| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From * MRC Human Reproductive Sciences Unit and
the Division of Reproductive and Developmental
Sciences, University of Edinburgh Centre for Reproductive Biology, The Queen's
Medical Research Institute, Edinburgh, United Kingdom; and the
Division of Pathway Medicine, University of
Edinburgh, Edinburgh, United Kingdom.
| Correspondence to: Dr Rosemary A. L. Bayne, MRC Human Reproductive Sciences Unit, Centre for Reproductive Biology, The Queen's Medical Research Institute, 47 Little France Crescent, Edinburgh EH16 4TJ, United Kingdom (e-mail: r.bayne{at}hrsu.mrc.ac.uk). |
| Received for publication September 12, 2007; accepted for publication December 11, 2007. |
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Key words: Microarray analysis, steroidogenesis, testicular function
LH receptors are present on Leydig cells, where they control steroid biosynthesis and thus the production of testosterone. Testosterone and FSH receptors are found on Sertoli and peritubular cells, but there are no receptors for any of these hormones on germ cells (Themmen and Huhtaniemi, 2000; Collins et al, 2003), with the exception of mature human sperm, which have recently been shown to have functional androgen receptors (Aquila et al, 2007). This indicates that the effects of these hormones on immature germ cells must be indirect, in that all are required for quantitatively normal spermatogenesis (Matsumoto et al, 1983, 1984).
Gonadotropin suppression results in suppression to azoospermia in most but
not all men (Anderson and Baird,
2002). Suppression can be achieved by the administration of
testosterone alone or with a gonadotropin-releasing hormone (GnRH) antagonist,
but the addition of a progestogen increases spermatogenic suppression and is
currently the most promising approach toward a hormonal male contraceptive
(Nieschlag et al, 2003). It
has been suggested that the degree of spermatogenic suppression in some
progestogen regimens is greater than can be accounted for by gonadotropin
suppression, which could indicate that progestogens act directly on the testis
(McLachlan et al, 2004). Both
nuclear and membrane progesterone receptors have been found on spermatozoa and
on Sertoli and some Leydig cells in the human testis
(Shah et al, 2005) and
progestogens have direct inhibitory effects on Leydig cell function in a
murine cell line (El-Hefnawy et al,
2000). We have recently provided direct evidence for a
progestogenic effect on 5
-reductase expression in the testis
(Walton et al, 2006) that
might reduce the amplifying effect of conversion of testosterone to
dihydrotestosterone and thus contribute to greater suppression of
spermatogenesis.
To identify a more comprehensive picture of the effects in humans of suppression of gonadotropins by administration of a GnRH antagonist and of the addition of progestogen, we performed microarray analysis of gene expression on testis biopsies and show that the pharmacological effects are focused on a select number of pathway-specific genes indicating stringent regulation of testicular gene expression.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
One subject withdrew for personal reasons. After 4 weeks of drug treatment, further blood and semen samples were analyzed, and testis biopsies were carried out on the remaining 29 men under local anesthetic at the time of vasectomy with a 14-gauge needle (Tru-Cut, Allegiance Healthcare Corporation, McGraw Park, Illinois). Tissue samples were immediately frozen and stored at –80°C.
RNA Samples
Total RNA was extracted from testis biopsies as described
(Walton et al, 2006). Quality
was confirmed on RNA 6000 Nanochips in the Agilent 2100 Bioanalyzer (both from
Agilent Technologies UK Ltd, West Lothian, United Kingdom). Only very high
quality RNA (RNA integrity number >7.5) preparations were considered for
microarray screening. Five RNA samples from each of the control, CT, and CTD
groups were used for screening the microarrays, along with the 5
"nonsuppressors" from the treatment groups (3 CT, 2 CTD) who did
not show a reduction in sperm concentrations to below the reference range
(Walton et al, 2006).
Microarray Hybridization and Analysis
Affymetrix Human genome U133 plus 2.0 microarrays (Affymetrix UK Ltd, High
Wycombe, United Kingdom) containing 54 675 probes representing approximately
39 000 genes, were hybridized to 15 µg fragmented biotin-labeled cRNA
prepared from each RNA and spiked with eukaryotic hybridization control in a
volume of 200 µL for 16 hours. After hybridization, the arrays were washed
on the Genechip Fluidics Station 450 and scanned with the Genechip Scanner
3000 all according to the manufacturer's protocols (Affymetrix UK Ltd).
Raw microarray data were processed and analyzed with R 2.1.1 statistical environment (http://www.R-project.org) and Bioconductor 1.6 microarray libraries (Gentleman et al, 2004). Numerical quality control steps were applied to each of the 20 samples. Arrays were normalized and converted to probe-level data with Robust Multi-Array Average (RMA; Irizarry et al, 2003). To reduce statistical multiple testing and algorithmic problems in downstream analysis, a nonspecific filter to remove genes expressed at low levels (signal < log2 100) in all of the samples was applied. This nonspecific filter is set quite stringently to remove low-signal genes and thus reduce the number of false positives, given the relatively low sample size in each group. It reduces the number of genes available for comparison from about 19 000 that pass the basic Affymetrix quality criteria (MAS5 Present/Marginal/Absent calls) to 5824 after nonspecific filtering but gives more reliable results.
To identify candidate genes that change in response to the various treatments, differential expression (by at least 2-fold) and its statistical significance for each gene was calculated on the basis of an empirical Bayes test (Smyth, 2004), which is robust for small sample sizes. Analysis involved 3 different stages. Stage 1 testing compared gene expression between controls and each of the treatment groups (control n = 5, CT n = 7, CTD n = 7), as well as between the 2 treatment groups. Stage 2 testing repeated these comparisons with nonsuppressed patients removed (CT-ns n = 4, CTD-ns n = 5). Stage 3 testing explicitly compared all nonsuppressed patients (ns n = 5) to the "suppressed" patients in each of the 2 treatment groups (CT-s n = 4, CTD-s n = 5). A reduction in the potential number of false positive results was achieved by correcting the significance test P values in relation to the number of genes tested simultaneously (n = 5824) (Benjamini and Yekutieli, 2005).
Functional characterization of significantly expressed transcripts was
performed with the resource tool DAVID from the National Institute for Allergy
and Infectious Diseases (NIAID)
(http://niaid.abcc.ncifcrf.gov/).
DAVID examines the overrepresentation of gene ontology pathways within a set
of transcripts and provides statistical evaluation of the presence of these
transcripts. Statistical significance was based on P 
.05.
Analyses were made with the use of the filtered set of 5824 genes.
Clustering analysis was next performed: By tagging the statistically significant genes and then isolating genes that cluster alongside them, it is possible to identify additional genes that are potentially coregulated. These can be missed by statistical group means testing but could be biologically interesting. This was carried out with the open source Java application Biolayout Express3D (Goldovsky et al, 2005; Freeman et al, in press), which is specifically designed for the construction, visualization, clustering, and analysis of transcription networks generated from microarray datasets. The networks created by the software consist of nodes (representing transcripts) that are connected by edges (which represent similarities in expression profiles across multiple conditions). The analysis was performed on the entire RMA normalized data set (54 675 gene probes) from all of the sample arrays (n = 20). The statistically significant genes in the original differential expression list were flagged, allowing them to be identified in the network graphs produced in Biolayout Express3D. Clustering graphs where then produced with a Pearson threshold of 0.9 and a Markov Clustering Algorithm (MCL) inflation of 1.7, and those containing the flagged gene set were identified. Additional coregulated genes were thus identified, and the extended probe lists from these clusters were then analyzed with DAVID pathway analysis software to identify enrichment of gene ontologies within the lists.
Finally, an additional nonparametric means of verification of the data was obtained by assessing differential gene expression of the original data with the independent rank product method (Breitling et al, 2004). Briefly, this technique ranks each gene in each of the replicate samples. The product of these ranks is then calculated. Statistical significance is assigned by comparing the observed rank product value with a bootstrapped null distribution of rank product values for each gene, and the genes are sorted, with the highest ranking genes being placed at the top of the list.
Real-Time Quantitative Reverse Transcriptase Polymerase Chain Reaction
First-strand cDNA (+/– reverse transcriptase), prepared from all
testis biopsy samples as described previously
(Walton et al, 2006), was used
to confirm array results by real-time quantitative reverse transcriptase
polymerase chain reaction (qRT-PCR). This was performed with the Lightcycler
(Roche Diagnostics, East Sussex, United Kingdom). Reverse-transcribed RNA
samples were diluted 1:10 in nuclease-free water (Promega Ltd, Southampton,
United Kingdom). Diluted first strand cDNA (1 µL) was added to a final
volume of 10 µL containing 50 µg/mL BSA and 0.5 µM each of forward
and reverse primer in 1x Platinum SYBR Green qPCR SuperMix UDG
(Invitrogen, Paisley, United Kingdom) in duplicate. Primers
(Table 1) were either
previously published or designed with the use of online Primer3 software. mRNA
concentration was calculated relative to that of the ribosomal protein RPL32.
RPL32 had previously proved to be the most consistent reference gene
for these samples (Walton et al,
2006).
|
Results of these analyses are presented
(
± SEM). Treatment effects on
gene expression data were initially compared by analysis of variance (ANOVA).
Significant treatment group effects suggested by ANOVA were further
investigated by unpaired t tests, with cube root transformation to
normalize the distribution. For all comparisons, P < .05 was
considered significant.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Preliminary Array Analysis
Five high-quality RNA samples were chosen from each treatment group
(control, CT, and CTD) along with RNA from each of the 5 nonsuppressors for
analysis on Affymetrix microarrays; thus, a total of 20 samples were analyzed.
After data processing and analysis, 1 sample from the CT group showed a poor
RNA digestion plot and was excluded from further analysis. After data quality
control, normalization, and nonspecific filtering, 5824 probes—10.6%
from the total 54 675 probes on the chip—showed expression of greater
than log2(100) for at least 1 sample.
These 5824 probes (supplementary Table 1) represent the predominant genes expressed in the adult human testis. Assignment of gene ontology pathways to each gene allowed them to be grouped and for biological processes statistically overrepresented in the testis to be identified (Table 2). Not surprisingly, these included genes involved in sexual reproduction, with cell death genes also highly represented. Other significant processes were growth regulation, enzyme regulation, cell physiology, and metabolism.
|
A striking finding on initial analysis was that interindividual variability was remarkably low. The median coefficient of variance within each group was 6.5% for controls, 6.7% for the CT group, and 7.8% for the CTD group, with 95th percentile CV values of 14%, 12.9%, and 15.2%, respectively. The average correlation between samples in a group was greater than 0.99 for control and CT groups and greater than 0.98 for the CTD group. This level of variation is in marked contrast to other tissue biopsy types, which show far greater variability (Whitney et al, 2003; Radich et al, 2004; Critchley et al, 2006; P.G., unpublished data) and supports the biological significance of the array data.
Analysis of the Effect of Gonadotropin Suppression
Expression of very few genes was altered between the groups, as
demonstrated by volcano plots (Figure 1a
and b). Few genes showed significant 2-fold or greater differences
between controls and treatment groups, although a greater number of genes
showed significant differences less than 2-fold. No significant differences
were identified between CT and CTD treatment groups at this stage
(Figure 1c).
|
|
|
Analysis by Degree of Spermatogenic Suppression
Stage 2 testing (excluding men showing poor suppression of spermatogenesis)
identified a further 5 genes (SC4MOL, GSTA2, CHGA, PRPS2, CYB5;
Table 3, upper section) that
were significantly down-regulated at least 2-fold in the CT group relative to
controls. Similar down-regulation in the CTD group did not reach significance,
but there was no difference between CT and CTD groups.
|
No statistically significant differences were identified by the stage 3 test comparing sperm nonsuppressors and suppressors in the 2 treatment groups. However, 4 genes (PROK1, PGRMC1, APOE, RHOB; Table 3, lower section) that had shown small (<2-fold) but statistically significant reduction in expression between controls and the CT group, but not the CTD group, after stage 1 testing also showed slightly higher expression in CTD sperm suppressed samples compared with nonsuppressed samples. Although these array-based differences were not significant, this potentially intriguing effect of added progestogen was investigated further by qRT-PCR analysis on the full set of 29 samples (Figure 3B). This demonstrated that PROK1, PGRMC1, and RHOB in fact showed significant 2–3-fold down-regulation between control and all treated samples from both groups, with no difference between nonsuppressors and suppressors in either. APOE showed a 2–3-fold down-regulation that was only significant in the CT group (P = .04), as in the arrays, but did not differ significantly between nonsuppressors and suppressors within the CTD group (P = .8).
Given the dominance of GnRH suppression over progestogen-mediated effects and the absence of differences between the CT and CTD groups in terms of global gene expression, a final round of testing compared control samples with pooled CT and CTD groups. This identified no additional genes.
TAF4B is a germ cell–specific transcription factor that has recently been shown in the ovary to regulate expression of a number of genes, including INHA and RHOB (Geles et al, 2006). Because both INHA (Walton et al, 2006) and RHOB (this study) were down-regulated in the treatment groups, it seemed a likely possibility that it occurs through regulation of TAF4B. Although not identified in the array analysis (it neither passed the nonspecific filter nor showed an at least 2-fold change between groups), with the use of qRT-PCR (Figure 3C), we demonstrated a small (30%) but significant reduction in TAF4B expression in both treatment groups.
Clustering Analysis
Clustering analysis was also performed on the entire RMA-normalized dataset
with Biolayout Express3D
(Goldovsky et al, 2005;
Freeman et al, in press). The networks produced by the software consist of
nodes that represent individual transcripts connected by edges representing
similarities in their expression profiles across multiple conditions. This
approach allows the rapid identification of biological relationships that
could be missed by conventional analysis techniques (ie, statistical 2-sample
tests). Several small cluster graphs and 1 major graph of genes that contained
19 out of the 20 probes from the original gene list were produced. Following
clustering with MCL, the statistical hits (x20 probes) separated out
into 2 distinct clusters. The first cluster (cluster 1,
Table 4) consisted of 66
probes, including 4 of the original statistically significant probes
(CT45-1 [x2], SYCP1, and DPEP3). The second
cluster (cluster 2, Table 5)
consisted of 44 probes, including 9 of the original statistically significant
probes (INSL3 [x3], DHCR24, PAPSS2, CYP11A1, STAR,
CYP17A1, and HSD17B6). With the use of DAVID pathway analysis
software, cluster 1 showed enrichment for genes involved in spermatogenesis,
male gamete generation, and gametogenesis, including early meiotic pathways
(Table 6), whereas cluster 2
showed enrichment for genes involved in steroid, lipid, and cholesterol
biosynthesis/metabolism (Table
7).
|
|
|
|
Rank Product Analysis
Finally, as an independent means of assessing differential gene expression,
the original data were subjected to rank product analysis
(Breitling et al, 2004). The
gene list generated by this method (data not shown) was very similar to the
expanded gene lists generated from Biolayout Express3D, indicating
that the additional genes identified are likely to have some biological
significance. A summary of the notable genes identified in the screens, along
with potential functions and sites of expression within the testis, is
provided in Table 8.
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
The striking homogeneity both between individuals and between treatment groups could paradoxically reflect heterogeneity within the testis. Marked histological variation in spermatogenic activity between nearby tubules has been observed in men treated with regimens similar to those here (McLachlan et al, 2002). Thus, the biopsies will contain tubules with varying suppression that are then averaged during processing, potentially leading to smaller apparent changes in gene expression within the tubule compartments. Sertoli cell and spermatogonial genes are also likely to be highly underrepresented in RNA from total testis biopsies because the major cellular component will be postmeiotic germ cells, thus reducing the sensitivity to changes in these cells. This is a limitation of the use of whole-testis biopsies. Meanwhile, more consistent changes will be observed in genes confined to Leydig cells and thus will be more pronounced. In addition, some genes might be expressed in many cell types but only regulated by gonadotropins in, for example, Leydig cells. This will lead to a robust expression level that partially masks changes in one cell type between control and treated samples.
Previously (Walton et al,
2006), we identified a small number of genes by qRT-PCR in which
expression was reduced in at least 1 of the treatment groups, but only 1 of
these (CYP17A1) overlapped with the genes identified here by array
analysis. Probes for all of the other genes were present on the arrays, but in
a number of cases (SRD5A1, HSD3B2, MAGEA4), their expression level
was below the filtering threshold; therefore, they were excluded from the
analysis. The remaining genes (INHA, PEPP-2, ACRBP) were present in
the filtered gene set but did not show a greater than 2-fold change in
hybridization and did not show statistical significance better then P
.05. Although a very useful tool, it is well known that arrays are less
sensitive than qRT-PCR in detecting small changes in gene expression,
especially with modest numbers of clinical samples, and many of the changes we
observed by qRT-PCR were only around 2-fold and therefore close to the cut-off
level. In addition, these genes are expressed in the tubule compartment so
intertubule variation in suppression as mentioned above could have had an
effect here.
Overall, the most significant changes identified were in expression of genes involved in cholesterol biosynthesis/metabolism and steroidogenesis, which are directly sensitive to the concentration of LH acting on Leydig cells. Also within this group, SC4MOL, which is involved in cholesterol biosynthesis (Li and Kaplan, 1996), and CYB5, the product of which binds and allosterically modulates CYP17 in Leydig cells (Dharia et al, 2004), showed small changes in expression. APOE has many functions, but its role in cholesterol homeostasis (Levi et al, 2005) could also be relevant in this context. The observation that APOE was significantly down-regulated in the CT but not the CTD group suggests that progestogens, known to regulate Leydig cell steroidogenesis (El-Hefnawy et al, 2000), might affect its expression.
The only other gene observed to show a dramatic reduction in expression was INSL3, which is also expressed in Leydig cells. Serum concentrations were not affected by hCG administration to normal men (Bay et al, 2005); however, LH suppression resulted in reduced INSL3 concentrations (Bay et al, 2006). INSL3 is known to be induced by the transcription factor NR4A1 (Nur77) (Robert et al, 2006), which is expressed at low levels but is rapidly and robustly induced by LH and cyclic adenosine monophosphate analogs and is thought to mediate the dynamic expression of INSL3 in response to LH (Robert et al, 2006). The steroidogenic enzyme genes 3BHSD2, CYP17A1, and STAR are also activated by NR4A1 (Zhang and Mellon, 1997; Martin and Tremblay, 2005). Thus, reduced LH will lead to reduced NR4A1 (expression of which was below the threshold to detect changes in the array), which will produce lower levels of 3BHSD2, CYP17A1, STAR, and INSL3, all of which we have observed here or reported previously (Walton et al, 2006). Recently, it has been demonstrated that serum INSL3 concentrations are higher in nonsuppressors than suppressors on male contraceptive regimens (Amory et al, 2007). In this context, it is notable that we also observed higher levels of INSL3 transcripts in nonsuppressors compared with suppressors in the CT group (but not the CTD group).
The receptor for INSL3, LGR8, is expressed in both Leydig cells and in meiotic and postmeiotic germ cells but not in Sertoli or peritubular cells (Kawamura et al, 2004; Anand-Ivell et al, 2006). Insl3 knockout mice show cryptorchidism and INSL3 regulates apoptosis in germ cells in both male and female mice (Kawamura et al, 2004). Because germ cells do not express gonadotropin receptors, germ cell genes cannot be directly regulated by gonadotropins. INSL3 might therefore provide a link between Leydig cells and spermatogenesis, which could contribute to the changes in gene expression levels observed here in the germ cell compartment, such as for SYCP1, a gene essential for chromosomal synapsis at prophase I of meiosis and detected in the stage 1 screen, as well as SYCP3, DMC1, DAZ1, MSH5, TEX11, TEX14, TEX101, and HOR-MAD1 and 2, which were identified from the cluster analysis. We detected previously (Walton et al, 2006) a reduced expression of ACRBP, a spermatocyte-specific marker in the CTD and suppressor groups, and of MAGEA4, which is present in spermatogonia and primary spermatocytes, in the suppressor group. In contrast, expression of PRM1, a postmeiotic germ cell marker, remained unchanged. Together with these new early meiotic genes, this further implicates an effect of the treatments on early stages of spermatogenesis. What is currently less clear is whether the reduced expression of germ cell markers after treatment is due to small changes in gene expression within germ cells or loss of a percentage of early germ cells, as has been observed at the histological level under similar regimens (McLachlan et al, 2002). Both would yield the same result. This could be resolved by histochemical analysis of biopsies for these gene products, but insufficient material was available to us for this.
Further genes known to be expressed by Leydig cells and showing significant down-regulation during gonadotropin suppression were CHGA and PROK1. Expression of CHGA correlates with LH levels (Ortega et al, 2004). However, in the bovine testis, CHGA expression is also high in spermatogonia (Payan-Carreira et al, 2006) with a reducing gradient of expression with maturation up to the round spermatid stage. Elongating and elongated spermatids do not express CHGA. Thus, reduced expression of CHGA in the treated men might also reflect suppression of spermatogenesis. PROK1, a prokineticin, encodes EG-VEGF, which is expressed in Leydig cells and is thought to be involved in the integrity and proliferation of blood vessels in the testis (Samson et al, 2004). Knock-out in mice of 1 of its receptors, Pkr2, causes severe atrophy of the reproductive system (Matsumoto et al, 2006), an effect partly due to low concentrations of gonadotropins associated with lack of GnRH but also thought to have a direct testicular component. In the ovary, expression of PROK1 positively correlates with granulosa cell steroidogenesis (Kisliouk et al, 2003).
PGRMC1 (HPR6) encodes a component of the putative membrane progesterone receptor and is able to activate cytochrome P450 enzymes (Crudden et al, 2005). A reduction in its expression might therefore contribute to reduced expression of such enzymes. The array data suggested suppression in the CT but not CTD groups, but qRT-PCR analysis demonstrated similar suppression in both groups.
Although small changes in expression could simply be due to loss of a proportion of some germ cell types, alteration in the dynamics of Sertoli cell–germ cell adherens junctions was suggested by changes in expression of RHOB, a member of the RAS gene family (Lui et al, 2003). The breakdown of adherens junctions is necessary for germ cells to detach from Sertoli cells at spermiation, suggesting the possibility that reduction in expression of RHOB and its associated regulatory transcription factor TAF4B (Geles et al, 2006) could block this process as a mechanism for the failure of spermiation observed during male contraceptive regimens (Matthiesson and McLachlan, 2006). RhoB expression at adherens junctions is increased by 1-(2, 4-dichlorobenzyl)-indazole-3-carbohydrazide (adjudin), which interferes with adhesion of spermatids and spermatocytes to Sertoli cells, and hence their maturation, and is a promising approach to reversible male contraception (Lui et al, 2003; Mruk et al, 2006). The WNT1 pathway is also involved in the regulation of adherens junctions (Wechezak and Coan, 2003), and WISP2, a component of the WNT1 signaling pathway (Banerjee et al, 2003), was reduced in both treatment groups. A further potential gene associated with regulation of adherens junctions and the only gene identified as increased in both treatment groups is CST9L, a cysteine protease inhibitor (Siu and Cheng, 2004).
Of the remaining genes in which levels were observed to fall in the treated groups, PAPSS and PRPS2 are genes involved in nucleotide metabolism, and GSTA2 and DPEP3 have a function in glutathione metabolism. Both of these biological processes were found to be statistically overrepresented in the testis when higher level gene ontology and KEGG (Kyoto Encyclopedia of Genes and Genomes) analysis was performed on the expressed gene set (P = .03 and .05, respectively), indicating their importance in the testis.
Very few genes were differentially suppressed between the 2 treatment groups. Some were only significantly different from controls in one group (eg, CT group: APOE, HSD17B6, STAR, PRPS2, GSTA2; CTD group: SYCP1, DPEP3), but qPCR analysis showed significantly different expression in CT compared with CTD in only 1 gene, FLJ35767. This is an expressed sequence tag (EST) isolated from a human testis library in which the predicted coding sequence shows similarity to 2 mouse proteins, Tex19 and a hypothetical protein. Tex19 is expressed in spermatogonia (Wang et al, 2001). The finding that FLJ35767 is down-regulated further by desogestrel than by a GnRH antagonist and testosterone alone suggests a progestogenic effect, as we have previously reported for SRD5A1 (Walton et al, 2006). This provides further evidence for direct progestogenic effects on the testis. However, it is striking that only 1 such gene was identified in this study, indicating that it might be of limited biological significance. The absence of any significant differences between suppressors and nonsuppressors (other than for INSL3) within the treatment groups was disappointing. Given that the data were so tight between individuals, it seems likely that an array approach will require more refined tissue sampling to be of value in this key issue.
In this study, we identify the 10% of genes most highly expressed in the human testis. Only a small subset of genes was observed to show changes in expression after 4 weeks of gonadotropin suppression. Functional grouping indicates that genes associated with steroidogenesis are most markedly regulated by short-term gonadotropin suppression, with other Leydig cell genes also highly regulated, most notably INSL3. INSL3 might be a candidate for future contraceptive targeting. Within the tubule compartment, genes involved in early meiosis showed reduced expression, with regulation of a number of genes that might be involved with adherens junctions also identified. Such changes could underlie the observation that disruption of spermiation is an early component of the response to gonadotropin withdrawal. Future studies examining histological expression of these genes should help define whether they are down-regulated in individual cells or whether certain early germ cell stages expressing them are lost because of a lack of gonadotropin support.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Anand-Ivell RJ, Relan V, Balvers M, Coiffec-Dorval I, Fritsch M,
Bathgate RA, Ivell R. Expression of the insulin-like peptide 3 (INSL3)
hormone-receptor (LGR8) system in the testis. Biol
Reprod. 2006;74: 945
–953.
Anderson RA, Baird DT. Male contraception. Endocr
Rev. 2002;23: 735
–762.
Aquila S, Middea E, Catalano S, Marsic S, Lanzino M, Casaburi I,
Barone I, Bruno R, Zupo S, Ando S. Human sperm express a functional androgen
receptor: effects on PI3K/AKT pathway. Hum Reprod. 2007; 22: 2594
–2605.
Banerjee S, Saxena N, Sengupta K, Tawfik O, Mayo MS, Banerjee SK. WISP-2 gene in human breast cancer: estrogen and progesterone inducible expression and regulation of tumor cell proliferation. Neoplasia. 2003;5: 63 –73.[Medline]
Bay K, Hartung S, Ivell R, Schumacher M, Jurgensen D, Jorgensen N,
Holm M, Skakkebaek NE, Andersson AM. Insulin-like factor 3 serum levels in 135
normal men and 85 men with testicular disorders: relationship to the
luteinizing hormone-testosterone axis. J Clin Endocrinol
Metab. 2005;90: 3410
–3418.
Bay K, Matthiesson KL, McLachlan RI, Andersson AM. The effects of
gonadotropin suppression and selective replacement on insulin-like factor 3
secretion in normal adult men. J Clin Endocrinol
Metab. 2006;91: 1108
–1111.
Benjamini Y, Yekutieli D. Quantitative trait Loci analysis using
the false discovery rate. Genetics. 2005; 171: 783
–790.
Breitling R, Armengaud P, Amtmann A, Herzyk P. Rank products: a simple, yet powerful, new method to detect differentially regulated genes in replicated microarray experiments. FEBS Lett. 2004; 573: 83 –92.[CrossRef][Medline]
Chalmel F, Rolland AD, Niederhauser-Wiederkehr C, Chung SSW,
Demougin P, Gattiker A, Moore J, Patard J-J, Wolgemuth DJ, Jégou B,
Primig M. The conserved transcriptome in human and rodent male gametogenesis.
Proc Natl Acad Sci U S A. 2007; 104: 8346
–8351.
Collins LL, Lee HJ, Chen YT, Chang M, Hsu HY, Yeh S, Chang C. The androgen receptor in spermatogenesis. Cytogenet Genome Res. 2003;103: 299 –301.[CrossRef][Medline]
Critchley HO, Robertson KA, Forster T, Henderson TA, Williams AR, Ghazal P. Gene expression profiling of mid to late secretory phase endometrial biopsies from women with menstrual complaint. Am J Obstet Gynecol. 2006;195: 406 , e1–16.[CrossRef][Medline]
Crudden G, Loesel R, Craven RJ. Overexpression of the cytochrome p450 activator hpr6 (heme-1 domain protein/human progesterone receptor) in tumors. Tumour Biol. 2005; 26: 142 –146.[CrossRef][Medline]
Dharia S, Slane A, Jian M, Conner M, Conley AJ, Parker CRJr.
Colocalization of P450c17 and cytochrome b5 in androgen-synthesizing tissues
of the human. Biol Reprod. 2004; 71: 83
–88.
El-Hefnawy T, Manna PR, Luconi M, Baldi E, Slotte JP, Huhtaniemi I.
Progesterone action in a murine Leydig tumor cell line (mLTC-1), possibly
through a nonclassical receptor type. Endocrinology. 2000; 141: 247
–255.
Freeman TC, Goldovsky L, Brosch M, van Dongen S, Maziére P, Grocock RJ, Freilich S, Thornton J, Enright AJ. Construction, visualisation and clustering of transcription networks from microarray expression data. PLoS Comput Biol. 2007; 3(10): e20 .[CrossRef][Medline]
Geles KG, Freiman RN, Liu WL, Zheng S, Voronina E, Tjian R.
Cell-type–selective induction of c-jun by TAF4b directs ovarian-specific
transcription networks. Proc Natl Acad Sci U S A. 2006; 103: 2594
–2599.
Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, et al. (2004). Bioconductor: open software development for computational biology and bioinformatics. Genome Biol. 2004; 5 : R80.[CrossRef][Medline]
Goldovsky L, Cases I, Enright AJ, Ouzounis CA. BioLayout (Java): versatile network visualisation of structural and functional relationships. Appl Bioinformatics. 2005; 4: 71 –74.[CrossRef][Medline]
Habib GM, Shi ZZ, Cuevas AA, Lieberman MW. Identification of two
additional members of the membrane-bound dipeptidase family. FASEB
J. 2003;17: 1313
–1315.
Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, Speed TP. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics. 2003; 4: 249 –264.[Abstract]
Kawamura K, Kumagai J, Sudo S, Chun SY, Pisarska M, Morita H,
Toppari J, Fu P, Wade JD, Bathgate RA, Hsueh AJ. Paracrine regulation of
mammalian oocyte maturation and male germ cell survival. Proc Natl
Acad Sci U S A. 2004;101: 7323
–7328.
Kisliouk T, Levy N, Hurwitz A, Meidan R. Presence and regulation of
endocrine gland vascular endothelial growth factor/prokineticin-1 and its
receptors in ovarian cells. J Clin Endocrinol Metab. 2003; 88: 3700
–3707.
Levi O, Lutjohann D, Devir A, von Bergmann K, Hartmann T, Michaelson DM. Regulation of hippocampal cholesterol metabolism by apoE and environmental stimulation. J Neurochem. 2005; 95: 987 –997.[CrossRef][Medline]
Li L, Kaplan J. Characterization of yeast methyl sterol oxidase
(ERG25) and identification of a human homologue. J Biol
Chem. 1996;271: 16927
–16933.
Lui WY, Lee WM, Cheng CY. Sertoli–germ cell adherens junction
dynamics in the testis are regulated by RhoB GTPase via the ROCK/LIMK
signaling pathway. Biol Reprod. 2003; 68: 2189
–2206.
Martin LJ, Tremblay JJ. The human 3beta-hydroxysteroid
dehydrogenase/Delta5-Delta4 isomerase type 2 promoter is a novel target for
the immediate early orphan nuclear receptor Nur77 in steroidogenic cells.
Endocrinology. 2005; 146: 861
–869.
Matsumoto AM, Karpas AE, Paulsen CA, Bremner WJ. Reinitiation of sperm production in gonadotropin-suppressed normal men by administration of follicle-stimulating hormone. J Clin Invest. 1983; 72: 1005 –1015.[Medline]
Matsumoto AM, Paulsen CA, Bremner WJ. Stimulation of sperm
production by human luteinizing hormone in gonadotropin-suppressed normal men.
J Clin Endocrinol Metab. 1984; 59: 882
–887.
Matsumoto S, Yamazaki C, Masumoto KH, Nagano M, Naito M, Soga T,
Hiyama H, Matsumoto M, Takasaki J, Kamohara M, Matsuo A, Ishii H, Kobori M,
Katoh M, Matsushime H, Furuichi K, Shigeyoshi Y. Abnormal development of the
olfactory bulb and reproductive system in mice lacking prokineticin receptor
PKR2. Proc Natl Acad Sci U S A. 2006; 103: 4140
–4145.
Matthiesson KL, McLachlan RI. Male hormonal contraception: concept
proven, product in sight? Hum Reprod Update. 2006; 12: 463
–482.
McLachlan RI, O'Donnell L, Stanton PG, Balourdos G, Frydenberg M,
de Kretser DM, Robertson DM. Effects of testosterone plus medroxyprogesterone
acetate on semen quality, reproductive hormones, and germ cell populations in
normal young men. J Clin Endocrinol Metab. 2002; 87: 546
–556.
McLachlan RI, Robertson DM, Pruysers E, Ugoni A, Matsumoto AM,
Anawalt BD, Bremner WJ, Meriggiola C. Relationship between serum gonadotropins
and spermatogenic suppression in men undergoing steroidal contraceptive
treatment. J Clin Endocrinol Metab. 2004; 89: 142
–149.
Mruk DD, Wong CH, Silvestrini B, Cheng CY. A male contraceptive targeting germ cell adhesion. Nat Med. 2006; 12: 1323 –1328.[CrossRef][Medline]
Nieschlag E, Zitzmann M, Kamischke A. Use of progestins in male contraception. Steroids. 2003; 68: 965 –972.[CrossRef][Medline]
Ortega HH, Lorente JA, Mira GA, Baravalle C, Salvetti NR. Constant light exposure causes dissociation in gonadotrophin secretion and inhibits partially neuroendocrine differentiation of Leydig cells in adult rats. Reprod Domest Anim. 2004; 39: 417 –423.[CrossRef][Medline]
Payan-Carreira R, Rodrigues P, Carvalho PR. Chromogranin-A expression in the bovine testis. Anim Reprod Sci. 2006; 96: 146 –153.[CrossRef][Medline]
Radich JP, Mao M, Stepaniants S, Biery M, Castle J, Ward T, Schimmack G, Kobayashi S, Carleton M, Lampe J, Linsley PS. Individual-specific variation of gene expression in peripheral blood leukocytes. Genomics. 2004;83: 980 –988.[CrossRef][Medline]
Robert NM, Martin LJ, Tremblay JJ. The orphan nuclear receptor
NR4A1 regulates insulin-like 3 gene transcription in Leydig cells.
Biol Reprod. 2006; 74: 322
–330.
Samson M, Peale FV, Jr., Frantz G, Rioux-Leclercq N, Rajpert-De
Meyts E, Ferrara N. Human endocrine gland-derived vascular endothelial growth
factor: expression early in development and in Leydig cell tumors suggests
roles in normal and pathological testis angiogenesis. J Clin
Endocrinol Metab. 2004;89: 4078
–4088.
Shah C, Modi D, Sachdeva G, Gadkar S, Puri C. Coexistence of
intracellular and membrane-bound progesterone receptors in human testis.
J Clin Endocrinol Metab. 2005; 90: 474
–483.
Siu MK, Cheng CY. Interactions of proteases, protease inhibitors,
and the beta1 integrin/laminin gamma3 protein complex in the regulation of
ectoplasmic specialization dynamics in the rat testis. Biol
Reprod. 2004;70: 945
–964.
Smyth GK. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol. 2004;3: article 3 .[Medline]
Themmen APN, Huhtaniemi IT. Mutations of gonadotropins and
gonadotropin receptors: elucidating the physiology and pathophysiology of
pituitary-gonadal function. Endocr Rev. 2000; 21: 551
–583.
Walton MJ, Bayne RA, Wallace I, Baird DT, Anderson RA. Direct
effect of progestogen on gene expression in the testis during gonadotropin
withdrawal and early suppression of spermatogenesis. J Clin
Endocrinol Metab. 2006;91: 2526
–2533.
Wang PJ, McCarrey JR, Yang F, Page DC. An abundance of X-linked genes expressed in spermatogonia. Nat Genet. 2001; 27: 422 –426.[CrossRef][Medline]
Wechezak AR, Coan DE. Subcellular distribution of Wnt-1 at adherens junctions and actin-rich densities in endothelial cells. Exp Cell Res. 2003;288: 335 –343.[CrossRef][Medline]
Whitney AR, Diehn M, Popper SJ, Alizadeh AA, Boldrick JC, Relman
DA, Brown PO. Individuality and variation in gene expression patterns in human
blood. Proc Natl Acad Sci U S A. 2003; 100: 1896
–1901.
Zhang P, Mellon SH. Multiple orphan nuclear receptors converge to
regulate rat P450c17 gene transcription: novel mechanisms for orphan nuclear
receptor action. Mol Endocrinol. 1997; 11: 891
–904.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
