|
|
||||||||
INVITED REVIEW
Departments of Internal Medicine and Physiology and Biophysics, University of Iowa, Iowa City, Iowa 52242
| ABSTRACT |
|---|
|
|
|---|
kidney androgen-regulated protein promoter; P1 artificial chromosome; bacterial artificial chromosome
While the genes involved in many of these disorders have been identified, their physiological roles and interactions often remain to be elucidated. Indeed, genome-sequencing projects have provided a wealth of information on which to base further studies. The obvious next step is to examine gene function in an in vivo setting. For this reason, the ability to generate transgenic animals specifically targeting gene expression to the kidney has become an invaluable tool for the study of normal and pathological states. Targeting tissue-specific gene expression can be a powerful tool for the delineation of a gene's physiological role because these techniques can offer a degree of specificity that cannot be achieved pharmacologically. Since the inception of transgenic techniques well over a decade ago, strategies have been developed to target gene expression to specific cells and tissues. We will review some of the nonclassic methods used to accomplish kidney-specific gene targeting.
| CLASSIC GENE TARGETING |
|---|
|
|
|---|
-glutamyl transpeptidase was used to express
-galactosidase restricted to proximal tubules (28). The epithelial cells of the thick ascending limb of the loop of Henle and early distal convoluted tubule were targeted to express green fluorescence protein by using the promoter for Tamm-Horsfall protein (41). The mouse aquaporin-2 promoter (1) and kidney-specific cadherin (Ksp-cadherin) (15) promoter were used to target reporter gene expression to epithelial cells of the collecting duct. The specificity of expression directed by these promoters offers great potential for the generation of transgenic animals to better understand the contributions of the nephron segments to overall renal function. In addition to physiological studies, cell-specific promoters have also been evaluated for the renal production of therapeutics in transgenic animals. The Tamm-Horsfall protein promoter was utilized to direct human
1-antitrypsin (40) and human erythropoietin (39) expression to the thick ascending limb of the loop of Henle and early distal convoluted tubules of mice for collection of the protein from urine. | CHIMERIC GENES |
|---|
|
|
|---|
The KAP promoter has been particularly useful in studying the intrarenal renin-angiotensin system. This promoter was successfully used to target overexpression of human angiotensinogen (hAGT) to proximal convoluted tubules in mice (10). Approximately 1.5 kb of the KAP promoter were fused to a hAGT genomic clone consisting of exons II-V (Fig. 1A). Expression of the transgene was restricted to proximal convoluted tubules in males and was strongly inducible in females by androgens (10). This inducibility allowed for evaluation of the model with the induced or uninduced transgene in females and castrated males (11). hAGT protein was not released into the systemic circulation but rather was secreted into the tubular lumen (9, 10). KAP-hAGT mice were bred with a transgenic mouse expressing human renin systemically and also in the kidney, resulting in increased arterial pressure. This likely occurred from an increase in tubular ANG II because we detected a large increase in urinary AGT (9). Importantly, the increase in blood pressure occurred despite normal circulating levels of ANG II, strongly implicating a renal mechanism downstream of increased intrarenal ANG II.
|
Given the success of the above transgene, we obviously concluded that 1.5 kb of the KAP are sufficient to accurately target androgen-regulated expression of a target gene to the proximal tubule. However, when the KAP promoter was used to generate gene-targeted or transgenic mice with other genes such as angiotensin-converting enzyme (ACE) (6), luciferase, or Na/H+ exchanger type 3 (NHE3), no expression of the transgene was observed. This prompted us to reevaluate the reason for the success of the KAP-hAGT construct. Unlike simple fusions between promoters and cDNAs, the KAP-hAGT construct contains the genomic sequence for hAGT extending from exon II through the poly(A) addition site. This construct has two features that likely lend to its successful expression. First, KAP-hAGT contains several spliceable introns in their normal position with respect to hAGT. Introns are known to increase the expression of chimeric transgenes (2, 5). Because of this, we have developed a vector to facilitate the cloning of cDNAs and tissuespecific promoters in a plasmid containing a chimeric intron and poly(A) addition site (34). However, the presence of chimeric intron/poly(A) sequences in KAP-luciferase, KAP-NHE3, and KAP-ACE constructs suggests that this feature, on its own, is not responsible for better expression of the KAP-hAGT construct.
More important than the presence of introns was the observation that because AGT is endogenously expressed in proximal tubule cells, it is possible that extrapromoter regulatory elements within the hAGT gene or 3'-untranslated region were sufficient to facilitate (or cooperate with) tissue-specific expression by the KAP promoter. Indeed, there is evidence to support the localization of tissue-specific enhancer elements in the fifth exon and 3'-untranslated region of the hAGT gene (22, 23). These observations lead to the idea that the structure of a well-characterized transgene with appropriate expression properties could be used as a backbone to target expression of other genes of interest. Accordingly, a vector for targeting genes to proximal tubule cells was generated based on the gene structure of the original chimeric KAP-hAGT construct. This vector, called KAP2, is identical to the KAP-hAGT construct described above with two modifications. First, the deletion of nucleotides 1814 (coordinates with respect to exon II) was made that removed most of the AGT-coding potential contained in exon II, including 271 amino acids of AGT protein containing the start codon, the secretory peptide, and the ANG I and ANG II peptides. Second, a NotI restriction site was engineered that allowed for easy insertion of any cDNA (Fig. 1B). Downstream of the Not1 site are the remaining 36 nucleotides of exon II (to facilitate splicing with hAGT exon III) and the remainder of the hAGT structural gene, including intron II through the poly(A) addition site. The construct encompasses the endogenous 3'-enhancer element d612, which has been shown to confer tissue-specific enhancement of reporter gene expression (22). This construct retains the best features of the KAP-hAGT construct; the KAP promoter, the exon-intron structure of hAGT, and the 3'-enhancer. In theory, the replacement of the AGT-coding region with any cDNA (inserted at the NotI site) should result in proximal tubule-specific and androgen-regulated gene expression.
The new KAP2 backbone was used to generate mice expressing rat NHE3. KAP2-NHE3 transgenic mice expressed NHE3 primarily in the kidney with some ectopic expression in the brain (Fig. 2A). Like KAP-hAGT, NHE3 expression was regulated in males and inducible in females by androgens (Fig. 2B). The cDNA for human renin was also inserted into KAP2. KAP2-hRen mice again showed kidney-specific, androgenregulated expression of the transgene (data not shown). These constructs demonstrated the utility of the KAP2 vector for targeting gene expression to proximal convoluted tubules. They also validate the concept of using the gene structure of a chimeric construct previously shown to exhibit an attractive tissue-specific expression profile.
|
It was serendipitous that the KAP promoter had the attractive property of being dependent on androgen, thus providing a mechanism for induction/repression of the transgene. Of course, we will not be as fortunate with all promoters. A number of inducible expression systems have been developed, including the tetracycline regulatory system (13), the ecdysone-regulated gene switch (25), as well as constructs based on mutant transcription factors such as the tamoxifen-responsive ligand-binding domain of the estrogen receptor (8, 17). Puttini et al. (27) reported the use of the tetracycline system to stimulate regulated expression of a reporter gene in the collecting duct, and selective gene disruption in glomerular podocytes was reported using a tamoxifen-regulated Cre recombinase transgene (3). Employing both cell-specific promoters active in the kidney and any one of these expression systems will provide a means of regulating production of exogenous transgenes in the kidney.
| P1 ARTIFICIAL CHROMOSOMES AND BACTERIAL ARTIFICIAL CHROMOSOMES AS LARGE PROMOTERS |
|---|
|
|
|---|
position effect
and in some cases can call into question the validity of a model (30). How can position effects be avoided? One strategy for accomplishing this is through the use of large genomic constructs. Large genomic sequences can be contained in a bacterial artificial chromosome (BAC) or a P1 artificial chromosome (PAC). These constructs, used for genomic mapping and positional cloning, can be employed to make transgenic mice by pronuclear injection. Because of their size, they generally contain ample 5'-regulatory sequences to limit expression of the genes they encode to appropriate sites of expression. Moreover, because these constructs contain all the necessary elements to initiate and regulate gene expression, they are less susceptible to position effects than cDNA-derived constructs and also exhibit copy number-dependent expression (31). It is thought that BAC and PAC transgenes are immune from position effects and exhibit copy number proportional expression because they contain matrix attachment sites, dominant control regions, or insulators that allow them to manipulate DNA at the level of chromatin and be shielded from influences of nearby loci.
An example of this strategy is the generation of transgenic mice expressing the human renin gene from 140- or 160-kb PACs (31). These PACs contained 35 or 75 kb of the 5'-untranslated region (Fig. 3). Human renin was only expressed in the kidney (Fig. 4A) and was restricted to juxtaglomerular cells (Fig. 4B), the same cells as endogenous mouse renin (Fig. 4C). Regulation and stimulation of the human renin gene were identical to those for the endogenous gene, suggesting tight physiological regulation. This was evidenced by 1) the large induction in both endogenous mouse renin and transgenic human renin expression in response to treatment with the ACE inhibitor captopril, 2) higher human renin mRNA expression in mice placed on a low-sodium diet compared with mice fed a high-salt diet, 3) downregulation of human renin mRNA in response to both subpressor and pressor doses of ANG II infused through a minipump, and 4) downregulation of the human renin gene in double transgenic mice containing the PAC construct and the hAGT transgene. The downregulation of human renin in the double transgenic mice correlated with a decrease in arterial pressure compared with double transgenic mice expressing an unregulated human renin transgene. Finally, there was no evidence of positional effects among lines, and transgene expression levels were strongly correlated with transgene copy number (31).
|
|
One limitation of this strategy is that the size of PACs and BACs makes traditional cloning time consuming and problematic, and thus modifications to these clones are more difficult than with standard plasmids. As a result, several plasmid- and phagebased recombination strategies have been developed to modify genes contained in BAC and PAC vectors. These include Chi-stimulated recombination, RAC prophage-based ET cloning, and bacteriophage
-based Red recombination (reviewed in Ref. 7). These techniques make it possible to modify and manipulate these large vectors in Escherichia coli. Our laboratory has recently utilized the Chi-stimulated recombination strategy to delete the putative renin enhancer and insert a
floxed
antibiotic selection cassette into the 5'-regulatory region of the human renin gene contained in a PAC vector (PAC160) (24). A plasmid construct was designed in which the renin enhancer was replaced with a floxed chloramphenicol resistance cassette. The selection cassette was flanked by 500 bp of sequence homologous to the DNA flanking the deletion and then flanked by properly oriented Chi sites (24). Chi is a DNA sequence (5'-GCTGGTGG-3') found
1,000 times/bacterial genome. These sites facilitate increased homologous recombination by the RecBCD pathway. RecBC enzyme travels through linear double-strand DNA, unwinding and rewinding the strands. When the enzyme approaches a Chi site from the 3'-side, it is able to cut the Chi-containing strand 46 bp 3' of the Chi site (38). The enzyme continues to unwind the strands, generating an invasive singlestrand tail. RecA is able to synapse this tail with a homologous duplex to facilitate homologous recombination (26). A schematic representation of homologous recombination in bacteria is shown in Fig. 5.
|
The Chi-flanked construct was removed from the plasmid and cotransfected with PAC160 into bacteria. After selection of recombinants containing the antibiotic selection cassette and purification of the altered PAC160, we were able to remove the selection cassette by passing the construct through a bacterial strain expressing Cre constitutively. This left a single loxP footprint at the site of the deletion, thus providing a vector for the evaluation of the renin enhancer element in the regulation of the human renin gene (24). Details on Cre-mediated recombination will be provided below. The modification of BAC and PAC vectors by Chistimulated and other homologous recombination strategies allows for the simple generation of functional mutants, alternate alleles, or the manipulation of 5'-regulatory regions. These modified vectors can be used to generate transgenic mice either by pronuclear injection or by gene targeting in mouse embryonic stem cells.
BAC and PAC vectors can also be utilized as an alternate strategy for targeting gene expression to a specific tissue. Because these vectors contain all the necessary regulatory elements for gene-specific expression, they have the potential to act as large targeting vectors, or in essence very large tissue- and cell-specific promoters. A portion or the entire coding region of a gene encoded on the PAC or BAC can be replaced with a cDNA of interest by homologous recombination, theoretically resulting in targeted expression of the cDNA to tissues where the original gene is normally expressed. The fidelity of this strategy was demonstrated using a BAC containing the genomic sequence for murine renin (Ren-1d) by Mullins and colleagues (20). Exons III and IV of the Ren-1d gene were replaced with a LacZ reporter gene (20). Transgenic mice generated with this BAC had
-galactosidase activity specifically targeted to cells endogenously expressing Ren-1d throughout development. Moreover, expression of the transgene was copy number dependent. This work thus shows the utility of these large targeting vectors for tissue-specific expression and developmental studies.
| CELL-SPECIFIC ABLATION OF GENE EXPRESSION |
|---|
|
|
|---|
Although in practice gene ablation is technically more difficult, the strategy is similar to that used to target gene expression. In this case, tissue-specific promoters can be used to target elimination of gene expression using the Cre/loxP system. The Cre/loxP system utilizes the bacteriophage P1 enzyme Cre recombinase to facilitate site-directed recombination between homologous loxP sites (14). In this strategy, a mouse is generated with the expression of Cre recombinase under control of a tissue-specific promoter. This strain is bred to a second mouse line, in which the gene of interest has been replaced with an allele containing loxP sites inserted into the intronic regions on either side of an essential exon. Cre facilitates the removal of the DNA between the loxP sites, eliminating functional protein expression only in cells expressing Cre recombinase.
Our first demonstration of the effectiveness of the Cre-loxP system came from studies using a floxed human AGT transgene and delivery of Cre recombinase by adenovirus (Adcre) (32). Systemic delivery of Adcre resulted in effective elimination of AGT mRNA in the liver and a reduction in circulating AGT protein. Double transgenic mice expressing the floxed AGT gene along with the human renin gene are chronically hypertensive. Blood pressure was significantly reduced after elimination of hepatic hAGT by systemic delivery of Adcre (33). We are presently working toward the development of a model expressing Cre recombinase under the control of the KAP promoter using the KAP2 promoter strategy detailed above.
In the kidney, the Cre-loxP strategy has been shown to be effective with the use of promoters specific for podocytes (12, 19), proximal tubules (16), thick ascending limb of the loop of Henle (29, 37), and principal cells of the collecting ducts (21). Regulated gene expression in the collecting duct was reported using a modified Cre protein containing a tamoxifen-response ligandbinding domain from the estrogen receptor (3). Further details of kidney-specific deletion using this technology has been previously reviewed (35).
| CONCLUSIONS |
|---|
|
|
|---|
outside the box.
| DISCLOSURES |
|---|
Address for reprint requests and other correspondence: C. D. Sigmund, Depts. of Internal Medicine and Physiology and Biophysics, 3181B Medical Education and Biomedical Research Facility, Roy J. and Lucille A. Carver College of Medicine, Univ. of Iowa, Iowa City, IA 52242 (E-mail: curt-sigmund{at}uiowa.edu).
| REFERENCES |
|---|
|
|
|---|
piggy-BAC
recombination of the mouse ren locus. J Biol Chem 275: 4037840384, 2000.
-glutamyl transpeptidase type II promoter contains sufficient cis-acting elements for kidney-restricted expression in transgenic mice. J Biol Chem 272: 1195911967, 1997.This article has been cited by other articles:
![]() |
H. Li, X. Zhou, D. R. Davis, D. Xu, and C. D. Sigmund An androgen-inducible proximal tubule-specific Cre recombinase transgenic model Am J Physiol Renal Physiol, June 1, 2008; 294(6): F1481 - F1486. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. A. Fenton, A. Shodeinde, and M. A. Knepper UT-A urea transporter promoter, UT-A{alpha}, targets principal cells of the renal inner medullary collecting duct Am J Physiol Renal Physiol, January 1, 2006; 290(1): F188 - F195. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. L. Lavoie, K. D. Lake-Bruse, and C. D. Sigmund Increased blood pressure in transgenic mice expressing both human renin and angiotensinogen in the renal proximal tubule Am J Physiol Renal Physiol, May 1, 2004; 286(5): F965 - F971. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |