Assistant Professor, Molecular Biology
Name: Ileana Cristea
Position: Assistant professor in molecular biology, Princeton University, 2008 to present
Background: PhD, University of Manchester, 2002; post-doc fellowship in the Laboratory of Mass Spectrometry and Gaseous Ion Chemistry at Rockefeller University, 2003 to 2008
Ileana Cristea, an assistant professor in molecular biology at Princeton University, has been using mass spectrometry to study chromatin and its modulation by viruses.
Earlier this month, she was named one of three recipients of the National Institute on Drug Abuse’s new Avant-Garde awards, given to researchers whose work may lead to groundbreaking insight into the prevention and treatment of HIV/AIDS in drug abusers.
As part of that award, Cristea will receive $500,000 each year for five years to support her research, titled “Proteomic tools to uncover the role of chromatin remodeling in HIV-1 infection.”
Cristea also recently began research on virus-host protein interaction using Thermo Fisher Scientific’s MALDI LTQ Orbitrap XL instrument.
ProteoMonitor recently spoke with Cristea about her research. Below is an edited version of the conversation.
Is the MALDI platform new technology for you?
It’s recent in my lab, especially as a new configuration of the [Thermo Fisher Scientic] Orbitrap, but we have been routinely using it in the past in conjunction with time-of-flight or ion trap mass analyzers.
Have you been using ion spray mass spectrometers as well?
Yes, sure, both while I was working on my PhD in the laboratory of Simon Gaskell and during my postdoctoral research in the laboratory of Brian Chait [at Rockefeller University]. We also have it available here, as the LTQ Orbitrap XL instrument in my lab has an interchangeable MALDI/ESI source.
Is there a clear delineation in the way you’re using the two technologies, the ion spray and the MALDI?
Yes, there is. Each technology has its merits and shortcomings, and one can choose the one most appropriate for particular studies. However, many times incorporating various techniques within the same project is an advantage as it can provide access to different types of questions. … I’m particularly interested in understanding cellular processes that are involved in regulating the outcome of viral infections.
We developed methodologies for rapidly isolating protein complexes and we aim to uncover virus-host protein interactions during the course of infection. One focus of these studies is to determine the role of chromatin remodeling complexes during viral infection and we study these complexes in both uninfected and infected cells.
In several studies carried out in my lab, we work with very low levels of material and it can take us quite a long time to get the amount required for a detailed study. One example is the early stage of an infection, which can sometimes be problematic to study due to the low expression levels of viral proteins. As another example, a PhD student in my lab works on a protein present only at approximately 50 copies per cell, so it takes us a while to culture the necessary cells and get the amount for isolating and characterizing the protein complex. The stoichiometry within the complex can also introduce difficulties, even more so when starting with a low level of your target protein.
For us to use electrospray, it would mean that we would have a limited time of analyzing that sample. When we spot our sample on a MALDI target, we have this capability of storing it for sequential analyses. So, we can quickly inquire and analyze the sample to gain an overall view of it. But after that we can formulate some better questions, some more targeted questions, [then] go back to the sample and … get a more in-depth analysis. So, we can avoid the redundant identification of co-precipitated abundant proteins and reveal those present at lower levels, which may be closer to the noise signal and may escape the threshold of a first analysis. That’s why MALDI for us is quite valuable.
When I was in Brian Chait’s laboratory, we would commonly make use of complementary mass spectrometric configurations. We would use a MALDI ion trap in conjunction with a MALDI time-of-flight instrument. We could use the same target in these different instruments and make use of their individual merits — let’s say, obtain the mass measurement accuracy from the time-of-flight device, and then the sensitivity of the ion trap.
By combining the MALDI linear ion trap with the Orbitrap, one benefits from the sensitivity of the linear ion trap and the dynamic range and accuracy of the Orbitrap.
Is this storage capacity the main advantage for your work?
It’s one of the advantages, I would say, but it’s not the only one. Another advantage is the robustness and ease of use of MALDI, which can come in handy for a starting PI who is mentoring students new to the field of mass spectrometry.
One additional advantage is its suitability to automation and high throughput. This is something very useful to us, especially since the number of samples that we deal with has increased systematically. And, an advantage that I have already briefly mentioned is the rapidity of this methodology. Another MALDI characteristic that people have made use of in the past is the reduced complexity of the resulting spectra. Some years ago, it used to be a frequent approach to split a sample and perform the MS analysis using a MALDI time-of-flight instrument, because of the reduced complexity of a spectrum with mainly singly-charged ions, and then perform the tandem mass spectrometric analyses using an electrospray configuration with, let’s say a Q-TOF or a triple quadrupole.
Until fairly recently, there has been limited MS/MS done on singly-charged ions just [because] the appropriate configurations were not … commercially available. Also, singly-charged ions don’t fragment as well as the multiply-charged ions and were not and are still not favored. When I joined the laboratory of Brian Chait … I had more experience and preferred electrospray configurations for peptide fragmentation.
But we had in the Chait lab this MALDI ion trap configuration … that worked very well and it was valuable to learn more about the fragmentation of singly-charged ions and to observe that it actually happens quite well and reliably. It occurs very specifically at particular amino acids, so you can almost use it as a fingerprint of fragmentation. It’s quite easy to confirm [whether] a protein is there or not because the relative intensity of a particular fragment can tell you if it’s at the site where you would expect to see a major cleavage. I benefited a lot from these two experiences in the Gaskell and Chait labs as I learned about a variety of complementary technologies in mass spectrometry.
How would you do that with an electrospray mass spec?
With an electrospray source, you obtain multiply-charged ions and have a mobile proton that allows for a better fragmentation. So, although the peptide sequence plays important roles in the fragmentation process, the fragmentation is not signature driven [to] the extent of the MALDI.
Peptide fragmentation is usually more complete for multiply-charged ions. At the same time, the analyte signals are split in electrospray, as the same peptide may be present as singly-, doubly-, triply-charged, which may be a disadvantage when working with very low levels. However, this disadvantage is compensated by the better detection of multiply-charged ions.
I think that the MALDI and the electrospray process both have advantages and disadvantages and are both needed. Sometimes peptides do not fragment well as singly-charged ions, and sometimes too highly-charged peptides can also be problematic for fragmentation. One obtains a much larger number of spectra from HPLC, but this also creates a stronger dependence on software and automation. Users don’t usually look manually through thousands of spectra from an HPLC run, but you don’t have to miss anything from a MALDI-generated spectrum, as you can see all the data without using too much time.
With the sample storage, how would you do that with electrospray?
It depends on how one performs the experiment. If an HPLC is directly coupled to an instrument, which can provide a very useful separation before the analysis, the samples can’t really be preserved. It’s similar for nanospray — I used to be a big fan of nanospray and I still like it very much — but of course, every methodology has some steps where a percentage of the sample may be lost.
For nanospray, some peptides are lost on the needle because the glass absorbs a little. For storage, samples can be of course preserved prior to the electrospray process or, if using an offline HPLC system, the different fractions can be also stored prior to the analysis.
Overall, my point of view is that these technologies, MALDI and the electrospray process, are complementary and that one needs both. I don’t agree with the “fashion-driven” selection of instrumentation that seems to happen sometimes lately. It would be a pity if such a valuable and powerful tool as MALDI would become less utilized because of this phenomenon. I think that the methodology should be selected according to a scientific rational and the actual need for that technology, not because of the latest “fashion.”
Describe for me what you’re working on now, what specific disease you’re looking at, and what stage your work is at.
I’m interested in learning from protein interactions that occur within the context of critical cellular processes during viral infection. The complexity of interactions at the protein level is quite overwhelming and we still know very little about their dynamics and functions. The interactions are much more [complex] at this level than at the genome level or at the transcriptome level and that’s also true for virus-host interactions. Viruses use such interactions to manipulate very complex organisms. If we can understand the interactions of a virus during the course of a viral infection, we can learn about cellular processes and diseases and hopefully identify new targets for therapeutic interventions. That’s why these studies of virus-host interactions have become such a key driving force in research of infectious diseases during this post-genomic era.
We developed approaches for isolating protein complexes and studying virus-host protein interactions. One of the important findings from our studies on cells infected with HIV or human cytomegalovirus was that viral proteins associate with chromatin remodeling complexes … that make use of quite a complex system of site-specific enzymes to coordinate post-translational modifications on histone proteins and some non-histone proteins. These modifications can include acetylation, methylation, phosphorylation, ubiquitination.
One particular group of enzymes that we are interested in is the family of histone deacetylases that remove the acetyl group from histones and a set of non-histone proteins. They have risen to prominence lately because they are associated with a variety of human diseases and studies have shown that HDACs also play key roles in regulating the latent stages of infections with HIV or herpes simplex virus 1.
We’re trying to characterize these histone deacetylases, find out exactly what protein complexes they are part of and determine their roles during viral infections. We’re studying several viruses and one big focus for us is HIV for which we collaborate with Mark Muesing [at Aaron Diamond AIDS Research Center].
Have you been able to find out how the HIV is able to manipulate proteins?
We found several interacting partners that seem to be important during the HIV infection. The time-consuming part is not finding the interactions because our methodology is robust and works very well for accurately identifying these protein-protein interactions. The bottleneck is finding the biological significance of a particular interaction, which cellular process or mechanism is actually affected or involved. This is what we are after and what we are currently working on.
Are these new interactions or interactions that have previously been observed?
Both. Our study so far confirms some known interacting partners, but also identified novel ones.
The things that you’re seeing so far, do they have broad implications to all virus-protein interactions, or are they very specific to the disease area that you’re looking at?
Our findings can have broad implications and may aid our knowledge regarding other related pathogens. Viruses may manipulate these chromatin-remodeling complexes to control gene expression and the outcome of an infection. It would not be surprising to find that these are common processes that occur in infections with several different viruses. We have previously observed common features between various viral infections.
Some of our earlier studies on Sindbis, in collaboration with Charlie Rice and Margaret MacDonald at Rockefeller University, identified G3BP as an interacting partner of the nsP3 Sindbis virus protein. We later observed that G3BP may be implicated in infections with other alphaviruses, such as the Ross River virus.
Some of our more recent studies on human cytomegalovirus in collaboration with Tom Shenk [at Princeton] identified the HCMV protein responsible for regulating the mammalian target of rapamycin pathway. This finding is not only related to observations reported from studies on other viruses, but has also clinical relevance, as it has been recently reported that sirolimus may inhibit the HCMV replication in patients.
Are you doing other proteomics-directed work aside from virus-protein interactions?
Yes, we also study several aspects of chromatin remodeling without the context of viral infection and other families of proteins with antiviral and immunomodulatory activities. We also focus on the development of new proteomic-based methodologies. I’m very interested in pushing the boundaries of these current technologies for studying protein complexes.
I think that we are quite far yet from identifying these complexes as they are present in vivo in the cell. We look under the microscope to observe a protein of interest. I think that many unwanted processes occur between that moment [when we are observing the molecule] and the moment we isolate a protein complex. I usually say that these complexes are not really designed for us to hold on a bead or column … so, I think that there are further developments required to gain a spatial-temporal resolution of protein interactions and I hope that I can contribute to that.