Funded Grants

Secreted molecular fingerprints for glioma diagnosis and a novel approach for In-Vivo identification of disease-causing network perturbations

Gliomas are primary tumors of the central nervous system and account for 80% of adult primary brain tumors. The prognosis for patients with an advanced glioma (glioblastoma multiforme) is very poor, with a median survival of less than one year. As noted by the McDonnell foundation, “Brain cancer survival statistics for the deadliest of tumors such as gliomas have not improved significantly over the past two decades.” Clearly, new approaches are needed if we are to alleviate this tragedy. Systems approaches to brain cancer promise to accelerate the pace of progress through new approaches to early diagnostics, a deeper understanding of the pathophysiology of the disease and ultimately more powerful approaches to therapy.

Systems biology promises to transform the practice of medicine from a reactive discipline (responding after the patient is sick), to a predictive, preventive, and personalized discipline. The systems approach to medicine derives from a simple idea — the difference between normal and diseased cells lies in one or more disease-perturbed networks. The initial perturbation could be the result of disease-causing DNA mutations and/or disease-inducing environmental signals. The disease-perturbed networks alter the levels of proteins whose expression they control—and this leads to the symptoms of the disease. Some of these proteins are expressed only in the diseased organ (organ-specific) and perhaps 10-20% of these are secreted into the blood and hence constitute a molecular fingerprint whose protein concentrations report the status of the corresponding organ— healthy or diseased—and if diseased different levels of proteins will correlate with distinct diseases. Brain-specific transcripts can be identified by comparing the brain transcriptome (level of expression for all genes) against the transcriptomes of more than thirty different tissues. Transcripts encoding potentially secreted proteins may be determined by several different computational programs. Since the disease-perturbed networks initiate a disease such as cancer, altered blood protein levels will permit early disease diagnostics. Blood circulates through the body and contains organ-specific, secreted proteins from all tissues and, accordingly, the analysis of the levels or these proteins will permit one to survey health and disease in all organs. As we learn to read these protein fingerprints, blood will become a highly informative window into human health and disease, providing a basis for diagnosing all types of human disease.

Two attributes of the brain make it a prime candidate for the use of brain-specific blood diagnostics. First, it is perhaps the most difficult human organ to biopsy, so tissue-based diagnoses are impractical. Second, it is protected by the blood brain barrier (BBB). At first glance, this fact would seem to be a significant impediment to identifying brainspecific blood-borne markers. When intact, the BBB blocks secretion of some – but not all – proteins secreted into the blood. However, the BBB also provides some unique advantages to us. Most brain proteins are sequestered from the blood when the BBB is intact, so there are not the same levels of background fluctuations in protein concentrations that are seen for blood-borne markers of cancers in other organs. Since the BBB becomes permeable under acute disease, the mere presence of a range of brainspecific proteins can signal acute disease that disrupts the BBB. In previous work, we have identified several brain-specific blood proteins that are potential diagnostic markers in the case of mouse prion disease (a neurological disorder). We believe we can do the same for brain cancers. The development of blood markers to detect, diagnose, and monitor therapies for gliomas are of utmost importance, and may be easily discoverable using the approach we used for prion disease.

In addition to the medically important problem of identifying candidate markers for early diagnosis and the stratification of the various stages and subtypes of gliomas, we also believe that the brain-specific fingerprints will permit us to follow responses to treatments—either successful or unsuccessful. Moreover, we propose a new systems approach to move beyond the identification of markers for diagnosis to a more challenging and, in the long run, more important problem of utilizing blood measurements to track and identify the key perturbations in the biological networks that cause the disease. Developing the capability to identify in vivo network perturbations through blood measurements will open up a new avenue to drug target identification and will provide a novel means to discover the perturbed subnetworks, the further study of which is most likely to yield fundamental mechanistic insights into glioma pathology, and ultimately improved therapies.

The goals of our project are to 1) identify candidate markers for early diagnosis, stratification of the types of gliomas, assessment of the stages of glioma progression and an assessment of responses to therapy, 2) develop a novel systems approach for identifying network perturbations in gliomas, and 3) discover a set of molecular fingerprints for specific perturbations to glioma cell lines. We will also be working with collaborators to test the candidate markers in the sera of patients. Upon successful completion of these aims, we will be prepared to use these markers in conjunction with our novel approach for identifying network perturbations from blood measurements in patients.