| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Article |
a Department of Pharmacology, East Carolina University School of Medicine, Greenville, North Carolina 27858
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The kinesin superfamily of motor proteins is composed of more than 100 proteins that participate in a wide variety of movements within cells [5–7]. One reason for the large number of kinesin motors is to allow for modulation of function during development. Most differentiation programs, such as spermatogenesis, involve rearrangement of the cytoskeleton and probably require numerous molecular motors at specific times during the acquisition of the final cellular phenotype. In order to successfully achieve the final differentiated state, the activity of numerous motors must be tightly regulated to participate in various complexes during development.
All kinesins have a conserved motor domain containing the sites for ATP hydrolysis and microtubule binding required for the generation of force against microtubules. The varying placement of the motor domain and its interaction with other parts of the protein are partially responsible for the direction in which a particular motor moves along microtubules [8]. In addition to variation in motor domain placement, the homology of superfamily members to one another decreases dramatically outside the motor domain. The divergent "tail" domains of these proteins are thought to specify the cargo to which they attach and move. Although alternative messenger RNA processing is a common method for generating diversity in structure and function in a variety of proteins, this mechanism does not appear to be commonly used by kinesin superfamily members. Only kinesin-associated proteins, such as KAP3, associated with KIF3A/3B [9], and the kinesin light chains [10, 11] have been shown to be the products of alternatively spliced genes.
We have previously identified six members of the kinesin superfamily that are expressed in the seminiferous epithelium of the rat [3]. Here we describe the cloning and initial characterization of one member of this group from the mouse, KIFC5A. We reasoned that the unique microtubule rearrangements of spermatogenesis might require specific motor activities. To investigate this idea, we have used molecular techniques to identify isoforms of a mitotic motor in the mouse testes. This study has identified another level of diversity that applies to motors related to KIFC5A and may be important for modulating their function during spermatogenesis. KIFC5A is part of a closely related group of motor proteins that comprise a small multigene family within the larger kinesin superfamily. The diversity of this group may be explained by both alternative splicing and the presence of separate genes. Strikingly, immunolocalization of related proteins in testicular cells reveals that some may function in several microtubule complexes in germ cells, including elements of the meiotic spindle, spermatid manchette, and sperm flagella.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Recombinant lambda gt10 clones containing KIFC5A sequence were identified by hybridization screening of a mouse 15-day embryonic library (Clontech, Palo Alto, CA). Primary screening was done by Lofstrand Labs Inc. (Gaithersburg, MD) according to established techniques using the radiolabeled rat KIFC1 homolog, KRP1 head domain cDNA, as a probe for hybridization [3]. Twelve potential positives were identified in the initial screen. The insert size for each clone was determined by polymerase chain reaction (PCR) using oligonucleotide primers to either the vector sequences flanking the cloning site or the vector and head domain sequences. One clone was identified with an insert size of about 2 kilobases (kb), sufficient to encode KIFC1, a related motor from the mouse, and was further purified by two rounds of infection, plating, and hybridization to the KRP1 head probe.
The KIFC5A cDNA was sequenced, in both directions, with overlapping oligonucleotide primers (Gibco-BRL, Rockville, MD), either directly from the phage DNA or after subcloning into Bluescript (Stratagene, La Jolla, CA), using the fmol DNA Sequencing System from Promega (Madison, WI). Sequence analysis was performed using both Apple Macintosh (Cupertino, CA) and PC computers with Seqpup (public domain software) and Dnasis (Eastman Kodak, Rochester, NY) software packages, respectively. Alignments were also generated against sequences in the Genbank databases using the National Center for Biotechnology Information (NCBI, Bethesda, MD) Blast server. Multiple sequence alignment of KIFC5A and other kinesin motors, including KIFC1, KIFC4, CHO2, and HSET, was performed using both the DNAsis (Pharmacia LKB Biotechnology, Piscataway, NJ) and the Clustal W program [12].
DNA Manipulations
PCR primers were designed to amplify the divergent region in the KIFC5A sequence. The 5' primer (KIFC1a: ATCGTCCTGCCTGCTGCCTTTG) hybridizes to nucleotides 10–31 of both the KIFC5A and the KIFC1 cDNAs, while the 3' primer (KRP1reva: GCAGCCGCTCTCCCTGAAGCTC) hybridizes to nucleotides 736–757 of KIFC5A and 547–558 of KIFC1. Recognition sites for HindIII (5' primer) or XbaI (3' primer) were included to permit subsequent subcloning of amplification products. Reverse transcription (RT)-PCR reactions were conducted essentially as previously described [3]. Total RNA, from various tissues, was first converted into cDNA using the RT-PCR kit from Stratagene. Briefly, 10 µg RNA was mixed with random primers, annealed by slow cooling to room temperature from 65°C, and extended with Moloney murine leukemia virus (MMLV) reverse transcriptase at 37°C in the provided buffer. After heat denaturation, the cDNA was ready for amplification.
Typical PCR reactions contained 1–3 µl of the cDNA reaction mixture, 20 mM Tris-HCl (pH 8.4), 50 mM KCl, 1.5 mM MgCl2, 0.2 mM each deoxynucleoside triphosphate (dNTP), and 1 µM each primer in a volume of 25 µl. The reactions were preheated to 95°C for 5 min, and 2.5 U Taq polymerase (Gibco-BRL) added as the reaction cooled to 60°C for 5 min. Cycling conditions were 72°C, 3 min, with a 6-sec extension per cycle; 96°C, 45 sec; and 55°C, 4 min, for 30 cycles. Amplification products were separated by electrophoresis in low-melting point agarose (Gibco-BRL), and the DNA fragments were purified by extraction from the gel using the Qiaex kit from Qiagen (Valencia, CA). For subcloning, the purified DNA was digested with the appropriate restriction enzymes and ligated to Bluescript II SK. Recombinants were sequenced with universal primers using the fmol sequencing kit from Promega. For RT-PCR experiments using isoform specific oligonucleotides, the melting temperature (Tm; temperature at which 50% of a population of oligonucleotide is removed from its complementary sequence) was calculated for hybridization to the desired isoform versus other isoforms. For example, the approximate Tm for the KIFC5B oligo (see Figure 4) is 67°C, while the Tm for the portion of this oligo complementary to KIFC5A was calculated to be 47°C. Annealing was done at 60°C in this case.
|
Northern Analysis
Messenger RNA was prepared from mouse tissues using the Poly(A)Pure Kit from Ambion (Austin, TX) and separated by electrophoresis through a denaturing formaldehyde gel. RNAs were transferred to Nytran nylon membrane (Schleicher and Schull, Keene, NH) and hybridized to one of three probes: 1) motor probe, an approximately 500-base pair (bp) fragment amplified by PCR from the head domain of KIFC5A; 2) KIFC5B (gaga) probe, a 23-residue oligonucleotide (CGCAGAGAAAGGGAAGGGAAGGG) specific for the GA-rich sequence found in the KIFC5B isoform; or 3) a 905-nucleotide (nt) fragment from the mouse glyceraldehyde 3-phosphate dehydrogenase gene (GAPDH; Ambion). The specificity of the KIFC5B oligonucleotide was tested against RNA transcribed in vitro from KIFC5A, KIFC5B, and KIFC5C templates. This oligonucleotide recognized only KIFC5B on Northern blots of in vitro-transcribed KIFC5A, KIFC5B, and KIFC5C RNA (data not shown). The motor and GAPDH probes were labeled by the random hexamer protocol using the kit from Boehringer Mannheim (Indianapolis, IN) and [
32P]dCTP while the oligo probe was labeled with [32P]dATP using terminal transferase (Boehringer Mannheim). Blots were hybridized to probe in Church and Gilbert solution overnight at 68°C for the motor probe and 50°C for the oligonucleotide [13]. After hybridization, blots were washed as described previously for the motor probe [3], and at 50°C with increasing stringency, with a final wash of 0.05-strength SSC (single-strength SSC is 0.15 M sodium chloride, 0.015 M sodium citrate), 0.1% SDS.
Cell Preparations and Indirect Immunofluorescence
Several different spermatogenic cell preparations were used for immunofluorescent localization of KIFC5-like proteins. Enriched populations of elongating spermatids and other cell types were prepared by extrusion of seminiferous tubules of defined stages according to the method of Parvinen and Hecht [14]. Testes were obtained from sexually mature rats (8–10 wk, Sprague-Dawley; Harlan, Indianapolis, IN), and the tunica albuginea was removed. The seminiferous tubules were then gently dissected apart using stainless steel probes. Tubules were viewed by transillumination, and sections approximately 2 mm in length representing stages VIII–XII, containing elongating spermatids, were identified as described by Parvinen and Hecht and removed, and the germ cells were extruded onto a glass slide for processing. For staining of meiotic cells, segments identified as between stages XIV and I were selected by transillumination. For staining of spermatozoa, the epididymis was removed from the testes, cut into 5–10 pieces, and placed in a small volume of Dulbecco's modified Eagle's medium (DMEM). After incubation for 15 min at room temperature, the fluid was removed, and sperm were isolated by centrifugation at 14 000 x g for 5 min.
KIFC5-like proteins were localized within different spermatogenic cell types by indirect immunofluorescence using a monoclonal antibody against the CHO2 kinesin-related protein (kind gift of Dr. Ryoko Kuriyama [15], Department of Cell Biology and Neuroanatomy, University of Minnesota, Minneapolis, MN) according to established procedures. Spermatogenic cells isolated as above were attached to polylysine-coated coverslips or slides coated in Vectabond (Vector Laboratories, Burlingame, CA) in PBS, pH 7.4, and fixed for 15 min at room temperature in 3% paraformaldehyde in 0.1 M cacodylic buffer (pH 6.2) with 1 mM CaCl2 and 1 mM MgCl2 followed by extensive washing in PBS [16]. The cells were then blocked in 2% BSA in TBST (20 mM Tris, pH 7.5, 154 mM NaCl, 2 mM EGTA, 2 mM MgCl2, 0.1% Triton X-100) and incubated sequentially with primary and secondary antibodies. Fixed cells were incubated with the CHO2 monoclonal antibody at a 1:1000 dilution in blocking buffer for 30 min and then rinsed 3 times in TBST. The primary antibody was detected with donkey anti-mouse IgG conjugated to Texas Red (1:100 dilution; Jackson ImmunoResearch Laboratories, West Grove, PA), and microtubules were colocalized with anti- tubulin conjugated to fluorescein isothiocyanate (FITC; 1:50 dilution; Sigma, St. Louis, MO). The intracellular localization of the CHO2 antigen was observed with a Nikon (Melville, NY) E600 fluorescence microscope fit with appropriate filters. Controls for these experiments included a nonspecific antibody (anti-HIS6 monoclonal; Qiagen), omission of primary antibody, and preincubation of the CHO2 antibody with KIFC5A protein. KIFC5A fused to an HIS6 tag was expressed in bacteria and purified by Nickel affinity (Qiagen). Ten micrograms of purified protein was incubated overnight at 4°C with 1 µl CHO2 antibody in 200 µl TNT (10 mM Tris, pH 8.0, 150 mM NaCl, 0.05% Tween-20) according to the method of Tres and Kierszenbaum [17]. For use in immunofluorescence, the preincubated antibody was diluted to 1:1000.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The KRP1 head domain cDNA fragment was previously identified in an RT-PCR screen of rat testis RNA [3] and was found to be homologous to the C-terminal motors, CHO2 [15], KIFC1 [18], and HSET [19]. In fact, the rat KRP1 head domain is 91% identical to that of mouse KIFC1, suggesting that KRP1 might be the rat ortholog of KIFC1 [18]. The rat KRP1 cDNA encoding part of the motor domain was used to screen a lambda gt10 library constructed from Day 15 mouse embryos. Because of its strong homology to mitotic motors, KRP1 was a good candidate for a motor involved in the cell divisions of spermatogenesis. Of the 14 positive clones detected, one of just over 2 kb was chosen for further analysis. This size was sufficient to encode a protein of similar size to the KIFC1 and CHO2 polypeptides. Sequence analysis showed that this clone, which we termed KIFC5A, encodes a polypeptide with a C-terminal motor domain, like its relatives CHO2 and KIFC1, along with other sequences characteristic of kinesin-related motor proteins.
Relationship of KIFC5A to CHO2 and KIFC1
We found that KIFC5A is closely related to both CHO2 and KIFC1; however, sequence alignment revealed additional features concerning the relationship between these cDNAs (schematized in Fig. 1A and shown in detail in Fig. 1B; [15, 18]). The KIFC5A nucleotide sequence was almost identical to that of KIFC1 throughout most of its length. We were able to detect a handful of single base-pair differences between the head domain of KIFC5A and the published sequence of KIFC1. However, the most striking difference appeared when the tail domains of these two cDNAs were compared: KIFC5A contains two approximately 100-bp inserts missing in the KIFC1 sequence. Both sequence inserts are highly homologous to sequences found in the CHO2 cDNA (Fig. 1B; [15]). It is also interesting to note, when comparing these sequences, that there is a deletion of 27 bp between insert 1 and 2 in the CHO2 cDNA compared to the other two cDNAs (Fig. 1B). These observations suggest that the KIFC5A and KIFC1 messages, and perhaps also CHO2, might be derived from alternative splicing of a common gene.
|
The KIFC5A sequence retains the Kozak consensus translation start site present in the KIFC1 message (filled arrowheads Fig. 1, A–C; [20]). However, 126 bp with high homology to the 5'-end of the CHO2 gene are inserted immediately after the KIFC1 start site (insert 1, Fig. 1C). Insert 1 is 80% identical to CHO2 at the amino acid level and 73% identical at the nucleotide level. The peptide encoded by this insert has a very basic pI predicted to be 11.5; 9 of the 41 amino acids are lysine or arginine (underlined in Fig. 1C). The concentration of these amino acids in this short peptide is suggestive of a basic nuclear localization signal (NLS; [21]). No sequence within the first insert matches the consensus NLS exactly; however, one 4-amino acid peptide is the precise reverse of the consensus (Fig. 1C, asterisk). The second insert of 63 bp at position 532 of the KIFC1 cDNA is also homologous to CHO2, but less so than insert 1. The second insert has no striking structural features, although both inserts contain putative protein kinase C phosphorylation sites.
KIFC5A Is a Member of a Closely Related Group of Motor Proteins
The cloning and characterization of the KIFC5A cDNA revealed sequence heterogeneity in the tail domain relative to other family members. In order to determine whether this region is the site of other sequence variants, we conducted an RT-PCR experiment using primers flanking this region to amplify RNA from different tissues (Fig. 2A). PCR products amplified from this region differed in size between different tissues, such as brain and testes, and multiple species were detected within an individual tissue (Fig. 2, B and C). These two panels show results representative of numerous amplification reactions. Some variability in the number of bands visualized was seen (compare Fig. 2B, primer 1+2, with Fig. 2C, primer 1+2). To identify the sequence composition of these amplification products and to compare them to related species, PCR products were gel-purified, cloned, and sequenced. A total of 3 distinct sequence variants were detected by sequence analysis of the amplification products, including the tail domain of the original KIFC5A sequence, and are shown schematized in Figure 3A and aligned in Figure 3, B and C. We have chosen to call these species KIFC5A, KIFC5B, and KIFC5C. This nomenclature is designed to avoid confusion with the various kinesin-related head domains and related genetic loci that have been detected recently until the correct "head" can be connected to the correct "tail" (see Discussion and [22]).
|
|
The KIFC5A tail sequence was identified in clones from brain cDNA, and the original clone was isolated from an embryonic library. Other sequences detected were very similar to that of KIFC5A; all contained the inserts not present in KIFC1. KIFC5B was detected in brain and spleen and is notable for the replacement of 12 bp in the first insert with a 9 nucleotide GA-rich sequence. KIFC5B is 89% identical to KIFC5A at the nucleotide level and 87% at the amino acid level. This species has a deletion of 54 bp at nucleotide 255 of the KIFC5A cDNA in addition to base substitutions throughout the amplified region. Another species called KIFC5C, amplified from liver, spleen, and testes RNA, is 92% and 83% identical to KIFC5A at the nucleotide and amino acid levels, respectively. In none of the KIFC5 variants do the base substitutions appear to be clustered, except for the deletion and the GA-rich region in KIFC5B.
KIFC5 Variant Tail Domains Are Connected to a Putative Head Domain
In order to determine whether the sequence variants detected in our RT-PCR experiments might reflect full-length kinesin-related transcripts, we designed oligonucleotide primers specific to the tail region of selected isoforms and to the KIFC5A head domain to amplify transcripts from spleen, brain, and testes (Fig. 4A). Because of the considerable homology between KIFC5 isoforms, we were able to make oligonucleotides specific only for the GA-rich region of KIFC5B (Fig. 4B). We also designed a KIFC1 "junction" oligonucleotide to amplify both the KIFC1 transcript, by hybridization across the insert junction, and the KIFC5 variants, through hybridization of the 3'-end of the oligonucleotide (see Fig. 4C). The annealing conditions used in this experiment were designed to greatly favor amplification of specific messages over that of other KIFC5 variants. Indeed, both oligos, in conjunction with the KIFC5A head primer, were able to support amplification of fragments of approximately 1.8 and 1.6 kb. These size fragments are consistent with those expected for amplification from KIFC5 variants and KIFC1 cDNA, respectively.
To further investigate whether the isoforms we detected by RT-PCR represent transcribed species and to determine whether any isoform is testes-specific, we used a sequence found only in KIFC5B to design an oligonucleotide probe for Northern analysis. The KIFC5 isoforms are extremely similar in sequence; however, KIFC5B contains an inserted GA-rich sequence that is unique to this message (see Fig. 3B). The KIFC5B-specific probe recognized an approximately 1.9-kb transcript enriched in the testes compared to brain (Fig. 5). This size is sufficient to encode a KIFC5 isoform. The KIFC5A motor domain probe recognized a species in testis RNA approximately the same size as that of the KIFC5B signal.
|
CHO2 Antigen Is Part of the Meiotic Spindle
We found that KIFC5A bears extensive homology to the mitotic motors CHO2 [15] and XCTK2 [23]. These motors are concentrated at spindle poles and have demonstrated microtubule bundling activities. On the basis of these observations, we were particularly interested in the localization of the CHO2 antigen in meiotically dividing cells of the testes. A monoclonal antibody raised against the CHO2 protein (a generous gift of Dr. Ryoko Kuriyama) was used to localize this protein, and related proteins, in mixed populations enriched in specific spermatogenic cell types. This antibody recognizes bacterially expressed KIFC5A (unpublished results). We were able to identify several cells in our preparation with spindle structures (Fig. 6A). These cells were isolated from stage XIV tubules and likely represented meiotic cells. In these cells, the CHO2 monoclonal clearly stains centrosomes and spindle fibers in meiotic cells (Fig. 6A, arrowheads), with an apparent concentration toward spindle poles.
|
CHO2 Antigen Is a Constituent of Two Other Microtubule-Based Structures
Several additional microtubule complexes are essential for the formation of viable sperm, including the spermatid manchette and the flagella. In order to investigate the localization of KIFC5-like proteins in these structures, germ cells were isolated from seminiferous tubule fragments of stages VIII to XII containing elongating spermatids and were triple-stained for the CHO2 antigen, -tubulin, and DNA. The CHO2 antigen is present in several cell types and colocalizes with microtubules in spermatocytes and spermatids. The most striking finding, however, was the localization of this group of proteins to the spermatid manchette and sperm flagella.
A conspicuous localization of the CHO2 antigen with microtubules of the spermatid manchette was observed in intermediate-stage spermatids, about step 12, isolated by transillumination aided microdissection (Fig. 6B). In these cells, the nucleus is quite elongated and the cell has little residual cytoplasm. The CHO2 antigen is seen to colocalize with manchette microtubules (Fig. 6B, center panel). This staining persists through all stages in which the manchette is present, transforming from a cap-like structure in early spermatids to a band at the distal end of the spermatid nucleus in more mature spermatids (data not shown). In order to determine whether KIFC5 variants are also found in spermatozoa, the CHO2 antibody was used to stain spermatozoa isolated from rat epididymis (Fig. 6, E and F). The CHO2 antigen was found throughout the flagellar tail of the sperm.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Comparison of the full-length KIFC5A sequence identified in our screen with two other members of this subfamily, KIFC1 and CHO2, revealed an interesting relationship between these motors. The KIFC5A sequence is identical to KIFC1 except for the presence of two additional sequence blocks in the 5'-end of the KIFC5A cDNA and a number of nucleotide differences in their head domains. Because the origin of these different species is still unclear, we will refer to the two sequence blocks found in KIFC5A as "inserts" relative to the KIFC1 sequence. The genetic relationship between these cDNAs is complicated by the presence of a small multigene family of KIFC5-like motors (see below).
KIFC5A is also homologous to two additional cDNA fragments found in the database, HSET [19] and KIFC4 [22]. KIFC5A is closely related to KIFC4, showing 98% sequence identity to the head domain of this protein. The relationship between these proteins will be clarified when the entire coding sequence of KIFC4 is reported. The KIFC4 head domain sequence may, in fact, be genetically linked to one of the KIFC5-related sequences reported here. However, on the basis of the number of new tail sequences, we expect additional KIFC5-related genes to be cloned. The entire human HSET gene was recently cloned as part of a genome-sequencing effort, and the amino terminus is available for comparison [24]. KIFC5A is much less identical to HSET than it is to KIFC4, being 79% identical in the head domain, with this homology dropping off to 59% in the tail.
The relationship between the KIFC5A cDNA and the KIFC1 cDNA suggested initially that they might be alternatively spliced products of a common precursor RNA. Such an arrangement would predict that probes specific for KIFC5A (inserts 1 and 2 in Fig. 1) would map to the same genomic location as sequences common to both KIFC1 and KIFC5A. Both probe types do, in fact, localize to identical locations on chromosomes 13 and 17 (N.G. Copeland and N.A. Jenkins, personal communication). Although cosegregation of these markers supports alternative splicing, this interpretation is complicated by the finding of at least 4 loci encoding related cDNAs [22, 25]. Our mapping studies, along with those of others, demonstrate the presence of a small multigene family of highly related motor genes: two on chromosome 17, one on each of chromosomes 10 and 13. Indeed, sequencing of the KIFC5A cDNA clone revealed several base-pair changes when compared to the published KIFC1 and KIFC4 sequences, suggestive of separate genetic loci. Confirmation of these differences in the KIFC1 and KIFC4 sequences, and the development of specific probes, will be necessary to clarify the relationship between the KIFC1, KIFC4, and KIFC5 mRNAs and to determine their chromosomal location.
That multiple genetic loci are detected with KIFC5A probes is consistent with our finding of several divergent tail domain sequences by RT-PCR. Each isoform differs from the KIFC1 cDNA, previously reported, by the presence of inserts at the same position as those of KIFC5A. In addition, these species have numerous nucleotide substitutions in the approximately 750-nt region we examined, consistent with the encoding of these species by separate genes. However, one identified sequence variant, KIFC5B, differs from KIFC5A in both the insertion and deletion of sequence blocks. Both alterations occur before the beginning of the predicted coiled-coil region at approximately nucleotide 432 or amino acid 144 in the predicted KIFC5A ORF. KIFC5B may represent a distinct KIFC5-like gene that is itself alternatively spliced.
The KIFC5-related coding sequences described here are very similar except for insertions, deletions, and nucleotide substitutions with respect to each other and to the KIFC1 sequence. That these changes occur in the tail domain of these motors has important implications for their function, as this region is important for cargo binding. The highly basic first insert present in the KIFC5 isoforms and absent in KIFC1 might mediate interaction with acidic proteins, thereby targeting these motors to specific microtubule complexes. Another possible role of the first insert is as a nuclear localization signal. The presence of a such a signal in members of this subfamily of motors is particularly intriguing given our finding that the CHO2 antigen is a constituent of both cytoplasmic and nuclear microtubule structures during spermatogenesis.
Several members of this subfamily have deletions in their tail domains compared to other members. This suggests that the spacing of sequence elements in the tail domains of these motors might be important for their function. For example, KIFC5B contains a 19-amino acid deletion, while CHO2 lacks 8 amino acids compared to other motor sequences. Both deletions occur in the tail domain prior to sequences with a high probability of forming a coiled-coil structure. These spacing differences might modulate the ability of different motors to interact with their intended target complex. For instance, proper interaction of microtubules in the manchette might require motor proteins with a different arrangement of microtubule binding sites than is necessary for formation of the spindle.
This idea that the KIF5B transcript is alternatively spliced is supported by examination of a genomic locus containing a KIFC1-related gene on chromosome 17 that was entered recently into the database [26]. We have aligned the KIFC5A cDNA against this genomic sequence (unpublished observations). It is notable that the deletion in KIFC5B compared to the other cDNAs is encoded entirely by one exon in the genetic locus, consistent with formation of KIFC5B by alternative splicing. Comparison of KIFC1 and KIFC5A with the genomic locus revealed extensive sequence identity with about a 5% difference in their coding sequence. Furthermore, 16 expressed sequence tags (ESTs) were detected in a search of the mouse EST database (unpublished observations). These sequences, although extremely related, fell into at least 4 classes. This is further evidence of a small multigene family in the mouse.
The sequences unique to the KIFC5A message show significant identity to sequences in the 5'-end of the CHO2 cDNA. Besides raising the possibility of a common genetic precursor for these messages, the identification of the KIFC5A cDNA makes the prediction of the translation start site for CHO2 and KIFC5A ambiguous. The CHO2 cDNA contains two, in-frame, ATG codons, 117 bp apart, with the second identified previously as the start site of translation for the CHO2 protein [15]. Neither of these start codons is part of a Kozak consensus [20]. In KIFC5A, however, a sequence similar to the CHO2 sequence between the two methionine residues is placed just downstream of the ATG for the KIFC1, which is a Kozak consensus start site. This insert is a distinguishing feature of the KIFC5 cDNAs and may be functionally significant.
Alternative splicing offers one possible mechanism for the regulation of motor protein attachment to intended targets. It would provide a means of combining head domains with similar force-generating properties with different tail domains. Furthermore, there is a precedent for the use of alternative splicing as a mode of regulation of protein function in the testes. The CREB (cAMP response element-binding protein) precursor mRNA is differentially spliced in mid to late pachytene spermatocytes, resulting in the conversion of this protein from a transcriptional activator to an inhibitor [27]. This cell-type specific splicing event results in the insertion of a 5' exon, causing a translation switch to a downstream ATG. This is especially pertinent to the results reported here, in which the insertion of a 5' exon might introduce an upstream ATG to begin the KIFC5A open reading frame. In the case of KIFC5A, such differential splicing might result in insertion of a highly basic peptide that may have functional importance.
We have localized a group of putative minus-end directed motors to multiple microtubule complexes in spermatogenesis. Several microtubule networks in the testes might require unique motors, including the meiotic spindle, the spermatid manchette, the flagella, and elements of the Sertoli cytoskeleton. These motors we have identified are almost identical except for their cargo-binding tail domains. The finding that the manchette contains motors related to those shown to be important for spindle stability is particularly intriguing. The manchette, like the spindle, displays complex motility that most likely relies on numerous motor activities [4, 28]. During its brief lifetime, the manchette assembles around the nucleus of step 7 spermatids, moves along the nuclear membrane towards its caudal end, constricts during nuclear elongation, and finally disassembles in step 14 spermatids. The fact that other members of the COOH terminal subfamily—such as XCTK2 [23], CHO2 [15], and ncd [29]—have been shown to bundle microtubules and to promote the stability of microtubule networks leads us to propose that the motors described here may play a similar role in various microtubule complexes, including the manchette, during spermatogenesis. The divergent tail domains might serve to modulate their association with, and motor function in, different complexes. Further examination of the role of members of the CHO2 subfamily in the biogenesis and function of the spindle, manchette, and flagella should reveal commonalities of motor function as well as explaining how different motor environments can result in such different forms of motility.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work was supported, in part, by a grant from the East Carolina University School of Medicine. 
2 Correspondence: Ann O. Sperry, Dept. of Pharmacology, East Carolina University School of Medicine, 600 Moye Blvd., Greenville, NC 27858. FAX: 252 816 3203; asperry{at}brody.med.ecu.edu
Accepted: December 16, 1999.
Received: August 12, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. S Vaid, J. A Guttman, R. R Singaraja, and A. W. Vogl A Kinesin Is Present at Unique Sertoli/Spermatid Adherens Junctions in Rat and Mouse Testes Biol Reprod, December 1, 2007; 77(6): 1037 - 1048. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Nath, E. Bananis, S. Sarkar, R. J. Stockert, A. O. Sperry, J. W. Murray, and A. W. Wolkoff Kif5B and Kifc1 Interact and Are Required for Motility and Fission of Early Endocytic Vesicles in Mouse Liver Mol. Biol. Cell, May 1, 2007; 18(5): 1839 - 1849. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Christodoulou, C. W. Lederer, T. Surrey, I. Vernos, and N. Santama Motor protein KIFC5A interacts with Nubp1 and Nubp2, and is implicated in the regulation of centrosome duplication J. Cell Sci., May 15, 2006; 119(10): 2035 - 2047. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W.-X. Yang, H. Jefferson, and A. O. Sperry The Molecular Motor KIFC1 Associates with a Complex Containing Nucleoporin NUP62 That Is Regulated During Development and by the Small GTPase RAN Biol Reprod, April 1, 2006; 74(4): 684 - 690. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. R. Tovich, P. Sutovsky, and R. J. Oko Novel Aspect of Perinuclear Theca Assembly Revealed by Immunolocalization of Non-Nuclear Somatic Histones During Bovine Spermiogenesis Biol Reprod, October 1, 2004; 71(4): 1182 - 1194. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. C. Ems-McClung, Y. Zheng, and C. E. Walczak Importin {alpha}/{beta} and Ran-GTP Regulate XCTK2 Microtubule Binding through a Bipartite Nuclear Localization Signal Mol. Biol. Cell, January 1, 2004; 15(1): 46 - 57. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Chennathukuzhi, C. R. Morales, M. El-Alfy, and N. B. Hecht The kinesin KIF17b and RNA-binding protein TB-RBP transport specific cAMP-responsive element modulator-regulated mRNAs in male germ cells PNAS, December 23, 2003; 100(26): 15566 - 15571. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W.-X. Yang and A. O. Sperry C-Terminal Kinesin Motor KIFC1 Participates in Acrosome Biogenesis and Vesicle Transport Biol Reprod, November 1, 2003; 69(5): 1719 - 1729. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Escalier New Insights into the Assembly of the Periaxonemal Structures in Mammalian Spermatozoa Biol Reprod, August 1, 2003; 69(2): 373 - 378. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Zou, C. F. Millette, and A. O. Sperry KRP3A and KRP3B: Candidate Motors in Spermatid Maturation in the Seminiferous Epithelium Biol Reprod, March 1, 2002; 66(3): 843 - 855. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Miki, M. Setou, K. Kaneshiro, and N. Hirokawa All kinesin superfamily protein, KIF, genes in mouse and human PNAS, June 19, 2001; 98(13): 7004 - 7011. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Yang, E. A. Roberts, and L. S. B. Goldstein Functional Analysis of Mouse C-Terminal Kinesin Motor KifC2 Mol. Cell. Biol., April 1, 2001; 21(7): 2463 - 2466. [Abstract] [Full Text]
|
|
|
|
|
|
Z. Yang, C.-h. Xia, E. A. Roberts, K. Bush, S. K. Nigam, and L. S. B. Goldstein Molecular Cloning and Functional Analysis of Mouse C-Terminal Kinesin Motor KifC3 Mol. Cell. Biol., February 1, 2001; 21(3): 765 - 770. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
