Head, HLA typing section/senior researcher, Laboratory of Experimental Immunology, National Cancer Institute
• Medical research associate, immunology, Duke University — 1985-1989
• Postdoc, microbiology, University of North Carolina — 1984-1985
• PhD, immunobiology, Iowa State University — 1982
Researchers from the National Cancer Institute, in collaboration with the Massachusetts Institute of Technology, Harvard University, and Massachusetts General Hospital, this month reported the discovery of a microRNA that plays a role in HIV control.
According to a paper published in Nature, miR-148a was found to regulate the expression of HLA-C, which plays a key role in immunity, with a variant having been associated with HIV load and progression to AIDS.
This week, Gene Silencing News spoke with Mary Carrington, the paper's senior author, about the findings.
Let's start with some background on your lab.
We're a genetics lab and we concentrate on a group of genes on chromosome 6 within a locus called the major histocompatibility complex. Some of those genes include the [human leukocyte antigen] class I and class II genes, which encode histocompatibility antigens. These antigens are important to match if you want to do a kidney transplant or a bone marrow transplant from one person to an unrelated person, so that the transplant is not rejected. These genes are hugely polymorphic genes and present antigenic peptides to T cells to initiate an immune response. A result of their polymorphism is that it gives us, as a species, the ability to resist a great variety of infectious pathogens.
And that has translated into an interest in HIV/AIDS?
Right. These HLA class I loci are the most important genome-wide in terms of determining how rapidly an individual will progress to AIDS, how well they control HIV viral load, and so on. The forms of these genes you have — HLA-A, -B, and -C — will determine, in part, how well an individual will control HIV after becoming infected, more so than any other locus throughout the human genome.
The research published marks your lab's foray into the microRNA field. How did that come about?
In two of the genome-wide association studies, it was found that there was a variant upstream of the HLA-C gene that associated very significantly with control of HIV viral load. This single-nucleotide polymorphism … is probably marking something rather than having a direct effect, and since it's close to HLA-C, we thought it was probably marking the different forms of HLA-C that would confer protection versus susceptibility.
The people who published the original GWAS paper, [which came out of David Goldstein's lab at Duke University], noted that that SNP, which we call -35, also associated with levels of HLA-C messenger RNA from previous work. So individuals who have one form of this SNP have high levels of HLA-C mRNA, and the other form of the SNP associates with low levels of HLA-C messenger RNA.
We didn't think that SNP is causing the high or low expression of HLA-C directly, and a post-doc in my lab, [Smita Kulkarni], started looking in other regions around HLA-C and honed in on the 3' UTR of HLA-C and identified a microRNA-binding site, which is also polymorphic — an insertion-deletion polymorphism right in the region that binds the microRNA, miR-148a. That's how we ended up studying microRNAs; we had never done any work in [the field] before. But it is really interesting because that microRNA-binding site is affecting which HLA-C alleles are going to be expressed at a high level, escaping miR-148a down-regulation, and which are expressed at low levels due to miR-148a regulation.
Can you give an overview of the work published in Nature?
There are many forms, or alleles, of [HLA-C] … so Smita went through the different forms of [the gene] and sequenced the 3' UTRs. The first goal was to define the level of variability in that region. Then she used in silico methods to identify any potential microRNA-binding sites. We actually were concentrating on [a different miRNA] but could not find it expressed in any of the cells we were studying. Then she concentrated on the miR-148-binding site … and that microRNA was expressed pretty much ubiquitously across cell types tested.
She and Ram Savan then did classical miRNA experiments, putting the 3' UTR in a reporter vector and determining the level of luciferase expression, depending on differences in the 3' UTR inserted into that vector. If she took a 3' UTR from a high-expression HLA-C allele, the luciferase was expressed at a high level in cells that were transfected. If she inserted the 3' UTR into the luciferase vector from a low-expression HLA-C allele, then the luciferase was expressed at a significantly lower level.
Smita and Savan did mutation analysis to show that the differential regulation of the reporter gene is due specifically to the variation in the region that binds miR-148a. They then did experiments to show that endogenous HLA-C could be down-regulated on the cell surface with the addition of a mimic of miR-148, and that the HLA-C allele was expressed at a higher level with the addition of an inhibitor, but only in cells that had endogenous HLA-C alleles that were low-expression and had the binding site for miR-148a. When she added the mimics or inhibitors to cells where the endogenous HLA-C was a high-expression allele and did not have the binding site, there was no effect.
Those were the basic methods used to pin down which microRNA was important, where the binding site is, [and] which variants in the binding site are important in miR-148. They were able to show, for the first time, that a classical HLA class I gene is being regulated by a microRNA. That adds to the diversity of the HLA locus; it's very polymorphic, and now on top of that, [we see] that it's expressed at different levels across different allelic types.
What are the plans to follow up on these findings?
We're looking at the microRNA gene, miR-148a … [which] itself is polymorphic, and we're trying to determine whether that adds another level of complexity to the regulation of HLA-C. In terms of [how this work could translate to the clinic], I think we have a fair bit more work to do before we could think of a therapy. The variant in the 3' UTR associates very strongly with HIV control, but we don't know whether it's having a true, direct effect on HIV control, and I think that needs to be answered before we could even begin to think about this in terms of a therapy.