We have permanently reversed the lethal phenotype in the cystic fibrosis (CF) transmembrane conductance regulator (CFTR)-deficient (knockout) mouse after in utero gene therapy with an adenovirus containing the cftr gene. The gene transfer targeted somatic stem cells in the developing lung and intestine, and these epithelial surfaces demonstrated permanent developmental changes after treatment. The survival statistics from the progeny of heterozygote-heterozygote matings after in utero cftr gene treatment demonstrated an increased mortality in the homozygous normal pups, indicating that overexpression during development was detrimental. The lungs of these pups revealed accelerated secretory cell proliferation and differentiation. The extent of proliferation and differentiation in the secretory cells of the lung parenchyma after in utero transfer of the cftr gene was evaluated with morphometric and biochemical analyses. These studies provide further support of the regulatory role of the cftr gene in the development of the secretory epithelium.
- cystic fibrosis transmembrane conductance regulator
this laboratory has developed a practical and reliable technique for in utero transfer of genes via the amniotic fluid. This method results in sustained expression and high-efficiency adenovirus-mediated transfer to the fetus. Transfer to the developing epithelium is accomplished by direct injection of a replication-defective adenovirus into individual amniotic sacs of rodent fetuses (28). Gene transfer to the fetus in this manner circumvents the inflammatory response that plagues adenoviral-mediated gene transfer after birth (28).
Gene transfer is performed at a stage of lung and intestinal development comparable to that of a 10- to 20-wk-gestation human. At the time of infection, the rodent airways are lined with multipotential columnar and cuboidal stem cells; differentiation and further growth occur after infection and gene transfer. These undifferentiated epithelial cells are the targets of the adenovirus (28).
Cystic fibrosis (CF) has been a major target for corrective gene therapy and an ideal candidate for in utero gene therapy targeting the developing lung. The CF transmembrane conductance regulator (cftr) gene was cloned without the prior knowledge of the structure and function of the protein (26,27). Structural analysis suggested that its main function was that of a cAMP-regulated chloride channel. Since this discovery, the protein has been shown to have diverse regulatory abilities, yet CF disease pathology is still defined in terms of a lack of a chloride channel. CFTR regulates other secretory channels (12), mediates vesicular trafficking (3), and affects glycosylation (19, 38). The two nucleotide binding domains (NBD1 and NBD2) of CFTR share structural homology with the conserved sequences of G protein (21). CFTR is degraded by the ubiquitin-proteasome pathway, which is usually reserved for regulatory proteins that require rapid degradation (35).
CFTR expression is greater in the fetal lung than in the adult lung. CFTR mRNA and protein show temporal and tissue-specific distribution during development. Protein expression is greatest in the lung during the first and second trimesters (9, 13,22, 23, 34).
The significance of the regulatory properties and expression of the CFTR during development was unclear until our laboratory demonstrated the complete reversal of the lethal phenotype of the CFTR-deficient [cftr(−/−); knockout] mouse after transient in utero expression of the gene. After gene therapy with a first-generation adenovirus vector carrying the cftr gene (Av1CF2), low levels of the cftr gene were detectable in the fetal gut up to 72 h but not after birth. In utero gene therapy did not permanently replace the CFTR-encoded cAMP-dependent chloride channel, and continuous functioning of CFTR was not required for correction of the intestinal obstruction incftr(−/−) mice (20).
The rescued knockout mice exhibited phenotypic and functional changes in their intestine and lungs (6). Intestines from the in utero cftr gene-treated cftr(−/−) mice did not develop the intestinal pathology of crypt dilation or goblet cell hypertrophy that proves fatal for the untreated knockout mice. The intestinal crypt cells in the untreated knockout mice were deficient in both intracellular uptake of Calcium Green-AM and purinergic receptors. Both of these deficiencies were partially corrected in the rescued knockout mice, and the restored cells were clonally distributed along the villi (6).
The phenotypic changes in the lungs were predominantly in the airways. Clara cells, which are the predominant secretory cell in the terminal airways, showed marked differences in ultrastructure in animals treated in utero with the cftr gene (6). Clara cells in knockout mice treated in utero with the cftr gene showed significant increases in vesicle number and dilated smooth endoplasmic reticulum. Untreated knockout mice had an increase in secreted glycoconjugates containing α-(2,6)-sialic acid and fucose (both implicated in the pathogenesis of CF) compared with that in control animals. Lectin histochemistry located these glycoconjugates to the airway epithelial cell surface. The in utero treated knockout mice had an increase in this material as well, but it was contained in intracellular vesicles. These data suggested that thecftr gene restored regulated secretion in both the lung and intestine and was required for normal differentiation of secretory cell populations.
Here, we examine the survival of our treated and untreated adult populations from heterozygote-heterozygote matings. We describe the lethal effects of excessive CFTR expression during normal development of the lung in additional experiments with homozygous normal [cftr(+/+)] populations. We present data that shows that CFTR expression during lung development increases epithelial proliferation and enhances secretory cell differentiation. These findings are consistent with the CFTR protein possessing a secretory cell missing (scm) function that is required for normal development of secretory cells in the lungs, intestines, and other organs (6).
The S489X cftr-mutant UNC mouse strain was obtained from Jackson Laboratories as a fifth-generation backcross. All animals were housed under standard vivarium conditions and fed mouse chow (Purina) after being weaned at 28 days of age. Wood chips were used as bedding, and the mice were not protected in any manner from exposure to pathogens. Under these conditions, the S489X cftr-mutant mice develop intestinal obstruction, and <5% survive into adulthood. Death occurs from intestinal obstruction in the mice homozygous for the mutant allele during one of two periods postnatally. The first period is within 5 days of birth (∼50%), and the second period occurs around weaning (30,31).
Samples of tail DNA were obtained from surviving animals between 10 and 15 days of age for PCR. Animals were then tagged and followed until 75 days of age. Experience with the colony has shown that animals that survive until 75 days of age [including in utero-treatedcftr(−/−) mice] survive past 1 yr of age. A χ2 analysis was used to analyze the significance between observed and expected survival frequencies in the lacZ gene- and cftr gene-treated populations. A P value of <0.05 was considered significant.
The fetuses were treated at 15–16 days gestation with either Ad5.CMVlacZ or Av1CF2. Initial experiments were planned around the progeny of heterozygote-heterozygote matings; therefore, treated groups included knockout [(−/−)], heterozygous [(−/+)], and homozygous normal [(+/+)] animals. Subsequent overexpression experiments utilized homozygous normal-homozygous normal matings. With amniotic fluid volume as a standard, the vector was injected in 10% of the known amniotic fluid volume with a fine-gauge needle. A concentration of 109 plaque-forming units/ml of each vector was injected into each fetal amniotic sac, resulting in a final concentration of 108 plaque-forming units/ml.
Genotyping of Heterozygote-Heterozygote Matings
In addition to the genotyping of live infants, the deceased fetuses were collected and genotyped when possible. Only animals in their entirety were used for typing. Wild-type CFTR and the knockoutS489X allele were amplified in parallel reactions with conditions provided by Jackson Laboratories. Primers CF1248R and CF944 were used for detection of the normal allele. Primers Neo150R and CF944 were used for detection of the mutant allele. Primer sequences (provided by Jackson Laboratories) were 5′-CTTTGATAGTACCCGGCATAATC-3′ for CF1248R, 5′-TCGAATTCGCCAATGACAAGAC-3′ for Neo150R, and 5′-TGAACCTTAGTCCTATGTTGCC-3′ for CF944.
Examination of Homozygous Normal Fetuses After In Utero cftr Treatment
Lung morphology and morphometry light level.
At 20–21 days gestation, the lungs and intestines were removed and weighed. The lethal dose of pentobarbital sodium that the mothers received inhibited any fetal respiratory effort. Lung weight-to-total body weight and intestine weight-to-total body weight ratios were determined. Volume densities of saccular air space and tissue, airway air space and tissue, and vessel lumen and tissue wall were estimated by point-counting morphometry. Light micrographs of the fields were transferred to the computer by densitometry, and a 121-point multipurpose test grid was applied. A minimum of 20 fields were evaluated from each of 12 animals from 4 separate litters. Differences between groups were evaluated by Student's t-test. Statistical significance was assigned at P < 0.05.
Ultrastructural Morphometry and Morphology
Isolated lungs were dissected and immersion fixed with gentle agitation in 100 ml of 0.1 M sodium cacodylate-buffered 2.5% glutaraldehyde at ambient temperature. Tissue slices were fixed overnight in fresh fixative, then rinsed three times for 30 min each in chilled 0.1 M cacodylate buffer. The lungs were then cut into 1-mm cubes and postfixed in aqueous 1.0% osmium tetroxide, en bloc stained with 0.5% aqueous uranyl acetate, dehydrated in acetone, and infiltrated and embedded in Polybed 812 (Polysciences). Thin sections were obtained on a Reichert-Jung Ultracut E ultramicrotome equipped with a diamond knife (Diatome), collected on 300-mesh copper grids, and poststained with lead citrate. They were examined with a JEOL 1210 transmission electron microscope (EM) at 60 kV.
Ten tissue blocks from each individual animal were thin sectioned as described in Tissue preparation. A ribbon of sectioned material was located under low magnification in the EM. The entire contents of an area outlined by a set of grid bars was recorded on Kodak 4489 electron-microscopic film by taking overlapping images at an initial magnification setting of ×2,000. Negatives were photographically enlarged to achieve a final magnification of ×6,000 and constructed into montages for visual analysis. Montages were analyzed for the presence of alveolar type I and type II cells, undifferentiated epithelial cells, endothelial cells, and interstitial cells.
Type II cell differentiation.
The size of the morphometric test lattice was determined previously on the basis of the size of the organelle of interest (mitochondria and secretory granules) and the frequency of its appearance in sectioned tissue (36, 37). A lattice consisting of intersections 1.0 cm apart was used in determining the volume percent of mitochondria and cytoplasmic granules. Lightly stippled cytoplasm was identified as glycogen. Volume densities for mitochondria, lamellar bodies, and percent glycogen (Vmt,l,g) are expressed as percent volumes and were calculated by dividing the number of points [grid intersections (Pi)] falling on a structure of interest by the total number of points (Pt) within that cell [(Vmt,l,g = Pi/Pt) × 100]. A ribbon of sectioned material was located under low magnification in the EM. As described inNuclear counts, the entire contents of an area outlined by a set of grid bars were recorded on Kodak 4489 electron-microscopic film by taking overlapping images at an initial magnification of ×2,000. Negatives were photographically enlarged to achieve a final magnification of ×19,500 and constructed into montages. Thirty cells were counted from each treatment group. Differences between volume densities in groups were evaluated by Student's t-test. Significance was assigned atP < 0.05. All morphometric analyses were done blinded by three independent investigators.
DNA was isolated from tissue with the commercially available method of Trisolv (Biotecx Laboratories, Houston, TX). The tissues were homogenized in the reagent, and chloroform was added, which yielded a top aqueous phase, an interphase, and a bottom organic phase. DNA was precipitated from the interphase and organic phase and dissolved in 8 mM NaOH after being washed. The DNA concentration was measured by spectrophotometry (5).
Statistical Comparison of Biochemical and Morphometric Analysis
t-Tests were applied for comparison of observations between homozygous normal animals treated in utero with thelacZ gene and homozygous normal animals treated in utero with the cftr gene. A P value < 0.05 was considered significant. Data are expressed as means ± SD.
Effects of In Utero cftr Gene Therapy on Surviving Population Statistics of Heterozygote-Heterozygote Matings
Uniform correction of a lethal autosomal recessive defect such as CF would result in a normal adult population distribution, with an expected Mendelian ratio of 1:2:1. Adult animals (75 days) from heterozygote-heterozygote matings treated in utero at 15–16 days gestation with the cftr gene were genotyped to evaluate the effect of transgene expression on genotype-specific survival. These data are presented in Table 1.
The population ratios of adult survivors (>75 days) after in utero treatment with the cftr gene varied significantly from expected (Table 1). Only 8% of the surviving adult population wascftr(+/+) (expected percentage of 25%; P < 0.001). The intrauterine treatment of fetuses with cftrresulted in twice as many surviving knockout mice as treated homozygous normal genotypes. A χ2 analysis of the two surviving populations found them to be significantly different (P< 0.001). After rescue of the 11 surviving knockout mice shown above, this laboratory has produced >30 additional knockout adults, including some that are offspring from knockout-heterozygote breeding. Thus transient expression of the cftr gene during development reversed the lethal phenotype in the cftr(−/−) mice but was lethal to the cftr(+/+) mice.
Mice dying during the perinatal period were genotyped when possible; many appeared to be cyanotic. The predominant genotype of mice that died within 72 h after birth and that had received in utero cftr gene treatment was cftr(+/+) (71%). This high perinatal mortality explained the decreased proportion of treated homozygous normal mice surviving to adulthood (Table 1). Only 9% of the deceased genotyped mice were homozygous knockouts [cftr(−/−)], and the remaining 21% were heterozygotes. Although perinatal mortality did occur in the control (Ad5.CMVlacZ-infected) mice, the genotype distribution of the deceased pups suggested a more random effect. The genotype of perinatal deceasedlacZ gene-treated mice was 23% homozygous normal, 73% heterozygous, and 4% knockout.
Comparison of survivors of litters that received adenovirus containing the lacZ gene in utero and litters that were born without surgical intervention showed that the surgical procedure itself had no impact on the homozygous normal perinatal mortality. A surviving adult population carrying a lethal autosomal recessive disorder consists of a normal population distribution of two-thirds heterozygous and one-third homozygous normal. Both untreated (no surgery) and controllacZ gene-treated populations had a higher percentage of homozygous normal animals than would be expected in a lethal autosomal recessive disorder (P < 0.05; Table 1). The homozygote-to-heterozygote ratio approached 1:1 in our colony and demonstrated a selection advantage for homozygous normal survival in this strain. Surgery did not change this proportion. Therefore, the large perinatal loss seen in the cftr(+/+) population treated in utero with the cftr gene was specifically due to the overexpression of the gene during development.
The proportion of rescued homozygous knockout animals was also lower than expected for the number of surviving heterozygotes (Table 1). The low percentage of cftr(−/−) animals in both the surviving mice and the mice that succumbed during the perinatal period indicated that some loss occurred prenatally and intrauterine rescue was not 100% efficient with this protocol. Although we could not label individual fetuses in utero, a number of fetuses were identified by inspection during injections as growth retarded. These litters subsequently produced confirmed knockout [(−/−)] mice. The in utero cftr gene-treated knockout mice were smaller at weaning and remained so throughout adulthood. These observations as well as the numerous resorbed fetuses observed during surgery suggested an earlier requirement for the cftr gene in the fetus that was not corrected by replacement of the gene at 15–16 days gestational age.
Somatic Growth Parameters of In Utero cftr Gene-Treated cftr(+/+) Mice
Previous work by this laboratory (28) with the reporter gene lacZ demonstrated that injection of adenovirus into the amniotic fluid targets the fetal intestine and the lung epithelium specifically. It would be expected that these would be the primary organs to reflect toxicity. In our studies of heterozygote-heterozygote matings, many of the cftr(+/+) animals treated in utero with the cftr gene had appeared cyanotic. This suggested that the cause of death was cardiorespiratory.
To evaluate possible toxic effects of in utero cftr gene therapy on the lungs and intestines, animals homozygous for the normalcftr allele were bred, and the fetuses were treated at 15–16 days gestation with either Ad5.CMVlacZ, serving as control, or Av1CF2. Only cftr(+/+) mice were used because these animals exhibited the greatest sensitivity to CFTR in utero toxicity. Treated mice were evaluated at 21 days gestation immediately before birth because of the high mortality in the perinatal period. Both the lung and intestine were evaluated.
Somatic and organ growth in the in utero cftr gene-treatedcftr(+/+) mice varied markedly from the in uterolacZ gene-treated cftr(+/+) mice. The body weights of the in utero cftr gene-treated cftr(+/+) mice were significantly decreased compared with those in the control animals (Table 2). Absolute lung weights were also significantly decreased. Despite the lung weight decrease, the lung weight-to-body weight ratios in the in utero cftrgene-treated cftr(+/+) mice were significantly increased. The intestine weight and intestine weight-to-body weight ratio were slightly increased in the in utero cftr gene-treatedcftr(+/+) mice, but the differences did not approach significance. These data suggested that lung growth was either spared or accelerated after in utero cftr gene therapy in thecftr(+/+) mice, and their somatic growth suffered.
DNA Content of Lungs and Intestines After In Utero cftr Gene Treatment
DNA content was measured in the lung and intestine as an indicator of cell number because these two characteristics have been shown to be directly related to each other (40, 41). There was a significant increase in the amount of DNA per milligram of lung weight of the in utero cftr gene-treatedcftr(+/+) mice (Table3). These findings agreed with the increase in lung weight-to-body weight ratio in the same group and suggested that the increase in the lung weight-to-body weight ratio was due to increased cellularity in these lungs. Thus the treated mice did not have pulmonary hypoplasia as defined by a decrease in organ size due to a decreased cell number (7). In fact, the increase in lung weight-to-body weight ratio and increased DNA were consistent with a hyperplastic lung.
The DNA content in the intestines between the two groups showed no significant difference, and this corresponded to the unchanged intestine weight between the two groups. The intestine DNA content-to-body weight reached significance because somatic growth was so substantially decreased in the in utero cftr gene-treatedcftr(+/+) animals.
Volume Proportion of Lung Structure
Morphometric analysis was used for statistical evaluation of lung structure. When growth alterations such as hyperplasia occur, morphometry can delineate the specific changes in different tissue compartments of the lung. The percentage of saccular air in the animals treated with cftr was significantly decreased, and the percentage of saccular wall was significantly increased (Table4). Normally, as the lung develops, the volume density of air increases and the mesenchyme and epithelia become more attenuated. The epithelial proliferation that occurred as a result of the cftr gene treatment did not allow this attenuation to happen.
The volume proportion of the remaining lung structures (airways and vessels) was unchanged between the two groups (Table 4). At the time of in utero cftr therapy, the large conducting airways were already formed and differentiation of the proximal portions of the airways had begun. The most rapid proliferation was occurring in the developing respiratory part of the lung, i.e., the distal tubules that would develop into saccules and eventually alveoli. The morphometric analysis established that the cftr gene had its most discernible effect on the cells that were rapidly proliferating at the time of infection and not on those that were already undergoing differentiation. The overexpression of the cftr gene in utero resulted in increased cellularity in the parenchyma (air-exchanging area of lung) but in decreased complexity. Although not hypoplastic by the criteria of cell number, lung development was defective and met the definition of structural hypoplasia (7).
Evaluation of Parenchymal Cell Type
A nuclear count was obtained as an index of the number of specific cell types in the proliferating air-exchanging parenchyma (2). This ultrastructural morphometric analysis was done because at this stage of gestation, many cell types share cellular markers, making histochemical and immunohistochemical techniques of little value. The parenchyma was analyzed for alveolar type I, type II, undifferentiated, endothelial, and interstitial cells. These results are summarized in Table 5. In the control (LacZ gene-treated) lungs, 22% of the cells had large pools of lightly stippled material identified as glycogen in their cytoplasm, had no other differentiating features, and were defined as undifferentiated. In the treated animals, there were no glycogen-containing cells in the parenchyma (P < 0.001). Although the proportion of type II cells remained the same within the two groups (21.2 and 20.1%, respectively), there was an increase in type I cells from 10.1–20.1% in the in uterocftr-treated animals (P = 0.004). Together, the type I and type II cells comprise the epithelial surface of the lung parenchyma. Type I cells are thought to be the terminally differentiated cells of the epithelium, and type II cells are their precursors (1). This suggested that the cftrgene played a role in differentiation as well as in proliferation of the lung epithelium. There was also an increase in interstitial cells in the treated animals, again indicating a role in proliferation. There was no difference in the proportion of vascular endothelial cells in the groups.
Evaluation of Epithelial Cell Maturity
To substantiate the role of the cftr gene in epithelial cell differentiation, the type II cells were examined by ultrastructure. Because the type II cells are the primary secretory cells in the parenchymal epithelium and produce lung surfactant, their differentiation has been examined in detail. With differentiation, the volume proportion of lamellar bodies (containing surfactant) and mitochondria increases and the volume proportion of glycogen decreases (2, 32, 43). The results of the ultrastructural morphometric analysis are described in Table6. The volume proportion of lamellar bodies increased from 5% in the control mice to 9.5% in the in utero cftr-treated animals (P = 0.008). Glycogen, thought to be a substrate for future surfactant production and an indicator of an immature cell, represented 12% of the volume in the control cells and none of the volume in the in uterocftr gene-treated cells (P < 0.001). Figure1, A and B, are type II cells from control animals (lacZ gene treated) showing large pools of glycogen and a paucity of lamellar bodies. Figure 1, C and D, are type II cells from in utero cftr gene-treated animals demonstrating the lack of glycogen and an increase in lamellar body volume proportion. These data confirmed that the cftr gene affects differentiation of the secretory epithelium in addition to proliferation.
The effects of overexpression of CFTR in utero on epithelial cell proliferation were readily apparent on histological examination. Figure2 A demonstrates the normal saccular formation of a homozygous wild-type mouse at 21 days gestation after in utero treatment with the lacZ gene. The lung of a homozygous wild-type mouse treated in utero with the cftrgene (Fig. 2 B) exhibited the hypercellularity documented by the biochemical and morphometric analyses above. The proliferating cells filled the saccular spaces and impeded adequate air exchange. This unchecked proliferation of the epithelium inhibited normal lung development and caused structural hypoplasia. This lung pathology was the most likely cause of the high perinatal mortality in the in utero cftr gene-treated cftr(+/+) mice.
Previous studies (reviewed in Ref. 18) of the pathogenesis of CF have alluded to the lack of correlation between disease pathology and the pattern of CFTR expression in the respiratory tissues. It has been observed that CF pathology starts in the small airways where there is little CFTR expression. CFTR is present, however, in the distal airway epithelium during lung development, and expression follows temporal and tissue-specific patterns. Lung development begins proximally with the formation of the large conducting airways and proliferation of the epithelium, which progresses distally. Once the proliferation has diminished, differentiation occurs in this same cephalocaudal pattern (15, 33). In developing airways, CFTR distribution follows this same cephalocaudal pattern of maturation and differentiation (9). The cellular distribution of CFTR protein and mRNA changes with the degree of differentiation of the epithelial cell. In more immature lungs, CFTR is not polarized to one domain of the cell but is diffusely expressed. As the lung matures and the cells become differentiated, the localization of CFTR mRNA and protein shifts to a more apical distribution and bears a similarity to that seen in the adult lung (9,23). This suggests that the role of CFTR changes with the differentiated state of the cell.
At the time of infection, these lungs consisted of blind airways lined with multipotential columnar and cuboidal stem cells (33). The airway cells were the primary targets of the adenovirus because the respiratory portion of the lung didn't yet exist (28). Yet the volume proportion of the lungs indicated that most of the cellular proliferation from in utero treatment with the cftrgene occurred in the respiratory portion of the lung. These data indicate that the cftr gene works in trans, altering the cellular environment. This is also evidence that CFTR affects proliferation of predominantly undifferentiated cell types. The airways showed mild cellular proliferation as evidenced by a decrease in airway lumen. However, the volume proportion of airway was unchanged between the two groups because airway structural development had occurred before infection. The upper airway cells were beginning to differentiate at the time of infection and were largely unaffected by CFTR. In contrast, the largest effects of the gene were in the respiratory epithelium, which was undifferentiated and proliferating at the time of infection.
These data explain the findings of Whitsett et al. (39), who used the cftr gene under the direction of the surfactant protein (SP) C and 10-kDa Clara cell (CC10) promoters in transgenic mice and evaluated the effects of expression in utero. These investigators found that somatic growth and lung development of the CFTR transgenic mice were identical to that of their nontransgenic littermates. The use of the SP-C and CC10 promotors ensured CFTR expression in more differentiated cell types because each of these proteins become specific markers of their respective cell types (type II and Clara cell) as they differentiate (25,29, 42). Although expression of chimeric genes with the SP-C promoter was observed as early as 11 days gestation in the mouse (10, 39), rodent fetal lungs at 15 days gestation do not demonstrate any detectable messages for SP-C (14). A dramatic increase in SP-C expression occurs as gestation progresses (14, 39). Likewise, a progressive increase in CC10 protein is apparent only from day 18 gestation in the rodent. The most marked increase occurs from birth to 1 wk of age (4, 29).
Whitsett et al. (39) acknowledged that CFTR expression in their transgenic mice was diffuse and not confined to the membrane surface. They stated that the lack of toxicity observed may have resulted from failure of the respiratory cells to insert or activate excess CFTR protein. Our studies suggest that their findings of unchanged growth, differentiation, or function were most likely due to the timing of the their hybrid cftr gene expression. Under the direction of the SP-C and CC10 promoters, the exogenous CFTR was largely expressed in the transgenic lineages after epithelial cell differentiation. Expression of CFTR in differentiated cells produced none of the effects described in our studies.
In this study, we examined the changes in the lung parenchymal cell types (type I and type II) because their proliferation and differentiation adversely affected survival. Cohen et al. (6) have previously shown that CFTR affected Clara cell differentiation in the terminal airways of the mouse lung. The effect of CFTR on divergent cell types becomes clearer with the understanding that these cells are thought to share a common progenitor. The Clara and type II cell lineages share many common markers until late gestation, and some functions, such as SP-A expression, are shared throughout their lifespans. At the time of infection, these cell lineages were largely undifferentiated; thus their common progenitor is the target of cftr gene transfer.
The increase in proliferation and decrease in complexity seen in the in utero cftr gene-treated mice are consistent with altered differentiation of the parenchyma in these lungs. The prevailing hypothesis in the pathogenesis of CF is that loss of the CFTR-encoded chloride channel results in altered salt and fluid concentrations and altered barrier immunity. Yet no one has been able to adequately restore this function with either gene replacement or any other effector molecule. In contrast, we have been able to permanently reverse the lethal phenotype in the cftr(−/−) mouse after in utero gene therapy as well as demonstrate permanent developmental changes in the lung and intestine. It has been suggested that the CF knockout mouse is not a good model of human disease because it does not develop overt lung disease. However, CF knockout mice have been shown to have impaired ability to control lung infection, increased inflammatory cell recruitment, and defective mucociliary clearance (8, 11, 16). These characteristics are similar to those found in human lungs early in the course of their disease. In infants discovered by early screening programs, studies have determined that airway inflammation is present as early as 4 wk of age (17) and before clinically apparent disease (18). In addition, fetuses examined after a prenatal diagnosis of CF have been shown to have tracheal epithelium atrophy with cells devoid of cilia (24). These data would suggest that early functional changes are present in the lungs of CF patients despite the lack of observable pathology. In the above studies, we quantified the extent of proliferation and differentiation in the secretory cells of the lung parenchyma after in utero transfer of the cftr gene. These studies further define the regulatory role of the cftr gene in the development of the secretory epithelium.
We are indebted to Carol L. Thouron and Michele F. St. Onge for excellent technical assistance with preparation of tissues, to Cathy C. Vial for excellent assistance in preparation of tissues for electron microscopy, and to Michael L. Johnson for assistance with ultrastructural morphometry.
Address for reprint requests and other correspondence: J. E. Larson, Laboratory of Molecular Genetics, Ochsner Medical Foundation, 1516 Jefferson Highway, New Orleans, LA 70121 ().
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society