’ approach, which means the list of chemicals being evaluated is NOT predetermined. An untargeted approach strives to be all-encompassing, providing investigators with an opportunity to cast a wide net in their search for relevant environmental risk factors (including both harmful and protective factors). By design, Petrick and Rappaport are measuring several chemicals which have not been considered as possible leukemia risk factors previously; indeed, some of the chemicals are still mysteries, as their exact structures remain unknown. About 30,000 chemicals are measured in tiny drops of blood collected from children with leukemia and looking for clues about what might cause (or prevent) the disease.

How CIRCLE Measures the Blood Exposome?

Researchers have shown that the development of childhood leukemia is a two-step process, which is often initiated before birth, followed by postnatal events leading to the disease. Both steps of the process are triggered by environmental factors such as exposure to microbes or toxins. To describe the neonatal blood exposome, Petrick and Rappaport are using a special resource called neonatal blood spots. These blood spots have been routinely collected from California newborns since 1982. The State of California uses the blood to test for metabolic defects at birth and has archived leftover spots for subsequent use in any epidemiological investigation with the potential to benefit the public health. Using just a fraction of a single drop of blood from each child, CIRCLE investigators can measure thousands of chemicals that were present at birth. Each laboratory test starts with a 4.7-millimeter punch of filter paper containing about 8 microliters of blood (the equivalent of one tenth of a single drop). Roughly 30,000 chemicals were present in each of the blood spots tested, of which 900 of the most-commonly found molecules were analyzed, including hundreds of different lipids and fatty acids.

Finding Causes of Childhood Leukemia

The goal of the CIRCLE Exposome Project is to identify agents that cause childhood leukemia, especially among those chemical exposures which have not been previously evaluated as risk factors for the disease. To this end, Petrick and Rappaport characterized the neonatal blood exposomes of 332 children with acute lymphoblastic leukemia (ALL) and 324 healthy controls matched by date of birth, sex and race/ethnicity. The study population was divided into two age groups – one for early diagnoses (ages 1–5 years) and one for late diagnoses (ages 6-14 years) – and leukemia patients were compared to healthy controls. In both groups, there were differences observed between the exposomes of the cases and controls, including a set of 9 discriminating chemicals among the younger group and 19 discriminating chemicals among the older group. Interestingly the list of chemicals did not overlap between the two age groups, suggesting that different factors were involved in the development of early- and late-onset ALL. The late-diagnosis group provided the most interesting findings – with a group of chemicals derived from the essential nutrients, linoleic acid and linolenic acid, being found at higher levels among the childhood ALL patients, suggesting that these fatty acids may be linked to a higher risk for the disease. Levels of metabolites found in our study were correlated to newborn nutrition (breastfeeding) and with the mother’s body mass index before pregnancy. Drs. Petrick and Rappaport are uncovering an intriguing web of connections between childhood leukemia and small lipid molecules related to newborn and maternal nutrition. This opens the door to hypotheses and follow-up studies that can validate the initial findings and pinpoint the underlying causal pathways. Identifying the causes of childhood leukemia will make it possible to intervene in ways that reduce the risks of this devastating disease.

Goal: Characterizing Prenatal Exposures, Constitutive Genetics, DNA Methylation & Childhood Leukemia Risk

Team members: Dr. Joseph Wiemels and Dr. Adam de Smith (University of Southern California). Dr. Wiemels’ seminal papers on the natural history of childhood leukemia, demonstrated that the disease is often initiated in utero. As such, a major focus of CIRCLE is to discover the prenatal events (such as epigenetic modifications) that cause leukemia later in childhood.

What is epigenetics?

Epigenetics literally means “above” or “on top of” genetics. Epigenetics are modifications to the outer surface of a DNA molecule that turn gene expression “on” or “off.” Epigenetic modifications do not change the DNA sequence; rather, they affect how genes are transcribed into proteins. One example of an epigenetic change is DNA methylation — the addition of a methyl group to the DNA structure — which can alter the expression of nearby genes. Epigenetic modifications can have a very strong influence on phenotypes, indeed, the functional inactivation of a gene by DNA methylation can be as powerful as a gene mutation.

Epigenetics and Childhood Leukemia

DNA methylation is critical during cell development so that genes which are not needed are permanently repressed in a cell that becomes specialized for a specific function. In cancer, DNA methylation processes can go awry; cancer cells are typically globally less methylated than their normal cell counterparts while having more DNA methylation in the control regions of key tumor suppressor genes. Like all cancers, childhood leukemia is characterized by DNA methylation changes in its target organ, blood. CIRCLE has reported that childhood leukemia tumor cells are profoundly altered from their pre-B cell precursors with regards to DNA methylation. Environmental risk factors for childhood leukemia, including polycyclic aromatic hydrocarbons and folic acid (protective), are also known to affect DNA methylation. CIRCLE is investigating the relationships between in-utero environmental exposures, DNA methylation at birth, genetics, and childhood leukemia risk.

CIRCLE Findings

CIRCLE investigated DNA methylomes of pediatric B-cell acute lymphoblastic leukemias using whole-genome bisulfite sequencing and high-definition microarrays, along with RNA expression profiles. To be able to characterize an abnormal departure from normal developmental states investigators compared genome-wide DNA methylomes of B-cell acute lymphoblastic leukemias (tumor cells) to their normal B-cell precursors from the same individual. Methylation profiles of 227 children with leukemia whose DNA were analyzed by Illumina 450k array. A hierarchical clustering analysis was used to characterize the 500 DNA loci with the most variable methylation into four distinct clusters. Some clusters were dominated by specific cytogenetic groups; cluster I by hyperdiploid, cluster II by others, cluster III by ETV6/RUNX1, and cluster IV by hyperdiploid/others. These initial findings of the CIRCLE Epigenetics Project demonstrate that modifications in DNA methylation influence childhood leukemia risk and that different subtypes of the disease have separate patterns of epigenetic alterations. CIRCLE will continue to investigate the causes of these methylation changes, with the ultimate goal of preventing future incident cases of childhood leukemia.