More... Share to Twitter Share to Facebook

A Sharper Tool for Genome Engineering

A Sharper Tool for Genome Engineering

New method for editing genomes may be faster, safer and more efficient than recent breakthroughs

Fri, 2013-01-11

Green fluorescent cells show a genetic sequence that was repaired using a new method for genome engineering. Image courtesy of the Church Lab.

A new method for more easily and precisely engineering the human genome may help scientists pin subtle changes in DNA to disease, moving the needle from correlation to causality and potentially improving gene therapy techniques.

Two groups recently reported in Science their success using a tool borrowed from a bacterial immune system called Cas, short for CRISPR-associated systems, which in turn stands for Clustered Regularly Interspaced Short Palindromic Repeats. In bacteria the Cas9 enzyme system uses short stretches of RNA to target and then cut invading viral DNA. Scientists have customized this system to work in human cells, creating an RNA-guided editing tool that allows them to integrate DNA changes into the genomes of living cells, a process called genome engineering.

One of the teams, led by George Church, HMS Robert Winthrop Professor of Genetics, also tested the method in induced pluripotent stem cells (iPSCs), an important milestone on the path to human genome engineering. These cells, taken from a child or adult, have been modified to mimic embryonic cells, which means they can develop into any adult cell type. In experiments, iPSCs offer greater clarity than traditional cell lines as researchers explore gene function in different cell types.

The other team, led by Feng Zhang of the Broad Institute of Harvard and MIT, independently showed the effectiveness of Cas9.

While the Cas9 method is still a long way from the clinic, Church is hopeful.

“We need a lot more experience and optimization, but it looks very promising,” he said. “This is much easier than any previous human genome engineering method, and it is relevant to testing the flow of ideas from GWAS [genome-wide association studies] and the Personal Genome Project.”

The Cas9 approach, first developed by Harvard University graduate Jennifer Doudna of the University of California, Berkeley, could end up supplanting a technique that Science just last month named one of the top 10 scientific breakthroughs of 2012. For this technique, a class of proteins called TALENs (Transcription Activator-Like Effector Nucleases) zero in on a particular region of the genome, where they can precisely cut DNA and insert or delete a gene. TALENs followed a method from the 1990s that used enzymes called zinc finger nucleases to target specific parts of genomes.

TALENs were easier to design, but, like zinc fingers, they require about 2,000 bases of messenger RNA to encode an enzyme that would cleave DNA at a specific site in a genome.  Cas9 needs 1/100 as much RNA; as few as 20 variable bases, embedded in short constant guide RNAs, are sufficient for precise targeting, the scientists showed in human cell lines.

According to Church, also testing Cas9 in iPSCs is particularly noteworthy.

“We need to have human stem cells in culture so we can manipulate the genome and see how the cells differentiate, or fit into tissues,” he said. “To change from correlation and speculation to causality, you introduce single changes, one by one, into a test genome and ask which of those or how many of those do you need in order to see the trait. That turns correlation into causation and moves it closer to the gold standard.”

The more compact Cas9 system opens the door to engineering multiple changes in different genes and then testing them simultaneously to see what role they play in complex diseases. Large genome-sequencing studies can find a variety of genes active in people with a particular disease, but it takes the kind of multiplex testing Cas9 allows to establish which ones actually matter, Church said.

A decade ago early attempts at gene therapy failed because delivery of a new, corrected gene could inadvertently promote cancer or provoke a harmful immune system response. Newer methods, in addition to having more precise delivery, are also designed to have lower toxicity and generate only a tolerable immune response in the case of a rare off-target event.  These newer approaches also make it possible to target an old gene with new information, either knocking it out or changing it precisely.

Cas9 has the potential to accelerate these improvements in gene therapy, but much more testing is necessary to see if it will continue to display the greater efficiency and very low toxicity shown in these early experiments. “But it looks like a very promising starting point,” Church said.

This work was supported by NIH grant P50 HG005550. Church and the Science paper’s first author, research fellow Prashant Mali, have applied for a patent based on the findings of this study. 


More... Share to Twitter Share to Facebook

Cancer Spread's Family Tree

Cancer Spread’s ‘Family Tree’

Rapid test reveals relationships between primary tumors and metastatic cells

Wed, 2014-04-23

Image: iStock

The process of metastasis—a tumor’s ability to spread to other parts of the body—is still poorly understood. It has not been easy to determine whether metastasis began early or late in the development of the primary tumor, or whether individual metastatic sites were seeded directly from the original tumor or from an intermediate site. Now a research team has developed a simple test that can reveal the evolutionary relationships among various tumor sites within a patient, information that may someday help with treatment planning.

“If we could build a ‘family tree’ of all cancer nodules in a patient, we could determine how different tumors are related to each other and reconstruct how the cancer evolved,” said Kamila Naxerova, Harvard Medical School research fellow in radiation oncology in the Steele Laboratory for Tumor Biology at Massachusetts General Hospital. She is corresponding author of the report, published in PNAS Early Edition. “Usually that would require extensive genetic analysis with complex sequencing methods, but our methodology achieves that goal quickly and with minimal experimental effort.”

Cancer researchers are just beginning to investigate the extent and significance of genetic differences among tumor cells, either among cells within a discrete tumor or between a primary tumor and metastases in other parts of the body. The authors note that there are two different models of metastasis. In one model, an advanced primary tumor disseminates metastatic cells late in its development, which would predict little genetic difference between primary and metastatic cells. In the other model, metastasis occurs early in tumor development, which would predict significant genetic differences in metastatic cells that have evolved separately from those in the primary tumor. Some studies have suggested that the two models apply to different types of cancer, but patient data so far has been limited.

Answering important clinical questions, such as whether genetic diversity is a risk factor for aggressive tumor development and how it relates to treatment resistance, requires analyzing samples from many patients with different types of cancer. Sequencing the whole genome or just the protein-coding portion of the genome (the exome) requires specialized equipment and advanced data analysis, making the process relatively expensive.

The approach developed by the Mass General team focuses on small areas of the human genome called polyguanine (poly-G) repeats that are particularly susceptible to mutation, with genetic “mistakes” occurring frequently during cell division. While these mutations do not directly relate to the development or progression of a tumor, they can reveal its lineage—how individual tumor cells are related to each other.

Poly-G repeat analysis was initially developed to study lineage relationships between single cells in mice, but in the current paper, the authors adapted it to study human cancer for the first time. Analyzing the poly-G profiles of primary and metastatic colon cancer samples from 22 patients revealed that the way the primary and metastatic tumors related to each other was different for each patient.

In some individuals there were significant genetic differences between tumor sites, suggesting early metastatic spread; in others, there was little difference between a primary tumor and its metastases. The investigators also identified instances in which the genetic profiles of metastases were similar to those of only some cells in the primary tumor, suggesting that those cells were the source of the metastases, and other cases in which the genetic profiles of metastases from the same primary tumor differed depending on their location.

“We found that there are several paths that can lead to metastatic disease,” said Naxerova. “We are now applying this methodology to address specific clinically relevant questions about the biology of metastasis in larger numbers of patients. The method is fast and inexpensive and should be applicable to other types of tumors than colon cancer.”

Co-author Elena Brachtel, HMS assistant professor of pathology at Mass General, noted that archival tissues from the files of the department were used for this study. “After diagnostic studies on tissue removed during a patient’s operation are completed, the formalin-fixed paraffin tissue blocks are stored for several years. Increasingly, new molecular tests can be performed on tissue that was removed from a patient several years earlier, at a time when these tests were not yet available.”

Rakesh K. Jain, the A. Werk Cook Professor of Radiation Oncology (Tumor Biology) at HMS and Mass General, director of the Steele Lab and senior author of the paper, added, “The assay has many potential clinical applications. For example, it could be used to reliably and quickly distinguish a metastasis from a second, independent tumor. Or it could identify the primary tumor in situations where multiple lesions are present and it is ambiguous which one is responsible for seeding metastases.”

The work was supported by Department of Defense grants W81XWH-10-1-0016 and W81XWH-11-1-0146.

Adapted from a Mass General news release.


More... Share to Twitter Share to Facebook

Connecting Crohn's and Gut Microbes

Connecting Crohn’s and Gut Microbes

Microbiome analysis reveals increase in pathogenic bacteria, drop in beneficial microbes in newly diagnosed patients

Wed, 2014-03-12

Image: iStock

A multi-institutional study led by Harvard Medical School investigators at Massachusetts General Hospital and scientists from the Broad Institute has identified how the intestinal microbial population of patients newly diagnosed with Crohn’s disease differs from that of individuals free of inflammatory bowel disease. In their paper in the March 12 issue of Cell Host and Microbe, the researchers report that Crohn’s patients showed increased levels of harmful bacteria and reduced levels of the beneficial bacteria usually found in a healthy gastrointestinal tract. 

Several studies have suggested that the excessive immune response that characterizes Crohn’s may be associated with an imbalance in the normal microbial population, but the exact relationship has not been clear. The current study analyzed data from the RISK Stratification Study, which was designed to investigate microbial, genetic and other factors in a group of children newly diagnosed with Crohn’s disease or other inflammatory bowel diseases. At 28 participating centers in the U.S. and Canada, samples of intestinal tissues were taken from 447 participants with a clear diagnosis of Crohn’s and from 221 control participants with noninflammatory gastrointestinal conditions. The researchers also analyzed samples from an additional group of about 800 participants in previous studies, for a total of more than 1,500 individuals.

Advanced sequencing of the microbiome—the genome of the entire microbial population—in tissue samples taken from sites at the beginning and the end of the large intestine revealed a significant decrease in diversity in the microbial population of the Crohn’s patients, who had yet to receive any treatment for their disease. The samples revealed an abnormal increase in the proportion of inflammatory organisms in Crohn’s patients and a drop in noninflammatory and beneficial species, compared with the control participants. The imbalance was even greater in patients whose symptoms were more severe and in those who had markers of inflammatory activity in tissue samples.

“These results identifying the association of specific bacterial groups with Crohn’s disease provide opportunities to mine the Crohn’s disease-associated microbiome to develop diagnostics and therapeutic leads,” said senior author Ramnik Xavier, the HMS Kurt J. Isselbacher Professor of Medicine in the Field of Gastroenterology at Mass General.

Other key findings of the study include:

  • identifying the niches of specific microbial strains in Crohn’s disease and their effects on other microbial community members
  • finding that rectal biopsies can indicate the presence of disease early in the course of Crohn’s, regardless of which intestinal segments are affected
  • showing that fecal samples collected at the onset of disease do not reflect changes in the bacterial communities of the intestinal lining

Antibiotics are often prescribed for symptoms suggestive of Crohn’s before a diagnosis is made. In participants who happened to be taking antibiotics at the time samples were taken, the microbial imbalance was even more pronounced, suggesting that the antibiotic use could exacerbate symptoms rather than relieve them, the authors note. Next steps will be to uncover the function of these microbes and their products and to learn how the microbiome and microbial products interact with the patient’s immune system, with the possibility that these interactions could represent the molecular basis for the disease.

“Identifying which microbial products are key to disease onset and to inflammation resolution in inflammatory bowel disease and establishing which can be effectively targeted are our best hope to uncover the first microbiome-based therapies in inflammatory bowel disease,” said Xavier, who is chief of the Mass General Gastrointestinal Unit and director of the Mass General Center for the Study of Inflammatory Bowel Disease.

Dirk Gevers of the Broad Institute is the lead author of the report and Curtis Huttenhower of the Broad and the Harvard School of Public Health is also a co-author.

The study was supported by the Crohn’s and Colitis Foundation of America, the Helmsley Charitable Trust, by Army Research Organization grant W911NF-11-1-0473, the MGH Center for the Study of Inflammatory Bowel Disease and by National Institutes of Health grants U54DE023798, R01HG005969 and R01DK092405.

Adapted from a Mass General news release.


More... Share to Twitter Share to Facebook

Testing the Whole Haystack

Testing the Whole Haystack

Genetic analysis reveals architecture of obsessive-compulsive disorder and Tourette syndrome


Thu, 2013-10-24

Image: iStock

An international research consortium led by investigators at Massachusetts General Hospital and the University of Chicago has provided the first direct confirmation that both obsessive-compulsive disorder (OCD) and Tourette syndrome are highly heritable. Their work, which appears in the October issue of PLOS Genetics, also reveals major differences in the genetic makeup of the two conditions that sometimes occur together.  

“Both Tourette syndrome and OCD appear to have a genetic architecture of many different genes—perhaps hundreds in each person—acting in concert to cause disease,” said Jeremiah Scharf, HMS assistant professor of neurology at Mass General and senior corresponding author of the report. “By directly comparing and contrasting both disorders, we found that OCD heritability appears to be concentrated in particular chromosomes, particularly chromosome 15, while Tourette syndrome heritability is spread across many different chromosomes.”

An anxiety disorder characterized by obsessions and compulsions that disrupt the lives of patients, OCD is the fourth most common psychiatric illness in the United States. Tourette syndrome is a chronic condition known for causing motor and vocal tics. Usually beginning in childhood, Tourette syndrome is often accompanied by OCD or attention-deficit hyperactivity disorder. Because OCD and Tourette syndrome often develop in close relatives of affected individuals, they have been considered heritable, but identifying specific genes that confer risk has been challenging. 

Scharf and several co-authors of the current study published two papers last year in the journal Molecular Psychiatry describing genome-wide association studies (GWAS) of thousands of people affected by OCD or Tourette syndrome. While those studies identified several gene variants that appeared to increase the risk of each individual condition, none of the associations was strong enough to meet the strict standards of genome-wide significance. 

The GWAS approach is designed to identify relatively common gene variants. Because OCD and Tourette syndrome might be influenced by a number of rare variants, the research team adopted a different method. Called genome-wide complex trait analysis (GCTA), the approach allows simultaneous comparison of genetic variation across the entire genome, rather than the GWAS method of testing sites on the genome one at a time. GCTA also estimates the proportion of disease heritability caused by rare and common variants.

“Trying to find a single causative gene for diseases with a complex genetic background is like looking for the proverbial needle in a haystack,” said Lea Davis, a research assistant professor of genetic medicine at the University of Chicago and co-corresponding author of the PLOS Genetics paper. “With this approach, we aren’t looking for individual genes. By examining the properties of all genes that could contribute to Tourette syndrome or OCD at once, we’re actually testing the whole haystack and asking where we’re more likely to find the needles.”

Using GCTA, the researchers analyzed the same genetic datasets screened in the Molecular Psychiatry studies: almost 1,500 people with OCD compared with more than 5,500 controls, and nearly 1,500 people with Tourette syndrome compared with more than 5,200 controls. To minimize variations that might result from slight differences in experimental techniques at more than one location, all genotyping was done by collaborators at the Broad Institute of Harvard and MIT, who generated the data at the same time using the same equipment. Davis analyzed the resulting data on a chromosome-by-chromosome basis, charting the frequency of the identified variants and the function of variants associated with each condition.

The researchers concluded that the degree of heritability for both disorders captured by GWAS variants is actually quite close to what previously was predicted based on studies of families affected by the disorders. 

“This is a crucial point for genetic researchers, as there has been a lot of controversy in human genetics about what is called ‘missing heritability,’” Scharf explained. “For many diseases, definitive genome-wide significant variants account for only a minute fraction of overall heritability, raising questions about the validity of the approach. Our findings demonstrate that the vast majority of genetic susceptibility to Tourette syndrome and OCD can be discovered using GWAS methods. In fact, the degree of heritability captured by GWAS variants is higher for Tourette syndrome and OCD than for any other complex trait studied to date.”

Nancy Cox, a professor of medicine and human genetics at the University of Chicago and co-senior author of the PLOS Genetics report, added, “Despite confirming there is shared genetic liability between these two disorders, we also show there are notable differences in the types of genetic variants that contribute to risk. Tourette syndrome appears to derive about 20 percent of genetic susceptibility from rare variants, while OCD appears to derive all of its susceptibility from variants that are quite common, which is something that has not been seen before.”

About half the risk for both disorders appears to come from variants already known to influence the expression of genes in the brain. Further investigation of those findings could identify the affected genes and show how changes in their expression contribute to the development of Tourette syndrome and OCD. Additional studies in even larger patient populations, some of which are in the planning stages, could reveal the biologic pathways disrupted in the disorders, potentially leading to new therapeutic approaches. 

The study reflects a collaboration between two consortia representing 43 institutions across 12 countries: the Tourette Syndrome Association International Consortium for Genomics and the International OCD Foundation Genetics Collaborative. Scharf is co-chair of the Tourette Syndrome Association International Consortium for Genomics steering committee and a member of the International OCD Foundation Genetics Collaborative steering committee. 

Support for the study includes National Institutes of Health grants U01 NS40024, R01 MH101820 and K23 MH085057; and grants from the Tourette Syndrome Association and the David Judah Fund.

Adapted from a Mass General news release.


More... Share to Twitter Share to Facebook

Tracking Brain Tumors

Tracking Brain Tumors

Noninvasive molecular biology tool detected a common tumor-associated genetic mutation in cerebrospinal fluid

Wed, 2013-07-24

Researchers have tracked brain tumors by searching for evidence of mutations in cerebrospinal fluid. Image: iStock

Borrowing a tool from molecular biology, HMS researchers at Massachusetts General Hospital have detected a tumor-associated genetic mutation in the cerebrospinal fluid (CSF) of a small number of patients with brain tumors.  In a paper published in the open-access journal Molecular Therapy – Nucleic Acids, the investigators described using digital versions of the gene-amplification technology polymerase chain reaction (PCR) to analyze bits of RNA carried in membrane-covered sacs. They found a common tumor-associated mutation in a gene called IDH1, a biomarker whose presence could potentially influence patient care.

“Reliable detection of tumor-associated mutations in cerebrospinal fluid with digital PCR would provide a biomarker for monitoring and tracking tumors without invasive neurosurgery,” said Xandra Breakefield, HMS professor of neurology at Mass General and corresponding author of the paper. “Knowing the IDH1 mutation status of these tumors could help guide treatment decisions, since a number of companies are developing drugs that specifically target that mutant enzyme.”

Both normal and tumor cells regularly release membrane-covered sacs called extracellular vesicles. Found in blood, CSF and other body fluids, they contain segments of RNA, DNA or proteins. A 2008 study from the Mass General team identified a relatively large tumor-associated mutation in extracellular vesicles from the blood of brain tumor patients, but most current diagnostic technologies that analyze CSF do not capture molecular or genetic information from central nervous system tumors.  

In addition, “Tumor-specific extracellular vesicles make up only a small percentage of the total number of extracellular vesicles found in either blood or cerebrospinal fluid, so finding rare, single-nucleotide mutations in a sample of blood or CSF is very challenging,” explained Leonora Balaj, an HMS research fellow in neurology and co-lead author of the paper. “These digital PCR techniques allow the amplification of such hard-to-find molecules, dramatically improving the ability to identify tumor-specific changes without the need for biopsy.”

The current study used two forms of digital PCR—BEAMing and Droplet Digital PCR—to analyze extracellular vesicles in the blood and in the CSF of brain tumor patients and healthy controls. The scientists were searching for the presence of a single-nucleotide IDH1 mutation known to be associated with several types of cancer. Both forms of PCR detected the presence and abundance of mutant IDH1 in the CSF of 5 of the 8 patients known to have IDH1-mutant tumors. 

Two of the three mutation-positive tumors that had false negative results were low grade and the third tumor was quite small, suggesting a need for future studies of more samples to determine how the grade and size of the tumors affect the ability to detect mutations. The failure to detect tumor-associated mutations in blood samples with this technology may indicate that CSF is a better source for extracellular vesicles from brain tumors. 

The ability to noninvasively determine the genetic makeup of brain tumors could have a significant impact on patient care. “The current approach for patients who may have a brain tumor is first to have a brain scan and then a biopsy to determine whether a growth is malignant,” said Fred Hochberg, HMS associate professor of neurology and a study co-author. “Patients may have a second operation to remove the tumor prior to beginning radiation therapy and chemotherapy, but none of these treatments are targeted to the specific molecular nature of the tumor.”

Having this kind of molecular diagnostic assay—whether in spinal fluid or blood—would allow clinicians to immediately initiate treatment that is personalized for a particular patient without the need for surgical biopsy, Hochberg said. 

“For some patients, the treatment could shrink a tumor before surgical removal. For others, it may control tumor growth to the point that surgery is not necessary, which in addition to keeping patients from undergoing an unnecessary procedure, could save costs,” he said. “We still have a long way to go to improve survival of these malignancies, so every improvement we can make is valuable.”

Mass General has applied for a patent on the use of BEAMing PCR to analyze RNA from extracellular vesicles. Support for the study includes National Institutes of Health grants CA069246, CA141226, CA156009 and CA141150 and grants from the Brain Tumor Funders’ Collaborative  and the American Brain Tumor Association.

Adapted from a Mass General news release.


More... Share to Twitter Share to Facebook

Gene variant linked to weight-loss surgery success

Gene variant linked to weight-loss surgery success

Genetic factors appear to determine a patient’s response to gastric bypass

Thu, 2013-05-02

Image: CDC/Debora Cartagena

Massachusetts General Hospital researchers have identified a gene variant that helps predict how much weight an individual will lose after gastric bypass surgery, a finding with the potential both to guide treatment planning and to facilitate the development of new therapeutic approaches to treating obesity and related conditions like diabetes. The report, published online in The American Journal of Human Genetics, is the first to identify genetic predictors of weight loss after bariatric surgery.

“We know now that bypass surgery works not by physically restricting food intake but primarily through physiological effects—altering the regulation of appetite to decrease hunger and enhance satiety and increasing daily energy expenditure,” said Lee Kaplan, HMS associate professor of medicine at Mass General and director of the hospital’s Obesity, Metabolism and Nutrition Institute. He is a senior author of the report. “Genetic factors appear to determine a patient’s response to gastric bypass, and the identification of markers that predict postoperative weight loss could provide important insight into those physiological mechanisms.”

The research team conducted genome-wide association studies of more than 1,000 patients who had bypass surgery at Mass General from 2000 to 2011, analyzing almost 2 million gene sites for associations between specific variants and the percentage of weight lost after surgery. One specific variant at a site on chromosome 15 was most closely associated with weight loss. Individuals with two copies of the beneficial version of the gene lost an average of almost 40 percent of their presurgical weight, while those with only one copy lost around 33 percent. The single individual in the study group who had no copies of the beneficial variant lost less than 30 percent of presurgical weight.

Expression of one of the genes closest to the site of this variant was also able to predict the percentage of weight lost. In addition, experiments in a mouse model of gastric bypass indicated that expression of the corresponding version of that human gene, as well as another gene adjacent to the variant site, was altered by bypass surgery. Additional gene variants not as strongly associated with the response to bypass surgery are candidates for further study in larger groups of patients.

Two predictive models developed by Kaplan and his team have had promising initial results. One of these combines the chromosome 15 genetic variant with clinical factors such as age, gender, the presence of diabetes and exercise behaviors to predict surgical outcomes; the other includes 12 additional gene variants the investigators are studying to determine their usefulness in treatment planning. 

Notably, none of the predictive gene sites identified in this study is involved in pathways previously known to influence the development of obesity, suggesting that different genes contribute to the benefits of bypass. Development of drugs that target the activity of those genes might produce some of the same benefits without the need for surgery, Kaplan said.

“The fact that genetics appears to play such an important role in how well bypass surgery works in an individual patient gives us even more evidence that obesity results from dysfunction of the biological mechanisms that regulate fat mass and body weight and not solely from aberrant behavior or limited willpower,” he adds. “Identifying the involved genes opens up the potential for new classes of antiobesity therapies that mimic or exploit the molecular mechanisms so effectively used by gastric bypass.”  

The study was supported by National Institutes of Health grants DK093257, DK088661 and DK090956, along with grants from Merck Research Laboratories and Ethicon Endo-Surgery.

Adapted from a Mass General news release.


More... Share to Twitter Share to Facebook

Clearing the Way for Alzheimer's Prevention

Clearing the Way for Alzheimer’s Prevention

Scientists find protective gene variant that removes toxic amyloid beta protein from the brain

Thu, 2013-04-25

Characteristics of Alzheimer’s disease. Image: National Institute on Aging

Harvard Medical School investigators at Massachusetts General Hospital have determined that one of the recently identified genes contributing to the risk of late-onset Alzheimer’s disease regulates the clearance of the toxic amyloid beta (A-beta) protein that accumulates in the brains of patients with the disease. 

In their report published in Neuron, the researchers describe a protective variant of the CD33 gene, which promotes clearance of A-beta from the brain. They also show that reducing expression of CD33 in immune cells called microglia enhances their ability to clear away A-beta protein, raising the possibility that blocking CD33 activity could help the brain’s immune system remove A-beta.

“Our findings show, for the first time, a ‘switch’ that controls how fast microglial cells can clear A-beta protein from the brain as we age. CD33 is the key,” said Rudolph Tanzi, Joseph P. and Rose F. Kennedy Professor of Child Neurology and Mental Retardation at Harvard Medical School and senior author of the Neuron paper. “If we can find a way of safely inactivating CD33 on microglia, we should be able to slow the accumulation of A-beta in aging brains and hopefully reduce risk for Alzheimer’s disease.”

In 2008, as part of the Alzheimer’s Genome Project, Tanzi, who is also director of the Genetics and Aging Unit in the Mass General Department of Neurology, and his team identified four novel genes containing variants that increased the risk of late-onset Alzheimer’s, the most common form of the devastating neurological disorder. One of these was CD33. The CD33 protein produced by the gene was known to play a role in regulation of the innate immune system—the body’s first line of defense against infection—but how it might function in the brain and possibly contribute to Alzheimer’s risk was not known. 

In the current study, the researchers first found that CD33 activity was significantly higher in microglia cells in brain samples from Alzheimer’s patients than in cells from non-demented controls. Moreover, they showed that the presence of a version of the gene that protected against Alzheimer’s disease reduced CD33 protein levels in the brain. Importantly, the same protective version of CD33 was found to reduce levels of A-beta 42—the primary constituent of the amyloid plaques that characterize the disease. Greater numbers of CD33-containing microglia also were associated with higher levels of A-beta 42 and more plaques overall.

In an Alzheimer’s mouse model, knocking out the CD33 gene improved the ability of microglia in the brain to clear away A-beta 42 and reduced the presence of amyloid plaques. Experiments with cultured microglia showed that increasing CD33 expression on the cells’ surface inhibited their ability to take up A-beta 42, while reducing CD33 activity led to greater clearance of A-beta 42.

“Collectively these experiments indicate that CD33 directly modulates the ability of microglial cells to clear A-beta 42 from the brain,” said Tanzi. “Our findings raise the possibility that inhibiting CD33 activity in the brain could represent a potentially powerful new approach to treating and possibly preventing Alzheimer’s disease.”

Primary support for the study includes grants from the Cure Alzheimer’s Fund and National Institutes of Health grants R37MH060009, P01AG15379, R01AG08487 and P50AG05134. 

Adapted from a Mass General news release.