0 (S)-(-)-Blebbistatin levels are representative of mice leukemic burden The percentage of ALL cells in peripheral blood may not necessarily reflect bone marrow and secondary organ infiltration by leukemia. Usually, the beginning of exponential increase of ALL cells in peripheral blood corresponds to the end of log phase or even plateau phase in bone marrow. To address whether Hsp90 would be a useful marker PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19704093 for earlier detection of engraftment and to monitor the dynamic growth of leukemia in mice, groups of non-irradiated NOD/SCID mice injected with ALL were weekly monitored for plasma Hsp90 levels and sequentially sacrificed once the threshold of 0.1 ng/mL was reached. These animals were used to investigate leukemia cell proportion in bone marrow, liver, spleen and blood, in comparison to plasma Hsp90 levels. A scheme of this experiment is shown in Fig 2A. As expected, Hsp90 levels in ALL-transplanted mice increased in a time dependent manner. All three B-cell precursor ALL reached Hsp90 levels above 0.1 ng/mL in the first week after transplantation. At first sight, Tcell ALL had slower engraftment than BCP-ALL. However, a closer examination showed that the percentage of T-ALL cells in bone marrow, liver, spleen and blood, at time point 1, tended to be equal or even higher in T-ALL than BCP-ALL. Even though previously reported western blot analysis showed no difference in Hsp90 expression by BCP-ALL versus 4 / 13 Plasma Hsp90 for In Vivo Monitoring of ALL Fig 1. Analysis of three potential leukemia plasma biomarkers. Peripheral blood B2M, IGFBP-2 and Hsp90 levels in NOD/SCID mice transplanted with a primary human T-ALL or healthy controls. Animals were divided into groups according to the percentage of ALL cells in peripheral blood by FACS. Data points correspond to individual samples, and horizontal bars correspond to median. P<0.05; Mann-Whitney U test. Correlation between Hsp90 levels and % ALL cells in peripheral blood. Data points correspond to individual samples. Data were transformed to log10 and analyzed by Pearson's correlation. Cut-off value of Hsp90 as determined by adding 2 times SD to mean. Animals transplanted with different BCP-ALL and T-ALL, sacrificed at the earliest time point were included in the analysis. Leukemia engraftment was confirmed by FACS analysis. Data points correspond to individual samples. PB, peripheral blood; SD, standard deviation. doi:10.1371/journal.pone.0129298.g001 5 / 13 Plasma Hsp90 for In Vivo Monitoring of ALL Fig 2. ELISA of plasma Hsp90 levels for earlier engraftment confirmation. Experimental PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19705070 workflow. Cryopreserved primary ALL cells were thawed and injected into 3 mice for expansion. Fresh xenograph ALL cells were then injected into 6 to 12 secondary animals for experiments. For RS4;11 and TALL1, cultured cells were directly injected in animals for the experiments. Animals were monitored weekly and sacrificed as indicated. Kinetic of Hsp90 plasma levels over time, following ALL injection. Data points represent mean of 3 animals. The cut-off Hsp90 value is indicated. Asterisks represent starting week of sequential sacrifice of animals. Some groups did not have enough animals for the experiment to be carried out until the third week of sacrifice. Percentage of ALL cells in bone marrow and peripheral blood at the different time points. Data points represent mean of 3 animals. Dotted line represents the usual cut-off for ALL detection by flow cytometry in blood samples. Different ALL cells are represented by colo