Protein patterns of developmentally totipotent mouse teratocarcinoma cells and normal early embryo cells

Protein patterns of mouse teratocarcinoma stem cells were compared, by two-dimensional gel electrophoresis, with those of early embryo cells. These malignant cells were known from previous experiments (B. Mintz and K. Illmensee, 1975, Proc. Nat. Acad. Sci. USA72, 3585–3589) to be capable of conversion to normalcy and of contributing to embryogenesis when introduced into a blastocyst. The protein comparisons were intended to reveal whether totipotent teratocarcinoma cells most nearly resemble normal totipotent cells of a specific stage, as a possible clue to their developmental origins. A simple method was devised for the purpose of generally facilitating comparisons of two-dimensional gels, among which technical variations commonly alter the absolute positions of individual proteins. This variation was normalized by the use of a reference constellation, or a network of lines connecting shared landmark proteins identified in all the gels. Whereas the network may undergo topological change from one gel to another, it continues to provide a readily recognized standard of reference. Protein patterns displayed many similarities and some differences, hence nonidentity, between teratocarcinoma cells and all normal preimplantation embryo stages tested, as well as between the various embryo stages themselves. The results also unexpectedly disclosed, however, that changed physiological states or posttranslational alterations may contribute significantly to some of the protein differences irrespective of the developmental status or potentialities of the cells. For example, in the OTT 6050 teratocarcinoma transplant line, pure teratocarcinoma cell groups (“cores”) found in the ascites fluid synthesized several proteins not expressed when the cores were enveloped (in embryoid bodies) by a yolk saclike epithelium; yet the core cells from both sources form comparable tumors if injected subcutaneously and are able to undergo differentiation if injected into blastocysts. In another comparison, some proteins that were present in inner cell masses isolated from blastocysts were absent in intact blastocysts, possibly because of their modification by the surrounding trophoblast in the latter case. These observations imply that protein differences between embryo regions or stages, however real, are not necessarily relevant for an evaluation of their developmental prospects.