Functional elimination of calmodulin within the nucleus by targeted expression of an inhibitor peptide.

Genetic manipulation has proven valuable in identifying the role of specific genes in cellular function. Genomic disruption of genes that are expressed during embryonic development or in multiple tissue types, however, complicates phenotypic analysis. We demonstrate that targeted expression of an inhibitor peptide derived from myosin light chain kinase can neutralize the function of calmodulin. We have shown that elimination of the nuclear function of Ca(2+)-calmodulin causes disruption of the nuclear structure. Targeted expression of this calmodulin inhibitor gene in the lung epithelium of transgenic mice leads to cellular death and dysfunctional lung development. This approach is a strategy to modify the activity of a targeted protein within a specific organelle in order to evaluate its role in cellular and tissue function.

zymes. In the present study, the MLCK CaM binding domain was used to design a synthetic concatemer gene that would target and functionally neutralize CaM in the nuclei of cultured cells and in lung epithelium of transgenic mice.

EXPERIMENTAL PROCEDURES
Cloning and Expression of MLCK Peptide Gene in E. coli-A synthetic gene (72 bp) containing an ATG initiation codon and XmaI restriction enzyme sites at each end was designed to encode the rabbit skeletal muscle MLCK CaM-binding sequence, KRRWKKNFIA-VSAANRFKK. The respective coding and complimentary oligonucleotides were annealed and cloned into the XmaI site of pBluescript SK (Stratagene) using T4 DNA ligase. Proper orientation of the MLCK peptide gene was confirmed by dideoxy DNA sequencing (United States Biochemical Corp.). A BamHI-EcoRI fragment containing the coding sequence was released from pBSK and fused to the 3Ј end of the glutathione S-transferase (GST) gene using the BamHI and EcoRI sites of the bacterial expression vector pGEX-2T (Pharmacia Biotech Inc.). The GST-MLCK peptide fusion protein (30.6 kDa) was induced in Escherichia coli SURE cells (Stratagene) with 100 M isopropyl-1-thio-␤-D-galactopyranoside for 2 h and purified by glutathione-Sepharose affinity chromatography according to the manufacturer's instructions (Pharmacia).
Purification and Assay of CaM-dependent Protein Kinase II-CaM kinase II was purified from rabbit brain using Ca 2ϩ -dependent binding to CaM-Sepharose (12). Enzyme reactions (50 l) were performed at 30°C in 50 mM HEPES (pH 7.5) containing 0.1 mM calcium chloride, 10 mM magnesium acetate, 0.1 mg/ml bovine serum albumin, 2 Ci of [␥-32 P]ATP (DuPont NEN), 0.1 g of CaM kinase II, and 0.5 g of myelin basic protein, with or without 0.5 g of calmodulin as indicated. Synthetic MLCK peptide or purified GST-MLCK peptide fusion protein at the indicated concentrations were added. The reaction was terminated after 30 s with 10% trichloroacetic acid. The mixture was centrifuged at 12,000 rpm for 15 min and the resulting pellets dissolved in 150 l of SDS sample buffer and separated by 12% SDS-PAGE. The gel was dried and autoradiographed.
Production and Affinity Purification of Anti-MLCK Peptide Antibodies-Initial injections into four individual rabbits were performed using KRRWKKNFIAVSAANRFKK coupled to keyhole limpet hemocyanin (KLH, Sigma). The quality of the peptide was confirmed by fast atom mass spectroscopy. Six milligrams of the peptide were mixed with an equal weight of KLH and suspended in distilled water (pH 5.0); 5 mg of carbodiimide (Pierce) was dissolved in the mixture. A second 6 mg:6 mg mixture of the peptide and KLH was suspended in 125 mM borate (pH 8.4) containing 75 mM NaCl and 5 mg of dimethylsuberimidate. Both mixtures were allowed to react overnight in the dark at room temperature. The reactions were terminated by the addition of 100 l of 0.1 M ethanolamine. For the first injection, each rabbit received a 2-mg peptide equivalent of the mixture emulsified in complete Freund's adjuvant. The second injection, 30 days later, contained 1 mg of peptide equivalent emulsified in 1/10 complete Freund's adjuvant. The third and fourth injections were on days 45 and 60, respectively. Each boost contained 1 mg of free peptide in incomplete adjuvant. The animals were bled on day 75 and every 2 weeks thereafter.
Affinity columns were constructed using 5 mg of the free peptide dissolved in 125 mM borate (pH 8.4) containing 75 mM NaCl and 5 ml of CNBr-activated Sepharose (Pharmacia Biotech Inc.) Immune sera were centrifuged (10,000 ϫ g for 15 min) and 5 ml applied to the affinity resin. The column was washed with 100 ml of 0.1 M NaHCO 3 (pH 8.4) containing 0.5 M NaCl. The bound IgG was eluted with 0.1 M glycine (pH 2.7) and immediately neutralized to pH 7.4 with 1 M Tris (pH 9.0).
Expression and Localization of the MLCK Peptide Gene in Mammalian COS-7 Cells-5Ј-CGTACTCGGGATGAAGAGAAGATGG-3Ј and 5Ј-ACTACCCGAGCTTCTTGAATCTGTT-3Ј were designed to incorporate an AvaI restriction enzyme site "CTCGGG" at each end of the monomer gene in pBluescript SK using the polymerase chain reaction (PCR). Following AvaI digestion, the amplified monomer genes (66 bp) were self-ligated using T4 DNA ligase in a head-to-tail manner. The ligation mixture was blunt-ended using Klenow and cloned into the SmaI site of pBluescript SK (Stratagene). Clones that contained differ-ent lengths and proper orientation of concatemer genes were confirmed by restriction enzyme mapping and DNA sequencing. Amino acids Leu and Gly, which are encoded by the AvaI site, CTCGGG, acted as a spacer between each monomer peptide. The Kozak initiation sequence (13) and an XbaI site were incorporated into the 5Ј PCR primer 5Ј-CTAGCTAGCGCCACCATGGGACCCCTCGGGATGAAG-3Ј. A SmaI site and a stop codon were designed into the 3Ј PCR primer 5Ј-CCCCTC-GAGATTCGTACCGTACCTGCAGCCCCCGAG-3Ј. A four-repeat MLCK peptide gene (318 bp), which is a XbaI-SmaI fragment, was subcloned into the XbaI and SmaI sites of the eukaryotic expression vector pSVL (Pharmacia). This vector contains the SV40 late promoter, a VP1 intron, and the SV40 polyadenylation signal. COS cells were maintained in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) containing 10% fetal bovine serum at 37°C in a 5% CO 2 humidified incubator. Recombinant DNA (20 g) was purified using a Qiagen Maxi Kit and was transiently transfected into COS-7 cells by the calcium-phosphate precipitation method (31). Transfected cells were grown on 22 ϫ 22-mm coverslips. At 24, 48, or 72 h after transfection, cells were fixed in 4% paraformaldehyde and permeabilized in Ϫ20°C acetone. The fixed cells were incubated for 1 h with affinity-purified rabbit anti-MLCK peptide antibody, washed with phosphate-buffered saline, and incubated with fluorescein-conjugated goat-anti-rabbit IgG. Double-stained cells were also incubated for 1 h with 10% normal rabbit serum, followed by affinity-purified sheep anti-calmodulin antibody and CY3-conjugated rabbit anti-sheep IgG (Jackson Laboratories). The coverslips were then mounted and photographed using Kodak Elite 200 Ektachrome film with a Nikon Microflex UFX exposure attachment mounted on a Nikon Optiphot epifluorescence microscope and the appropriate Nikon fluorescence filter cubes.
Targeted Expression of MLCK Peptide Gene to the Lung Epithelium of Transgenic Mice-The transgene is composed of the 3.7-kb HindIII-HindIII fragment of human SP-C promoter, 318-bp NheI-XhoI fragment of a four-repeat MLCK peptide gene, and 850-bp XhoI-BamHI fragment of SV40 intervening sequence and polyadenylation signal. The 4.8-kb transgene was excised from pBluescript KS by SalI and NotI digestion, separated on 1.0% agarose gel, and purified by Geneclean (Bio101 Inc.). This DNA was dissolved in 10 mM Tris (pH 8.3) and 0.1 mM EDTA and microinjected into donor eggs of mouse strain FVB/N by the University of Cincinnati Transgenic Animal Core Facility. Founder mice were identified by PCR amplification of DNA isolated from tail-clippings using primer 5Ј-AGCTTGATATCGAATTCCTG-3Ј (specific for human SP-C) and the 3Ј primer, 5Ј-GCAAGCAAGAGTTCTATTCA-3Ј (specific for the 3Ј untranslated region of SV40). The PCR product was a 703-bp fragment of the MLCK peptide transgene. Four surviving founder mice were identified from 83 pups delivered by surrogate mothers. Embryos obtained from the X-linked transgenic females were harvested and fixed in 4% paraformaldehyde for 48 h at 4°C. The fixed tissues were then embedded in paraffin, and 4-m cross-sections were cut. Deparaffinized sections were incubated with 2 g of the respective affinitypurified rabbit anti-MLCK peptide antibody for 2 h, washed three times in Dulbecco's phosphate-buffered saline, incubated with fluoresceinconjugated goat anti-rabbit IgG for 1 h, and washed again. Sections mounted in glycerol/phosphate-buffered saline (9:1) containing 0.1% p-phenylenediamine were viewed on a Nikon epifluorescence microscope.

RESULTS AND DISCUSSION
We used the rabbit skeletal muscle MLCK 19-amino acid CaM-binding sequence to design a gene, which, when expressed, would eliminate the cellular function of Ca 2ϩ -CaM by competing with endogenous target proteins. In order to test whether the synthetic gene product would inhibit CaM, the MLCK peptide gene was expressed in bacteria as a fusion protein. The fusion protein, like the chemically synthesized peptide, inhibits Ca 2ϩ /CaM-dependent protein kinase II in a concentration-dependent manner (Fig. 1).
CaM is composed of two highly structured globular Ca 2ϩbinding domains connected by an ␣-helical peptide (14). In the presence of Ca 2ϩ , the lobes engulf the target peptide like a vise (15). The inhibitor peptide gene for in vivo studies was designed to include two properties, high binding capacity and intracellular targeting. Due to the fact that calmodulin is abundant in cells, the monomer gene was self-ligated into four tandem repeats in order to increase the calmodulin binding capacity of each transcript and to target localization to the nucleus. The sequence of the first two repeats of the concatemer is: MKRRWKKNFIAVSAANRFKKLGMKRRWKKNFIAVSAAN-RFKKLG. This sequence contains several basic amino acid clusters and the bipartite lysine-lysine motif, KK . . . . . . . KK (underlined). These are potential nuclear localization signals common in many karyophilic proteins such as the SV40 Tantigen and transcription factors (16). In order to evaluate the cellular targeting of the CaM inhibitor gene and its effects in intact cells, COS-7 cells were transiently transfected with the concatemer gene. The cells were examined by indirect immunofluorescence using affinity purified antibody produced against the inhibitor peptide. As shown in Fig. 2a, approximately 10% of the cells expressed the MLCK peptide concatemer, which was concentrated in the nuclei of transfected cells (Fig. 2, c and d). During early stages of expression, the peptide staining pattern was "donut-like" (Fig. 2e). By 72 h, the nuclei became extremely large and multi-lobed, and contained condensed chromatin indicating disruption of nuclear function (Fig. 2f). It did not appear that the MLCK peptide-transfected cells entered mitosis. A splicing isoform of CaM kinase II contains the nuclear targeting sequence, KKRK, and is localized in the nucleus (3). Microinjection of a CaM kinase II inhibitory peptide into sea urchin eggs causes a delay in nuclear envelope breakdown (17). Targeted overexpression of the MLCK peptide in the nucleus may inhibit CaM kinase II activity preventing cells from entering mitosis. Double immunostaining of transfected cells demonstrates coincident localization of CaM and the MLCK peptide in the nucleus. CaM is also present throughout the cytoplasm, indicating that only nuclear functions of CaM were being neutralized (Fig. 2, g and h).
In order to evaluate the consequences of expression of the MLCK peptide during cell growth and differentiation, the CaM inhibitor gene was targeted to a specific cell type in transgenic mice. Due to the potent inhibitory effects of the peptide observed in the CaM kinase II assay and in cultured cells, the lung was selected as the target organ since the animals would not require respiratory function until birth. The 3.7-kb 5Ј promoter region of the human SP-C gene dictates specific expression to type II cells of the distal airway epithelium (18). Activation of gene expression by the 3.7-kb SP-C promoter occurs

Elimination of CaM by Expression of an Inhibitor Peptide 30246
approximately on day 11 of fetal life, when a definable lung epithelium first appears.
Our transgene was composed of the SP-C promotor, the four-repeat MLCK peptide concatemer gene, and SV40 RNA processing motifs (Fig. 3a). Offspring of transgenic founder mice displayed varying degrees of lung pathology, which corresponded to congenital cystic adenomatoid malformation. In this condition, the terminal bronchioli continue to develop in the absence of alveolar epithelial proliferation. At birth, trapped air causes cystic dilation of the bronchioles. Alveoli that appeared normal were found among the cysts in some transgenic animals. One founder, a female, demonstrated no obvious physical symptoms. Her litters contained two transgenic phenotypes. Some pups displayed cyanosis and died within 15 min of birth (Fig. 3b). All of the surviving transgenic animals were females. They demonstrated retarded post-natal development as exhibited by a 10 -20% reduction in body size and displayed delayed hair growth, eye opening, and tooth eruption (Fig. 3c). Affinity-purified antibodies produced against the MLCK peptide showed that expression of the transgene product occurs by approximately day 13.5 of embryonic development (Fig. 3d). Cellular differentiation and bronchial branching began to decrease. Dilated tubules appeared at day 15.0 (Fig. 3e). There was no anti-MLCK peptide staining in the lung of day 16.5, 17.5, and post-natal male transgenic animals. Histological examination of tissue sections revealed that the transgenic male animals that died at birth had not developed a lung epithelium or lower airways (Fig. 3f). These results indicate that the type II cells that express the inhibitor peptide did not survive early development of the lung epithelium. The trachea and all major organs were histologically normal. Analysis of subsequent F 1 , F 2 , and F 3 generations confirmed that all of the transgenic males died at birth and that all of the surviving transgenic animals were female (Fig. 4).
The transgene is lethal in males, suggesting an X-linked dominant trait. During morula development of females, there is random inactivation of one of the X-chromosomes, which form a Barr body. All of the cells in a given lineage inactivate the same X-chromosome throughout the life of the animal (19). In the present study the "Lyonized" females develop a mosaic lung epithelium. In some proliferating type II cells, and their clonal descendants, the CaM inhibitor peptide gene would remain silent, enabling some of the heterozygotic transgenic fe-

Elimination of CaM by Expression of an Inhibitor
Peptide 30247 males to survive. Most of the transgenic females lived to adulthood, reproduced, and had a normal lung morphology. These studies confirm that progenitor type II cells are important for normal lung development. Genetic manipulation of organisms at all levels of complexity has proven valuable in identifying the role of specific genes in cellular function. Gene "knock-outs" by genomic homologous recombination in embryonic stem cells allows for the selective disruption of targeted genes (20). Ablation of genes that are critical for the function of many cell types influences fetal development and can prove lethal. Indeed, disruption of fungal CaM genes demonstrates that CaM is essential for survival (21)(22)(23). Our study demonstrates that binding peptides can be used to selectively neutralize the function of a targeted protein such as CaM in a specific organelle of a defined cell type. Our findings also suggest that the mosaic inactivation of the MLCK peptide gene in lung allowed the unaffected clonal type II cells to proliferate and compensate for early loss of the cells expressing the concatemer peptide. These findings demonstrate that dominant lethal genes can be sustained in the germ-line of animals by targeting transgene integration into the X-chromosome. Male progeny display the most severe phenotype, while females demonstrate a broad spectrum of phenotypes depending upon the ratio of X-chromosome inactivation of the transgene during early embryogenesis.
Organism organization is based upon recognition between individual biochemical components. Cell adhesion, assembly of organelles, and signal transduction pathways are each, for example, a complex series of molecular interactions. Properties of individual proteins are altered through covalent modifications and ligand binding. Stimuli initiate a cascade of protein conformational changes. These changes, in turn, modify the properties of other proteins in sequential order. These physical interactions transmit signal bifurcation and propagation throughout the cell, ultimately leading to a cellular response. Analysis of protein-protein and protein-nucleic acid associations has shown that the interactive domains are highly structured and are often composed of less than 25 amino acids (24 -30). These domains of recognition are attractive target sites for the design of peptides that modify specific cellular pathways. Identification and design of binding peptides that modify a targeted protein activity can be obtained by several techniques including biochemical fragmentation, three-dimensional structural analysis, mutagenesis of interacting species, two-hybrid transcriptional systems, and libraries of random peptides (27). Once identified, the binding properties of a selected peptide can be further refined by mutagenesis and reselection. Due to the fact that cellular processes involve a cascade of protein-protein interactions, the interference of molecular recognition by the use of binding peptides should be applicable to modulate any targeted cellular function.

FIG. 4. Pedigree of transgenic mice
Carrying the X-linked dominant MLCK peptide concatemer gene. This pedigree shows F 1 , F 2 , and F 3 generations of X-linked founder mouse (F 0 ) in which f } ϭ transgenic males that died at birth; F } ϭ transgenic females that died at birth; F ϭ surviving transgenic females; Ⅺ ϭ non-transgenic male; E ϭ non-transgenic female.