Myeloma RPMI-8226 cells. Cell uptake of 64Cu-CB-TE1A1P-LLP2A (0.1 nM), in human RPMI-8226 cells at 37uC in the absence (red bar) and presence (blue bar) of excess LLP2A (P,0.0001). doi:10.1371/journal.pone.0055841.g008 Figure 7. Tumor histology and SPEP (Serum Protein Electrophoresis) analysis on the serum of KaLwRij mice. A. Hematoxylin and eosin (H E) stained slide of a representative 5TGM1 s.c. tumor tissue. The tumor cells show irregularly shaped nuclei and increased mitosis consistent with myeloma pathogenic features. B. SPEP gel showing qualitatively the c-globulin (M protein) in tumor bearing (lanes 5, 6 7) and non-tumor bearing (lanes 1, 2, 3 4) mice. The 5TGM1 tumor bearing mice (lanes 5, 6 7) 
were analyzed two weeks post tumor cell inoculation. The top arrow represents the M Protein band and the lanes represent serum SPEP for each mouse. C) Quantitative representation of the total c-globulin g/dl in the mice (1?). doi:10.1371/journal.pone.0055841.gSPEP analysis, and ex-vivo tissue biodistribution and tumor histology. The 5TGM1 cell binding of 64Cu-CB-TE1A1P-LLP2A was significantly reduced in the presence of excess LLP2A at 37uC and 4uC respectively, signifying Epigenetic Reader Domain receptor mediated endocytosis. The Epigenetics saturation binding assays were performed in the presence of Mn2+ ions and produced high Bmax values (136 pmol/mg (619)), which is key to a desirable imaging outcome resulting from the high binding potential (Bmax/Kd) ratio [41]. Cell uptake studies performed in the absence of receptor activating cations, in this case, Mn2+, resulted in much lower binding (data not shown). Lam and co-workers have also shown independently that LLP2A binds to the activated form of a4b1 integrin [38,42]. Additional studies will be needed to describe the changes in VLA-4 activation status in response to different stimuli [43]. Binding of 64Cu-CB-TE1A1PLLP2A was evaluated in the human MM cell line, RPMI-8226. 64 Cu-CB-TE1A1P-LLP2A demonstrated high binding to RPMI8226 cells, which was significantly blocked in the presence of excess LLP2A (Figure 8). The ex vivo biodistribution and 
1407003 in vivo imaging studies in the 5TGM1 mouse models of MM support the fact that 64Cu-CBTE1A1P-LLP2A has desirable in vivo pharmacokinetics forachieving excellent tumor to background ratios. Three 5TGM1 tumor models were evaluated in this study to demonstrate the versatility of the PET imaging probe. In the mouse bearing s.c. tumor cells without matrigel, PET imagining readily detected the presence of VLA-4 expressing cells localized to the femur, tibia, and spine (Figure 4B, arrows). While this signal is presumed to indicate dissemination of myeloma cells to the bone, further ex vivo analyses will be required to characterize the extent that inflammatory or stromal cells may also be imaged as components of the pre-metastatic niche. Regardless, this data demonstrates that VLA-4-targeted PET imaging has the ability to image cells within mineralized bone. In vivo 1317923 tumor blocking of 64Cu-CB-TE1A1PLLP2A was observed in both the biodistribution and small animal PET imaging studies. Small animal PET images collected at 2 h post-injection had high tumor to muscle ratios (2866) (Figure 5 and supplemental image 1), with improved ratios observed at 24 h time point due to the clearance from non-targetorgans (161641). In summary, here we have described for the first time use of a VLA-4 targeted PET imaging probe, 64Cu-CB-TE1A1P-LLP2A, to image extra-osseous MM tumors in animal models using sm.Myeloma RPMI-8226 cells. Cell uptake of 64Cu-CB-TE1A1P-LLP2A (0.1 nM), in human RPMI-8226 cells at 37uC in the absence (red bar) and presence (blue bar) of excess LLP2A (P,0.0001). doi:10.1371/journal.pone.0055841.g008 Figure 7. Tumor histology and SPEP (Serum Protein Electrophoresis) analysis on the serum of KaLwRij mice. A. Hematoxylin and eosin (H E) stained slide of a representative 5TGM1 s.c. tumor tissue. The tumor cells show irregularly shaped nuclei and increased mitosis consistent with myeloma pathogenic features. B. SPEP gel showing qualitatively the c-globulin (M protein) in tumor bearing (lanes 5, 6 7) and non-tumor bearing (lanes 1, 2, 3 4) mice. The 5TGM1 tumor bearing mice (lanes 5, 6 7) were analyzed two weeks post tumor cell inoculation. The top arrow represents the M Protein band and the lanes represent serum SPEP for each mouse. C) Quantitative representation of the total c-globulin g/dl in the mice (1?). doi:10.1371/journal.pone.0055841.gSPEP analysis, and ex-vivo tissue biodistribution and tumor histology. The 5TGM1 cell binding of 64Cu-CB-TE1A1P-LLP2A was significantly reduced in the presence of excess LLP2A at 37uC and 4uC respectively, signifying receptor mediated endocytosis. The saturation binding assays were performed in the presence of Mn2+ ions and produced high Bmax values (136 pmol/mg (619)), which is key to a desirable imaging outcome resulting from the high binding potential (Bmax/Kd) ratio [41]. Cell uptake studies performed in the absence of receptor activating cations, in this case, Mn2+, resulted in much lower binding (data not shown). Lam and co-workers have also shown independently that LLP2A binds to the activated form of a4b1 integrin [38,42]. Additional studies will be needed to describe the changes in VLA-4 activation status in response to different stimuli [43]. Binding of 64Cu-CB-TE1A1PLLP2A was evaluated in the human MM cell line, RPMI-8226. 64 Cu-CB-TE1A1P-LLP2A demonstrated high binding to RPMI8226 cells, which was significantly blocked in the presence of excess LLP2A (Figure 8). The ex vivo biodistribution and 1407003 in vivo imaging studies in the 5TGM1 mouse models of MM support the fact that 64Cu-CBTE1A1P-LLP2A has desirable in vivo pharmacokinetics forachieving excellent tumor to background ratios. Three 5TGM1 tumor models were evaluated in this study to demonstrate the versatility of the PET imaging probe. In the mouse bearing s.c. tumor cells without matrigel, PET imagining readily detected the presence of VLA-4 expressing cells localized to the femur, tibia, and spine (Figure 4B, arrows). While this signal is presumed to indicate dissemination of myeloma cells to the bone, further ex vivo analyses will be required to characterize the extent that inflammatory or stromal cells may also be imaged as components of the pre-metastatic niche. Regardless, this data demonstrates that VLA-4-targeted PET imaging has the ability to image cells within mineralized bone. In vivo 1317923 tumor blocking of 64Cu-CB-TE1A1PLLP2A was observed in both the biodistribution and small animal PET imaging studies. Small animal PET images collected at 2 h post-injection had high tumor to muscle ratios (2866) (Figure 5 and supplemental image 1), with improved ratios observed at 24 h time point due to the clearance from non-targetorgans (161641). In summary, here we have described for the first time use of a VLA-4 targeted PET imaging probe, 64Cu-CB-TE1A1P-LLP2A, to image extra-osseous MM tumors in animal models using sm.