|
FOUR COLOR CLARITY:Data summary of a four-color flow cytometric analysis of 40 healthy blood donors (arrayed from left to right). The intensity of the bars represents the frequency of cells (around 0.1% to 10.0%) that responded to classes of stimuli (arrayed from top down). The responses include the production of IFN-γ, IL2, or TNFα among CD4+ or CD8+ T cells. White bars represent missing data and yellow represents <0.1% responders. Responses to CMV peptides (top) cluster among CMV-seropositive donors (left).
|
Faith in the promise of pharmacogenomics has changed the plan for primary healthcare. We expect, someday, that doctors will discern what medicines to prescribe based in part on the mixture of genes sitting on the paper-covered exam tables in their offices.
But while I'm delighted about personalized medicine and its attendant technical and business innovations, I doubt that pharmacogenomics will drive healthcare much. I liken it to trying to understand Los Angeles by reading its phonebook. You will quickly recognize some important names, and with sufficient cross-referencing to other databases (school, hospital, credit, and criminal records, genealogy charts, grocery shipments, etc.), you might develop a sense of what's going on and why some neighborhoods are safer than others, but it will be difficult because the information is too granular and too static.
I'm more encouraged about gene-expression and proteomic profiling. Here, at least, are dynamic assessments of an organism's current state, where reactions to trauma or degeneration are likely visible, though perhaps very subtle. The biggest challenge here will be sorting wheat from chaff. Most mRNA and protein species will not be changing coordinate with disease or therapy. Changes that can be informative are likely to be cell-type specific, and so lost in a sea of complex tissue or body fluid. Nonetheless, big changes will be available for analysis, and biomarkers that highlight specific cellular changes could vastly alter the course of both drug development and patient management.
IMMUNOMODULATORY THERAPIES
Immunomodulatory therapies serve as a good example. Clinically managing immunosuppressives has long been a somewhat crude art. These drugs are dosed empirically: If the patient shows opportunistic infection or liver or kidney toxicity, back off; if the patient shows rejection of the transplant, dose up. Both situations are expensive and dangerous, and a biomarker that could enable critical and dynamic dosing would save lives.
On a smaller but still informative scale, the recent withdrawal of natalizumab, an anti-VLA4 therapeutic antibody, is illuminating. Three patients among thousands undergoing treatment with this novel class were diagnosed with progressive multifocal leukoencephalopathy, a rare and often fatal disease associated with failure to control a common virus. Biomarkers that would allow clinicians to recognize such dangers will be critical.
Or, consider the greatest healthcare challenge of our generation, the development of vaccines to HIV. This vicious infection is destroying lives by the millions, yet some of us have immune systems that let us live disease free for many years. As an immunologist I am ashamed for not knowing why. A biomarker for protective immunity is required.
A STORY FROM CELLS
My favorite analytes for personal health are the cells of our immune system. They do an amazing job integrating information from the body and responding to changes, especially changes that represent a threat. Far from static, immune cells not only reflect immediate status, but also contain elements of memory that can report crucial aspects of personal history.
John Dunne
|
Our research group and a growing network of collaborators and colleagues focus on a family of flow cytometry assays that probe this immunological memory in beautiful detail. [ 1-3]
In general, the idea is to expose a blood sample to a collection of stimuli: some of them pharmacological, such as phorbol esters and ionophores; others polyclonal but cell-type specific, such as bacterial lipopolysaccharides or T-cell receptor antibodies; and most interestingly, peptide cocktails which together recall immunogenic proteins from viral or bacterial pathogens, such as cytomegalovirus or HIV. With this last class of stimuli, we can recognize those rare T cells circulating in blood that are sentinels of recurring or chronic infection.
We recognize them because within a few hours of in vitro stimulation, they respond to the peptide stimuli by making characteristic cytokines. We use a protein secretion inhibitor to trap the cytokines inside the cells. Then when we fix and permeabilize the cells, we can identify them from nonresponding cells using fluorescently labeled antibodies in a flow cytometer.
In hundreds of publications to date, our community is confidently measuring these often-rare cells (frequencies commonly <0.2% of T cells) and beginning to link their function to personal history and immunomodulatory therapy.
The figure on the preceding page depicts the strong bias for detection of CMV-responsive T cells from blood of CMV-seropositive donors. Responses to flu peptides show a much more heterogeneous distribution among healthy donors, reasonably interpreted as dynamic responses to acute asynchronous infection. Responses to disease-related antigens such as HIV-peptide cocktails and tumor-associated antigens are sometimes seen with this assay format, but of much lower magnitude and more dispersed distribution among healthy donors. More prospectively, we and others have shown that these response measurements can track therapeutic intervention in vaccines for HIV, [ 4-6]
other infectious diseases, [ 7]
and cancer. [ 8, 9]
THE NEXT DIMENSIONALITY
The data represent a state-of-the-art implementation of flow technology. Each blood sample was distributed into a 96-well plate containing two replicas of 48 different experimental stimuli, and allowed to respond for six hours. The samples were then stained with two four-color cocktails of monoclonal antibodies, allowing the detailed characterization of the magnitude and various aspects of the normal donors' responses. For each stimulus we show the number of CD4+ and CD8+ cells that make various combinations of interferon γ, interleukin 2, and/or tumor necrosis factor.
UPPING THE DIMENSIONALITY:Gating strategy for assessing T-cell responses in eight-color flow cytometry. First, particles that scatter light not typical of lymphocytes are removed from subsequent analysis. Then, lymphocytes are clustered into two groups of interest, T cells with bright fluorescence characteristic of CD4+ (red) or CD8+ (blue) cells. For each of these, we define cells that make either IFNγ alone or IFNγ and IL2 (pale blue and green, respectively, for the CD8+). Then we differentiate these four kinds of responsive cells among those cells positive or negative for T-cell differentiation markers, CD45RA, CD27, and CD28. This analysis generates, for every blood sample, the frequency of 32 different biologically meaningful responsive phenotypes, a complex network of "who makes what."
|
These cytokine combinations represent the nature of the decision that each of these cells made in response to the various stimuli, and in aggregate tell a rich story about how these healthy people perceive clinically relevant threats. These 4-color measurements are now typical of flow cytometry. The turn of the century is seeing yet another generation of capabilities developing with the installation of routine clinical systems of five and six colors, and research groups publishing eight, 12, and even 17-color analyses.
To get a sense of what that does to the information content of flow-cytometric mapping studies, consider the figure on this page from a recent study by our group. Here, in an eight-color analysis, we measure two cytokines and six differentiation antigens on all the lymphocytes in just one run of a blood sample. The cytometer is simultaneously measuring eight different fluorescence intensities (quantitatively related to the binding of eight different monoclonal antibodies), on several thousand cells per second, with precision over about four decades of useful dynamic range. The result is a very detailed description of who is making what, an informatics challenge of very unusual dimensionality.
We can now automatically generate the simple answers from the raw data with standard flow cytometric analysis strategies, but how these and a thousand other patterns in eight or more dimensions will map into a valuable set of biomarkers will require the development of novel data-management and visualization tools. Our biggest gap now, is informatics.
To truly realize the promise of personalized medicine, biomarkers of immunomodulatory pharmacodynamics, safety, and efficacy are required. I believe that blood cell-based assays deserve an industrial mapping exercise. The critical blood sample is already well-handled in routine clinical practice. The analytical platform (flow cytometry) is sophisticated and broadly engaged. The assays are widely published and the clinical implementation is beginning. Let the mapping of blood-cell responses to immunomodulatory therapies begin in earnest.
John Dunne is associate scientific director at BD Biosciences, leading a group working to apply flow cytometry technologies to the mapping of immunological responses. The author gratefully acknowledges his colleagues at BD Biosciences, particularly Skip Maino, Holden Maecker, Margaret Inokuma, and Laurel Nomura for sharing dedication and data. He can be contacted at john_dunne@bd.com.
References
| 1. | | Maecker HT:
"Cytokine flow cytometry,".
In Flow Cytometry Protocols. Edited by: Hawley TS, Hawley RG. Totowa, NJ: Humana Press; 2004.
|
| | | Return to citation in text:
[1]
|
| |
| 2. | | Maino VC, Maecker HT:
"Cytokine flow cytometry: A multiparametric approach for assessing cellular immune responses to viral antigens,".
Clin Immunol 2004, 110:222-31. [Publisher Full Text]
|
| | | Return to citation in text:
[1]
|
| |
| 3. | | Ghanekar SA, et al.:
"Immune monitoring using cytokine flow cytometry,".
In Manual of Clinical Laboratory Immunology. Edited by: Rose NR. Washington, DC: ASM Press; 2005.
|
| | | Return to citation in text:
[1]
|
| |
| 4. | | Maino VC, et al.:
"Enhancement of HIV type 1 antigen-specific CD4+ T cell memory in subjects with chronic HIV type 1 infection receiving an HIV type 1 immunogen,".
AIDS Res Hum Retroviruses 2000, 16:539-47. [Publisher Full Text]
|
| | | Return to citation in text:
[1]
|
| |
| 5. | | Jacobson MA, et al.:
"Results of a cytomegalovirus (CMV)-specific CD8+/interferon-γ+ cytokine flow cytometry assay correlate with clinical evidence of protective immunity in patients with AIDS with CMV retinitis,".
J Infect Dis 2004, 189:1362-73. [Publisher Full Text]
|
| | | Return to citation in text:
[1]
|
| |
| 6. | | Komanduri KV, et al.:
"Restoration of cytomegalovirus-specific CD4+ T-lymphocyte responses after ganciclovir and highly active antiretroviral therapy in individuals infected with HIV-1,".
Nat Med 1998, 4:953-6.
[Publisher Full Text]
|
| | | Return to citation in text:
[1]
|
| |
| 7. | | Dunn HS, et al.:
"Dynamics of CD4 and CD8 T-cell responses to cytomegalovirus in healthy human donors,".
J Infect Dis 2002, 186:15-22. [Publisher Full Text]
|
| | | Return to citation in text:
[1]
|
| |
| 8. | | Hobeika AC, et al.:
"Enumerating antigen-specific T-cell responses in peripheral blood: A comparison of peptide MHC tetramer, ELISpot, and intracellular cytokine analysis,".
J Immunother 2005, 28:63-72.
[Publisher Full Text]
|
| | | Return to citation in text:
[1]
|
| |
| 9. | | Smith JW, et al.:
"Adjuvant immunization of HLA-A2-positive melanoma patients with a modified gp100 peptide induces peptide-specific CD8+ T-cell responses,".
J Clin Oncol 2003, 21:1562-73. [Publisher Full Text]
|
| | | Return to citation in text:
[1]
|
| |
|
Register for FREE Online Access
Subscribe to the Magazine