Search

Congenital stationary night blindness 2A (CSNB2A) is an X-linked retinal disorder, characterized by phenotypically variable signs and symptoms of impaired vision. CSNB2A is due to mutations in CACNA1F, which codes for the pore-forming α1F subunit of a L-type voltage-gated calcium channel, Cav1.4. Mouse models of CSNB2A, used for characterizing the effects of various Cacna1f mutations, have revealed greater severity of defects than in human CSNB2A. Specifically, Cacna1f-knockout mice show an apparent lack of visual function, gradual retinal degeneration, and disruption of photoreceptor synaptic terminals. Several reports have also noted cone-specific disruptions, including axonal abnormalities, dystrophy, and cell death. We have explored further the involvement of cones in our ‘G305X’ mouse model of CSNB2A, which has a premature truncation, loss-of-function mutation in Cacna1f. We show that the expression of genes for several phototransduction-related cone markers is down-regulated, while that of several cellular stress- and damage-related markers is up-regulated; and that cone photoreceptor structure and photopic visual function – measured by immunohistochemistry, optokinetic response and electroretinography – deteriorate progressively with age. We also find that dystrophic cone axons establish synapse-like contacts with rod bipolar cell dendrites, which they normally do not contact in wild-type retinas – ectopically, among rod cell bodies in the outer nuclear layer. These data support a role for Cav1.4 in cone synaptic development, cell viability, and synaptic transmission of cone-dependent visual signals. Although our novel finding of cone-to-rod-bipolar cell contacts in this mouse model of a retinal channelopathy may challenge current views of the role of Cav1.4 in photopic vision, it also suggests a potential new target for restorative therapy.

Voltage-gated calcium channels (VGCC) are key to many biological functions. Entry of Ca2+into cells is essential for initiating or modulating important processes such as secretion, cell motility, and gene transcription. In the retina and other neural tissues, one of the major roles of Ca2+-entry is to stimulate or regulate exocytosis of synaptic vesicles, without which synaptic transmission is impaired. This review will address the special properties of one L-type VGCC,CaV1.4, with particular emphasis on its role in transmission of visual signals from rod and cone photoreceptors (hereafter called “photoreceptors,” to the exclusion of intrinsically photoreceptive retinal ganglion cells) to the second-order retinal neurons, and the pathological effects of mutations in theCACNA1Fgene which codes for the pore-formingα1Fsubunit ofCaV1.4.

X-linked congenital stationary night blindness (CSNB) is a recessive non-progressive retinal disorder characterized by night blindness, decreased visual acuity, myopia, nystagmus and strabismus. Two distinct clinical entities of X-linked CSNB have been proposed. Patients with complete CSNB show moderate to severe myopia, undetectable rod function and a normal cone response, whereas patients with incomplete CSNB show moderate myopia to hyperopia and subnormal but measurable rod and cone function. The electrophysiological and psychophysical features of these clinical entities suggest a defect in retinal neurotransmission. The apparent clinical heterogeneity in X-linked CSNB reflects the recently described genetic heterogeneity in which the locus for complete CSNB (CSNB1) was mapped to Xp11.4, and the locus for incomplete CSNB (CSNB2) was refined within Xp11.23 (ref. 5). A novel retina-specific gene mapping to the CSNB2 minimal region was characterized and found to have similarity to voltage-gated L-type calcium channel alpha1-subunit genes. Mutation analysis of this new alpha1-subunit gene, CACNA1F, in 20 families with incomplete CSNB revealed six different mutations that are all predicted to cause premature protein truncation. These findings establish that loss-of-function mutations in CACNA1F cause incomplete CSNB, making this disorder an example of a human channelopathy of the retina.

Mutations in CACNA1F, which encodes the Ca(v)1.4 subunit of a voltage-gated L-type calcium channel, cause X-linked incomplete congenital stationary night blindness (CSNB2), a condition of defective retinal neurotransmission which results in night blindness, reduced visual acuity, and diminished ERG b-wave. We have characterized two putative murine CSNB2 models: an engineered null-mutant, with a stop codon (G305X); and a spontaneous mutant with an ETn insertion in intron 2 of Cacna1f (nob2).Cacna1f ( G305X ): Adults were characterized by visual function (photopic optokinetic response, OKR); gene expression (microarray) and by cell death (TUNEL) and synaptic development (TEM). Cacna1f ( nob2 ): Adults were characterized by properties of Cacna1f mRNA (cloning and sequencing) and expressed protein (immunoblotting, electrophysiology, filamin [cytoskeletal protein] binding), and OKR.The null mutation in Cacna1f ( G305X ) mice caused loss of cone cell ribbons, failure of OPL synaptogenesis, ERG b-wave and absence of OKR. In Cacna1f ( nob2 ) mice alternative ETn splicing produced ~90% Cacna1f mRNA having a stop codon, but ~10% mRNA encoding a complete polypeptide. Cacna1f ( nob2 ) mice had normal OKR, and alternatively-spliced complete protein had WT channel properties, but alternative ETn splicing abolished N-terminal protein binding to filamin.Ca(v)1.4 plays a key role in photoreceptor synaptogenesis and synaptic function in mouse retina. Cacna1f ( G305X ) is a true knockout model for human CSNB2, with prominent defects in cone and rod function. Cacna1f ( nob2 ) is an incomplete knockout model for CSNB2, because alternative splicing in an ETn element leads to some full-length Ca(v)1.4 protein, and some cones surviving to drive photopic visual responses.

Background: Mutations in CACNA1F, which encodes the Cav1.4 subunit of a voltage-gated L-type calcium channel, cause X-linked incomplete congenital stationary night blindness (CSNB2), a condition of defective retinal neurotransmission which results in night blindness, reduced visual acuity, and diminished ERG b-wave. We have characterized two putative murine CSNB2 models: an engineered null-mutant, with a stop codon (G305X); and a spontaneous mutant with an ETn insertion in intron 2 of Cacna1f (nob2).

Tuberous sclerosis (TSC) is a multisystem autosomal dominant hamartosis whose genetics is complicated by reduced penetrance and widely varying clinical expression. Results of linkage analyses have variously suggested two different locations for a TSC gene. A collaborative dataset has been assembled to clarify the issue of genetic heterogeneity. We have now analyzed the data from a combined sample of 111 families. Using Ott's HOMOG programs, we completed three tests of homogeneity: (1) for chromosome 9q, (2) for chromosome 11q, and (3) for the combined 9q and 11q data. For test 1 the chi-square (1 df) was 21.54 (p less than 0.001), for test 2 the chi-square (1 df) was 0.13 (p greater than 0.35), and for test 3 the chi-square (2 df) was 37.61 (p less than 0.0001). Additionally, we examined the combined data for evidence that a third, as yet unlinked locus exists. Results of this last test were suggestive but not significant. Clearly loci for TSC are present on both chromosomes 9q and 11q. The maximum likelihood estimate of the proportion of chromosome 9q-linked families is 0.38, for chromosome 11q-linked families is 0.47, and for the unlinked type 0.15. Alternative explanations for these latter families include chance sampling of recombinants, nongenetic phenocopies, or misclassification.

Incomplete X-linked congenital stationary night blindness (CSNB) is a recessive, non-progressive eye disorder characterized by abnormal electroretinogram and psychophysical testing and can include impaired night vision, decreased visual acuity, myopia, nystagmus, and strabismus. Including the 20 families previously reported (Bech-Hansen et al. 1998b), we have now analyzed patients from a total of 36 families with incomplete CSNB and identified 20 different mutations in the calcium channel gene CACNA1F. Three of the mutations account for incomplete CSNB in two or more families, and a founder effect is clearly demonstrable for one of these mutations. Of the 20 mutations identified, 14 (70%) are predicted to cause premature protein truncation and six (30%) to cause amino acid substitutions or deletions at conserved positions in the alpha1F protein. In characterizing transcripts of CACNA1F we have identified several splice variants and defined a prototypical sequence based on the location of mutations in splice variants and comparison with the mouse orthologue, Cacnalf.

Congenital stationary night blindness (CSNB) is a nonprogressive retinal disorder characterized by night blindness, nystagmus, myopia, a variable decrease in visual acuity, an abnormal electroretinographic response, and a disturbance in dark adaptation. Two forms of X-linked CSNB have been defined, complete CSNB in which rod function is extinguished, and incomplete CSNB in which rod function is reduced but not extinguished, as seen by electroretinography and dark adaptometry. In studying a large family of Mennonite ancestry, we have confirmed linkage between the locus (CSNB2) for incomplete CSNB and genetic markers in the Xp11 region. In particular, lod scores of 12.25 and 15.26 at zero recombination were observed between CSNB2 and the markers DXS573 and DXS255. Detailed analysis of critical recombinant chromosomes in this extended family have refined the minimal region for the CSNB2 locus to the interval between DXS6849 and DXS8023 in Xp11.23.

The identification by fundus examination of those females carrying an X-linked retinitis pigmentosa (RP) gene can reportedly be as high as 87%. In genetic counselling sessions with young females with a 50% risk of being a carrier who wished to know their status, it has not been possible to achieve such a level of success. A review and reanalysis of previous reports indicated that if a tapetal-like reflex was not present in those age 35 years or less, the likelihood of identifying a carrier by fundus examination was small. A family with 7 females with a 50% risk of being a carrier of X-linked RP was evaluated using haplotype analysis in an attempt to identify the X chromosome carrying the RP gene. In the family described, it was possible to establish that a mutation in the RP3 locus most likely causes the disease. This has permitted the determination of the carrier status in each of the females with a high degree of certainty.

X-linked congenital stationary night blindness (CSNB1) is a hereditary retinal disorder in which clinical features in affected males usually include myopia, nystagmus, and impaired visual acuity. Electroretinography demonstrates a marked reduction in b-wave amplitude. In the study of a large Mennonite family with CSNB1, three of five sisters in one sibship were found to have manifestations of CSNB1. All the sons of these three sisters were affected. Each of the two nonmanifesting sisters had at least one unaffected son. Analysis of Xp markers in the region Xp21.1-Xp11.22 showed that the two sisters who were unaffected had inherited the same maternal X chromosome (i.e., M2). Two of the daughters who manifested with CSNB had inherited the other maternal X chromosome (M1). The third manifesting sister inherited a recombinant X chromosome with a crossover between TIMP and DXS255, which suggests that the CSNB1 locus lies proximal to TIMP. One of the affected daughters' sons had inherited the maternal M1 X chromosome, a finding consistent with that chromosome carrying a mutant CSNB gene; the other affected sons inherited the grandfather's X chromosome (i.e., P). Molecular analysis of DNA from three sisters with manifestations of CSNB is consistent with their being homozygous at the CSNB1 locus and with their mother being a carrier of CSNB1.

A sibship is reported in which two of three children developed a symmetrical polyarthropathy associated with a mild spondyloepiphyseal dysplasia. Although the physical findings resembled juvenile arthritis, laboratory investigations for inflammatory disease were entirely negative. Molecular studies in members of this family showed that none of them had any structural rearrangements or other major abnormality of the type II procollagen gene (COL2A1). The laboratory findings in this family with "pseudo-rheumatoid arthritis" do not exclude the possibility of a minor mutation of the type II procollagen gene or a defect in the processing of articular cartilage collagens.

Linkage between X-linked congenital stationary night blindness (CSNB1) and seven markers on the X chromosome was investigated in a large four-generation Albertan kindred. We detected significant linkage between the CSNB1 locus and the locus DXS255 (maximum lod score = 6.73 at a recombination fraction of 6%; confidence interval of 1% to 18%), which anchors the CSNB1 locus to the proximal region near p11.22 on the short arm of the X chromosome.

L.L. Baumbach*, D. Dressman, H. Basterrechea, M.E. Ahearn, N.T. Bech-Hansen, R.D. Clark, A. Meindl, E. Hoffman Dept of Pediatrics, Univ Miami Sch Medicine, PO Box 016820, Miami, FL., 33136; Children's National Medical Center, Washington, D.C; Univ. of Pittsburgh, Pittsburgh, PA; University of Calgary, Calgary, Alberta; USC Cancer Center, Los Angeles, CA; Ludwig Maximilians Universtat, Munich, Germany * lbaumbac@med.miami.edu

X-linked congenital stationary night blindness (CSNB) is a clinically and genetically heterogeneous non-progressive retinal disorder characterized by impaired night vision, decreased visual acuity, myopia, nystagmus, and strabismus. Two loci for CSNB exist on the X chromosome. The locus for complete CSNB (nonrecordable scotopic b-wave and lack of rod dark adaptation) has been mapped to Xp11.4 (Boycott et al. AJHG 62:865-875, 1998), while the gene responsible for incomplete CSNB (subnormal scotopic b-wave and mildly elevated rod adaptation), CACNA1F, has been identified in Xp11.23 (Bech-Hansen et al. Nature Genet. 19:264-267). Our analysis of this retina-specific L-type calcium channel α1-subunit gene has identified a total of 17 different mutations (two-thirds of which are predicted to cause a loss-of-function) in 36 families with incomplete CSNB. One of these mutations, L1045insC, is seen in 15 families of Mennonite ancestry from Western Canada. Clinical variability was examined in 66 patients from these families in terms of night blindness, myopia, visual acuity, congenital nystagmus and strabismus. In 80% of the patients at least one of the main features of CSNB (night blindness, myopia, and nystagmus) was absent. The only clinical feature present in all 66 patients tested was impaired visual acutiy. Among these Patients who shared the common CACNA1F mutation, considerable variability in clinical expression is evident and suggests the presence of genetic modifiers. This research was supported in part by the RP Research Foundation (Canada), the Alberta Heritage Foundation for Medical Research and the Roy Allen Endowment.

OBJECTIVE--To assess the efficiency, reliability, and ease of use of DNA diagnosis for Duchenne and Becker muscular dystrophies (DMD/BMD) using the polymerase chain reaction (PCR). DESIGN--DNA from the patients was screened for deletion mutations using multiplex PCR, and the results were compared with those obtained by Southern blot analysis. The PCR multiplex reaction detects nine specific "hot-spot" exons in the dystrophin gene while the Southern analysis detects 66 specific dystrophin gene restriction fragments. The multiplex reaction requires 50-fold less DNA than Southern analysis and thus is considerably more sensitive. SETTING--Fourteen university-affiliated and private genetic disease diagnostic laboratories. PATIENTS--Male patients with clinical signs of DMD/BMD. Cases were selected for analysis randomly, without knowledge of whether a deletion was present within the dystrophin gene. MAIN OUTCOME MEASURES--The percentage of cases that were detectable by multiplex PCR in comparison with Southern analysis, the frequency, extent, and location of the detected deletion mutations. In some cases, duplication mutations were monitored. RESULTS--The accuracy of a single PCR multiplex amplification (nine exons) was compared with Southern analysis with 10 cDNA probes that cover the full length of the gene. The multiplex PCR analytic method detected 82% of those deletions detected by Southern analysis methods. In one of 745 analyses, the multiplex method suggested a single exon deletion, which was not confirmed by Southern analysis, representing a false-positive rate of 0.013%. CONCLUSIONS--Multiplex PCR represents a sensitive and accurate method for deletion detection of 46% of all cases of DMD/BMD. The method requires 1 day for analysis, is easy to perform, and does not use radioactive tracers. As such, multiplex PCR represents an efficient and rapid method for prenatal or postnatal diagnosis of DMD/BMD.

Colchicine resistant (CHR) mutants of CHO cells with reduced permeability to colchicine display extensive cross-resistance to a number of apparently unrelated compounds including puromycin, daunomycin, emetine, ethidium bromide and gramicidin D. A positive correlation was observed between the level of cross-resistance and the relative hydrophobicity of these compounds. The mutants also showed increased (collateral) sensitivity to local anaesthetics (procaine, tetracaine, xylocaine and propanolol), steroid hormones (1-dehydrotestosterone, corticosterone and 5beta-pregnan-3,20-dione) and some Triton X compounds. In general, the degree of the pleiotropic response (cross-resistance or collateral sensitivity) correlated with the degree of colchicine resistance in mutant lines. These results are consistent with the pleiotropic phenotype being the result of the same mutation(s) which confer colchicine resistance and support a model for resistance in which the reduced permeability is assumed to be the result of an alteration in the modulation of the fluidity of the surface membrane.

An unusual property of GR25a [rho+] was the production of 20 to 30 percent [rho−] zygote colonies when crossed to a tester strain lacking mitochondrial DNA. Spontaneous [rho−] isolates of GR25a [rho+] were observed to be highly suppressive and to contain mitochondrial DNA of a parental buoyant density (1.685 g/cm3). Three ethidium bromide induced neutral petites of GR25a [rho+] did not have detectable mitochondrial DNA and were neutral in crosses to [rho+] strains. Seven [rho−] zygote colony isolates obtained from crossing GR25a [rho+] to a neutral petite were shown to contain abnormal mitochondrial DNA. Six zygote colony isolates had mitochondrial DNA of a buoyant density less than, or equal to, GR25a (1.682–1.685 g/cm3), whereas one isolate had a buoyant density greater than GR25a (1.688 g/cm3). It was suggested that abnormal mitochondrial DNA is generated during the mating reaction.

The relative abundance of 7 messenger RNAs extracted from Alzheimer and control neocortex were examined by Northern and quantitative dot blot analysis. The average yield of mRNA coding for NF-L, the 68-kDa moiety of neurofilament protein, was reduced to 27% of control when expressed as the percentage of total RNA or 14% when expressed per gram of neocortex. In contrast, the yields of 6 other messenger RNAs fell into two categories: those which were statistically significantly reduced to about 65% of control and those which were not reduced when expressed as percentage of total RNA. The anomalous low abundance of neuron specific NF-L mRNA, coding for the lowest molecular weight moiety of neurofilament proteins, in cerebral cortex of Alzheimer's disease cannot be adequately accounted for by a non-specific effect of brain damage, neuron cell loss or neurons with neurofibrillary degeneration. We speculate that this mRNA decrease is related to a functional deficit of gene expression in Alzheimer's disease, perhaps related to the non-random increase in chromatin compaction previously reported from this laboratory. The inability of neurons to maintain homeostatic amounts of NF-L transcription products may be linked to the accumulation of abnormal filamentous components characteristically associated with the diseased cytoskeleton.

The deleterious effects of many environmental agents--ultraviolet (UV) light, ionizing radiation, and polycyclic aromatic hydrocarbons, to cite a few--are primarily attributed to their ability to react with, and thereby structurally modify the deoxyribonucleic acid (DNA) of the living cell1. Given the appreciable level of damage believed to occur naturally in DNA2 and the need to maintain its fidelity as the repository of the genetic script, it s understandable that all living organisms, humans included, should possess multiple cellular processes whose combined actions promote the repair of damage to DNA3. An increasing number of genetically transmitted disorders in man is being found to be associated with enhanced sensitivity to extrinsic DNA-damaging agents. Patients afflicted with any one of these particular Mendelian single-gene traits are typically cancer-prone and suffer from progressive neurodegeneration, and, at least for a few disorders, their cells when cultured in vitro exhibit anomalies in one or more DNA epair processes4–8. The majority of these hereditary diseases can be divided into two broad groups: (i) disorders associated with hypersensitivity to UV light and UV-mimetic chemicals; and (ii) disorders associated with hypersensitivity to ionizing radiation and radiomimetic chemicals. Exemplary diseases of the former group--namely, xeroderma pigmentosum (XP), Bloom syndrome (BS), and Cockayne syndrome--are reviewed elsewhere in this volume. Here we present an overview of the clinical and laboratory features of the latter group of disorders with emphasis on the prototype disease, ataxia telangiectasia (AT)7,8, and comment briefly on current insight into the pathogenesis of these disorders and its implications for theories on environmental carcinogenesis and neurodegenerative processes in man.

The immunoglobulin heavy-chain genes are located in a cluster on chromosome 14. The simultaneous absence of the human IgG1, IgG2, IgG4, and IgA1 subclasses was previously reported in a healthy Tunisian Berber and was later shown to be due to a multigene deletion. We now describe a serological and molecular study of a different deletion observed in a healthy Tunisian. Blot hybridization analysis of the proband's DNA using gamma, epsilon, alpha, and mu switch probes showed that the deletion involves a large region of the immunoglobulin heavy-chain gene cluster: C psi epsilon, C alpha 1, C psi gamma, C gamma 2, and C gamma 4. Incidentally, we showed that the restriction enzyme EcoRI alone can be used with the alpha probe to differentiate A2m types. The deletion described, present in a person homozygous for GM-Am haplotypes (Gm1,17;..;5,14,11,13,10 A2m2), is consistent with previous location, by association analysis, of C psi gamma between C alpha 1 and C gamma 2. There is evidence to suggest that deletions may be more common in the Mediterranean region than in North American Caucasians.

We have demonstrated that restriction fragment length polymorphisms (RFLPs) produced by BamHI can be used as markers for constant (C) region heavy chain genes C psi gamma (C gamma pseudogene), C gamma 2, and C gamma 4. These RFLPs were found nonrandomly associated in the population sample studied. Of the eight combinations (haplotypes) of RFLPs theoretically possible, only two accounted for a total of 88% of the 116 chromosomes examined, a value greater than the total of 25% expected from random segregation of alleles. This indicates considerable linkage disequilibrium between C psi gamma, C gamma 2, and C gamma 4. Quantitative assessment of the degree of association between C gamma gene RFLPs, Gm markers, and switch region RFLPs adjacent to C mu and C alpha 1 revealed that C psi gamma is most tightly associated with C gamma 2 (r = 0.81 and 0.95 for the two common haplotypes), suggesting that C psi gamma maps to a position lying between C alpha 1 and C gamma 2. The association analysis used here should have general applicability for studying the genomic organization of other multigene families.

We have detected restriction fragment length polymorphisms associated with the immunoglobulin heavy chain C gamma genes. DNA from both parents of an individual having an unbalanced rearrangement of the long arm of chromosome 14, region q32 [Cox, D. W., Markovic, V. D. & Teshima, I. E. (1982) Nature (London) 297, 428-430], revealed distinctive patterns of BamHI fragments which hybridized with cloned probes from the C gamma 2-C gamma 4 gene cluster. The number of hybridizing fragments in both cases (five) equaled the number of known C gamma genes. Pedigree and densitometric analyses indicated that the proband did not have any maternal complement of C gamma gene-hybridizing fragments. Included on the deleted chromosomal segment was a C gamma gene having properties of the previously reported C gamma pseudogene. We also examined DNA from this family with a probe for the highly polymorphic locus D14S1, which recently was demonstrated to be tightly linked to the C gamma 1 gene locus [Balazs, I., Purrello, M., Rubinstein, P., Alhadeff, B. & Siniscalco, M. (1982) Proc. Natl. Acad. Sci. USA 79, 7395-7399]. EcoRI and EcoRI-BamHI fragments from both parents hybridized with a probe for this locus in DNA from the proband, indicating that, unlike the C gamma gene family, D14S1 was not deleted from the abnormal chromosome. Thus, the chromosomal breakpoint in the proband lies within region 14q32 between the two tightly linked markers, D14S1 and the C gamma 1 heavy chain gene locus. The D14S1 locus must lie proximal to the centromere relative to the C gamma gene family. The genetic variability detected with C gamma gene probes may prove useful for genetic analysis of structural rearrangements involving this region of chromosome 14.

Colonic polyposis and carcinoma developed in a woman with Turcot's syndrome at the age of 31 years; astrocytoma developed when she was 37. Her brother and sister had died of astrocytoma at the ages of 18 and 33 years, respectively. Progressive neutropenia developed in the patient three months after radiotherapy for her brain tumor and acute myelomonocytic leukemia 19 months after treatment. Three laboratories independently evaluated cultures of her skin fibroblasts for in vitro sensitivity to cell killing (loss of colony-forming ability) by x-rays. Survival assays consistently revealed slight but significant radiosensitivity in an early-passage (six to 10 doublings) fibroblast subculture. A later subculture (21 to 29 doublings) showed no abnormality, a possible effect of selective in vitro loss of radiosensitive cells.

The gamma-ray sensitivity of skin fibroblasts from six members of a cancer family was investigated using a colony-forming assay. Fibroblasts from the three members with cancer (two sisters with acute myelogenous leukemia and the mother with cervical carcinoma) showed a significant (p less than 0.05) increase in radiosensitivity, while three members without cancer (the father and two sons) showed a normal radioresponse. The possibility that the increased gamma-ray sensitivity was due to defective DNA repair was investigated using assays for DNA repair replication, single-strand break rejoining, and removal of enzyme-sensitive sites in gamma-irradiated DNA. Results of these assays indicate that the kinetics of enzymatic repair of radiogenic DNA damage in general, and the rejoining of single-strand scissions and excision repair of base and sugar radioproducts in particular, were the same in the cell lines from the sensitive and clinically normal family members.

Assays that assess the ability of cells to incorporate labeled precursors into acid-precipitable material in the presence of adriamycin, daunorubicin, puromycin, vinblastine, melphalan, or methotrexate were investigated as an approach to the detection of resistant cells in human tumor samples. Each assay was evaluated with suitable drug-resistant Chinese hamster ovary cell lines and normal human fibroblasts to determine whether the assays reflected the drug sensitivity of these lines. Moreover, the ability to detect the presence of drug-resistance cells in a mixed population was evaluated. Validated assays were then used to measure the drug sensitivity of cell samples from pleural and peritoneal effusions of patients, mainly with carcinoma of the breast or ovary. Though the responsiveness of the majority of the samples in these assays was similar to that of a human fetal lung fibroblast line, 37 of 142 samples displayed responses consistent with the presence of a significant proportion of drug-resistant cells. Of these 37 nonresponsive samples, 12 displayed nonresponsiveness to three drugs.