We are using an animal model to explore the possibility that fetal ultrasound exposure in utero can be a contributing factor of autism. The Williams and Casanova Triple Hit Hypothesis suggests that autism is caused by a combination of genetic predisposition, a specific period of developmental vulnerability, and exposure to environmental factors during that time of vulnerability. We hypothesize that ultrasound is an environmental factor contributing to autism in those who are genetically predisposed. A previous study found that prenatal ultrasound exposure affects neuronal migration, indicating that ultrasound is not completely harmless to the developing brain. Our experiment will use an animal model of three strains of mice with varying degrees of predisposition to autistic-like behavior. Some pregnant mice will be anesthetized and exposed to a clinically relevant dose of ultrasound while control mice will be subjected to the same procedure with the ultrasound turned off. When the mice pups are 3 and 6 weeks of age, we will use a variety of behavioral tests to assay for autistic-like characteristics. Examples of autistic-like behaviors include reduced play, anxiety in open areas, little curiosity for a novel mouse, and affinity for darkness. Results of our study may relate ultrasound exposure to autism, which would lead to further animal studies and, eventually, recommendations on the use of ultrasound on pregnant women. A positive correlation between ultrasound and autism would suggest that pregnant women should be exposed to ultrasound only long enough to assay the fetus' health.

Ultrasonic neuromodulation is the process of stimulating different neural pathways in the brain through the use of transcranially delivered ultrasound in order to induce clinically relevant emotional, cognitive, physical or behavioral reactions. My research goal is to find out if it is possible to add anatomical specificity to current neuromodulation practice through the use of modulated focused ultrasound (mFU), where 'modulated' refers to adding complex temporal structure to the waveform to optimize neuromodulation. Since mFU has a small focus relative to the brain, the acoustic power from the ultrasound can be moved within the brain; previous experiments have used unfocused ultrasound which sends ultrasound through the entire brain. We hypothesize that we can map varying physical reactions through stimulation of different parts of the brain, with initial emphasis on the motor cortex because stimulation there produces readily observable results. With lightly anesthetized mice as our subjects, and always delivering ultrasound through the skin and skull, we began neuromodulation with unfocused transducers to transmit our ultrasound. This critical step was done to determine the optimal intensity, waveform shape and carrier frequency of ultrasound that created robust, observable effects. Next we translated these ultrasound parameters to our mFU transducer. The current stage of the process includes moving through different parts of each hemisphere and recording the associated motor response such as leg movements and tail flicks. We have already seen a change in activity between activating the front, middle, and rear regions of the brain, as well as dosing effects. Going forward, we intend to see if we could produce varied motor movements with greater spatial resolution using our focused ultrasound across a predefined grid atop the brain. Further studies will also change the applied stimulation intensity. Clinical implications for neuromodulation research include retraining brain function after injury and improving current therapeutics.

The insertion and placement of an external ventricular drain (EVD) addresses two important neurological problems with more than two hundred fifty thousand procedures per year. For example, ventriculostomies are performed when intracranial is elevated and either is immediately dangerous or will not respond to drug therapies. The success rate of EVD placement is sufficiently low to cause concern and few systems exist for guided assistance in EVD placement. We propose a handheld, disposable, ultrasound guided ventricular catheter to assist neurosurgeons in precise and accurate EVD placement. Utilizing existing catheter designs, we have replaced the current guide wire with one that contains a small (one mm in diameter) ultrasound transducer at the tip. The ventricles have a set of acoustic properties significantly different from brain tissue, allowing us to detect them with ultrasound. We processed the ultrasound backscatter signal from the ventricle walls using custom electronics to give feedback to the neurosurgeons while they are directing the catheter. We have collected initial data that supports the feasibility of this device. Results to date in porcine models have demonstrated an ability to detect ventricles and assist in the precise guidance of the catheter into the desired lateral ventricle. Outcomes have been repeatable for practicing neurosurgeons over all trials in five experiments and for others over 8/10 attempts in three experiments using a Sonosite MicroMaxx to verify accurate EVD placement. Moving forward we will incorporate pulsitility and Doppler imaging modalities to decrease the margin of error for false positive ventricle detection. We aim to increase the reliability of accurate EVD placement in both emergency room and operating room situations while providing a product that is less expensive and more accessible than current guided systems.

Traumatic brain injury (TBI) is a major issue in warzones and it is often misdiagnosed and therefore mistreated in the field due to unavailability of accurate imaging technologies and subjectivity of motor response tests. Soldiers who experience TBI and are not properly treated often deal with severe mental, physical, and emotional consequences for their remaining lifetime. As such, there is a need for a field-deployable modality for the detection of TBI. Proposed here is the use of ultrasound vibro-acoustography for detection, which has been shown to measure differences in mechanical properties of materials using intense focused ultrasound (iFU) to vibrate the material and record the acoustic emissions. I am working to quantify the emission differences of healthy and damaged tissue through in vivo rat testing.  By optimizing ultrasound parameters and experimental protocol, my work will support the end-goal design an affordable and portable device for TBI detection away from modern medical centers.

Annually, an estimation of 1.7 million people in the U.S suffer from traumatic brain injury (TBI), in which 50,000 people dies and another 85,000 people survive with long term disabilities. Moreover, TBI is reported to be the most common combat-related injury and accounts for 25% of combat casualties. CT and MRI are the current diagnostic tools for TBI, but due to their bulky size and complicated interpretation procedures, prevalent implementation of the devices is limited in undeveloped areas and near the battle field. Consequently, a size deployable diagnostic tool that is able to present quantitative evaluation of TBI is needed. The purpose of the project is to optimize an ultrasound elastography machine for detection and classification of mild and moderate TBI based on the theory that abnormal brain elasticity will be observed in presence of hemorrhage and edema. The goal was achieved in three steps: post-processing and stiffness maps were first generated in MATLAB for breast phantom to test and ensure the feasibility of coding, then data collected from moderate TBI rat brains were fed into the code with modified parameters to generate brain elasticity maps highlighting areas of damage, finally elasticity maps were be constructed for mild TBI rats. All the brains were imaged in alginate of same impedance. The quantified elasticity values in moderate and mild TBI brains allow classification of injury levels; the accuracy of constructed images will be ensured by comparisons between TBI, control and histology results. Successfully optimized ultrasound elastography machine displays quantitative analysis of brain elasticity which simplifies diagnostic procedures. Furthermore, implementation of the device in hospitals and battle fields enables early TBI detection, which will potentially reduce TBI-related deaths, lower TBI-related treatment costs and greatly enhance residents’ life quality.

Traumatic brain injury (TBI) is injury to the brain caused by external trauma. Approximately 1.7 million cases of traumatic brain injury are reported every year in the US, and contributing to an estimated 30.5% of injury related deaths. TBI is currently treated by managing symptoms, present treatment methods are incapable of reversing the damage done by the initial trauma. Gene therapy is a promising method of treatment for TBI; by transfecting cells such as neural stem or progenitor cells (NSCs and NPCs respectively) in the brain with plasmids which induce differentiation and migration damage to neurons can be alleviated or even reversed. This projects aims to develop and optimize a protocol which utilizes non-viral transfection vectors in conjunction with ultrasound to efficaciously target and transfect neural progenitor cells (NPC), i.e. cells capable of differentiation into neurons. By transfecting NPCs with plasmids which cause differentiation and/ or migration NPCs can be induced to become new neurons and migrate to areas of neuronal injury caused by traumatic brain injury. In this manner some of the dead and dying neurons might be replaced and the initial trauma mitigated. This is a novel method of treatment both in its use of non-viral vectors capable of targeting NPCs in conjunction with ultrasound mediated increase in transfection efficiency and in its potential for reversing the initial damage seen in TBI; something that no current method of treatment is capable of.

Traumatic brain injury (TBI) is a common injury sustained by soldiers and requires immediate and sensitive brain imaging in order to diagnose the extent of damage. Currently, conventional methods of brain imaging such as computed tomography and magnetic resonance imaging are not field-deployable and immediately available on the battlefield. Tissue pulsatility imaging (TPI) is a technique that measures the natural pulsatile motion of tissue as a result of blood flow. Preliminary brain tissue samples of mice with TBI observed under widefield microscopy have shown increased mitochondrial accumulation at the site of injury relative to comparable sites in the hemisphere contralateral to injury, suggesting that TBI disrupts normal mitochondrial migration patterns. It is hypothesized that biological responses such as this, as well as the development of hemorrhage and edema, cause a change in brain tissue pulsatilities of TBI animals that can be detected by TPI. In this project, TPI by ultrasound will be used to image the brains of mice with acute and chronic (24-hour) TBI to determine whether this method of ultrasound imaging has the potential to be used for initial detection of TBI. Animal trials with both bilateral and hemispheric cranial windows will be used to study TPI patterns in the brains of these mice. Among the issues we will address include the effects of the skull on image resolution. Subsequent data analysis and device parameter adjustment should increase the resolution of images in order to optimize a design for an ultrasound device based on imaging tissue pulsatility that can accurately detect TBI and may eventually serve as a field-deployable brain imaging device. 

The project I worked on was part of a larger project in which we examined the activity of O6-methylguanine-DNA methyltransferase activity, CpG promoter methylation, and progression-free survival following alkylating agent therapy in glioblastoma and anaplastic glioma brain tumors. For my part in the research I performed standard biochemical assays to quantify O6-methyalguanine-DNA methyltransferase (MGMT) in extracts of glioblastoma (GBM) and anaplastic glioma (AG) tumor samples from adult patients operated on with consent at the University of Washington Medical Center between 1991 and 2008. MGMT is a DNA repair protein that excises methyl adducts from the O6 position of guanine residues. Chemotherapy and other alkylating agents such as methylating triazene derivative temozolomide (TMZ) are responsible for methylating DNA, or damaging the DNA and thus destory the tumor cell. There has been poor prognosis for GBM due to the intrinsic resistance mechanisms in the MGMT that limit the efficacy of TMZ and other alkylators. GBM and other gliomas display a wide range of MGMT activity and thus the question for research was whether the amount of MGMT in tumors correlated with progression-free survival following alkylator-based chemotherapy. Overall the data strongly supported that MGMT activity in GBM and AG tissue mediates resistance to alkylator therapy.

Glioblastomas (GBMs) are grade four astrocytomas, composed of heterogeneous cells with unique molecular features. Using brain tumor and tumor stem cell lines, we are developing ex vivo methods to use on-chip, single cell characterization and mass spectrometry to better understand brain tumor heterogeneity in order to personalize drug therapy and optimize treatment of GBMs. We first performed an on-chip, single cell culture, using a biocompatible poly dimethyl-siloxane (PDMS) microchip that contains 50-100 nm subnanoliter wells. A diluted solution of cultured cells was loaded onto the microchip while the chip was submerged in DMEM, and the cells settled into the wells by gravity. The PDMS microchip was then subjected to hybridization conditions, to simulate the technique of assaying cytokines. Cell viability was monitored using a live/dead fluorescence assay. To pursue the goal of analyzing protein arrays using single cell mass spectrometry, we loaded a PDMS microchip; first with bovine serum albumin digest, then with hemoglobin, as controls to obtain a MALDI (Matrix-Assisted Laser Desorption Ionization) spectrum. We found that a solution of approximately 7.14x103 cells/mL ensured that one cell was delivered per well. The neuroblastoma cells did not exhibit typical dendritic morphology within the microchip. They remained round due to the non-adherent properties of PDMS. When the loaded microchip remained submerged in DMEM and incubated, the cells grew to confluence and remained viable up to seven days. Under hybridization conditions, the cells remained viable up to five days. The MALDI controls were assessed, but did not produce a spectrum. The soft and thick PDMS surface may inhibit the MALDI laser’s accuracy to pinpoint proteins. Also, PDMS is not electrically conductive and impedes analyte ionization. Therefore, the project has been continuing to optimize the surface for analysis. This work is still ongoing with promising results.

 Personalized medicine aims to improve patient outcome by tailoring cancer treatment to individuals based on the specific molecular makeup of tumors. This approach is showing particular promise in the treatment of gliomas, the most common type of primary brain cancer. The DNA repair protein O⁶-methylguanine-DNA methyltransferase (MGMT) removes O⁶-alkylguanine adducts from DNA produced by the therapeutic methylating agent temozolomide (TMZ). In glioblastoma multiforme (GBM) tumors, low levels or lack of MGMT activity is correlated with longer progression free survival (PFS) following TMZ treatment and other alkylating chemotherapy. However, in oligodendorgiomas and oligoastrocytomas the effect of MGMT activity on tumor resistance has not been established. Better clinical outcome is associated with 1p and 19q loss of heterozygosity (LOH), which occurs in an appreciable fraction of oligodendrogliomas. Other studies associate these deletions with CpG methylation of the MGMT promoter, an epigenetic indicator of low MGMT activity. However, there have been no studies linking MGMT activity to chromosomal deletion, and 1p/19q LOH is associated with better clinical outcome regardless of therapy, suggesting that other mechanisms may also be responsible for prolonged survival. To clarify the unknown role of MGMT in these tumors, we tested the hypothesis that 1p/19q LOH accompanies low MGMT activity by: [1] assaying MGMT activity by measuring the extent of transfer of tritium-labeled methyl groups at the O6 position of guanine to MGMT in substrate DNA extracts of deleted and non-deleted tumors; [2] assessing promoter CpG methylation status using methylation specific PCR; and [3] examining the correlation of MGMT activity to TMZ treatment response using a Cox regression analysis. This approach should help to clarify the relationship of MGMT activity and 1p/19q LOH.

Spinal Cord Injury (SCI) currently affects more than 250,000 people in the US alone, and approximately 11,000 new injuries are reported annually. SCI is most commonly caused by trauma, such as vehicular accidents, and often results in devastating paraplegia or tetraplegia. Among many points of interest in treating SCI is what happens to myelin post injury. In terms of electrical conduction, myelin sheaths are the insulation to the wire that is the neuron’s axon, and proper myelination is crucial for effective signaling in the nervous system. Previous findings in SCI research show abnormally thin and short myelin near injury site, supporting the current research dogma that oligodendrocytes (OLs), the glial cells that myelinate the central nervous system, undergo complete cell death following initial injury. Due to lack of methodology, little research has been done to support the idea of partial myelin degeneration. My current study seeks to shed light on the question of where the abnormally thin myelin is derived, if it is indeed newly developed, and explores the possibility that thin myelin found by previous researchers is intact degenerating myelin. Pre-labeled myelin in transgenic mice was injected with a Green Fluorescent Protein (GFP)-labeled virus vector, allowing clear visualization of which myelin sheaths are mature and which are newly generated. By histology and confocal microscopy, I will analyze morphology of neural processes and characteristics of myelinating OLs, such as distances between axon internodes and the axon g-ratio (the relationship of axonal diameter to total fiber diameter, a common measure of myelin thickness). If the prevailing research dogma is supported, I expect to find shortened internode lengths and smaller g-ratios in the green myelin. The findings help determine the significance of myelin change post-SCI and is hoped to contribute more effective therapies to improve axon conduction via pharmacological or cell-based approaches.