Histological evidence for the existence of germ cell tumor cells showing embryonal carcinoma morphology but lacking OCT4 expression and cisplatin sensitivity.

The biological basis for manifestation of chemotherapy resistance in metastatic testicular germ cell tumors (GCT) remains obscure and is of particular clinical interest. In nonseminomatous GCT (NSGCT) the pluripotent embryonal carcinoma (EC) cells are the precursors of the manifold differentiated structures but also drive the malignant growth. They are known to be hypersensitive towards DNA-damaging agents and to express the embryonal transcription factor OCT4. We recently characterized EC cells that lack OCT4 expression and show cisplatin resistance. In the present, immunohistochemical study we analyzed the composition of NSGCT with the focus on such OCT4-negative EC cells using a NSGCT xenograft model as well as patient-derived NSGCT samples. In the xenograft model, the cisplatin-sensitive cell line H12.1 gives rise to xenografts where EC structures are mainly composed of OCT4-positive cells, whereas xenografts from the resistant cell line 1411HP exclusively comprise OCT4-negative EC areas. We found that post-chemotherapy residual metastatic tumors of patients can be comprised of exclusively OCT4-negative EC, whereas the matched testicular primary tumor harbors OCT4-positive EC. Thorough histological analyses revealed a few examples of such OCT4-negative EC cells also in the testicular primary tumor as well as in xenografts from the cisplatin-sensitive NSGCT-cell line. For these cells we propose an identity as early extraembryonal progenitor cells directly derived from OCT4-expressing EC cells. This challenges the use of the term EC cell. The data also support our hypothesis that malignant growth of resistant NSGCT may be driven by this cell type.

Cytogenetics detects infiltrations of a primary cutaneous acute myeloid leukemia to the kidney.

Extramedullary manifestations of acute myeloid leukemia (AML) are rare and commonly involve one tissue. We report of a cutaneous acute myelomonocytic leukemia infiltrating the kidney next to the skin. A 61-year-old female patient with complex karyotype cutaneous AML FAB M4 underwent abdominal computed tomography scans. A lesion in her left kidney appeared suspicious of renal carcinoma as confirmed by histology. However, fluorescence in situ hybridization cytogenetics revealed a chromosome 11q23 abnormality in the nephrectomy specimen, which also appeared in the leukemic blasts of skin and bone marrow. Closer histomorphologic workup revealed an infiltration of the kidney with leukemia. This case report illustrates how modern diagnostic procedures can help to reveal rare sites of disease.

Monitoring of xenograft tumor growth and response to chemotherapy by non-invasive in vivo multispectral fluorescence imaging.

A continuous monitoring of the whole tumor burden of individuals in orthotopic tumor models is a desirable aim and requires non-invasive imaging methods. Here we investigated whether quantification of a xenograft tumor intrinsic fluorescence signal can be used to evaluate tumor growth and response to chemotherapy. Stably fluorescence protein (FP) expressing cell clones of colorectal carcinoma and germ cell tumor lines were generated by lentiviral transduction using the FPs eGFP, dsRed2, TurboFP635, and mPlum. Applying subcutaneous tumor models in different experimental designs, specific correlations between measured total fluorescence intensity (FI) and the tumor volume (V) could be established. The accuracy of correlation of FI and V varied depending on the cell model used. The application of deep-red FP expressing xenografts (TurboFP635, mPlum) was observed to result in improved correlations. This was also reflected by the results of a performed error analysis. In a model of visceral growing mPlum tumors, measurements of FI could be used to follow growth and response to chemotherapy. However, in some cases final necropsy revealed the existence of additional, deeper located tumors that had not been detected in vivo by their mPlum signal. Consistently, only the weights of the tumors that were detected in vivo based on their mPlum signal correlated with FI. In conclusion, as long as tumors are visualized by their fluorescence signal the FI can be used to evaluate tumor burden. Deep-red FPs are more suitable for in vivo applications as compared to eGFP and dsRed2.

Loss of Oct-3/4 expression in embryonal carcinoma cells is associated with induction of cisplatin resistance.

Although the majority of testicular germ cell tumors (TGCTs) are curable by cisplatin-based chemotherapy, in a few cases, the occurrence of cisplatin resistance results in a poor outcome. The biological basis of this differential cisplatin sensitivity in TGCTs remains largely unexplained. Embryonal carcinoma (EC) cells represent the presumptive tumor stem cells in nonseminomatous TGCTs and are known to express the embryonal transcription factor Oct-3/4 and to be hypersensitive to cisplatin. In the present study, we analyzed TGCT cell lines and nude mouse xenografts showing differential cisplatin sensitivity. Here we demonstrate that a lack of expression of Oct-3/4 in TGCT cells is associated with a higher apoptotic threshold and cisplatin resistance which is accompanied by an impaired caspase-9 activation, reduced caspase-3 activity and altered p53 accumulation. We were able to induce loss of Oct-3/4 in a cisplatin-sensitive EC cell line resulting in a secondary cisplatin-resistant cell type with retained EC cell characteristics and changes in apoptotic signaling identical to those in primary resistant cells. Furthermore, we show that EC cells are retained in their undifferentiated state by Oct-3/4 and that a complete and ultimate loss of Oct-3/4 followed by an early differentiation step is necessary to establish the cisplatin-resistant state. Our data suggest that loss of Oct-3/4 expression leads to induction of a higher apoptotic threshold and to cisplatin resistance in EC cells of nonseminomatous TGCTs. We hypothesize that in refractory TGCTs the original tumor stem cell population of Oct-3/4-positive, cisplatin-sensitive EC cells could be replaced by an Oct-3/4-negative, resistant population in a selection process. In contrast, the presence of the Oct-3/4-positive, highly sensitive EC cells as the tumor stem cell component in most TGCTs could explain the general high chemosensitivity and curability of these tumors.