UCSF

It is becoming increasingly possible to analyze detailed genotypic and phenotypic data in the study of complex disorders. One such disorder is brain injury and the accompanying cerebral edema, or brain water accumulation, caused by trauma. Traumatic brain injury (TBI) is a significant public health problem. Each year in the United States, approximately 1.4 million people experience a TBI. These injuries result in about 50,000 deaths, and survivors often face life-long cognitive and physical disability.

Here we describe cellular, computational, and clinical work to study brain injury phenotypes, defined by physiological vital signs and response to drug treatment, and genotypes, with emphasis on variation in Aquaporin-4 (AQP4), the primary water channel protein in the brain. Finally, we discuss early work to begin to study the association between outcomes after brain injury and race.

In our study of brain injury phenotypes, we first implemented a data-driven classification approach to construct multi-variate physiological data "profiles" to classify patients for diagnosis and treatment. We then focused on the specific problem of understanding the dose-response relationship of mannitol used when treating elevated intracranial pressure (ICP). We measured ICP continuously in patients with TBI who were given at least one dose of either 50g or 100g of mannitol. After 100 minutes, ICP had increased in the 50g group to nearly its initial value but was still lower in the 100g group. We also identified previous studies that quantitatively characterized the dose-response relationship of mannitol and ICP. Meta-regression found a weak linear relationship between change in ICP and dose.

In our study of brain injury genotypes, we focused on AQP4, a member of the aquaporin gene family involved in the physiological water balance essential for life. For human, naturally-occurring, non-synonymous point mutations, we developed a map of aquaporin variation and functional features. AQP4, a water channel expressed in the apical membrane of the foot processes of brain astrocytes, plays a significant role in brain water homeostasis. We analyzed DNA samples from an ethnically-diverse cohort of healthy volunteers to screen for AQP4 variants and identified four novel AQP4 nsSNPs. Cellular assays found that compared to wild type, the mutations reduced water permeability. We then compiled a data set of structures, interfaces, and functional mutations and applied structural modeling to describe the location and effect of SNPs. We found that nsSNPs were more likely to occur within 5 Angstroms of at least one interface than background.

Many factors are thought to be prognostic indicators for mortality and disability after TBI. However, the effect of race and ethnicity has been studied tangentially or in homogeneous settings. We retrospectively assessed outcome by race and found that, against dead versus alive outcome, Asians had an odds ratio of 2.25 for worse outcome.

This dissertation represents progress towards quantitatively defining physiological and drug responses after traumatic brain injury and cerebral edema as well as potentially important genetic factors that may influence outcomes. While larger, prospective studies will be needed to validate predictors as well as genotype-phenotype-outcome relationships, the methodologies of dose-response and physiological profile analysis as well as the identification and characterization of genetic variation provide an important foundation for future research.