Differential regulation of transcription of human 7 S K and 7 S L RNA genes.

Two functional human genes coding for 7 S RNA species K and L were analyzed for promoter requirements by in vitro transcription experiments with cytoplasmic S-100 extracts. Since accurate and efficient transcription of both genes is dependent on the presence of 5'-flanking sequences, hybrid genes representing crossover fusions between the 5' external control regions and the coding sequences of both genes were analyzed for their capacity to direct RNA synthesis in vitro. Differing results were obtained with both types of constructs. While the 5'-flanking L-7 S K gene fusion revealed no activity in the in vitro transcription assay, the 5'-flanking sequence of the 7 S K RNA gene did confer the ability for accurate in vitro transcription to the 7 S L coding sequence. However, a 5'-flanking L sequence element including the first 22 nucleotides of the 7 S L RNA coding sequence was active in promoting transcription of the 7 S K RNA gene. Together, these results demonstrated that the 7 S L promoter is located inside and outside the coding region, whereas the 7 S K RNA gene is exclusively controlled by an upstream promoter element.

Two functional human genes coding for 7 S RNA species K and L were analyzed for promoter requirements by in vitro transcription experiments with cytoplasmic 5-100 extracts. Since accurate and efficient transcription of both genes is dependent on the presence of 5"flanking sequences, hybrid genes representing crossover fusions between the 5' external control regions and the coding sequences of both genes were analyzed for their capacity to direct RNA synthesis in vitro. Differing results were obtained with both types of constructs. While the 5"flanking L-7 S K gene fusion revealed no activity in the in vitro transcription assay, the 5"flanking sequence of the 7 S K RNA gene did confer the ability for accurate in vitro transcription to the 7 S L coding sequence. However, a 5'flanking L sequence element including the first 22 nucleotides of the 7 S L RNA coding sequence was active in promoting transcription of the 7 S K RNA gene. Together, these results demonstrated that the 7 S L promoter is located inside and outside the coding region, whereas the 7 S K RNA gene is exclusively controlled by an upstream promoter element.
Among the three forms of eucaryotic RNA polymerases, the class I11 enzyme has been shown to be involved in transcription of several low molecular weight RNA species. These include 5 S and tRNA (1) as well as the adenoviral VA RNA species I and I1 (2) and the small nuclear RNA U6 (3). In addition, the two small cytoplasmic RNA species, 7 S K and 7 S L have been shown to be transcribed by RNA polymerase I11 in vivo (4) and in vitro (5).
Recently, functional genes coding for human 7 S L RNA (6, 7) and for 7 S K RNA (8,9) have been cloned and characterized. These studies revealed that efficient transcription of both 7 S RNA genes depends on the presence of 5'flanking sequences. In contrast, in various other class I11 genes such as 5 S RNA (IO), essential promoter elements have only been identified inside the coding region. 5' Deletion mutants of the 7 S L RNA gene revealed that an essential regulatory element is located within the first 37 nucleotides upstream from the transcription start site (11). A similar analysis of 7 S K DNA identified crucial promoter elements around position -60 (9) and -25 (12), respectively. In addition, recent findings demonstrated that 7 S K RNA transcription only requires an upstream sequence element (12, 13).
*This study was supported by a grant from the Deutsche Forschungsgemeinschaft (to B. J. B.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ T o whom correspondence and reprint requests should be addressed.
Although both genes do not share significant homology within their 5'-flanking sequences, they appeared to be regulated in a similar manner. Therefore, we wanted to know whether or not the two external control elements are functional in conjunction with the coding sequence of the other gene, respectively. Hybrid genes were constructed by crossover exchange of both 5"flanking regions and analyzed for accurate transcription in vitro.

RESULTS
7 S RNA Genes-Two functional human genes coding for the small cytoplasmic 7 S RNA species K and L, respectively, have been cloned in our laboratory. The 7 S K RNA gene has been characterized in detail previously (9). Efficient in vitro transcription depended on 5"flanking sequences up to position -60. The isolation, structural organization, and sequence analysis of the functional 7 S L RNA gene are described in Fig. 1 and in the text in the Miniprint.
Construction and Analysis of Hybrid Genes-Since transcription regulation of the two 7 S RNA genes seemed to depend on an interaction between external promoter elements and gene internal control regions, we wanted to know whether or not these regulatory elements are interchangeable without losing the capacity to promote transcription initiation. Therefore, based on these two functional 7 S RNA genes, crossover fusions between the 5'-flanking and coding sequences were constructed (see "Materials and Methods"). The specificity of this major transcript was verified by hybridization selection (Fig. 2B, lane 5), by SI nuclease protection (Fig. 2B,  lane 6) and by determination of the transcription start site (Fig. 3). Together, these results demonstrate that the 5'flanking K-L RNA gene fusion is functional in uitro. This finding is in good agreement with recent reports that the 7 S K upstream sequence alone is sufficient for accurate and efficient transcription initiation by RNA polymerase I11 in vitro (12) and in vivo (13).
' Portions of this paper (including "Materials and Methods," part of "Results," and Figs. [1][2][3][4][5] are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press. Analysis of 5"L-K RNA Hybrid Genes-Initially we obtained a hybrid gene which-due to the cloning procedurecontained additional 23 bp2 between the -1 position of the 5"flanking 7 S L sequence and the +1 nucleotide of the 7 S K coding sequence (map not shown). This construct turned out to be not functional since the RNA obtained with this hybrid gene (Fig. 4C, lane 3) revealed no additional product when compared to the transcription pattern of the PAT 153 vector alone (Fig. 4C, lane 1). One might argue, however, that this negative result reflects the "incorrect" spacing between the two sequence elements rather than a noncompatibility between external and internal promoter elements. Therefore, an almost perfectly matching hybrid gene was constructed which represents a -61-3 fusion between the 5"flanking 7 S L and the 7 S K coding sequence. As shown in Fig. 44, only two nucleotides (-5, -4) are missing. However, transcription analysis (Fig. 4C, lane 4 ) again revealed no functionality of this construct. Therefore, it appeared that in these fusions failure of transcription initiation was not due to incorrect spacing. This conclusion was supported by analysis of a 5'flanking L-L coding construct (Fig. 4 B ) which also contained an additional 18 bp. Yet, this construct was active in the in vitro transcription assay (Fig. 4C, lane 7), although reduced in efficiency if compared to the wild type gene (lane 6).
Therefore, we conclude that the 5"flanking L sequence does not represent a complete functional promoter but rather needs a gene internal complement which is absent in the coding region of the 7 S K RNA gene.
In order to localize this internal control element, deletions were introduced into the 5' half of the 7 S L RNA gene (Fig.   5A). These minigenes are actively transcribed in vitro although the rate of synthesis is lowered compared to the 7 S L wild type gene (Fig. 5B, lunes 1-3). However, a further deletion mutant representing a +7/145 fusion of the 7 S L RNA gene, was not active any more (data not shown). Therefore, it appeared that the gene internal part of the 7 S L RNA promoter is located within the first 23 bp of the coding sequence. In view of these findings, another 5'-flanking L-K coding fusion gene was set up. In this case, a 5'-flanking 7 S L sequence including the first 22 bp of the coding sequence was fused with the 7 S K RNA gene (3rd construct of Fig.   5A). The analysis of this fusion gene revealed that now the 7 S K DNA is transcribed under the control of the 7 S L promoter, resulting in the expected transcript of 379 nucleotides (Fig. 5B, lane 6). Again, transcription efficiency was reduced as compared to the 7 S L wild type gene (lune 5 ) but corresponded to the synthesis rate observed with wild type 7 S K DNA under optimal conditions (lune 4 ) .
In summary, transcription under the control of the 7 S K promoter requires only 5"flanking sequences, whereas functionality of the 7 S L promoter depends on sequence elements located outside and inside the coding region.

DISCUSSION
Originally, in vitro transcription experiments had demonstrated that initiation of RNA polymerase I11 is primarily regulated by gene internal sequence elements (14-19). This view was modified when evidence accumulated that 5'-flanking sequences play an important role at least in modulating the transcription efficiency of various class I11 genes such as tRNA and VA RNA genes (20, 21). In uitro transcription of two functional human 7 S RNA genes also depended on their 5'-flanking sequences (8,9, 11). More recent findings further demonstrated that transcription of human 7 S K DNA is exclusively under the control of 5"flanking sequences (12, 13). Therefore, it appears that the 5 S RNA gene transcription which is only dependent on the gene internal control region represents the exception rather than the rule among the class I11 gene transcription regulation.
In order to gain insight into a possible cooperation between external and gene internal regulatory elements of class I11 genes, we constructed 7 S RNA hybrid genes by a reciprocal exchange of 7 S K and 7 S L RNA flanking and coding sequences. The 5°K-L hybrid gene was accurately transcribed in vitro supporting the view of a RNA polymerase 11-type regulation with the 5"flanking region of the 7 S K gene alone containing all sequence elements necessary for accurate and efficient transcription initiation (12, 13). It should be noted that the in vitro transcription efficiency of the 5°K-L gene construct is comparable to that of the 7 S K RNA wild type gene (see lanes 1 and 3 of Fig. 3). This finding is particularly interesting since the construct did not contain an authentic transcription start site. Although the inserted polylinker sequence also provides a G nucleotide at the initiation site of the 7 S K wild type gene, initiation occurred two nucleotides upstream where another G is located at position -2 relative to the 7 S K flanking sequence. Thus, it appears that the RNA polymerase I11 does not simply scan a fixed number of nucleotides toward the coding sequence for initiation of transcription. The enzyme rather seems to select the first G of a purine cluster (GGA in the wild type gene uersus GAG within the construct) in the vicinity of the original start site.
In contrast, no specific transcript was obtained with a 5'L-K fusion gene. This finding suggests that both 7 S RNA genes are subjected to distinct mechanisms of transcription regulation. However, the negative result obtained with the 5'L-K gene fusion can be reversed by retaining the first 22 bp of 7 S L coding sequence within the construct. Now, such a hybrid is active in transcription initiation, indicating that the 7 S L promoter extends from the region around -30 (11) to an internal sequence located between +7 and +22. Since, within certain limits, the spacing between these two regions of the 7 S L promoter is not crucial and, on the other hand, an insertion of vector sequences also did not cause inactivation, we assume that the 7 S L control region represents a type of split promoter with one element being located inside and the other outside the coding sequence. Clearly, the 5"flanking sequence alone is not able to support transcription initiation, and the coding part of the gene is not transcribed in the absence of upstream sequences. Therefore, an interaction or cooperation is likely to exist between the two sequence elements. It appears that a detailed investigation on the transcription factor and/or RNA polymerase binding sites inside and outside the 7 S L RNA gene is a prerequisite for the understanding on how these internal and external control elements cooperate in transcription initiation.
In contrast, the structure of the external 7 S K RNA promoter is quite different. Here, at position -25, a TATAbox motif has been identified as being essential for efficient in uitro transcription (12). One should keep in mind, however, that upstream deletions of the 5"flanking 7 S K DNA revealed that the sensitive region of this external promoter extends to about position -60 (13). This is well beyond the limits of binding regions identified so far for TATA-box binding factors and may indicate that more than one factor is involved.