This web only provides the extract of this article. If you want to read the figures and tables, please reference the PDF full text on Blackwell Synergy. Thank you.
- Original Article -
Identification of a novel testis-specific gene and its potential
roles in testis development/spermatogenesis
Lan-Lan Yin, Jian-Min Li, Zuo-Min Zhou, Jia-Hao Sha
Key Laborary of Reproductive Medicine, Nanjing Medical University, Hangzhong Road, Nanjing 210029, China
Abstract
Aim: To identify and characterize a novel gene with potential roles in testis development and spermatogenesis.
Methods: A cDNA microarray was constructed from a human testis large insert cDNA library and hybridized with
probes of human or mouse adult and fetal testes. Differentially expressed genes were isolated and sequenced. RT-PCR
was used to test the tissue distribution of the genes of interest and
in situ hybridization was performed to localize the
gene expression in the mouse testis. A range of bioinformatical programs including Gene Runner, SMART, NCBI
Blast and Emboss CpGPlot were used to characterize the new gene's feature.
Results: A novel testis-specific gene, NYD-SP5, was differentially expressed in fetal and adult testes. The deduced protein structure of NYD-SP5 was
found to contain an IQ motif (a short calmodulin-binding motif containing conserved Ile and Gln residues), a
Carbamate kinase-like domain, a Zn-dependent exopeptidase domain and a lactate dehydrogenase (LDH) C-terminal-like
domain. RT-PCR analysis revealed that NYD-SP5 was predominantly expressed in the testis but not in other 15 tissues
examined. In situ hybridization and RT-PCR examinations revealed that the expression of NYD-SP5 was confined in
the male germ cell but not present in the somatic cell in the testes.
Conclusion: NYD-SP5 is a newly found
testis-specific gene with potential roles in testis development and spermatogenesis through a calmodulin-activated
enzyme.(Asian J Androl 2005 Jun; 7: 127-137)
Keywords: spermatogenesis; testis; calmodulin
Correspondence to: Dr Jia-Hao Sha, Key Laborary of Reproductive Medicine, Nanjing Medical University, Hanzhong Road, Nanjing 210029, China
Tel/Fax: +86-25-8686-2908
E-mail: shajh@njmu.edu.cn
Received 2004-09-28 Accepted 2005-02-04
DOI: 10.1111/j.1745-7262.2005.00041.x
1 Introduction
A central question in developmental genetics is how
a complex organism with structurally, morphologically
and functionally distinct tissues and organs can be
derived from a single-cell zygote. The formation of differ
ent organs and tissues is based primarily on differential
gene expression. While "housekeeping genes" that
contribute to basic structural or metabolic cellular functions
are expressed ubiquitously throughout the body,
"tissue-specific" genes that contribute to specialized functions
in differentiated cell types are expressed in a regulated
fashion. Spermatogenesis, a complex process leading to
the formation of male gametes, has been considered as a
model system for developmental analysis of regulatory
mechanisms associated with tissue-specific gene
expression [1] because spermatogenesis is characterized by the
expression of many genes that either are not expressed
in any other cell type or are expressed in only very few
other cell types. Most of these genes also exhibit
stage-specific expression during spermatogenesis, which could
be considered as spermatogenic cell type-specific since
the occurrence of different spermatogenic cell types is
also stage-specific during spermatogenesis. Therefore,
the spermatogenic cell lineage has provided a unique
opportunity for developmental analysis of tissue-specific
gene expression and the governing regulatory mechanisms.
In order to identify developmentally regulated genes
in the testis, we have constructed cDNA microarrays
from a human testis large insert cDNA library [2]. cDNA
probes from human fetal and adult testes were used to
hybridize the cDNA microarray. NYD-SP5 was one of the clones, which was identified as a differentially
expressed gene with higher intensity in the adult testis than
fetal testis. Additionally, the tissue distribution and
cellular localization, together with protein structure
prediction results, indicated that NYD-SP5 was a novel
testis-specific gene with a potential calmodulin-binding region.
Calcium plays a central role in spermatogenesis [3],
spermiogenesis [4] and following fertilization [5]. Many of
the calcium activating events are mediated by the
intracellular calcium receptor calmodulin (CaM), which when
bound to calcium can activate a variety of enzymes,
including protein kinases, phosphatases and
phosphodiesterases [6]. Because CaM is present in all tissues,
cell-type-specific functions are determined by the
complement of its downstream targets [7, 8]. The newly found
CaM binding protein, NYD-SP5, in a spermatogenic cell
would point to a new insight into the mechanism of
spermatogenesis regulation through CaM.
2 Materials and methods
2.1 Samples
Informed consent was received from either the
participants or their kin and the ethics committee of Nanjing
Medical University (China) granted research approval
prior to sample collection. Human adult testes (28 years
old and 37 years old) were obtained from the Body
Donor Center (Nanjing Medical University) and fetal testes
were obtained from accidentally aborted (as a
consequence of road accidents) 6-month-old fetuses (Clinical
Reproductive Center, Nanjing Medical University).
Testis tissue samples from five individuals with
Sertoli-cell-only syndrome (SCOS) were acquired via biopsy and
health volunteers with proven fertility and normal semen
quality (assessed by WHO criteria, 1999) donated ejacu
lated sperm.
2.1 Preparation of human testis cDNA microarray
The testis cDNA microarray was constructed as
described previously [2]. Briefly, this microarray contained
9216 cDNA clones that were derived from a human testis
5'-STRETCH PLUS cDNA library (Clontech, Palo Alto, CA,
USA [source of insert cDNA came from 25 Caucasians
aged from 20 to 65 years] ). The inserts were amplified by
PCR using 5'-CCATTGTGTTGGTACCCGGGAATTCG-3' (P1) as a forward primer and 5'-ATAAGCTTGC
TCGAGTCTAGAGTCGAC-3' (P2) as a reverse primer. PCR products were used to make the human testis cDNA
microarray. The microarray was hybridized with human
testis cDNA probes from deceased adults and accidentally
aborted 6-month-old fetuses. The human testis cDNA
microarrays were also hybridized with probes prepared from
the testes of 1- and 4-week-old mice to screen for
homologous genes in testis development.
2.3 Sequence identification and analysis
For differentially expressed genes, cDNA clones were
isolated for further analysis. The amplified cDNA
plasmids were isolated and purified (QIAprep Spin Miniprep
Kit, Qiagen, Hilden, Germany) and the inserts were
sequenced using the ABI 377 automatic sequencing
machine (Perkin-Elmer, Norwalk, USA). For each clone,
sequence homologies were searched in the databases of
GenBank. The nucleic and deduced amino acid sequences
were also analyzed using a range of bioinformatical
programs including Gene Runner
(http://www.generunner.com),SMART
(http://www.smart.embl-heidelberg.de),NCBI Blast
(http://www.ncbi.nlm.nih.gov/blast),and Emboss CpGPlot
(http://www.cbi.pku.edu.cn/tools/EMBOSS/cpgreport).
2.4 Analysis of NYD-SP5 gene expression in different
tissues by RT-PCR
After sequence identification and analysis, a novel
testis specific gene, named NYD-SP5, was found. The
expression profile of NYD-SP5 was determined by using PCR screening. Multiple tissue cDNA panels,
including testis, thymus, small intestine, colon,
spleen, leukocyte, prostate gland, ovary, pancreas, heart, kidney,
lung, placenta, liver, brain and skeletal muscle were purchased
from Clontech (#1420-1). The NYD-SP5 specific primers
were as follows: upstream: 5' CTCACCT TATACCTGACAAACG 3' (nt 2,627- nt 2,648)
and downstream: 5' GCCTTTCCTCAAGATCATAGC 3' (nt 2,853-nt 2,873). The PCR product was 247 bp in size.
G3PDH was used as the positive control. The reagents in
50 mL PCR reaction tubes were as follows:
H2O 33.6 mL, buffer 5 mL, 10 mmol/L dNTP
1 mL, Tag DNA polymerase 0.4 mL, upstream primer 5pmol 2.5 mL, downstream
primer 5 pmol
2.5 mL, and cDNA sample 5 mL. PCR
conditions were as follows: denaturation at 94
¡æ for 1 min, annealing at 56 ¡æ for 30 sec and extension at 72
¡æ for 30 sec. The first cycle had a denaturation period of 5 min;
the last cycle an extension period of 7 min. Thirty-five
cycles of PCR were performed. The PCR products were
analyzed by 1.5 % (w/v) agarose gel electrophoresis.
2.5 Cellular localization of NYD-SP5
2.5.1 Probe preparation
Mouse homologous fragment was prepared by RT-PCR. Mouse testis total RNA was isolated using TRIzol
Reagent (GIBCO BRL, Grand island, NY). Reverse
transcription was performed in 15 ¦ÌL of reaction mixture.
First 1 ¦ÌL of total RNA (about 3 ¦Ìg, 1 ¦ÌL random
primer(0.2 ¦Ìg/mL Sangon, Shanghai, China) and 7 ¦ÌL DEPC
water were mixed and incubated at 70¡æ for 5 min; then
3 ¦ÌL M-MLV RT 5 ¡Á buffer, 0.75 ¦ÌL dNTP (20 mmol/L), 0.35 ¦ÌL RNasin(50 U/¦ÌL), 1 ¦ÌL moloney murine
leukemin virus (M-MLV) Reverse Transcriptase (Promega, Shanghai, China), 1 ¦ÌL DEPC water were
added and incubated at 37¡æ for 1 h, and then
95¡æ for 5 min. The primer sequences for amplification of mouse
NYD-SP5 cDNA were: P1: 5'-CCGATATGCTGAATGTCC3'; P2: 5'-TGTCACAAAATGCTGTCC- 3'.
The desired fragment was 249bp. PCR reaction mixture and conditions
are the same as above except the annealing temperature
was lowered to 54 ¡æ. The PCR products were detected
by staining with ethidium bromide after electrophoresis
on a 1.5 % (w/v) agarose gel and purified by DNA gel
extraction kit (Biorad [Hercules, CA, USA]) according
to the instructions.
T7, Sp6 promoter sequences were added to the 5'
end and 3' end of mouse DNA fragment, respectively
also by PCR reaction. And the PCR products with T7,
Sp6 promoter sequences on both sides were purified and
used as the template in the in vitro transcription (DIG
RNA labeling kit, Roche [Indianapolis, IN, USA]). The
efficiency of thus obtained anti-sense and sense RNA
probes were evaluated by a standard direct detection as
described in The DIG System User's Guide for Filter
Hybridization (Roche)(http://www.roche-applied-science.
com/prodinfo_fst.htm?/PROD_INF/MANUALS/DIG_MAN/dig_toc.htm)
2.6 In situ hybridization
For preparation of paraffin-embedded sections, mouse
testes at 15 days, 30 days, and 60 days were cut and
fixed in 4 % (w/v) paraformaldehyde in Phosphate Buffer
Saline (PBS) at 4¡æ overnight. The fixed testes were
dehydrated with ethanol and embedded in paraffin.
Paraffin-embedded sections of 5 ¦Ìm thickness were cut and
collected on slides pretreated with polylysine.
Sections were deparaffinized in xylene for 10 min
(two changes) and rehydrated through 100 %, 70 %
ethanol (two changes), DEPC H2O and PBS (two changes).
Then, the sections were treated for 10_15 min at
37¡æ with 2 ¦Ìg/mL proteinase K in 10mmol/L Tris, pH 7.5.
After that, the enzyme digestions were stopped and
post-fixed by immersing the sections in 4 %
paraformaldehyde in PBS for 5 min. Finally, the sections were further
washed in PBS (two changes for 5min).
The sections were prehybridized and blocked in
hybridization buffer (DIG Easy Hyb, Roche) without probe
at 42¡æ for 2 h. The hybridization buffer was applied to
each section with an optimal concentration of 100
ng/mL labeled RNA probes. Hybridization was carried out at
58¡æ for 16 h in a humidified chamber. Subsequently,
the sections were washed in 4¡ÁSSC for 5 min, 2¡ÁSSC
for 30 min, 1¡ÁSSC and 0.5¡ÁSSC for 10 min, and twice
in 0.01 mol/L PBS for 10 min. The immunological
detection of the DIG labeled signal was performed as described
by the manufacturer (DIG Nucleic Acid Detection Kit,
Roche).
2.7 Analysis of NYD-SP5 mRNA in normal sperm and
testes of five patients with SCOS
Five male patients with SCOS were recruited in this
study. Tissues from their testes were obtained via
biopsy at the First Affiliated Hospital of Nanjing Medial
University (Nanjing, China) for section pathologic
diagnosis, and RNA was extracted using Trizol reagent.
Total RNA of ejaculated sperm was also extracted with
Trizol reagent. Then total RNA was
reverse-transcripted to cDNA with Avian Myeloblastosis
Virus (AMV) reverse transcriptase. Expression of NYD-SP5 was
determined as follows: The cDNAs were amplified with the
sequence specific primers (P1 and P2) as described above
and PCR products were resolved by electrophoresis; the
testis cDNAs were processed in a similar way to detect
the presence of b-actin mRNA.
3 Results
3.1 cDNA microarray hybridization
The hybridization of the constructed testis cDNA
microarrays with adult and fetal testis probes revealed a
series of clones that were highly expressed in adult but
not fetal testis. Of 9216 clones analyzed, 592 had
intensities at least three times higher for probes prepared from
adult tissue than those from fetus tissue, whereas 139
cDNA clones had at least three times higher signals for
probes prepared from the fetus testis than those from
adult testis. The reciprocal expression characteristics of
these genes indicated that different sets of genes were
involved in different developmental stages. The
hybridized signal intensities from human adult and fetal
testicular probes for one of the clones, named NYD-SP5, were
26.32 and 4.87 respectively, indicating a 5-fold higher
expression in the adult than that in the fetus (Figure 1A,
B). Also, in situ hybridization with mouse testis cDNA
probe the intensities from adult and fetal testes for NYD-
SP5 were 12.37 and 3.0, respectively (Figure1C, D),
showing the same expression time pattern with human
testis.
3.2 Structural features of the cDNA and deduced
protein
NYD-SP5 (GenBank Accession No. AY014282) was found to consist of 3598 nucleotides and contain an open
reading frame of 1027 amino acids (Figure 2). A
nucleotide blast search against the GenBank database revealed
a similar nucleotide sequence in mice (GenBank Accession No. AK019535), and a similar nucleotide sequence
in rats (GenBank Accession No. XM_236331).
Further, NYD-SP5 protein shares 68 % identity and
78 % positive amino acid sequence with its mouse
homolog protein BAB31783 encoded by AK019535, and shares 57 % identity and 67 % positive with its rat
homolog protein XP_236331 encoded by XM_236331 (Figure 3). These data
indicate high levels of homology between NYD-SP5 and mice and
rats at either the nucleotide or the protein level; therefore,
NYD-SP5 is a human_mouse_rat homologous gene. GenBank human
genome database searching mapped NYD-SP5 to chromosome 15q22.31 (NT_086827).
Simple modular architecture research tool (SMART)
predicted a high possibility of an IQ motif (368-390aa)
in the middle of NYD-SP5 protein sequence and three
enzyme domains within NYD-SP5 sequence (Figure 2).
These domains include a carbamate kinase-like domain
located before the IQ motif (269_328aa), a
Zn-dependent exopeptidases domain after the IQ motif
(424_470aa), and a LDH C-terminal domain-like sequence that
lies in the C-terminal (626-672aa) (Figure 2). In addition,
CpG island revealing program, Emboss CpGPlot, reported
two relatively high GC content regions occurred from
-57 to -269 bp and -411 to -629 bp in the upstream of
NYD-SP5 chromosome [8] (Figure 4).
3.3 Expression of NYD-SP5 in normal tissues
Tissue distribution studies using RT-PCR on a
human tissue kit showed that the designed 247 bp product
was expressed predominantly in the testis, but not in the
other 15 tissues examined, including the ovary _ the
female germ-cell-producing organ (Figure 5). Therefore,
NYD-SP5 was identified as a testis-specific gene.
3.4 Cellular localization of NYD-SP5
Using an in situ hybridization technique, we
examined the localization of NYD-SP5 mRNA at mouse testes.
The results showed that the expression of NYD-SP5 was
confined to seminiferous tubules. Strong hybridization
signals from the mouse probe was exclusively localized
in the spermatocytes and spermatids, and no signal above
the background level was detected outside the
seminiferous tubules or in Sertoli-cell at all ages examined, namely
days 15, 30 and 60 (Figure 6). It is suggested that mouse
NYD-SP5 is expressed in the male germ line cell but not
the somatic cell in mouse testis.
3.5 Analysis of NYD-SP8 mRNA in spermatozoa and
the testis of five patients with SCOS.
RT-PCR analysis showed that NYD-SP5 mRNA was also detected in spermatozoa but not in the testes of
patients with SCOS that is characterized histologically by
complete loss of the germinal epithelium in testicular
tubules, and clinically by aspermia (Figure 7). Therefore,
it has been suggested that NYD-SP5 is expressed in germ
cells but not somatic cells in human testes.
4 Discussion
Using cDNA microarrays constructed from the human testis large insert cDNA library, we have identified a novel testis-specific gene, NYD-SP5, which is differentially expressed in fetal and adult testes of humans and mice. The tissue distribution and cellular localization of NYD-SP5 mRNA suggests that it is a male germ line cell specific gene with potential roles in the process of spermatogenesis. The predominant expression of NYD-SP5 mRNA in adult but not fetal testes can be confirmed by the in situ hybridization analysis in mouse testes, showing restricted localization of NYD-SP5 in the seminiferous tubules with signals mainly in primary spermatocytes and advanced spermatogenic cells, as well as ejaculated spermatozoa, which are absent from fetal testes.
Because of the tissue and cell-line-specific expression pattern of NYD-SP5, CpG frequency of the chromosome upstream of NYD-SP5 was examined. There is
a CpG rich area in the promoter region of NYD-SP5,
which might be related to the expression regulation of
the gene. There is evidence indicating that methylation
of the CpG island inhibits the transcription of genes [9].
The regulation of tissue-specific expression of the
Pgk-2 gene, which is expressed only in spermatogenic cells
in eutherian mammals, has been described to be in such
a fashion that its expression is repressed in somatic cells
[1]. The tissue-specific and stage-dependent expression
of NYD-SP5 may also require demethylation of its CpG
rich segment. Further studies along this line with
NYD-SP5 may provide detailed mechanisms for
developmentally regulated gene expression.
As to the function of NYD-SP5 in spermatogenesis,
bioinformatical analysis provided some clues. NYD-SP5
contains three enzyme domains and one IQ motif, which
is in charge of CaM binding. CaM is recognized as a
major calcium sensor and orchestrator of regulatory
events through its interaction with a diverse group of
cellular proteins [6]. Because CaM is present in all tissues,
cell-type-specific functions are determined by the
complement of its downstream targets. In the testis, several
calmodulin binding proteins, such as calspermin, Ca
(2+)/calmodulin-dependent protein kinase IV (CaMKIV), and
testis-specific calcineurin B, were isolated and
demonstrated to be essential for spermatogenesis. For instance,
CaMKIV is expressed in spermatids and targeted to
chromatin and the nuclear matrix [12], and calspermin has
been speculated to play a role in binding and
sequestering CaM during the development of the germ cell [13].
NYD-SP5 might be another member of CaM targets involved in the process of spermatogenesis. The IQ motif,
which NYD-SP5 contained, is one of the three
recognition motifs for CaM interaction and reported as a
consensus for Ca2+-independent binding. Neuromodulin (GAP
43/P-57), neurogranin and Brush Border Myosin I (BBMI), all of which contain an IQ motif, interact with
CaM in the absence of Ca2+ [14]. Neuromodulin is a
major component of the motile "growth cones" that form
the tips of elongating axons and plays an important role
in regulation of axon growth and new connection
modulation [15]. Neurogranin is the most prominent substrates
of protein kinase C (PKC) in the mammalian brain [16].
BBMI is a major component of the actin assembly in the
microvilli of intestinal cells, and has also been reported
to have effects on membrane traffic in polarized
epithelial cells [17]. All of the three IQ motif-containing mem
bers take part in the regulatory events in the cell life.
Therefore, NYD-SP5 seems to operate as a regulator in
spermatogenesis. Furthermore, there are three enzyme
domains occurring near the IQ motif, which advanced
the possibility of regulatory function for NYD-SP5
through catalyzing some reactions in cell metabolism. In
hybridization signal analysis, clones with intensities of >10 were considered as positive signals to ensure that they
were distinguished from background with statistical
significance of >99.9 % [2]. Signal intensity of NYD-SP5
in adult microarray was just a little higher than the
threshold value, only 26.32 and 12.37 in human and mouse
adult testes, respectively, which was quite lower than
skeleton protein or cell structure protein. For example,
in adult microarray the signal intensity of outer dense
fiber protein 2 is 581.07, and the intensity for kinesin
family member 2 is 104.66. The comparatively low
intensity in microarray hybridization together with predicted
function domains in NYD-SP5 protein sequence indicated
that NYD-SP5 might play a regulatory role in the
spermatogenesis process.
In summary, NYD-SP5 is a newly found testis-specific gene with potential regulatory roles in human
spermatogenesis. The tissue-specific and stage-specific
expression of NYD-SP5 has suggested its importance in
the fundamental understanding of spermatogenesis.
Further research is required to determine the physical
function of NYD-SP5 protein in spermatogenesis.
Acknowledgment
The work was supported by grants from China National 973 (No. G1999055901) and National Natural
Science Foundation of China (No. 30170485). We thank Dr
Min Xu for valuable discussions and Dr Monica Antenos
for her critical reading of the manuscript.
References
1 McCarrey JR. Spermatogenesis as a model system for
developmental analysis of regulatory mechanisms associated with
tissue-specific gene expression. Semin Cell Dev Biol 1998; 9:
459-66.
2 Sha J, Zhou Z, Li J, Yin L, Yang H, Hu G,
et al. Identification of testis development and spermatogenesis-related genes in
human and mouse testes using cDNA arrays. Mol Hum Reprod
2002; 8: 511-7.
3 Santi CM, Darszon A, Hernandez-Cruz A. A
dihydropyridine-sensitive T-type Ca2+ current is the main
Ca2+ current carrier in mouse primary spermatocytes. Am J Physiol 1996; 271:
C1583-93.
4 Fernandes AP, Bao SN. Detection of calcium and calmodulin
during spermiogenesis of phytophagous bugs (Hemiptera:
Pentatomidae). Biocell 2001; 25: 173-7.
5 Breitbart H. Intracellular calcium regulation in sperm
capacitation and acrosomal reaction. Mol Cell Endocrinol 2002; 187:
139-44.
6 Stull JT. Ca2+-dependent cell signaling through
calmodulin-activated protein phosphatase and protein kinases. J Biol Chem
2001; 276: 2311-2.
7 Zhang T, Brown JH. Role of
Ca2+/calmodulin-dependent protein kinase II in cardiac hypertrophy and heart failure.
Cardiovasc Res 2004; 63: 476-86.
8 Grossman SD, Futter M, Snyder GL, Allen PB, Nairn AC,
Greengard P, et al. Spinophilin is phosphorylated by
Ca2+/calmodulin-dependent protein kinase II resulting in regulation
of its binding to F-actin. J Neurochem 2004; 90: 317-24.
9 Corpet F. Multiple sequence alignment with hierarchical
clustering. Nucleic Acids Res 1988; 16: 10881-90.
10 Larsen F, Gundersen G, Lopez R, Prydz H. CpG islands as
gene markers in the human genome. Genomics 1992; 13:
1095-107.
11 Bird AP. CpG-rich islands and the function of DNA
methylation. Nature 1986; 321: 209-13.
12 Wu JY, Means AR. Ca2+/calmodulin-dependent protein
kinase IV is expressed in spermatids and targeted to chromatin
and the nuclear matrix. J Biol Chem 2000; 275: 7994-9.
13 Ono T, Koide Y, Arai Y, Yamashita K. Heat-stable
calmodulin-binding protein in rat testis. Inhibition of
calmodulin-stimulated cyclic nucleotide phosphodiesterase activity. J Biol Chem
1984; 259: 9011-6.
14 Bahler M, Rhoads A. Calmodulin signaling via the IQ motif.
FEBS Lett 2002; 513: 107-13.
15 Chen B, Wang JF, Sun X, Young LT. Regulation of GAP-43
expression by chronic desipramine treatment in rat cultured
hippocampal cells. Biol Psychiatry 2003;
53: 530-7.
16 Baudier J, Deloulme JC, Van Dorsselaer A, Black D, Matthes,
HW. Purification and characterization of a brain-specific
protein kinase C substrate, neurogranin (p17). Identification of a
consensus amino acid sequence between neurogranin and
neuromodulin (GAP43) that corresponds to the protein
kinase C phosphorylation site and the calmodulin-binding
domain. J Biol Chem 1991; 266: 229-37.
17 Durrbach A, Raposo G, Tenza D, Louvard D, Coudrier E.
Truncated brush border myosin I affects membrane traffic in
polarized epithelial cells. Traffic 2000; 1: 411-24. |