Polycystin-1 transforms the cAMP growth-responsive phenotype of M-1 cells

BACKGROUND

Polycystic kidney disease (PKD) is characterized by the abnormal proliferation of tubular epithelial cells. It was recently shown that the growth of PKD cyst-lining cells is stimulated by cyclic adenosine monophosphate (cAMP), whereas the growth of normal human kidney cortex cells is inhibited.

METHODS

We have examined the effects of overexpressing the C-terminal cytosolic tail of mouse polycystin-1, as a membrane-targeted fusion protein, on cAMP-responsive cell proliferation in stably transfected M-1 cortical collecting duct cells. Two cell lines that express high levels of the polycystin-1 fusion protein and two control cell lines that do not express the fusion protein were tested.

RESULTS

Growth of parental M-1 cells and the control cell lines was inhibited by 8-Br-cAMP and by a variety of cAMP agonists. In contrast, growth of the polycystin-1-expressing clones was stimulated by cAMP. Consistent with this, the protein kinase A (PKA) inhibitor H-89 caused either a positive or a negative growth effect depending on the primary response to cAMP. PD98059 blocked the cAMP stimulation of cell proliferation, indicating that the pathway is MEK1 dependent.

CONCLUSIONS

Expression of the polycystin-1 C-terminal tail disrupts normal cellular signaling and transforms the stably transfected M-1 cells to an abnormal PKD cell proliferation phenotype.

The genetics and physiology of polycystic kidney disease

Autosomal dominant polycystic kidney disease (ADPKD) is a major, inherited disorder that is characterized by the growth of large, fluid-filled cysts from the tubules and collecting ducts of affected kidneys, and by a number of extrarenal manifestations including liver and pancreatic cysts, hypertension, heart valve defects, and cerebral and aortic aneurysms. Mutations in either of 2 different genes (PKD1 or PKD2) give rise to ADPKD. Most mutations identified in affected families appear to inactivate the PKD genes, and accumulating evidence suggests that a 2-hit mechanism, in which the normal PKD1 or PKD2 allele is also mutated, may be required for cyst growth. The protein products of the PKD genes (polycystin-1 and polycystin-2) are thought to function together as part of a multiprotein membrane-spanning complex involved in cell-cell or cell-matrix interactions. Polycystin-1 and polycystin-2 can initiate signal transduction, leading to the activation of a number of downstream effectors, including heterotrimeric G-proteins, protein kinase C, mitogen-activated protein kinases, beta-catenin, and the AP-1 transcription factor. In addition, polycystin-2 may function in mediating calcium flux. The pathogenesis of cyst formation is currently thought to involve increased cell proliferation, fluid accumulation, and basement membrane remodeling. It now appears that cyclic adenosine monophosphate (cAMP) metabolism is a central component of cyst formation, stimulating apical chloride secretion and driving the accumulation of cyst fluid. Recent evidence has shown that ADPKD cells also have an altered responsiveness to cyclic AMP. In contrast to normal kidney cells whose cell proliferation is inhibited by cyclic AMP, ADPKD cells are stimulated to proliferate. Thus, it is likely that an alteration in polycystin function transforms the normal cellular phenotype to one that responds to elevated cyclic AMP by an increased rate of cell proliferation and that the enlarging cyst expands by an increased rate of cyclic AMP-driven fluid secretion. Cyclic AMP and growth factors, including epidermal growth factor, have complementary effects to accelerate the enlargement of ADPKD cysts, and thereby to contribute to the progression of the disease. This knowledge should facilitate the discovery of inhibitors of signal transduction cascades that can be used in the treatment of ADPKD.

Identification of the major site of in vitro PKA phosphorylation in the polycystin-1 C-terminal cytosolic domain

Sequence analysis of the C-terminal cytosolic domain of human and mouse polycystin-1 has identified three RxS consensus protein kinase A (PKA) phosphorylation motifs. GST-fusion proteins containing the full-length and truncated C-terminal cytosolic domain of murine polycystin-1 were phosphorylated in vitro by the purified catalytic subunit of PKA. This identified a sequence of 25 amino acids, immediately downstream of a previously identified heterotrimeric G-protein activation sequence, as the major site of PKA phosphorylation. Phosphorylation of wild-type and alanine substituted synthetic peptides containing this motif demonstrated that alanine substitution of serine 4159 largely eliminated phosphorylation. Mutation of this residue in the fusion protein reduced phosphorylation by about 70%, whereas mutation of the other two conserved phosphorylation motifs had little effect. We conclude that serine 4159 is the major site of PKA phosphorylation in the C-terminal cytosolic domain of murine polycystin-1.

Matrix metalloproteinase-2 in a murine model of infantile-type polycystic kidney disease

It was previously found that elevated levels of matrix metalloproteinase (MMP)-2 (gelatinase A) and -9 (gelatinase B) were synthesized and secreted into the medium by cultured kidney tubules derived from cystic C57BL/6J-cpk mice. To determine whether increased synthesis and secretion occur in vivo in this mouse model of polycystic kidney disease, kidney protein extracts, mRNA, and tissue sections were compared for expression and activity of MMP-2 and -9. Although both MMP were detected in tissue extracts, the differences in expression levels and activity in normal and cystic kidneys were far greater for MMP-2. High levels of MMP-2 seemed to result from increased expression by the cystic kidneys predominantly in the second and third postnatal weeks (a time when the kidneys are undergoing rapid cystic enlargement). Much of the increased MMP was present in the inactive zymogen form, although active enzyme was readily detected by sodium dodecyl sulfate-polyacrylamide gel zymography and in situ zymography. MMP-2 was abnormally localized to the interstitium and to foci between cysts, suggesting that MMP-2 may regulate collagen accumulation at those sites, thus allowing cyst enlargement and limiting the severity of interstitial fibrosis.

Kid-1 expression is high in differentiated renal proximal tubule cells and suppressed in cyst epithelia

The cDNA coding for the transcriptional repressor protein Kid-1 was cloned in a screen for zinc finger proteins, which are regulated during renal development and after renal ischemia. Kid-1 mRNA levels increase in the course of postnatal renal development and decrease after acute renal injury caused by ischemia or administration of folic acid. We have raised a monoclonal anti-Kid-1 antibody and demonstrate that the Kid-1 protein is strongly expressed in the proximal tubule of the adult rat kidney. During nephron development, the Kid-1 protein appears after the S-shaped body stage concomitantly with the brush-border enzyme alkaline phosphatase. In two animal models of polycystic kidney disease, the expression of Kid-1 is downregulated. The loss of expression of Kid-1 in cyst wall cells correlates with the loss of alkaline phosphatase histochemical staining. Kid-1 mRNA levels are also reduced in rodent renal cell carcinomas, another condition characterized by epithelial cell dedifferentiation and increased proliferation. We propose that Kid-1 plays an important role during the differentiation of the proximal tubule.

The polycystic kidney disease-1 protein, polycystin-1, binds and activates heterotrimeric G-proteins in vitro

Analysis of the C-terminal cytosolic domain of human and mouse polycystin-1 has identified a number of conserved protein motifs, including a 20-amino-acid heterotrimeric G-protein activation sequence. A peptide specific for this sequence was synthesized and shown to activate purified bovine brain heterotrimeric Gi/Go in vitro. To test whether the C-terminal cytosolic domain of polycystin-1 stably binds G-proteins, GST-fusion constructs were used in pull-down and co-immunoprecipitation assays with purified bovine brain Gi/Go and rat brain lysates. This identified a 74-amino-acid minimal binding domain that includes the G-protein activation sequence. This region of polycystin-1, including the G-protein activation peptide and flanking amino acid sequences, is highly conserved in mouse, human, and puffer fish, lending further support to the functional importance of the minimal binding domain. These results suggest that polycystin-1 may function as a heterotrimeric G-protein coupled receptor.

Molecular genetics of polycystic kidney disease

Polycystic kidney disease (PKD) is a major, multigene disorder that is characterized by the growth of large, fluid-filled cysts from the nephrons and collecting ducts of affected kidneys. At least three different genes are thought to give rise to the autosomal dominant (ADPKD) form of PKD. Two of the genes (PKD1 and PKD2) have now been isolated and sequenced, and based on their predicted structures are thought to encode proteins (polycystin-1 and polycystin-2) that function together as part of a multi-component membrane-spanning complex involved in cell-cell or cell-matrix interactions. Most mutations identified in affected families appear to be inactivating for the PKD1 and PKD2 genes, and recent evidence has suggested that a two-hit mechanism, in which the normal PKD1 allele is also inactivated, may be required for cyst growth. The large number of genes showing abnormal expression in cystic kidneys in humans and rodent models suggests that cellular processes associated with signal transduction, transcriptional regulation, and cell-cycle control are involved in cyst formation and that the cellular defect in PKD directly affects the regulation of epithelial differentiation.

Matrix metalloproteinases and TIMPS in cultured C57BL/6J-cpk kidney tubules

Restructuring of basement membranes is a hallmark of the pathology of renal cystic disorders. Here, we present findings consistent with the view that basement membrane degradation by matrix metallo-proteinases (MMPs) may contribute to abnormal basement membrane structure in polycystic kidney disease. Cells from cystic kidney tubules embedded in collagen gels appeared to migrate through the gel. This migration through collagen indicated that these cells could degrade the matrix. To examine this activity, we cultured cystic kidney tubules derived from the C57BL/6J cpk/cpk mouse, a hereditary model of polycystic kidney disease, and assayed conditioned medium for the presence of MMPs and tissue inhibitors of metalloproteinases (TIMPs). The conditioned medium from the cystic tubules contained higher than normal levels of MMP-9, MMP-2, and MMP-3 as well as TIMP-1 and TIMP-2. A 101 kDa protease was present equally in cystic and control cultures and although inhibited by EDTA, it was not inhibited by TIMPs, nor activated by the mercurial compound APMA. These data suggest that cystic kidney tubules synthesize and secrete high levels of MMPs which may then participate in the restructuring of the tubular basement membrane.

The cystic fibrosis transmembrane conductance regulator mediates transepithelial fluid secretion by human autosomal dominant polycystic kidney disease epithelium in vitro

Transepithelial fluid secretion promotes the progressive enlargement of cysts in autosomal dominant polycystic kidney disease (ADPKD). Recent indirect evidence indicated that active chloride transport may drive net fluid secretion in cultures of epithelia derived from ADPKD cysts. We now report that forskolin, which stimulates adenylate cyclase, increased the efflux rate constant for 36Cl in monolayers of ADPKD cells in vitro from 0.23 +/- 0.02 min-1 to 0.44 +/- 0.05 min-1 (N = 4) and that diphenylamine 2-carboxylate (DPC), which blocks chloride channels, eliminated the forskolin-stimulated chloride efflux from these cells. To establish whether the cAMP-regulated chloride transporter, cystic fibrosis transmembrane conductance regulator (CFTR), may potentially be involved in the chloride transport and fluid secretion of ADPKD epithelia, we examined CFTR mRNA and protein in these cultures. Northern blot hybridization using a human (h) CFTR cDNA probe demonstrated the presence of an approximately 6.5 kb transcript in total RNA from polarized cultures of ADPKD, normal human kidney cortex (HKC), and T84 cells. Utilizing several antibodies to hCFTR, immunocytochemistry and confocal fluorescence microscopy localized an immunoreactive protein primarily in the apical region of forskolin-stimulated ADPKD cells grown on permeable supports. This immunoreactivity could be eliminated by preincubation of antibody with immunizing peptide. To determine the effect of CFTR abundance on the magnitude of net fluid secretion, polarized ADPKD cultures were treated with deoxyoligonucleotides that were either complementary (antisense), homologous (sense), or partially complementary (misantisense) to a sequence near the translation initiation site in hCFTR mRNA. Treatment with 5.0 microM antisense oligonucleotide resulted in a 73% reduction in forskolin-stimulated fluid secretion and a comparable reduction in the abundance of CFTR as detected by immunocytochemistry. By contrast, treatment with 5.0 microM sense oligonucleotide reduced fluid secretion by only 34% and had less of an effect on CFTR abundance, while the effects of 5.0 microM misantisense oligonucleotide on both fluid secretion and CFTR abundance were insignificant. On the basis of these results we suggest that CFTR is a major mediator of forskolin-stimulated chloride and fluid secretion by epithelial cells of human polycystic kidneys in vitro.

A mouse kidney- and liver-expressed cDNA having homology with a prokaryotic parathion hydrolase (phosphotriesterase)-encoding gene: abnormal expression in injured and polycystic kidneys

To investigate abnormalities in gene expression associated with cyst formation in polycystic kidney disease, differential cDNA library screening was carried out using RNA from normal and cystic kidneys of the C57BL/6J-cpk mouse. Among a number of genes found to be abnormally expressed was one (cDNA clone 56-1) that was significantly underexpressed in cystic kidneys. Hybridization analyses revealed that the 56-1 mRNA is expressed primarily in kidney and liver, and that the kidney expression begins postnatally and continues in the adult. Expression of this mRNA was found to be significantly decreased upon acute renal injury induced by a single intraperitoneal injection of folic acid, and to return to normal levels upon recovery of kidney function. Analysis of the cDNA sequence predicted a protein of 349 amino acids (aa), which was confirmed by in vitro translation of a sense-strand transcript, producing a protein of approx. 39 kDa. The aa sequence shows similarity to Flavobacterium sp. and Pseudomonas diminuta parathion hydrolase (phosphotriesterase or PTE), an enzyme that hydrolyzes toxic organophosphates and other phosphotriesters, and to the predicted product of an Escherichia coli open reading frame of unknown function (phosphotriesterase homology protein or PHP). Use of optimal alignment programs demonstrated a significant overall homology between the bacterial and mouse sequences, with greater than 50% aa sequence similarity. This cDNA represents the first eukaryotic sequence showing similarity to these prokaryotic genes. Based on this apparent homology, it has been named mpr56-1 (for mouse phosphotriesterase-related 56-1).

Growth characteristics of cells cultured from two murine models of polycystic kidney disease

Polycystic kidney disease (PKD) is characterized by multiple renal cysts that are lined by epithelium and filled with fluid. PKD may result from one of a number of factors, either inherited or environmental. In this study, we have compared two mouse models in which PKD results from a genetic cause. In the C57BL/6J-cpk model, the mutated gene is unknown. In the other model, an SV40 large T antigen transgene causes renal cysts. We examined cultured cells from the kidneys of these mouse models, comparing growth characteristics. Although several features of PKD lead one to expect that the epithelial cells lining the cysts would have an increased rate of proliferation in culture, we found that they did not. The implications of these findings are discussed.

Renal expression of a transforming growth factor-alpha transgene accelerates the progression of inherited, slowly progressive polycystic kidney disease in the mouse

Polycystic kidney disease (PKD) is a prevalent inherited disease in human beings. The pathogenesis of PKD is as yet unclear. The epidermal growth factor family of proteins has been implicated in PKD based largely on in vitro data. To determine whether these growth factors contribute to the progression of inherited PKD in vivo, we crossed mice with a transgene for human transforming growth factor-alpha (TGF-alpha, a member of the epidermal growth factor (EGF) family) and mice with the pcy gene (which causes a slowly progressive form of PKD very similar to human autosomal dominant PKD). Renal expression of the TGF-alpha transgene in cystic mice (homozygous for the pcy gene) accelerated the development of PKD as shown by an increased kidney weight as a percent of body weight and an increased volume density of renal cysts at 8.5 weeks of age. However, renal expression of the TGF-alpha transgene did not appear to precociously initiate cyst development (at 6.5 weeks), nor did it cause an increase in the final degree of renal enlargement (at 29 weeks). Thus TGF-alpha accelerated the enlargement of cysts once initiated. At 8.5 weeks of age, renal expression of the TGF-alpha mRNA correlated positively with the amount of renal enlargement. At all time points studied, cystic kidneys exhibited increased expression of c-myc mRNA as compared with phenotypic normal kidneys, consistent with PKD being a hyperplastic disease of renal tubules. However, the renal expression of c-myc in 8.5 week cystic kidneys, with or without the transgene, did not correlate with the degree of renal enlargement. The results of this study suggest that EGF-like proteins may accelerate the progression of inherited renal cystic disease. However, the final degree of cystic change is dictated by the primary disease process rather than by the continued presence of growth factor.

Analysis of differential gene expression in the kidney by differential cDNA screening, subtractive cloning, and mRNA differential display

It is becoming increasingly evident that significant changes in gene expression occur during the course of disease progression in both genetic and nongenetic kidney diseases. Knowledge of the differentially expressed genes may yield information about the abnormal biochemical events that occur in the initiation and pathogenesis of these diseases. The purpose of this review is to provide an overview of some of the current approaches for identifying and analyzing differentially expressed genes. The power of these techniques lies in their their ability to detect differences in the levels of specific mRNAs in the diseased compared to the nondiseased kidney without prior knowledge of their identity. The three basic techniques considered are differential cDNA library screening, subtracted cDNA libraries, and PCR-based differential display. Emphasis is placed on cDNA library construction and differential screening. Also reviewed are the analysis of differentially expressed cDNAs by Southern and Northern blot hybridization, S1-protection, RT-PCR, DNA sequencing, and DNA sequence analysis.

Primary and secondary genetic responses after folic acid-induced acute renal injury in the mouse

Folic acid-induced acute renal injury results in dramatic changes in gene expression. Among the genes affected by folic acid treatment are the primary response genes, c-fos and c-myc, which are thought to function to initiate cell cycle events. In this report, changes in the expression of three other genes in response to folic acid injury have been investigated: ornithine decarboxylase, epidermal growth factor (EGF), and sulfated glycoprotein-2 (SGP-2). Renal injury was found to cause a rapid decrease in EGF mRNA, which remained absent for several days after the initial injury, gradually returning to normal levels over an approximately 3-wk regeneration and recovery period. Ornithine decarboxylase mRNA showed a similar decrease. In contrast, folic acid caused a rapid increase in SGP-2 mRNA, which peaked several days after treatment, decreasing to normal levels over the 3-wk period. The mRNAs for the primary response genes were superinduced in the injured kidneys in the presence of the protein synthesis inhibitor cycloheximide. In contrast, the changes in EGF and SGP-2 mRNA levels were blocked by cycloheximide, indicating that these responses required new protein synthesis during the first few hours after folic acid injury. The opposite but parallel responses in the expression of the EGF and SGP-2 genes suggest that their regulation is coupled to the initial injury-induced dedifferentiation and subsequent return to the fully differentiated state.

Mouse plasma glutathione peroxidase. cDNA sequence analysis and renal proximal tubular expression and secretion

A mouse kidney cDNA isolated by differential screening was found to be highly homologous to rat, human, and bovine plasma glutathione peroxidase (GPx) sequences. Analysis of the full-length coding region sequence demonstrated an in-frame selenocysteine-encoding opal codon and putative signal sequence, suggesting that the sequence represents the mouse homolog of plasma GPx. The level of expression of plasma GPx in various mouse tissues and during development was investigated by Northern blot analysis. Plasma GPx mRNA was observed to be very abundant in kidney compared with placenta, epididymis, intestine, lung, heart, testis, ovary, salivary gland, spleen, thymus, stomach, brain, and fetal kidney and could not be detected in pancreas or in liver except from pregnant mice. In addition, plasma GPx mRNA levels were shown to increase during postnatal development of the kidney. In situ hybridization localized plasma GPx mRNA to proximal tubules, while primary cell culture demonstrated that plasma GPx is synthesized and secreted by proximal tubular epithelial cells. The relative abundance of plasma GPx mRNA in mouse kidney suggests that proximal tubules may be the primary source of the enzyme detectable in plasma and further suggests that plasma GPx has an important function in protecting the kidney from oxidative damage.

Renal cysts in transgenic mice expressing transforming growth factor-alpha

Transforming growth factor-alpha (TGF-alpha) is a member of the epidermal growth factor (EGF) family of proteins and, like EGF, elicits its cellular function by binding to the EGF receptor. EGF stimulation may have a role in several normal and pathologic processes in the kidney, and EGF has been implicated in the development of renal cysts in vitro and in human autosomal dominant polycystic kidney disease. We sought to determine whether renal expression of an EGF-like protein (TGF-alpha) could lead to the formation of renal cysts in vivo. We examined morphologic alterations to the normal kidney caused by renal expression of a TGF-alpha transgene linked to a mouse metallothionein promoter stably integrated into the genome of the CD1 mouse. TGF-alpha transgene expression was induced with exogenous zinc treatment starting at 4 weeks of age, and mice were killed at 8 weeks of age. The transgene was expressed at higher levels in female transgenic mice than in male transgenic mice. The augmented expression of the TGF-alpha transgene in females was associated with increased renal size and the development of renal epithelial cysts. Both male and female mice exhibited increases in glomerular size and mesangial volume density. These results provide evidence that stimulation by an endogenous EGF-like protein can lead to renal enlargement, glomerular mesangial expansion, and renal cyst formation.

Injury and development in polycystic kidney disease

The large diversity of genetic and nongenetic causes of polycystic kidney disease begs the question of whether there is a common denominator in the process of cyst formation. Accumulating evidence now indicates that cysts are made up of immature epithelium, which suggest that cystic transformation may be caused by an arrested developmental process that locks cells in a unique differentiated state. Although there may be any number of possible causes of this cystic change, it is proposed that one unifying mechanism may be subcritical injury-induced cellular dedifferentiation.

Autosomal-dominant polycystic kidney disease in the rat

Kaspareit-Rittinghausen described a rodent model of inherited polycystic kidney disease (PKD), the Han:SPRD rat [1, 2], in which heterozygotes develop renal cysts and renal failure (in males) over several months, whereas homozygous animals develop rapidly progressive renal enlargement that leads to death in a few weeks. In this study, we examined selected elements of the pathogenesis of this disease in heterozygotes and homozygotes from birth to advanced disease. Heterozygous male rats developed slowly progressive renal cystic disease with interstitial fibrosis and azotemia seen by six months of age. Female heterozygotes developed slowly progressive renal cystic disease, but did not develop interstitial fibrosis or azotemia. Epithelial cells lining cyst cavities showed various degrees of morphologic immaturity. Cyst walls also developed basement membrane thickening, especially in areas of cellular immaturity, suggesting an interrelationship between this basement membrane thickening and cellular dedifferentiation. Thickened basement membranes were associated with increased immunoreactivity for type IV collagen, laminin, and fibronectin. Homozygous rats developed massive renal enlargement, marked azotemia, and died near three weeks of age. Renal c-myc proto-oncogene expression was elevated in homozygous cystic infants and in adult heterozygotes. In situ hybridization showed high levels of c-myc mRNA in cyst epithelia, suggesting abnormal regulation of cellular proliferation in the cells lining cysts, as seen in other models of PKD. The Han:SPRD rat is the only well-documented animal model of inherited PKD with an autosomal-dominant inheritance pattern and appears to have several features which resemble human ADPKD.

Polycystic kidney disease: primary extracellular matrix abnormality or defective cellular differentiation?

Polycystic kidney disease (PKD) is inherited as a dominant or recessive trait or can be provoked by environmental factors. The disease is characterized by the growth of large epithelial-lined cysts derived from the nephrons and collecting ducts of affected kidneys. Cysts are thought to initiate as small dilations in renal tubules, which then expand into fluid-filled cavities of relatively large size. Cyst formation appears to involve increased cell proliferation, reversal of tubular epithelial polarity, and epithelial fluid secretion. In addition, a number of pronounced extracellular matrix changes have been found in the cystic kidneys of several animal models and in human autosomal dominant PKD. These abnormalities include thickened, laminated basement membrane, increased expression of alpha 1 type IV collagen and laminins B1 and B2, and changes in heparan sulfate proteoglycan and fibronectin. Some of these changes can also be seen in vitro, reflecting intrinsic abnormalities, and may be associated with abnormal tubular morphogenesis early in cyst formation as well as later in cyst expansion. We have been investigating gene expression in the C57BL/6J-cpk mouse, which has an autosomal recessive form of PKD, to determine the genetic basis of the abnormal tubule cell growth and morphology manifested during cyst formation.(ABSTRACT TRUNCATED AT 250 WORDS)

C-fos expression is hypersensitive to serum-stimulation in cultured cystic kidney cells from the C57BL/6J-cpk mouse

Cystic kidneys of the C57BL/6J-cpk murine model of polycystic kidney disease show a marked overexpression of the proto-oncogenes c-fos, c-myc, and c-Ki-ras, consistent with an increased rate of cell proliferation and an altered state of differentiation. To determine if cystic cells have increased responsiveness to stimulation with mitogenic agents, quiescent primary cultures from normal and cystic cpk kidneys were treated with fetal bovine serum (FBS), 8-bromo-cAMP (cAMP), or epidermal growth factor (EGF). The level of c-fos induction following stimulation by FBS was found to be dramatically higher in cystic cells than in normal cells; whereas induction by cAMP or EGF was essentially the same in both cell types and much less than that seen in FBS-stimulated cells. To determine if this serum hypersensitivity reflects an increased proliferative state in vivo, c-fos induction was examined in cultures derived from normal kidneys stimulated to regenerate by folic acid-induced acute renal injury. As with cystic kidneys, the folic acid-injured kidneys showed increased c-fos responsiveness to FBS in cell culture. These experiments suggest that cystic and regenerating kidneys have an altered phenotypic state in vivo that is manifested in cell culture by serum hypersensitivity. However, whereas the folic acid-injured kidneys ultimately reestablish normal kidney function, cystic kidneys further progress to renal failure, suggesting that cystic epithelial cells are locked in this altered state of differentiation.

Localization of overexpressed c-myc mRNA in polycystic kidneys of the cpk mouse

The C57BL/6J-cpk mouse has a form of autosomal-recessive polycystic kidney disease characterized by the rapid growth of large collecting duct cysts and the development of severe renal failure usually by three to four weeks of age. Previous studies had shown higher steady-state levels of proto-oncogene mRNA in these cystic kidneys. It is now shown using nuclear run-on transcription that the c-fos and c-myc proto-oncogenes are transcribed at higher rates in cystic kidneys, and thus that increased transcription, in part, may account for the increased mRNA levels. c-myc mRNA was detected by in situ hybridization in nephron anlagen and elongating tubules of normal and cystic kidneys during late fetal and early neonatal kidney development. Localization of c-myc expression in the normal kidney decreased with age over the three-week postnatal period. By contrast, c-myc mRNA was found in cysts as early as three days of age, with increased levels at two and three weeks. c-myc expression was also elevated in apparently normal, non-dividing proximal tubules in three-week-old cystic animals. On the basis of these findings, we suggest that c-myc expression is linked to the proliferation of cells engaged in the primary cystogenic process, and that expression of this gene in proximal tubule cells of severely azotemic animals reflects the compensatory response of residual tubular epithelial cells to progressive renal dysfunction.

Molecular approaches for analyzing differential gene expression: differential cDNA library construction and screening

Complementary deoxyribonucleic acid (cDNA) libraries can be used as a means to isolate and identify cell-specific messenger ribonucleic acid (mRNA) sequences. The basic elements of cDNA library construction and screening are reviewed in the context of analyzing differentially expressed mRNAs. A brief overview of the recombinant DNA systems applied to cDNA library construction and the principles of screening cDNA libraries by plaque hybridization are provided. Methods for comparing mRNA populations by differential screening and by competition hybridization are discussed, and methods for constructing subtracted cDNA libraries, enriched in differentially expressed sequences, are presented. Also reviewed are the analysis of differentially expressed cDNAs by Southern and Northern hybridization, RNase protection, polymerase chain reaction, and sequencing.

The SGP-2 gene is developmentally regulated in the mouse kidney and abnormally expressed in collecting duct cysts in polycystic kidney disease

Sulfated glycoprotein-2 (SGP-2) is a secreted, dimeric, glycosylated protein synthesized by a number of different epithelial cell types. Although its function is not yet understood, SGP-2 has been hypothesized to be involved in such diverse processes as the promotion of cell-cell interactions, spermatogenesis, modulation of the complement system, and programmed cell death. We have now found that the SGP-2 gene is developmentally regulated in the mouse kidney. SGP-2 gene expression is first detected in the condensing nephrogenic mesenchyme and is subsequently down-regulated during the maturation of the glomerular epithelia, proximal tubules, and collecting ducts. SGP-2 continues to be expressed in the mature kidney in distal tubules and in the urothelial lining of the calyx and papilla. We have also examined the expression of the SGP-2 gene in polycystic kidneys of the C57BL/6J-cpk mouse, a model of autosomal recessive polycystic kidney disease in which there is development of epithelial-lined cysts arising primarily from the collecting duct system. Abnormally high levels of SGP-2 mRNA were found in the cyst wall epithelium of polycystic kidneys. The expression of the SGP-2 gene in normal development suggests that it plays a role in differentiating epithelial structures; and the abnormally high levels of SGP-2 gene expression in polycystic kidneys suggests that the cells lining cysts are not fully differentiated. It is possible, therefore, that polycystic kidney disease is caused by a defective developmental process in which there is a delay in terminal differentiation.

Murine infantile polycystic kidney disease: a role for reduced renal epidermal growth factor

Polycystic kidney disease (PKD) represents a form of renal epithelial hyperplasia. The C57BL/6J-cpk mouse, which has an infantile form of PKD, has a dramatically reduced expression of renal prepro-epidermal growth factor (EGF) mRNA and immunoreactive protein. Since EGF promoted maturation of epithelia in the neonate, the relative lack of renal EGF may contribute to the development of the collecting duct cysts by delaying epithelial maturation.

Elevated proto-oncogene expression in polycystic kidneys of the C57BL/6J (cpk) mouse

Polycystic kidney disease in the C57BL/6J (cpk) mouse is an autosomal recessive disorder which leads to the rapid development of renal cysts and kidney failure during the first 3 to 4 postnatal weeks. Previously, we showed that the cystic kidneys of affected mice have abnormally elevated levels of c-myc mRNA. In the study presented here, it is shown that mRNAs for the proto-oncogenes c-fos and c-Kiras, as well as c-myc, are markedly elevated in cystic kidneys, suggesting that there is a more general abnormality in gene expression associated with the disease. It is also evident that there are two stages to this abnormal proto-oncogene expression. In the first stage, which occurs up through the second postnatal week, there are modest increases in proto-oncogene mRNA which parallel the increased cell proliferation that accompanies cyst growth at this time. In the second stage, which occurs after the second postnatal week, there are markedly elevated levels of proto-oncogene mRNA that are seen at a time when cell proliferation is declining. The development of this latter stage suggests either that there is a fundamental abnormality intrinsic to polycystic kidneys that leads to uncontrolled proto-oncogene expression later in disease progression or that there is a secondary response in the kidney to the progressive renal failure.

A hereditary model of slowly progressive polycystic kidney disease in the mouse

There are two known forms of hereditary polycystic kidney disease (PKD) in humans. Although both forms initiate early in life, autosomal recessive PKD is rapidly progressive to kidney failure shortly after birth whereas autosomal dominant PKD is slowly progressive, taking many years to end stage. Research in this field has been limited by the availability of suitable animal models of PKD. Recently the C57BL/6J-cpk mouse has been used to study the pathogenesis of rapidly progressive hereditary PKD. The study presented here describes a slowly progressive PKD in the DBA/2-pcy mouse. The disease trait is transmitted in an autosomal recessive pattern and was localized to chromosome 9 through linkage to the dilute coat color and transferrin genes. Whereas some cystic changes were seen in fetal and newborn affected mice, renal enlargement did not develop until after 8 weeks of age and azotemia did not develop until after 18 weeks of age. Renal cysts were identified in all segments of the nephron and collecting duct and progressively enlarged with age. Individual cysts were found to be lined by a single layer of epithelial cells in most areas, with focal polyps and mounds of cells principally in collecting duct cysts. Early stages of cyst formation were associated with some abnormalities of tubular and glomerular basement membranes and accelerated eruption of incisors. Late stages of the disease were characterized by azotemia and chronic renal interstitial inflammatory infiltrates in all affected animals and cerebral vascular aneurysms in a few. We conclude that the DBA/2-pcy mouse has a form of renal cystic disease that appears similar in many respects to that seen in the dominant form of human PKD.

Defective epidermal growth factor gene expression in mice with polycystic kidney disease

The C57BL/6J-cpk mouse has an inheritable form of polycystic kidney disease similar to the autosomal recessive disorder seen in humans. Between approximately 1 and 3 weeks of age, affected cpk mice develop numerous large cysts in the collecting tubule segment of kidney nephrons. The present study examined the ontogeny of renal and submandibular gland prepro-epidermal growth factor (preproEGF) gene expression in the cpk mouse using Northern blot hybridization and immunohistochemistry. There was a virtual absence of renal preproEGF gene expression in cystic kidneys over the 3-week postnatal period, during which time renal preproEGF mRNA and proEGF/EGF protein normally reach significant levels. PreproEGF mRNA was expressed in salivary glands of cystic mice; however, this mRNA could not be further elevated with testosterone suggesting that there are abnormalities in the regulation of the preproEGF gene in the submandibular gland, as well as in the kidney. Since renal preproEGF expression during the early postnatal period occurs when collecting duct cysts form, it is possible that a deficiency in renal proEGF or EGF contributes to the rapid development of collecting duct cysts and the concomitant renal failure in the C57BL/6J-cpk cystic mouse.

Renal epithelial fluid secretion and cyst growth: the role of cyclic AMP

Transepithelial fluid secretion has been postulated to account for the accumulation of fluid within hereditary and acquired renal cysts, but no such mechanism has been demonstrated in human kidney epithelium. It is shown here that transepithelial fluid secretion was stimulated by prostaglandin E1 (PGE1), forskolin, 8-Br-cyclic AMP, and 1-methyl-3-isobutylxanthine in polarized monolayers of established renal cell lines (MDCK and rat glomerular epithelial cells) and in monolayer cultures derived from the cyst walls of human autosomal dominant polycystic kidney disease and from epithelial cells of normal human renal cortex. Treatment with cyclic AMP agonists caused the same cells, when dispersed within a gel matrix of type I collagen (Vitrogen), to proliferate and form spherical fluid-filled monolayered cysts. Our findings suggest that increased intracellular cyclic AMP levels may have a critical role in the formation and expansion of hereditary and acquired renal cysts.

U3 small nuclear RNA can be psoralen-cross-linked in vivo to the 5' external transcribed spacer of pre-ribosomal-RNA

U3 small nuclear RNA is hydrogen-bonded to high molecular weight nucleolar RNA and can be isolated from greater than 60S pre-ribosomal ribonucleoprotein particles, suggesting that it is involved in processing of ribosomal RNA precursors (pre-rRNA) or in ribosome biogenesis. Here we have used in vivo psoralen cross-linking to identify the region of pre-rRNA interacting with U3 RNA. Quantitative hybridization selection/depletion experiments with clones of rRNA-encoding DNA (rDNA) and cross-linked nuclear RNA showed that all of the cross-linked U3 RNA was associated with a region that includes the external transcribed spacer (ETS) at the 5' end of the human rRNA precursor. To further identify the site of interaction within the approximately 3.7-kilobase ETS, Southern blots of rDNA clones were sandwich-hybridized with cross-linked RNA and then probed for cross-linked U3 RNA. These experiments showed that U3 RNA was cross-linked to a 258-base sequence between nucleotides +438 and +695, just downstream of the ETS early cleavage site (+414). The localization of U3 to this region of the rRNA precursor was not expected from previous models for a base-paired U3-rRNA interaction and suggests that U3 plays a role in the initial pre-rRNA processing event.

Sequential protooncogene expression in regenerating kidney following acute renal injury

Following loss of functional renal mass due to acute injury, there is significantly increased proliferation of tubular epithelium to replace injured and necrotic cells. In contrast, following uninephrectomy, the contralateral kidney increases in size primarily by hypertrophy, with little cellular proliferation. We and others have demonstrated only modest increases in renal protooncogene expression following uninephrectomy. In this study, we demonstrate markedly elevated expression of the protooncogenes c-fos, c-myc, c-Ki-ras, and c-Ha-ras following acute renal injury induced by a single large parenteral dose of folic acid. The expression of these genes occurs in a sequential pattern similar to that seen in proliferating cells in culture and in regenerating liver. In addition, we demonstrate elevated levels of histone H4 and beta-actin mRNAs consistent with increased cell proliferation. These data suggest that the molecular mechanisms regulating cell proliferation in the kidney are similar to those in regenerating liver and in cultured cells. In addition, it appears that these events are regulated normally after acute renal injury and following uninephrectomy, since in both instances the levels of protooncogene expression correlate with the degree of cell proliferation. This is in direct contrast to a pathologic renal condition, polycystic kidney disease, in which the level of protooncogene expression is out of proportion to the degree of cell proliferation. Further studies of the molecular correlates of acute renal injury may yield insight into the pathogenesis of this and other clinically important renal disorders.

Autosomal recessive polycystic kidney disease in a murine model. A gross and microscopic description

The C57BL/6J-cpk genetic murine model of autosomal recessive polycystic kidney disease was examined to gain insight into the pathogenesis of renal cystic disease. Fetal through 3-week-old offspring of heterozygote matings were used to study growth parameters and morphology of this genetic form of cystic disease. The kidneys were examined by light and electron microscopy and nephron segments were microdissected. Two phases of cystic disease development were morphologically identified. The first phase in fetal and newborn affected pups was characterized by proximal tubule enlargement and a general increase in the tubular mitotic index. The proximal tubules showed cytologic abnormalities along with an increased necrotic cell index. The later phase, in one through 3-week-old cystic pups, was characterized by progressive enlargement of the kidneys due mainly to cystic change of the collecting ducts and by development of azotemia. Secondary to the azotemia was a stunted body growth. Significant tubular epithelial hyperplasia was not found by mitotic index during the second phase, but an increase in collecting duct cellularity was present. Histone H4 gene expression, which is tightly coupled to DNA synthesis and thus an index of cell proliferation, showed only a minimal increase in cystic kidneys at 1, 2, and 3 weeks of age. Therefore, the degree of cell proliferation necessary to allow the observed tubular enlargement appears to be minimal.

Elevated c-myc protooncogene expression in autosomal recessive polycystic kidney disease

The polycystic kidney diseases (PKDs) are a group of disorders characterized by the growth of epithelial cysts from the nephrons and collecting ducts of kidney tubules. The diseases can be inherited or can be provoked by environmental factors. To investigate the molecular basis of the abnormal cell growth associated with PKD, c-myc protooncogene expression was studied in a mouse model for autosomal recessive PKD. Homozygous recessive C57BL/6J (cpk/cpk) mice develop massively enlarged cystic kidneys and die from renal failure shortly after 3 weeks of age. Quantitative dot blot and RNA blot hybridization experiments in which whole kidney poly(A)+ RNA was hybridized with a c-myc RNA probe showed a 2- to 6-fold increase in c-myc mRNA at 2 weeks, and a 25- to 30-fold increase in c-myc mRNA at 3 weeks of age in polycystic mice, as compared to normal littermates. c-myc expression was also examined under two conditions in which kidney cell growth was experimentally induced in normal adult mice: compensatory renal hypertrophy and tubule regeneration following folic acid-induced renal cell injury. While compensatory hypertrophy resulted in only a small (less than 3-fold) increase in c-myc, folic acid treatment gave rise after 24 hr to a 12-fold increase in c-myc mRNA. The induction of c-myc by folic acid is consistent with increased cellular proliferation in regenerating tubules. In contrast, polycystic kidneys show only a minimal increase in cellular proliferation over that seen in normal kidneys, while c-myc levels were found to be markedly elevated. Thus, the level of c-myc expression in cystic kidneys appears to be out of proportion to the rate of cell division, suggesting that elevated and potentially abnormal c-myc expression may be involved in the pathogenesis of PKD.

In-vivo secondary structure analysis of the small nuclear RNA U1 using psoralen cross-linking

Intramolecular base-pairing interactions have been probed in the small nuclear RNA U1 in vivo. HeLa cells were treated with the psoralen derivative aminomethyltrioxsalen, and cross-linking was carried out by irradiating the intact cells with light of 365 nm wavelength. Cross-linking resulted in a discrete shift in electrophoretic mobility of approximately 65 to 70% of the U1. This intramolecularly cross-linked U1 RNA, termed XU1, was purified and shown to co-migrate with uncross-linked U1 upon photo-reversal of psoralen cross-links with light of 254 nm wavelength. XU1 was also generated by the in-vitro cross-linking of deproteinized U1, suggesting that the secondary structure of U1 RNA in solution is similar to that of U1 ribonucleoprotein in the cell. A sequencing analysis was developed, based on partial enzymatic and alkaline cleavage of psoralen-treated RNA, to identify the position of psoralen cross-links and to distinguish between psoralen monoadducts and diadducts (cross-links). Sequencing of 3' and 5' end-labeled XU1 provided direct evidence for the presence of a unique intramolecular cross-link in XU1, located on uridine 116 (U116). This result is consistent with several secondary-structure models for U1 in which U116 is located in a base-paired stem. The proximity of uridine 96 (U96) to U116 on the opposite side of the base-paired stem suggested that U116 was cross-linked to U96. An additional U1 species having an electrophoretic mobility between those of U1 and XU1 was also generated by psoralen treatment. Analysis of this U1 species, termed U1M, revealed a psoralen monoadduct on U96. Further longwave (365 nm) irradiation of purified U1M resulted in its conversion to XU1 by completion of the U96-U116 cross-link. This suggested that cross-linking at the U96-U116 site occurred as a two-step process in which the psoralen first reacted with U96 and then with U116. Sequencing analysis also identified a psoralen monoadduct on uridine 45 (U45) of XU1. Efficient psoralen-adduct formation, which resulted in cross-linking at the U96-U116 site and monoaddition on U45, suggests that these regions are relatively accessible in the native U1 small nuclear ribonucleoprotein particle in vivo.

The heat shock response in HeLa cells is accompanied by elevated expression of the c-fos proto-oncogene

Several known inducers of the heat shock response (heat stress, arsenite, and heavy metals) were shown to cause a significant elevation of c-fos mRNA in HeLa cells. Heat stress resulted in a time- and temperature-dependent prolonged elevation in the level of c-fos mRNA, which was accompanied by increased translation of c-fos protein and its appearance in the nucleus. Elevated expression of c-fos during heat stress was paralleled by induction of hsp 70 mRNA, while levels of c-myc and metallothionein mRNAs declined. Treatment of HeLa cells with arsenite or heavy metals also resulted in increased levels of hsp 70, as well as c-fos mRNA. Although elevated expression of c-fos was prevented by inhibitors of RNA synthesis, analysis of relative rates of gene transcription showed that during heat stress there was a negligible change in c-fos transcription. Therefore, the enhanced expression of c-fos during the heat shock response is likely to occur primarily through posttranscriptional processes. Cycloheximide was also shown to significantly increase the c-fos mRNA level in HeLa cells. There results are consistent with the observation that these inducers of the heat shock response, as well as cycloheximide, repress protein synthesis and suggest that the increase in the level of c-fos mRNA is caused by an inhibition of protein synthesis. This supports the hypothesis that c-fos mRNA is preferentially stabilized under conditions which induce the heat shock response, perhaps by decreased synthesis of a short-lived protein which regulates c-fos mRNA turnover.

Human glutamate tRNA forms stable hybrids in vitro with 28S ribosomal RNA

A human glutamate tRNA has been shown to form stable hybrids with 28S ribosomal RNA. This tRNA was purified from HeLa cell cytoplasmic RNA by RNA-RNA solution hybridization followed by the isolation of tRNA-28S rRNA complexes by hybridization-selection with ribosomal DNA or by recovery of the 28S peak from formamide-sucrose gradients. The single hybridizing tRNA species was identified as tRNAGluCUC by sequencing: pU-C-C-C-U-G-G-U-G-m2G-U-C-phi-A-G-U-G-G-D-phi-A-G-G-A-U-U- C-G-G-C-G-C-U-C-U-C-A-C-C-G-C-G-G-C-m5C-m5C-G-G-G-Tm-phi-C-G-A- U-U-C-C-C-G-G-U-C-A-G-G-G-A-A-C-C-AOH. Computer analysis located a nucleotide sequence near the middle of human 28S rRNA which is complementary to 15-26 nucleotides between residues 20 and 50 of this tRNA. An interaction between this tRNA and 28S rRNA suggests that tRNAGluCUC may have functions in the cell in addition to translation.

Sequence analysis of the glutamate tRNA family: evidence for pseudogenes

Human tRNA(CUCGlu) has been isolated by direct hybridization of the tRNA to 28S ribosomal RNA. We now report the isolation of mouse tRNA(CUCGlu) using the same procedure. Partial sequence analysis of the mouse tRNA shows that it is identical to the human tRNA and to a cloned rat tDNA(CUCGlu) sequence. This mouse tRNA(CUCGlu), however, differs by one nucleotide from a previously cloned mouse tDNA(CUCGlu) sequence, suggesting that the tDNA may be a pseudogene. Further evolutionary comparison of these and other glutamate tRNAs and tDNAs has provided evidence to suggest that two other tDNA(Glu) sequences arose by mutation of functional tRNAGlu genes such that their anticodon sequences were converted from one glutamate isoacceptor to the other. These tDNA sequences may also represent pseudogenes.

U6 small nuclear RNA is transcribed by RNA polymerase III

A DNA fragment homologous to U6 small nuclear RNA was isolated from a human genomic library and sequenced. The immediate 5'-flanking region of the U6 DNA clone had significant homology with a potential mouse U6 gene, including a "TATA box" at a position 26-29 nucleotides upstream from the transcription start site. Although this sequence element is characteristic of RNA polymerase II promoters, the U6 gene also contained a polymerase III "box A" intragenic control region and a typical run of five thymines at the 3' terminus (noncoding strand). The human U6 DNA clone was accurately transcribed in a HeLa cell S100 extract lacking polymerase II activity. U6 RNA transcription in the S100 extract was resistant to alpha-amanitin at 1 microgram/ml but was completely inhibited at 200 micrograms/ml. A comparison of fingerprints of the in vitro transcript and of U6 RNA synthesized in vivo revealed sequence congruence. U6 RNA synthesis in isolated HeLa cell nuclei also displayed low sensitivity to alpha-amanitin, in contrast to U1 and U2 RNA transcription, which was inhibited greater than 90% at 1 microgram/ml. In addition, U6 RNA synthesized in isolated nuclei was efficiently immunoprecipitated by an antibody against the La antigen, a protein known to bind most other RNA polymerase III transcripts. These results establish that, in contrast to the polymerase II-directed transcription of mammalian genes for U1-U5 small nuclear RNAs, human U6 RNA is transcribed by RNA polymerase III.

A sequence related to 4.5 S RNA and the B1 family of repeated DNA in the 5' flanking region of the mouse beta-globin gene

A beta-globin gene probe containing 5' flanking DNA hybridizes to a small nuclear RNA from Friend erythroleukemia cells, which was subsequently identified as 4.5 S RNA. This hybridization is shown to reflect complementarity between the Alu consensus region of 4.5 S RNA and the DNA sequence between nucleotides -33 and -48 from the beta-globin gene transcription initiation site.

Small nuclear RNA U2 is base-paired to heterogeneous nuclear RNA

Eukaryotic cells contain a set of low molecular weight nuclear RNA's. One of the more abundant of these is termed U2 RNA. The possibility that U2 RNA is hydrogen-bonded to complementary sequences in other nuclear RNA's was investigated. Cultured human (HeLa) cells were treated with a psoralen derivative that cross-links RNA chains that are base-paired with one another. High molecular weight heterogeneous nuclear RNA was isolated under denaturing conditions, and the psoralen cross-links were reversed. Electrophoresis of the released RNA and hybridization with a human cloned U2 DNA probe revealed that U2 is hydrogen-bonded to complementary sequences in heterogeneous nuclear RNA in vivo. In contrast, U2 RNA is not base-paired with nucleolar RNA, which contains the precursors of ribosomal RNA. The results suggest that U2 RNA participates in messenger RNA processing in the nucleus.

Base-pairing interactions between small nuclear RNAs and nuclear RNA precursors as revealed by psoralen cross-linking in vivo

The psoralen derivative 4'-aminomethyl-4,5',8-trimethylpsoralen (AMT) reacts with base-paired regions of RNA and forms interstrand covalent cross-links. Since psoralens permeate living cells, they can be used to probe RNA-tRNA interactions in vivo. We used AMT to investigate whether small nuclear RNAs are base-paired with high molecular weight nuclear RNA in the cell. Intact HeLa cells were treated with AMT, and high molecular weight RNA was isolated under denaturing conditions from nuclei or from subnuclear fractions. The presence of base-paired snRNAs in the high molecular weight nuclear RNA was examined by electrophoresis after photochemical reversal of the cross-links. We found snRNA U3 and 5.8S rRNA to be cross-linked to nucleolar RNA. IN contrast, snRNA U1 was crosslinked to high molecular weight RNA in ribonucleoprotein particles containing hnRNA. The U1 base-paired to hnRNA was identified by its hybridization with a cloned U1 DNA sequence after reversal of the cross-links. These results demonstrate that U1 is base-paired with hnRNA in vivo, suggesting a role in mRNA processing.

Heterogeneous nuclear RNA double-stranded regions probed in living HeLa cells by crosslinking with the psoralen derivative aminomethyltrioxsalen

The psoralen derivative aminomethyltrioxsalen (AMT, 4'-aminomethyl-4,5',8-trimethylpsoralen) has been employed as a probe for heterogeneous nuclear RNA (hnRNA) double-stranded regions in experiments with living HeLa cells. hnRNA ribonucleoprotein (hnRNP) particles were purified from untreated or AMT-treated cells after irradiation with 365-nm light, and double-stranded hnRNA regions (dsRNA) were isolated by RNase A + T1 digestion of hnRNP, followed by preparative Cs2SO4 isopycnic centrifugation. The purified, hnRNP-derived dsRNA was then assayed for interstrand crosslinks by measurement of its "snapback" to RNase-resistant form after thermal denaturation. By this procedure, the amount of crosslinked dsRNA was found to be increased 3- to 7-fold in cells exposed to AMT in vivo. The levels of crosslinking in vivo compared favorably with those observed in model experiments with pure dsRNA in vitro. These results establish that double-stranded hnRNA regions exist in the living cell, and they further demonstrate that these base-paired regions are organized as rather accessible sites within the nucleus.

Photochemical cross-linking of secondary structure in HeLa cell heterogeneous nuclear RNA in situ 1

The psoralen derivative 4'-hydroxymethyl-4, 5', 8-trimethylpsoralen (hydroxymethyltrioxsalen) has been used in experiments with isolated HeLa cell nuclei to photochemically cross-link double helical regions in heterogeneous nuclear RNA in situ. Although there are other self-complementary sequences in hnRNA that can form base-paired structures upon phenol deproteinization of annealing, the present in situ cross-linking results demonstrate that some double-stranded regions are an authentic component of native hnRNA structure. Moreover, these special regions of secondary structure are apparently highly accessible to chemical probes within the intact cell nucleus, despite the fact that hnRNA possesses a ribonucleoprotein organization.

Secondary structure of heterogeneous nuclear RNA: two classes of double-stranded RNA in native ribonucleoprotein

Heterogeneous nuclear RNA (hnRNA) from HeLa cells contains intramolecular duplexes. Since hnRNA is associated with protein in vivo, it is possible that the double-stranded regions observed in deproteinized hnRNA form spontaneously upon the release of protein from single-stranded but potentially complementary sequences. We show here that this is not the case for a class of double-stranded sequences that is defined by resistance to RNases A + T(1) at high ionic strength. Exposure of HeLa hnRNA.ribonucleoprotein (hnRNP) particles to Escherichia coli RNase III, a double-strand-specific endoribonuclease, destroys most of the sequences resistant to RNases A + T(1). This effect is completely blocked when hnRNP is exposed to RNase III in the presence of an excess of purified double-stranded RNA. In addition, we show that there exist two classes of double-stranded RNA in hnRNP at a salt concentration of 0.13 M. These are distinguished by their relative resistance to RNases A + T(1). The more stable double-stranded sequences, which are resistant to RNases A + T(1) at 0.13 M, comprise 1.0-1.1% of the nucleotides in hnRNP. The less stable double-stranded sequences comprise an additional 1.5-2.0% of the nucleotides in hnRNP. These are sensitive to RNase III at 0.13 M, but are not resistant to RNases A + T(1) unless the salt concentration is raised to 0.63 M. The demonstration that double-stranded sequences resistant to RNases A + T(1) exist in native ribonucleoprotein and are not artifacts of deproteinization now makes it appropriate to seriously consider their possible functional role in hnRNA metabolism, perhaps as binding sites for regulatory proteins involved in mRNA processing.