| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Medical and Molecular Genetics Center-IRO, Hospital Duran i Reynals, 08907 L'Hospitalet de Llobregat, Barcelona, Spain
b Andrology Department, Fundació Puigvert, 08025 Barcelona, Spain
ABSTRACT
Cystic fibrosis transmembrane regulator (CFTR), multidrug-resistant (MDR)1, and multidrug resistance-associated (MRP) proteins belong to the ATP-binding cassette (ABC) transporter superfamily. A compensatory regulation of MDR1 and CFTR gene expression has been observed in CFTR knockout rodent intestine and in an epithelial cell line of human colon, whereas a high homology and similar anion binding site are shared by MRP and CFTR proteins. To provide better insight into the relationship among the expression behavior in vivo of the three genes in human testis, analysis of MDR1 and MRP gene expression in testicular biopsies was performed and related to the presence of CFTR gene mutations in congenital absence of the vas deferens (CAVD: n = 20) and non-CAVD (n = 30) infertile patients with azoospermia or severe oligozoospermia. A CFTR mutation analysis performed in both groups of patients supported the involvement of CFTR gene mutations in CAVD phenotype (85%) and in defective spermatogenesis (19%). Quantitative reverse transcription-polymerase chain reaction analysis of testicular tissue showed a CFTR-independent MDR1 and MRP gene expression in human testis, suggesting that the mechanisms underlying CFTR gene regulation in testis are different from those in intestine. These findings should contribute to the understanding of patterns of in vivo expression of CFTR, MDR1, and MRP genes in CFTR-related infertility.
gene regulation, male reproductive tract, testis
INTRODUCTION
The presence of mutations in the cystic fibrosis transmembrane regulator (CFTR) gene has commonly been associated with cystic fibrosis (CF) [1] and other CF-related disorders, such as the congenital absence of the vas deferens (CAVD) phenotype [2, 3]. Cystic fibrosis is the most common autosomal recessive disease in the Caucasian population. This disease affects respiratory, digestive, and reproductive systems as the result of the aberrant expression and/or function of cAMP-regulated chloride channel encoded by CFTR [4]. The gene product has been shown to be related structurally to the protein associated with drug resistance MDR1 or P-gp (multidrug resistance P-glycoprotein) and MRP protein (multidrug resistance-associated protein). MDR1, MRP, and CFTR are all members of the ATP-binding cassette (ABC) transporter superfamily [5]. The three proteins are localized to the plasma membrane and have 12 transmembrane domains and two ATP sites [5].
The MDR1 gene product is an active transporter that pumps hydrophobic drugs out of the cell, reducing the cytoplasmic concentration and toxicity, conferring resistance of cells and tumors to chemotherapy [6]. In addition, expression of MDR1 or P-gp also seems to be associated with Cl- transport [7]. The drug transport and chloride channel activities associated with P-gp are distinct and can be separated by their requirements for ATP hydrolysis. Initially, some authors defended the hypothesis that MDR1 was a volume-activated Cl- channel itself or a component thereof [8]. It seems clear now that P-gp is a channel regulator rather than possessing intrinsic channel activity [9]. MDR1, like CFTR, is expressed in tissular epithelium including pancreas, intestinal tract, and reproductive organs [10–14]. The relationship of CFTR and MDR1 gene expression is quite close. Several authors have demonstrated a complementary pattern of CFTR and MDR1 expression in epithelial intestine in vivo, suggesting a coordinated regulation of the expression of both genes [15]. Furthermore, selective downregulation of constitutively expressed CFTR was observed in vitro due to an induction of P-gp expression [16] and an upregulation of MDR1 due to the absence of CFTR gene expression in CFTR knockout mice [17], suggesting a compensatory regulation of the expression of both genes.
The MRP protein is homologous to MDR1 but is actually more closely related to the CFTR protein [18]. Recent experiments demonstrated that substrates of the MRP protein block the CFTR channel, suggesting that CFTR and MRP share a structurally similar anion binding site at the cytoplasmic face of the membrane [19]. MRP actively transports conjugated organic anions, glutation-dependent cotransport, out of the cell [18]. MRP cDNA appears to be ubiquitously expressed at low levels in all normal tissues, including CFTR gene-expressing tissues: digestive tract (colon), respiratory tract (lung), and the urogenital tract (testis) among others [20].
CFTR, MDR1, and MRP are known to be expressed in testicular cells; nevertheless, little is known about the behavior of the ABC superfamily expression in testis. Several experiments in rodents have shown that CFTR gene expression occurs mainly in round spermatids [14, 21] and at low levels in Sertoli cells [22]. MDR1 is expressed in the endothelial cells lining the blood capillaries and in immature germ cells; thus, it seems that MDR1 and CFTR gene expression, as occurs in colon [15], is also coordinated in germ cells from testis. Our previous data have demonstrated the association between CFTR gene expression in testis and the efficiency of the spermatogenic process in CAVD patients [23]. The presence of the 5T allele in the CFTR gene was mainly related to a reduction in mature spermatids per tubule in CAVD males [23]. Nevertheless, the heterogeneity of spermatogenic phenotype in CAVD and non-CAVD related infertile patients suggests the influence of additional genes coexpressed in this tissue.
Infertility affects approximately 1 in 25 men, but the cause of spermatogenic anomalies is still largely unknown [24]. Some defects in fertility have often been associated with varicocele, infection, immunologic factors, anatomic malformation, or chemical insults. The pattern of familial aggregation of male infertility suggests an important role of heritable factors [25]; nevertheless, little is known about the possible genetic etiologies. While spermatogenesis must require many gene products, no human mutations specifically disrupting spermatogenesis have been defined at the molecular level. Recent studies indicate that one of the genetic causes that could be important in some cases of testicular failure is linked to the Y chromosome. Among the men with azoospermia or severe oligozoospermia, between 10% and 16% have deletions in the Y chromosome [26, 27].
The purpose of this work was the study of the expression in vivo of the genes that belong to the ABC transporter superfamily in CAVD and non-CAVD infertility, by means of, first, the analysis of CFTR mutations in both groups. This information may increase our knowledge of the involvement of CFTR in non-CAVD male infertility. Moreover, the correlation between the presence of CFTR gene mutations (with the associated CFTR functional protein decrease) and MDR1 and MRP gene expression in testis from infertile patients could elucidate the presence or absence of a compensatory mechanism for the CFTR defective chloride transport efficiency in this tissue.
MATERIALS AND METHODS
Subjects of Study
Blood and testicular samples were obtained from patients with a phenotype consistent with azoospermia to severe oligozoospermia. All the patients were informed and gave written consent to the procedures of the study. Fifty men referred for couple infertility to the Andrology Department of Fundació Puigvert in Barcelona between January 1997 and May 2000 were selected and studied by anamnesis, physical examination, repeated semen analysis, hormonal determinations, and testicular biopsy [28]. Scrotal exploration initially suggested the CAVD diagnoses due to the absence of vas deferens, which was confirmed further with transrectal ultrasonography. Spermiograms were performed in accordance with World Health Organization guidelines [29] and included volume, pH, sperm concentration, motility, and fructose and citrate levels in seminal plasma. After a thorough clinical evaluation 20 men were diagnosed with CAVD (16 bilateral CAVD [CBAVD] and 4 unilateral CAVD [CUAVD]) and 30 had non CAVD-related infertility. Testicular failure (TF) was found in 22 patients with non-CAVD phenotype (11 showed secretory azoospermia [SA] and 11 severe secretory oligozoospermia [SSO], defined as a sperm count <5 x 106 spermatozoa/ml) and obstructive azoospermia (OA) was diagnosed in 8 patients (Table 1). Testicular biopsies were carried out under local anesthesia by open incision. Each specimen was divided into two aliquots, one fixed in Bouin solution and processed for histological analysis and the other immediately frozen in nitrogen vapor for gene expression experiments. The study was approved by the Institutional Review Board for Clinical Research of the Centre.
|
Associated risk factors with possible effects on fertility (varicocele, cryptorchidism, and acquired pathologies) are noted (Table 1).
Patients were also genetically characterized by karyotype analysis for chromosomal aberration detection. Screening for microdeletions of chromosome Y was performed via a Y Chromosome Deletion Detection System, version 1.1 (Promega Corporation, Madison, WI) in patients with testicular failure, except for patient nos. 22, 27, 31, and 34.
Nine CAVD and five TF patients have been described previously [23].
Analysis of CFTR Mutations
DNA was isolated from peripheral blood lymphocytes as previously described [30]. For the analysis of CFTR mutations, PCR-OLA Cystic Fibrosis Assay kit (Perkin Elmer, Foster City, CA) was first used. A complete CFTR screening was also performed by multiplex denaturing gradient gel electrophoresis (DGGE) [31] for 15 exons and by single-strand conformation analysis (SSCA) [32] (Multiphor; Amersham-Pharmacia Biotech, Buckinghamshire, U.K.) for the other 12 exons. Finally, direct automatic sequencing of all abnormal bands was performed. Analysis of alleles in the IVS8-6(T)n locus of intron 8 and the polymorphism M470V of exon 10 were determined as published elsewhere [3, 33].
RNA Analysis
Testicular biopsies for RNA studies were available in all CAVD and 22 non-CAVD-related patients. Fluorescent reverse transcription-polymerase chain reaction (RT-PCR) analysis was performed in order to quantify the MDR1 and MRP transcripts. Complementary DNA synthesis reaction was performed from 2 µg RNA in a 20-µl reaction according to the manufacturer's instructions (Superscript II RNAse H- Reverse Transcriptase; Gibco BRL, Life Technologies Ltd., Paisley, U.K.) at 42°C for 60 min. MDR1 and MRP cDNAs were quantitatively coamplified with GAPDH cDNA as internal control in each case. Primers are derived from different exons of MDR1, MRP [34–36], and GAPDH [37] to avoid DNA amplification. In each case the reverse primer was fluorescently labeled. The PCR amplifications were performed in 50 µl of reaction solution, and the cycling conditions were as follows: 94°C 30 sec, 55°C (MDR1)/58°C (MRP) 40 sec, 72°C 1 min. Replicate cDNA and PCR reactions were performed for each sample. The amplified products were size fractionated by 6% acrylamide gel electrophoresis. The quantification of the fluorescently labeled product was performed with the Applied Biosystem 672 Genescan software where the area of the peaks and the size of the products were obtained. The size of the fragments was automatically determined by comparison with an internal standard DNA ladder GS-1000 (Perkin Elmer). After amplification, MDR1 transcripts were 157 base pairs (bp) long, MRP transcripts 296 bp long. and GAPDH transcripts (internal control) 497 bp long. The results are expressed as the ratio between the specific area of the gene of interest and the area of the control gene (Table 2). In this way, it is possible to compare the data of different samples.
|
Studies for the validation of our RT-PCR protocol were performed as follows. Amplification by PCR was carried out for a number of cycles per amplification ranging from 20 to 35 cycles, increasing at a rate of 5 cycles per amplification in order to select the number of cycles where transcripts could be quantified but had not reached the stationary phase. Twenty-four cycles were selected to perform the experiments.
Statistical Analysis
Contingency two-by-two tables with significance calculation using the Yates correction of 
2 test and the Fisher test to the 5% limit were employed to test for independence between physiological variables and genetic mutations. MDR1 and MRP gene expression related to CFTR genotype statistical analysis was performed comparing the means of independent data and two independent groups. Due to the similarity of variances (Levene test, P > 0.05) a t-test for equality of variances was performed.
RESULTS
Characterization of Patients
Patients were classified as having CAVD, severe oligozoospermia, or azoospermia due to testicular failure and obstructive azoospermia based on the combined results of a comprehensive study. Concentrations of FSH reflected in general the findings of testicular histology, although some patients showing blockade of primary spermatocytes, hypospermatogenesis, or asymmetric lesions had normal FSH. Two individuals with CAVD showed elevated FSH and abnormal testis biopsy (Table 1).
Associated clinical conditions with possible contribution to infertility (varicocele, cryptorchidism, high alcohol consumption, inguinal herniorraphy, infections) are described for seven patients in Table 1.
Karyotype analysis identified a translocation (X:12) in one patient. A large deletion from STS SY152 to SY157 including the absence of the whole azoospermia factor region C (AZFc) was found in 2 out of 18 patients without an associated pathology, giving an incidence of 11%, in agreement with the data obtained from previous work [26, 27]. An additional patient with bilateral varicocele also presented the deletion Yq (Table 1).
CFTR Analysis
We have identified 14 different CFTR mutations (R117H, L206W, V232D, R258G, F508del, G542X, 621+1G>T, Q890R, S945L, Y1014C, Y1092X, D1270N, 2789+5G>A, IVS8–6[5T]) in 17 of 20 patients of the CAVD group, giving a CFTR mutation frequency of 85%. Only three CAVD (two CBAVD and one CUAVD) patients showed no CFTR gene mutations after completing the screening of the gene (Table 1).
As regards non-CAVD infertility, although CFTR gene study was performed in all cases (n = 30), statistical analysis for the frequency of CFTR mutations in non-CAVD infertility was determined in samples using minimal selection criteria: a normal 46 XY karyotype, no relationship to other patients in the database, no other severe congenital abnormalities (or abnormalities known to be associated to testicular malformations or obstruction), no microdeletion detected in chromosome Y, and no history of genitourinary surgery or infection (described to be associated with epididymal obstruction [38]). Following these criteria, 9 of 30 patients were excluded from the statistical CFTR gene analysis. In fact, no CFTR mutation was identified in these nonidiopathic patients. Twenty-one non-CAVD-related azoospermic patients (nos. 21–36, 43–47) were finally selected, because of their unknown etiology, for statistical analysis. Three different mutations were detected in patients who presented the TF feature: R75Q, R334W, and F508del accounting for 9.4% of chromosomes and 18.8% of patients (3 of 16) (Table 1). This result was compared with the CFTR mutation frequency in the general population. The incidence of CF in the Caucasian population is estimated to be 1 in 2500 [39], (carrier frequency 1 in 25). The mutation rate in our study was 97% [40]. With this assumption, one CF mutation in 25.8 (3.9%) individuals from the general population should be expected. The CFTR mutation carrier frequency in TF patients was found to be increased and significantly different from the CFTR mutation carrier frequency of the general population (2 = 11.05; P < 0.001). In the OA subgroup no mutation was identified.
The IVS8-6(5T) allele was found in 25% of CAVD patients (12.5% of chromosomes). No patient of the TF subgroup had the 5T allele; nevertheless, one patient of the OA subgroup was homozygous for the IVS8-6(5T) allele. None of the non-CAVD patients have two CFTR mutations or one mutation and the IVS8-6(5T) allele.
The analysis of the M470V polymorphism in infertile patients showed that the V470 allele was present in 16 of 40 chromosomes of patients with the CAVD (40%) and 26 of 40 (65%) with the idiopathic non-CAVD phenotype, being 63.3% for the TF group and 70% in the OA group. The V470 allele was found in 67% of chromosomes in the general population [41]. Statistical analysis showed that the lower incidence observed for this allele in infertile patients was significantly different in CAVD (2 = 14.65; P < 0.001), whereas it was not significant in non-CAVD patients (2 = 0.09; P = 0.76), as compared with the frequency in the general population. Homozygous for V470 variant was significantly more common among non-CAVD than in CAVD patients (40% versus 15%; 2 = 15.67; P < 0.001) (Table 1).
Qualitative and Quantitative Transcript Analysis of MDR1 and MRP Genes in Testicular Cells
In order to determine the relative level of MDR1 and MRP gene expression in testicular tissue, the 20 CAVD and 22 non-CAVD biopsies from severe infertile patients were analyzed by means of fluorescent RT-PCR analysis. Replicate cDNA and PCR reactions were performed for each sample (Table 2).
Samples of both groups were regrouped by their CFTR genotype in order to evaluate the correlation between CFTR gene analysis and the relative level of MDR1 and MRP transcripts (Table 2). Three groups were defined based on the number of CFTR mutations plus an additional fourth group comprising all patients with the 5T allele. The relative MDR1 mRNA ranged from 0.21 to 2.3 in the CFTR-genotype-related groups and MRP mRNA from 0.63 to 4.14. We observed considerable variability of MDR1 and MRP gene expression among individuals in each group. Statistical analysis, comparing the means of independent data in different groups (t-test for equality of means), showed no significant difference between groups (MDR1 gene expression: 2 CFTR mutations/no CFTR mutation P = 0.96; 1 CFTR mutation/no CFTR mutation P = 0.55; 5T/no CFTR mutation P = 0.51; MRP gene expression: 2 CFTR mutations/no CFTR mutation P = 0.62; 1 CFTR mutation/no CFTR mutation P = 0.44; 5T/no CFTR mutation P = 0.74).
DISCUSSION
The pathophysiological basis of CFTR mutation and CAVD infertility has been found to be based on the abnormal chloride channel leading to defective fluid transport in the male reproductive tract. Electrolyte/fluid epididymal secretion, essential for sperm maturation, is mediated by a cAMP-dependent Cl- channel, regulated by adrenergic and paracrine factors [42]. However, CFTR-related CAVD infertility is more widespread than previously thought, associated in part with a basic malfunction in the spermatogenic process [23]. Due to the potential role of CFTR during spermatogenesis [23, 43], a complete CFTR gene mutation analysis in non-CAVD severe infertility (n = 30) was performed, as well as in CAVD men (n = 20). The CFTR mutation frequency (85%) in the CAVD group does not differ from previous studies [40]. The absence of CFTR mutations in three CAVD patients may be due to other unknown genetic factors responsible for the CAVD phenotype. One of the most commonly identified mutations was IVS8-6(5T) as previously described [3]. It is well known that the presence of the 5T allele affects the normal splicing of exon 9, resulting in low levels of normal CFTR [44]. Due to the preponderance of the 5T variant and the fact that the 5T variant is uncommonly found in patients with classic CF, the 5T mutation has been called a specific allele for the absence of vas deferens. Referring to non-CAVD infertility, in 9 of the 30 males first selected, infertility could be explained by exogenous factors not related to CFTR, such as the presence of an infertility-associated pathology, chromosomal aberrations, and/or Y-chromosome microdeletions (Table 1). Among the remaining 21 males with idiopathic non-CAVD infertility, CFTR mutation analysis highlighted the presence of mutations in three patients with TF features implying a carrier frequency of 19%, significantly higher than that found in the general population (4%, P < 0.001). Our results are not in agreement with other studies [45–47] in azoospermic and oligozoospermic patients, whose CFTR mutation frequencies did not differ from those of the general population. These differences may be due to, first, the clinical criteria applied to select the patients. In addition, one of these studies [45] only performed the analysis of four CFTR mutations and the IVS8-6(5T) allele. Moreover, the exclusion of the cases that presented Y-chromosome microdeletions (which are detected in 10–16% of secretory azoospermia and oligozoospermia [26, 27]) may further explain the higher incidence of CFTR mutations in our study.
A 5T homozygous genotype was identified in one of five men in the idiopathic OA group. Previous studies have failed to detect 5T homozygotes in 41 individuals [41] and 232 individuals [48] from the general population. However, IVS8-6(5T) homozygotes have been described in CFTR-related phenotypes, as CBAVD patients [3, 40, 49–51] and females that presented different lung diseases: 1 in 144 asthmatic patients [41] and 1 in 120 patients with pulmonary disease [52].
Previous data demonstrated that some of the most common polymorphisms in the CFTR gene have consequences at the functional level, for instance, the V470 allele decreases the half-life of the CFTR protein, affecting its biogenesis [53]. The V470 allele was thought to play a role in the partial penetrance of the IVS8-6(5T) allele as a mutation causing disease, resulting in a less functional protein [53]. The V470 allele was present in 65% of chromosomes in the non-CAVD infertile group, a similar proportion to that found in our general population (67%) [41], differing from the results of a previous study [46] that found a significant difference between the two groups. However, the frequency of V470 was significantly decreased in the CAVD group (40%), in agreement with our previous study [40]. Consequently, the V470 homozygous genotype was significantly more frequent among non-CAVD infertile patients as previously described [46]. Given that we have only identified one V470 allele in five OA non-CAVD-related patients associated with IVS8-6(5T) in the genotype, further data would be necessary to draw any conclusion.
Little has been reported about CFTR, MDR1, and MRP gene expression in testicular tissue. In testis, CFTR gene expression is maximal in the round spermatids associated with spermatogenic stages VII and VIII. MDR1 cDNA was detected in spermatogonia or spermatocytes of stages IX–X of the cycle of the seminiferous epithelium. Expression of the two genes cannot be detected in the same cell at the same time [15], thus MDR1 and CFTR gene expression, as occurs in the colon [15], seems to be coordinated. MRP is also expressed in testis, but no data are available on the specific cellular location. The compensatory mechanism observed between MDR1 and CFTR gene expression in vitro, in an epithelial cell line of human colon [16] and in vivo, in intestine of CFTR knockout mice [17] led us to initiate a study on MDR1 gene expression in azoospermic human testis related to CFTR genotype. Due to the structural and functional similarities between MRP and CFTR genes, MRP gene expression experiments were also performed. The behavior of MDR1 and MRP gene expression related to the presence of CFTR mutations was analyzed in testicular tissue, in order to determine the presence or absence of a compensatory mechanism in testicular cells.
A high variability of MDR1 and MRP gene expression is observed in testicular biopsies. No statistical difference was observed in MDR1/MRP transcript analysis among the different groups related to the CFTR genotype, suggesting that MDR1/MRP gene expression is independent of the CFTR genotype and the associated CFTR function defect in human testis. Our previous results [23] on the relationship of the 5T allele and the efficiency of the spermatogenic process lead us to hypothesize the existence of other genetic factors that compensate for the CFTR function in patients who presented two CFTR mutations, but not in the presence of the 5T allele (alone or associated with a CFTR mutation). Our present data demonstrate the absence of an MDR1 and/or MRP compensatory mechanism for deficient CFTR expression in testis, independent of the CFTR genotype that the patient presents, and contrasts with the upregulation of MDR1 observed in the intestine of CFTR knockout mice [17]. These data suggest that the regulatory mechanism of CFTR deficiency is not the same among different tissues, and different genes may be involved.
At least 30% of infertile men suffer from idiopathic infertility, reflecting the fact that the underlying pathophysiology for the testicular defect has remained elusive until now. Relatively little research has focused on possible genetic etiologies, and only 10–16% of idiopathic secretory infertility has been related to the presence of microdeletions of the Y chromosome [26, 27]. Our present data provide additional details on the role of the ABC transporter superfamily in azoospermia and severe oligozoospermia in vivo. Due to the impossibility of obtaining CF testicular biopsies, CAVD pathology is the closest in vivo model to determine the MDR1/MRP gene expression alteration in response to the absence of CFTR gene expression in testis. The influence of CFTR gene expression, extensively demonstrated on the CAVD phenotype, seems to be important also in other causes of infertility, such as deficient spermatogenesis due to failure of seminiferous tubules. Nevertheless, due to the controversial data from other reported studies, a larger group of patients might be analyzed to determine the specific role of the CFTR gene in non-CAVD infertility. In contrast, MDR1 and MRP may not have a determinant role in spermatogenesis by compensating for the defect of CFTR function.
ACKNOWLEDGMENTS
We are indebted to the patients who participated in this study. We thank Helena Kruyer for her help with the manuscript and the European Community Concerted Action for Cystic Fibrosis for providing primers for DGGE analysis.
FOOTNOTES
First decision: 11 January 2001.
1 This work was supported by grants from Fondo de Investigación Sanitaria (FIS) (99/0654), Fundació La Marató TV3 (980410), and Institut Catalá de la Salut. S.L. is supported by Fundació La Marató TV3. 
2 Correspondence: Teresa Casals, Medical and Molecular Genetics Center-
IRO, Hospital Duran i Reynals, 08907 L'Hospitalet de Llobregat, Barcelona,
Spain. FAX: 34 93 260 77 76; tcasals{at}iro.es
Accepted: March 14, 2001.
Received: November 30, 2000.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  T. M Hagedorn, R. W Carlin, and B. D Schultz Oxytocin and Vasopressin Stimulate Anion Secretion by Human and Porcine Vas Deferens Epithelia Biol Reprod, September 1, 2007; 77(3): 416 - 424. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. W. Carlin, J. H. Lee, D. C. Marcus, and B. D. Schultz Adenosine Stimulates Anion Secretion Across Cultured and Native Adult Human Vas Deferens Epithelia Biol Reprod, March 1, 2003; 68(3): 1027 - 1034. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
