Reduction in Antibody Repertoire Complexity with Lymphoma slide

2015-01-13

slide
"C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\DR\2014\03-24-2014d0913\Reduction in Antibody Repertoire Complexity with Lymphoma.pptx"

some powerpoints this slide is found in
-"C:\Users\kurtw_000\Documents\kurt\storage\CIM Research Folder\kwhittem\Presentations\2014\PhD Oral Defense\Oral Defense Presentation Kurt Whittemore 4-11-14.pptx"

text on slide
{
Summary Points
-experiment performed by Bart Legutki
-about 25 normal dogs and 25 healthy dogs
-24 male and 15 female healthy dog samples from The Flint Animal Cancer Center at Colorado State University (median age of 6)
-22 male and 16 female B cell lymphosarcoma dogs (median age of 7.9)
-expected entropy to decrease when LSA sera applied to array
-CIM10Kv2 array
-results same as observed in spiking antibody experiment

Some text from dissertation
{
Are there changes in the AbStat measurement during the course of a lymphoma which reduces the complexity of the antibody repertoire?  During the course of a B cell lymphoma one particular antibody in the repertoire becomes dominant as one B cell proliferates uncontrollably and produces large amounts of the antibody.  Can this change be detected in the fluorescence intensity distribution produced from applying the sera to an array of peptides?  The anticipated result is that the entropy of the fluorescence intensity distribution will decrease when sera associated with a B cell lymphosarcoma (LSA) is applied to the array relative to normal sera.  The antibody repertoire for LSA sera will have a higher prevalence of an antibody against one specific target.  Therefore, this sera will resemble a monoclonal antibody more than normal sera, and this sera will bind to a few features with high intensity which contain peptides with a mimotope similar to the cognate epitope of the antibody.  Many of the other features on the array will have very low intensity ultimately resulting in a less complex fluorescence intensity distribution and a lower entropy value. 
The experiment was performed by Bart Legutki using sera from dogs with and without a B cell lymphosarcoma (LSA).  These sera samples were obtained from The Flint Animal Cancer Center at Colorado State University.  There were 24 different male healthy dogs and 15 female healthy dogs with an age range from 2 to 15 (median of 6) of the following breeds: Mix Breed (19), Golden Retriever (6), Labrador (3), Staffordshire Terrier (2), Australian Cattle Dog (2), Australian Shepherd, Dalmatian, Doberman, German Wire Haired Pointer, Std. Poodle, St. Bernard, and Rottweiler.  There were 22 male and 16 female B cell lyphosarcoma dogs with an age range from 2 to 13 (median of 7.9) of the following breeds: Mix Breed (10), Golden Retriever (5), Border Collie (4), German Shepard (2), Rottweiler (2), Scottish Terrier (2), Vizsla (2), Bassett Hound, Belgian Malinois, Boxer, Chesapeake Bay Retriever, Collie, Doberman, Labrador, Miniature Schnauzer, Sheltie, Staffordshire Terrier, and one non-classified breed.  The sera from these dogs was applied to the CIM10Kv2 array.  A dotplot of the entropy of the two classes is presented in Figure 29, and the p-value from a t-test of each measure for the two groups is presented in Figure 30.
Figure 29 Box and dotplot of entropy for normal (N) and lymphosarcoma (LSA) dogs
Figure 30 Significance of each measure from a t-test with normal and LSA dogs
The data for all of the measures for the two groups was input into a SVM algorithm, and the algorithm was able to correctly classify 78.8% of the instances with a Kappa statistic of 0.566 and ROC area of 0.779.  When the normal or LSA class was randomly assigned to each sample, the SVM could only correctly classify 52.3% of the instances with a Kappa statistic of 0.0348 and a ROC area of 0.517.
These results demonstrated that the entropy of the LSA group of dogs was indeed lower than the entropy of the normal dogs.  The reason for this is that in a B cell lymphoma one antibody comes to predominate the repertoire.  The solution comes to resemble more of a monoclonal antibody rather than a complex antibody mixture, and the entropy value is closer to that of a monoclonal antibody.  This result was also observed in the section in which a monoclonal antibody was artificially spiked into monoclonal sera (3.1.3 Spiking antibody into sera).  A similar phenomena is also observed with vaccines as was demonstrated with mouse vaccines and human vaccines (3.2 Mouse vaccines and infections and 3.3 Human vaccines).  The reason that this is the case is due to the fact that vaccines are also characterized by an increase of the number of antibodies against one specific epitope or a few epitopes.  In conclusion, these results show that the AbStat measures could be used to monitor dogs for the occurrence of B cell lymphomas.  Presumably, the same techniques would work for human sera as well.

}

azim58wiki: