Antiprotozoal activity of tricyclic compounds
per 2 million cells transfected with 10 ,ug of plasmid DNA, a value similar to that obtained with pNE03, a plasmid in which neo transcription is driven by the HSV-tk promoter. Twelve clones were randomly selected and cultured for the analysis of neo mRNA in exponentially growing cells (asynchronous culture at 35°C) and in cells arrested in GI (cultures shifted to 40.5°C) (Fig. 2 ). When the neo mRNA level of exponentially growing cells was compared with that of cells blocked in GI, increased
... in GI, increased amounts of neo transcripts were seen in asynchronously cycling cells. This observation for all the transformant clones tested indicates that position of the hybrid gene in the host genome was not a major factor in determining the differential neo expression between growing and GI-arrested cells. Nonetheless, the sites of integration and copy number may have some effect on the transcription of the neo RNA, since some transformants had more neo RNA than others. Therefore, the sequence elements present on a 1.1-kb Pvu II fragment of the hamster H3 gene may be important for its cell -cycle regulation. To further investigate whether the neo mRNA expression was regulated during the cell cycle as it is in the endogenous histone H3 genes, we studied three stable transformants that produced different amounts of neo mRNA for further analysis. The transformants were synchronized by serum deprivation, and upon addition offresh medium, were stimulated to progress through the cell cycle (3, 10) . The rate of DNA synthesis in these synchronized cells was monitored by [3H]thymidine incorporation. Neo transcripts were detectable at low levels during the G1 period (Fig. 3) . As cells entered the S phase, the neo RNA began to increase in parallel to the increase in histone H3 mRNA accumulation. In all three cases, at the peak, neo mRNA increased 6to 17-fold. In contrast, actin mRNA remained relatively constant throughout the cell cycle. We conclude that the 1.1-kb Pvu II fragment derived from this particular hamster H3 gene contains some elements necessary for regulated temporal expression of the gene during the cell cycle. It is not yet known whether the sequences encoding for the first 20 amino acid residues of the H3 gene play a role in this regulation. In studying regulated gene expression during the cell cycle, it is important to show that the gene is regulated temporally. (In contrast, growth might be induced by the addition offresh medium to starving synchronized cells.) We used two independent methods to address this is- 29 NOVEMBER 1985 sue. (i) When DNA synthesis was inhibited by the K12 ts mutation, the neo mRNA levels decreased substantially (Fig. 2) . Since the cells were not starved of serum, the result argued against the effect of nutrient on neo mRNA. In addition, when DNA synthesis was inhibited by hydroxyurea, similar results were observed. Together, these results reaffirm the coupling between DNA synthesis and the amount of neo mRNA. (ii) Where serum-synchronized cells were used, neo mRNA increased concurrently with the increase in host cell DNA synthesis and, as with histone mRNA, decreased as host cell DNA synthesis declined (Fig. 3) . This result shows that the hybrid gene is regulated in a cell cycledependent manner, and that changes were not a result of overall growth stimulation. However, it is also clear that the regulation of the neo RNA is not as stringent as that of histone mRNA (Fig. 3) , probably because of both transcriptional and post-transcriptional control in histone mRNA regulation; perhaps only transcriptional control is conferred by the sequences used here. Our results show in vivo that the 5' sequence of a mammalian histone gene is able to confer cell cycle regulation on expression of a heterologous gene to which it is fused in the same transcriptional orientation. This observation is in agreement with a recent report that in in vitro systems, promoter elements in the 5'-flanking sequence of a human histone H4 gene are specifically recognized by cellular factors produced in S-phase nuclear extracts (13). A combination of both approaches should yield informa-tion on the trans-acting regulatory factors and the cis-acting elements involved in the regulation. A further investigation into the cell cycle regulation of the histone genes can be achieved by more detailed analysis of the sequence and function of the 1. 1-kb fragment. Furthermore, with the isolation of this regulatory-promoter sequence, it may now be possible to direct the expression of any compatible heterologous gene in a cell cycle-regulated manner in mammalian cells.