NEW YORK (GenomeWeb) – Vanderbilt University researchers have developed a multi-modal imaging system for studying the molecular characteristics of host-pathogen interactions.
Described in a study published this week in Science Translational Medicine, the work combines MALDI imaging-based proteomics, mass spec-based metal imaging, bioluminescent imaging (BLI), and MRI to investigate multiple aspects of Staphylococcus aureus infection and host response.
The integration of these techniques provided a three-dimensional look at broad cellular and tissue changes due to infection and is potentially applicable to other disease systems, including studies of patient tumors and the tumor microenvironment, said Eric Skaar, professor of pathology, microbiology, and immunology at Vanderbilt and senior author of the paper.
Vanderbilt researcher and study co-author Richard Caprioli is a leader in MALDI mass spec imaging (MSI) and, Skaar said, he and Caprioli have been using MSI proteomic studies to explore models of infection.
Skaar added that his lab has been working independently on an inductively coupled plasma mass spec (ICP-MS) imaging method to look at cellular distribution of metals, which are key co-factors for many proteins and play a significant role in the outcome of infections. His lab also uses MRI to look at the anatomic disruptions caused by infections, as well as BLI, which uses luciferase linked to bacterial promoters to look at the activity of various genes of interest.
"We had the idea that maybe we could bring these [different modalities] together in a single experiment," Skaar said. "We could ask the question of 'how does the infection damage tissue?' using MRI, 'how does that lead to a redistribution of metalloproteins?' using MALDI imaging mass spec, 'how does that affect metal distribution in the tissue?' using ICP-MS imaging, and then, finally, 'how do the bacteria respond to these changes in metal distribution through gene expression changes? [as measured by BLI]."
S. aureus infection is a good model disease for this sort of imaging study because of the abscesses the infection forms that, Skaar said, " are very easily differentiable from otherwise healthy tissue."
"An imaging modality like this is really powerful when you apply it to something like a lesion you can see in an H&E stain and then compare the [imaging] profiles to the pathology," he said.
In the STM study, the researchers looked at mice infected with S. aureus, first imaging the animals with MRI while they were alive and then collecting a series of sections in groups of four throughout the entire animal — one for MALDI, one for ICP, one for H&E staining, and a fourth as a reserve.
"Then we do all those imaging modalities and computationally make a 3D rendering of all the different modalities and co-register them with each other," Skaar said.
The study led to a number of findings, "some of them very interesting but unexplainable," Skaar said, and several that provide clear points for follow-up experiments he and his colleagues are now planning.
Among the most interesting results was the observation of "a massive redistribution of metals throughout the tissue when the animals are infected," he said, adding that "there really aren't proteins known that can explain the degree of [metal] redistribution that we see."
"So the fundamental metal biology that is coming out of this is very interesting, and there is clearly a lot left to learn about how metals are moved around in response to infection and what the ramifications are to the disease process," Skaar said.
Another key finding was the heterogeneity of the infection-related abscesses. Different abscesses in the same tissue had different molecular compositions and the bacteria present in these abscesses showed different gene expression patterns, indicating a response to different environments.
Skaar said he and his colleagues believe these differences had to do with the different stages of abscess development. "We think that the abscess sort of has a life cycle, and that depending on how old the abscess is, the molecular environment in it will be different," he said, adding that they are now working to link specific molecular characteristics to different abscess timepoints.
"Is there a certain molecule or molecular profile that only exists in a three-day-old abscess or only in a five-day-old abscess," he said. "That is what we are trying to figure out."
Skaar added that this question of abscess heterogeneity is an example of the sort of finding the team's imaging approach yields that would likely be missed by other methods that don't include a spatial component.
"You can imagine how most people just grind up an organ and do RNA-seq or proteomics," he said. "We are arguing that you are missing a lot of information when you do that."
Perhaps the most curious result was the identification of an apparent gap within the abscess between the bacteria and the host innate immune proteins. This finding stemmed from improvements in MALDI imaging technology that allowed the researchers for the first time to image and identify bacterial proteins as distinct from host proteins, Skaar said.
"So what we could now do was image where the bacterial proteins were compared to where the host innate immune proteins were, and what we clearly saw was, there was a population of bacteria and then there was a set of innate immunity proteins surrounding the bacteria, but there was a gap between the two," he said. "There was no signal immediately around the bacteria. It was sort of like a demilitarized zone."
"The innate immune proteins, at least the ones we are imaging, are not getting all the way to the bacterial microcolony, and we think that might have a really profound impact on the host-pathogen interaction," he added. "Because if the innate immune proteins can't reach the bacteria, they can't do whatever they are supposed to do to inhibit them. So we are now interested in what is in that zone around the microcolony."
Skaar said moving forward, he hopes to further improve the platform to enable more highly multiplexed BLI measurements to get a broader sense of bacterial gene expression changes in response to the host environment.
He also hopes to expand the breadth of bacterial proteins detectable via MALDI imaging. "If we can find the bacterial proteins that are most abundant during infection in the host, I think that might provide some really valuable information regarding targets we could be going after for therapeutic development, vaccine design, et cetera," he said.
Skaar said that his research is focused mainly on biology, but he suggested the integrated imaging platform his team developed could be of use in other areas, cancer biology in particular.
"It's not a research direction I will move in, but I think this technology is perfectly suited for tumor biology," he said. Cancer is already a major focus for MALDI imaging-based proteomics, but Skaar suggested the ICP-MS-based metal distribution analysis could be useful in this area, too.
"Metals ae required cofactors for something like 30 percent of all proteins in nature, so they are hugely important to major biochemical processes, and redistribution of metal like this is going to have a big effect on any disease process," he said. "Presumably, similar redistribution of metal happens in diseases like cancer."