Colin Collins, PhD
Research Interests
Oncogenomics
The primary research focus in the Collins laboratory is oncogenomics,
the application of genomics to basic cancer biology and clinical
oncology, with a primary emphasis on understanding and treating
prostate cancer. Prostate cancer is diagnosed in approximately
180,000 men per year and kills 30,000. Collins is principal investigator
on two NIH funded grants, Project 2 of the UCSF Comprehensive
Cancer Center Prostate SPORE grant, and an RO1 award funded through
June 2010.
The first goal of these projects is to identify prognostic markers
with the power to separate prostate cancer patients into lower
and higher risk groups, to enable the selection of appropriate
therapy, and to identify novel drug targets. To accomplish this,
his laboratory is using Array Comparative Genomic Hybridization
to generate copy number genotypes of approximately 130 primary
and 25 metastatic tumors. They have identified approximately 40
DNA-based biomarkers that occur in primary tumors with a high
likelihood of recurring or metastasizing after surgery. The DNA
biomarkers are segments of DNA that represent the human genome,
bacterial artificial clones (BAC) called Genomic Evaluators of
Metastatic Prostate Cancer (GEMCaP) loci. The laboratory has shown
that GEMCaP loci are as good at predicting disease recurrence
as nomograms, the current gold standard. Collins’ laboratory
is developing technologies so that the GEMCaP assay can be performed
on tissue biopsies before surgery, to separate men into two groups:
those who need definitive therapy and those who would benefit
from active surveillance.
The second goal of these projects is to identify drug targets
encoded at the GEMCaP loci so that physicians can predict which
men need therapeutic intervention and determine the appropriate
therapy. It is significant that many of the GEMCaP loci contain
proteins with known or hypothesized roles in cancer metastasis.
The third goal of these projects it to test the hypothesis that
the GEMCaP loci can predict post surgical recurrence and metastasis
in other types of cancer, such as breast, brain, and ovarian cancer.
ZNF217 cancer
gene
The ZNF217 breast cancer gene, located on chromosome 20q13.2,
was discovered and characterized in Collins’ laboratory.
This discovery demonstrates his research process. Chromosome 20q13.2
is amplified in up to 40% of breast tumors and can occur in early-stage
disease. High-level amplification is associated with high-grade
breast cancer, and short disease-free survival.
Collins’ group hypothesized that amplification of a specific
gene or genes within the amplified area on Chromosome 20q13.2
would lead to abnormal gene expression as the tumor evolved to
a more malignant phenotype. In collaboration with the Department
of Energy’s Joint Genome Institute and Los Alamos National
Laboratory, Collins’ laboratory determined the sequence of the 1.2 Mb 20q13.2 locus for both human and mouse. Comparative sequence analysis and data visualization
was done using software developed in Collins’ laboratory;
this software is now available to the scientific community through
the UCSF Comprehensive Cancer Center. These studies led to the
identification of 10 potential genes and associated
regulatory elements.
ZNF217 was isolated from the potential genes and shown to be the
gene responsible for the amplification of chromosome 20q13.2.
This was demonstrated by showing that cancer tumors with extra
copies of the ZNF217 gene at the DNA level, also showed increased
expression of ZNF217 at the RNA level.
Studies in a number of laboratories have established that the
gain of an extra chromosome 20q in the cells of the tumor is associated
with epithelial cell immortalization, cells growing continually
unchecked. Chromosome 20q includes the ZNF217 locus. Collins’
laboratory hypothesis that ZNF217 is a pro-immortalization gene
and successfully demonstrated that it allows cultured mammary
epithelial cells to grow continually. His laboratory has demonstrated
that ZNF217 functions to weaken cell death and that it makes breast
and brain cancer cell lines resistant to doxorubicin, a cancer
treatment drug. Collins’
laboratory has shown that ZNF217 and Akt can exist in feedback
loop. This suggests that abnormal
levels of ZNF217 may predict the therapeutic usefulness of drugs
that inhibit the PI3K pathway. These findings are significant
because they unify genomic, cell biologic, and clinical data collected
over the last decade.
End Sequence Profiling
Collins invented, and was awarded a patent for, a new method
of analyzing the structure of tumor genomes called End Sequence
Profiling (ESP). He transformed ESP from a scientific idea into
a dynamic large-scale inter-institutional research project. ESP
is a direct result of the Collins laboratory’s growing expertise
in computational genomics. ESP is sequence-based analysis of tumor
genomes that can deliver approximately 90% of the information
of whole genome sequencing, the current standard, for about 1%
of the cost.
ESP was developed and refined theoretically using a computer
model. Collins’ laboratory then validated the ESP concept
with laboratory experiments; they cloned translocations and other
chromosomal rearrangements from the breast cancer cell line MCF7.
In these experiments, amplifications and
deletions were mapped at kilobase resolution. Direct evidence for
co-localization of amplified genomic DNA from multiple loci
has been obtained for the first time. This discovery may lead
to broad-spectrum diagnostics and therapeutics, if not to therapy
unique to each individual.
The Collins laboratory has performed ESP on primary breast, brain,
ovarian, and prostate tumors; on three additional cell breast
cancer lines; and has modified ESP for structural analysis of
RNA copied from DNA. Collins’ laboratory is validating
translocations and fusion transcriptions, combinations of genes
that would not normally be near each other on the chromosome,
while exploring the translational applications their discoveries.
ESP is one of the motivating factors for the proposed NIH/NCI
Human Tumor Genome Project.
return
to top
|
