Disruption of Kinesin II Function Using a Dominant Negative-Acting Transgene in Xenopus laevis Rods Results in Photoreceptor Degeneration

Abstract

purpose. Kinesin II is a motor protein that moves on microtubules and whose importance in ciliary and flagellar transport has been well documented. In the current study, the role of kinesin II in rod photoreceptors was examined by expressing a dominant negative-acting transgene that disrupts kinesin II function in Xenopus laevis rods of transgenic tadpoles.

methods. A previously characterized dominant negative-acting kinesin II transgene tagged with enhanced green fluorescent protein (EGFP) driven by the Xenopus rod opsin promoter was used to make Xenopus transgenic tadpoles to disrupt kinesin II function specifically in rod photoreceptors. Transgenic tadpole retinas were examined to ascertain transgene expression pattern and morphologic phenotype. Rod-to-cone ratios were determined in experimental and control retinas.

results. Visualized by its EGFP tag, the kinesin II transgene was expressed in rods in a mosaic pattern in the retina. Subcellular localization of transgenic kinesin II was similar to that of endogenous kinesin II subunit photoreceptor expression—that is, it was localized to the connecting cilium, inner segment, and synapse. However, in kinesin II transgene–expressing animals, fluorescence was transient. Ocular fluorescence was lost 6 days after its first detection. The disappearance of fluorescence was due to degeneration of rods expressing the transgene. Retinas of 7- to 9-day old kinesin II transgenic tadpoles had significantly fewer rods than did control retinas.

conclusions. The observation that rod degeneration is produced by expression of a dominant negative-acting kinesin II transgene in Xenopus rods is consistent with previous studies in mice, suggesting that kinesin II function is required for photoreceptor survival.

Kinesin II is a plus-end directed microtubule motor protein that has been characterized in a number of different systems and species. ¹ The holoenzyme exists as a heterotrimeric molecule consisting of two different subunits containing motor domains, KIF3A and KIF3B/C, and a third globular accessory subunit (KAP3), which is thought to be involved in cargo binding. Kinesin II is a highly conserved protein and has been found in species ranging from Chlamydomonas and Tetrahymena to humans. In addition, kinesin II is essential for mammalian development, in that mice with knockouts of either of the kinesin II motor subunits show several defects that cause embryonic lethality. ^{2 3}  

In several species, kinesin II is crucial in ciliary or flagellar formation and maintenance. ¹ In Chlamydomonas, kinesin II is necessary for proper flagellar formation, ⁴ whereas in mammals, kinesin II is crucial in the proper formation and function of the nodal cilium that generates left–right asymmetry in embryos. ^{2 3} Although it has functions central to embryogenesis, kinesin II is also expressed in a number of tissues later in development where it has roles in cellular transport processes that affect tissue morphogenesis or specialized cellular functions. In Caenorhabditis elegans, kinesin II has been localized to cilia of differentiated sensory cells. ⁵  

The photoreceptor is one type of vertebrate sensory cell that contains kinesin II. It has been localized to the photoreceptor inner segment, synapse, and connecting cilium between the inner and outer segments. ^{6 7 8} In Xenopus rod photoreceptors, kinesin II is especially concentrated in the connecting cilium. ⁸ The inner segment of photoreceptors contains the protein synthesis apparatus with a predominant product that is the essential photopigment opsin. Newly synthesized opsin must be transported from the inner segment distally through the connecting cilium to the outer segment where it functions to initiate the phototransduction pathway. Indeed, the microtubule polarity of the connecting cilium is such that the plus end is distally located, and opsin transport across the connecting cilium to the outer segment is likely to entail a plus end–directed microtubule motor such as kinesin. ⁹ Enormous amounts of opsin transport to the outer segment are required for both the initial formation of a functional outer segment and later in the mature photoreceptor, because there is continual addition of opsin to new outer segment disks at the proximal outer segment to balance the shedding of older disks at the distal end. Evidence supporting the hypothesis that opsin transport across the connecting cilium to the outer segment is mediated by kinesin II has come from CRE-lox mice containing a photoreceptor-specific knockout of the KIF3A subunit of kinesin II. ¹⁰ In the absence of KIF3A in the photoreceptors, photoreceptors aberrantly accumulate opsin in the inner segments and undergo apoptosis. It has been hypothesized that compromised kinesin II transport across the connecting cilium produces rod degeneration in these mice. 

In this study, we used a previously cloned dominant negative-acting Xenopus transgene to disrupt kinesin II activity in rod photoreceptors. ¹¹ This transgene encodes a fusion protein consisting of enhanced green fluorescent protein (EGFP) substituted for the motor domain of the Xenopus KIF3B subunit (EGFP/XKlp3B-ST) and was used in combination with an untagged, motorless Xenopus KIF3A subunit (XKlp3A-ST). Le Bot et al. ¹¹ used the EGFP/XKlp3B-ST construct for transfection into Xenopus cultured cells, where it caused a disruption in transport between the endoplasmic reticulum and Golgi apparatus. In addition, expression of the dominant negative-acting kinesin II transgene in Xenopus melanophores inhibited melanosome dispersion. ¹² In both cases, it was shown that the transgenic KIF3B subunit lacking the motor was able to form an inactive kinesin II molecule by its association with the other endogenous kinesin II subunits. We showed that using the Xenopus rod opsin promoter to drive EGFP/XKlp3B-ST expression in rod photoreceptors produced a reduction in the number of rods in transgenic tadpole retinas. This result, using a dominant negative-acting transgene to disrupt kinesin II motor activity in Xenopus, is consistent with results in which kinesin II activity in mouse photoreceptors was eliminated by using the CRE-lox strategy. These observations support the hypothesis that functional kinesin II is essential for rod photoreceptor survival in vertebrates. 

Materials and Methods

Constructs Used for Transgenesis

The XKlp3B-ST plasmid containing the rod opsin promoter driving the expression of a transgene containing the stalk and tail regions of XKlp3B fused to the C terminus of EGFP was generated by excision of XKlp3B-ST from the EGFP/XKlp3B-ST plasmid ¹¹ with XbaI after mutation of the XhoI site to XbaI and ligation into a vector containing the Xenopus rod opsin promoter (XOP) fused to EGFP. ¹³ The XbaI XKlp3B-ST fragment was also fused to a vector containing the modified Xenopus rod opsin promoter (XOP4/EGFP) that drives greater levels of transgene expression. ¹⁴ The XOP4 vector contains a deletion in the ∼500-bp Xenopus rod opsin promoter that induces increased levels of transgene expression that is restricted to rod photoreceptors. The stalk and tail regions for the Xenopus Xklp3A subunit were amplified by PCR from an XKlp3A plasmid ¹⁵ and cloned downstream of the XOP4 vector after the addition of a start codon and AgeI and NotI sites. Control EGFP tadpoles were generated with an expression vector containing only EGFP driven by XOP4. 

Xenopus Transgenesis

Linearized plasmid DNA containing the transgene was purified after restriction enzyme digest (High Pure PCR Product Purification Kit; Roche Diagnostics, Inc., Indianapolis, IN) and used to generate transgenic frogs. A modified procedure based on the technique developed by Kroll and Amaya ^{16 17} was used and is briefly described as follows: Sperm nuclei were prepared from a testis (∼10 × 3 mm) removed from an adult injected with 300 U human chorionic gonadotropin (HCG; Fujisawa, USA. Inc., Elmsford, NY) 15 hours before. The testis was crushed in a small glass homogenizer in 1 mL of nuclear preparation buffer (NPB) ¹⁷ and further homogenized by trituration through a micropipette tip. The prep was spun to pellet the sperm. The sperm pellet was resuspended in 400 μL NPB and 20 μL 10 mg/mL NP-40 in NPB was added to permeate the nuclear membrane. Bovine serum albumin (600 μL; BSA) in NPB was added to the sperm, and the mixture was spun briefly to pellet the permeated nuclei. The pellet was rinsed with 0.3% BSA in NPB and resuspended in 200 μL 40% glycerol-NPB. 

Heated egg extract (HEE) was prepared to treat the sperm nuclei before injection in a procedure similar to that of Amaya and Kroll, ¹⁷ except that the high-speed cytosol preparation was heated at 80°C for 10 minutes. The HEE was collected as the supernatant after a 1-minute spin in the microfuge. HEE (5 μL), 1.5 μL sperm nuclei, and 1 μL of DNA (100 ng) were mixed together and then diluted in 500 μL sperm dilution buffer (250 mM sucrose, 75 mM KCl, 0.5 mM spermidine trihydrochloride, 0.2 mM spermine tetrahydrochloride; pH 7.3–7.5). Injection of the sperm nuclei-DNA mixture was performed as described in Amaya and Kroll. ¹⁷ Approximately 2 to 5 nL of the sperm nuclei-DNA mix was injected per egg. The eggs were allowed to develop at 15°C for 3 to 4 hours, and then only the regularly dividing embryos were incubated at room temperature for the remaining time. 

Light and Electron Microscopy

Tadpoles that were generated by transgene expression were screened for EGFP fluorescence under a fluorescence dissecting microscope (Leica, Deerfield, IL) 4 to 5 days after fertilization. Tadpoles expressing green fluorescence in their eyes were selected, and the fluorescence was monitored daily. Eyes from wild-type and transgenic tadpoles that had been screened positively for EGFP were dissected at various times after fertilization. 

Tadpole eyes were dissected from the animal and punctured with a tungsten needle to allow penetration of the fixative. One eye was fixed for fluorescence microscopy in 4% paraformaldehyde diluted in phosphate-buffered saline (PBS) at 4°C overnight. The other eye was fixed overnight at 4°C for electron microscopy in 2% glutaraldehyde, 2% paraformaldehyde, 1 mM MgCl₂, and 1% sucrose in PBS. The bodies were stored at −80°C for genomic PCR analysis to confirm the presence of the transgene. 

Retinas fixed in paraformaldehyde were equilibrated in 15% sucrose and 2% paraformaldehyde in PBS for 1 hour and then in 30% sucrose and 2% paraformaldehyde in PBS for 2 hours. The retinas were then frozen in optimal cutting temperature (OCT) compound and cryosectioned. Sections (8–12 μm) were mounted on slides with gel mount containing 0.001% 4′,6′-diamino-2-phenylindole (DAPI). Digital images of retinal tissue sections viewed by fluorescence microscopy were recorded (Openlab software; Improvision; Lexington, MA). A Z-series of Xenopus retina sections were deconvolved using the iterative deconvolution software module (Openlab; Improvision). 

Retinas fixed for electron microscopic analysis were washed in PBS and postfixed in osmium tetroxide for 1 hour. Samples were washed with PBS and stained en bloc with 1% uranyl acetate overnight. After a wash in water, the tissue was dehydrated through an acetone series. The tissue was infiltrated with Epon-Araldite (50:50 with acetone) overnight, processed through three more changes of Epon-Araldite, and embedded in the last resin change at 60°C overnight. To determine the rod-cone ratio in a retina, three to five 0.5-μm sections of the retina were cut from different regions of the Xenopus eye and were stained with 1% methylene blue and 1% azure blue in 1% borax. The number of rods and cones from each section from the different regions were then counted under the light microscope and used to calculate the rod-cone ratio. The average rod-cone ratio of multiple founder tadpoles was determined for the different control and transgenic frogs and analyzed for differences in the rod-cone ratios between tadpoles containing the kinesin II fusion transgenes and control tadpoles (wild-type and EGFP tadpoles). The differences between the mean rod-cone ratios were examined statistically by analysis of variance (ANOVA). Sections (70 nm) were also cut and visualized to investigate the integrity of the retina of some kinesin II transgenic and control Xenopus tadpoles by electron microscopy. 

Analysis of the F1 Line Containing the EGFP/XKlp3B-ST Transgene

One male Xenopus containing the XOP/EGFP/XKlp3B-ST transgene was raised to sexual maturity and mated with a wild-type female. The resultant tadpoles were screened for transgene expression in their eyes by examination under the fluorescence microscope. F1 progeny tissue was processed for microscopy with the method used for the founder transgenic tadpoles, and the rod-cone ratio was compared to the ratio in wild-type and EGFP tadpoles of similar age. The presence of transgene in the F1 tadpoles was confirmed by genomic PCR of the DNA extracted from the bodies of the F1 tadpoles. 

Results

Expression of EGFP Transgenes

Transgenic tadpoles containing EGFP/XKlp3B-ST, EGFP/XKlp3B-ST and XKlp3A-ST, or EGFP alone were generated and viewed under fluorescence to evaluate transgene expression. Tadpoles were first screened 4 days after the initial injection, and green fluorescence was monitored daily. Green fluorescence was first detected from XOP promoters at approximately stage 40. ¹⁸ Tadpoles expressing EGFP alone contained green fluorescence confined to the cytosolic and nuclear compartments of rod photoreceptors and appeared to show no ill effects of transgene expression on photoreceptor morphology, as has been previously reported (Fig. 1C) . ^{13 14}  

Whereas expression of EGFP fusion transgenes was always rod-specific, expression of the EGFP/XKlp3B-ST construct alone or in combination with XKlp3A-ST was never as bright as EGFP alone. In addition, the fluorescence visualized in the retina was mosaic and also varied in intensity levels between rods that expressed the transgene. Figure 1A shows the mosaic pattern of EGFP/kinesin II transgene expression in subsets of rod photoreceptors. Mosaic transgene expression was also sometimes observed in animals expressing EGFP alone (Fig. 1C) , although retinas expressing the kinesin II transgene consistently displayed more mosaicism than those expressing the control EGFP transgene. Unlike the expression of EGFP alone, kinesin II transgene expression was found to be transient. Peak expression of EGFP/XKlp3B-ST was observed 5 days after injection (at approximately stage 46) after which ocular fluorescence sharply diminished. In most cases, fluorescence was undetectable in EGFP/XKlp3B-ST tadpoles 7 days after injection. Examination with the fluorescence microscope of retinal sections from tadpoles in which ocular fluorescence had faded or was very dim revealed that there were few, if any, rod photoreceptors that still expressed the transgene. Any rods expressing the transgene product in these eyes were confined to the lateral margins of the retina, suggesting they were the most recently differentiated photoreceptors. 

In retinas fixed soon after the transgene expression was activated, EGFP/XKlp3B-ST fluorescence was confined to the cytosolic compartments (inner segment, connecting cilium, and outer plexiform layer) of rods but was excluded from the photoreceptor outer segments and nuclei (Fig. 1A) . This pattern of transgene expression is identical with the pattern of endogenous kinesin II subunit expression that has already been documented for Xenopus. ⁸ In slightly older stages of tadpole retinas, in which fluorescence from the kinesin II transgene was still detectable, bright spots of transgenic fusion protein accumulated in the rod inner segment and/or synaptic region which probably represented the formation of apoptotic bodies (Fig. 1A , white arrows). No such fluorescent accumulations were ever observed in retinas of transgenic tadpoles expressing EGFP alone, despite the higher levels of transgene expressed in the EGFP control tadpoles. 

Analysis of Rod-Cone Ratios in Transgenic Kinesin II Tadpoles

One hypothesis for the transient nature of EGFP/XKlp3B-ST transgene expression in Xenopus rod photoreceptors is that disruption of kinesin II function within the cell by the dominant negative-acting transgene eventually causes cellular apoptosis and thus loss of transgene-expressing rods. To test this hypothesis, we examined 0.5-μm sections of wild-type and transgenic retinas from 7- to 9-day-old tadpoles containing the EGFP/XKlp3B-ST transgene, the EGFP/XKlp3B-ST and the XKlp3A-ST transgenes, or the EGFP transgene alone. Shown in Figure 2 are examples of control and transgenic retinal sections that were analyzed to compare the numbers of rods and cones. Note that in the section from the EGFP/XKlp3B-ST transgenic tadpole (Fig. 2A , between filled arrows) there was an area where there were fewer rods than cones compared with other areas in the same retina and with the retinas from an EGFP transgenic (Fig. 2C) or a wild-type (Fig. 2D) tadpole of the same age (areas between open arrows). The reduction in the number of rods was even more pronounced in transgenic retinas that contain both dominant negative-acting kinesin II subunit transgenes. In the double transgenic tadpole shown in Fig. 2B (between the filled arrows), there were very few remaining rod photoreceptors, whereas there were a significant number of intact cones. There was little evidence of cone degeneration or degeneration of cells in any of the other retinal layers, suggesting that the effect is specific only for the rods expressing the dominant negative-acting transgenes. 

The effect of kinesin II dominant–negative expression on the retinal rods was quantified by analyzing the rod-cone ratio of transgenic kinesin II retinas and comparing the average ratios with those of control retina. The average rod-cone ratios were determined by counting the numbers of rods and cones from three to five retinal sections from three to eight tadpoles fixed 7, 8, and 9 days after fertilization. Shown in Figure 3 are the combined data from tadpoles with retinas that were fixed 7 to 9 days after fertilization. Tadpoles expressing the EGFP/XKlp3B-ST had an average rod-cone ratio of 1.1 ± 0.019 (n = 21 tadpole retinas). The average ratio from the transgenic tadpoles containing both kinesin II subunit dominant negative proteins was even lower, 0.68 ± 0.13 (n = 13 tadpole retinas), whereas the ratios from transgenics containing EGFP alone (1.7 ± 0.095, n = 10 tadpole retinas) or from wild-type (1.6 ± 0.052, n = 18 tadpole retinas) were higher. The lower rod-cone ratios between the single or double kinesin II transgenic retinas and both EGFP and/or wild-type controls are statistically significant (P < 0.001). In addition, a statistical difference (P < 0.005) was found in the ratios between the retinas containing EGFP/Klp3B-ST alone and EGFP/XKlp3B-ST and XKlp3A-ST together, suggesting that expression of both dominant negative-acting subunits had a greater effect on reducing the rod-cone ratio than expression of only one subunit. The rod-cone ratio data from tadpoles fixed 7, 8, and 9 days after fertilization were combined, because no significant differences in the rod-cone ratios were detected between retinas fixed at 7, 8, or 9 days in either single or double kinesin II transgenic or control retinas. In addition, no significant difference was found between the data from control EGFP and wild-type retinas of any age, supporting evidence from previous studies that no deleterious effects on rod survival are produced by EGFP transgene expression. The presence of the transgene was confirmed by PCR of genomic DNA for all the tadpoles containing kinesin II transgenes. 

The dramatic effects of kinesin II dominant negative-acting transgene expression in the Xenopus tadpole retina are evident in one animal containing both EGFP/XKlp3B-ST and XKlp3A-ST transgenes, which was examined at the ultrastructural level (Fig. 4A) . Very few rods were detected, whereas the cones (indicated by an asterisk in the cone lipid droplet) occurred at regular intervals in the photoreceptor layer. The control tadpole retina expressing the EGFP transgene alone (Fig. 4B) , similar to the tadpole retina containing both dominant negative-acting kinesin II subunits, is from a stage-48 animal that was fixed 7 days after fertilization. Unlike the kinesin II transgenic animal, the retina expressing only EGFP contained more rods than cones, and the rods had developed long, orderly outer segments. In the one rod (Fig. 4A , arrow) found in the kinesin II transgenic retinal section, little of the outer segment remained, and what remained contained disordered disks. In addition, the inner segment of the kinesin II transgenic rod contained vacuoles and vesicles containing dense material that were not seen in the control EGFP rods. The nucleus of the kinesin II transgenic rod was apoptotic and highly condensed compared with the cone nuclei in the outer nuclear layer and with the nuclei in the inner nuclear layer below. Although the cones and the other layers of the retina in the kinesin II transgenic animal shown in Figure 4A looked relatively normal, retinas from a few kinesin II transgenic animals also displayed degeneration in cones and other retinal cell layers as well as rods that may reflect apoptotic effects caused by the degeneration of transgenic rods in later stages of tadpoles or in tadpoles containing a greater number of rods expressing the kinesin II transgene or containing a greater level of transgene expression. 

Effects of the XKlp3B-ST Transgene in F1 Tadpoles

In the preliminary Xenopus transgenic tadpoles generated with the unmodified (XOP) rod promoter driving the EGFP/XKlp3B-ST fusion, a few tadpoles that displayed green fluorescence in their retinas were selected at early stages. These founder tadpoles containing the kinesin II transgene were raised to sexually maturity, and one transgenic male was selected for breeding to a wild-type female frog. Approximately 500 F1 tadpoles were generated, of which a subset expressed green fluorescence in their eyes. Similar to the founder tadpoles previously described, the F1 progeny exhibited a pattern of transgene expression that was mosaic and transient. Several tadpoles with ocular green fluorescence were selected for analysis of the mean rod-cone ratio in their retinas, as described in founder tadpoles in the previous section (Fig. 5) . The presence of the transgene in the F1 tadpoles was confirmed by genomic PCR, using primers designed against the EGFP/XKlp3B-ST transgene. The rod-cone ratios from the F1 tadpoles (1.1 ± 0.075) are reduced compared with control EGFP (1.5 ± 0.15) and wild-type (1.6 ± 0.21) tadpoles of similar age. The average ratios from F1 EGFP/Xlp3B-ST tadpoles were similar to those from founder kinesin II tadpoles of the same age. Thus, the kinesin II transgene was transmitted faithfully from founder animal to its offspring and exhibited identical effects in both generations. 

Discussion

In this study, the effects of dominant negative-acting kinesin II transgenes were examined in rod photoreceptors. Kinesin II has already been found to play roles in the formation and maintenance of cilia and flagella in a number of different cell types at various stages of development. In the flagellated alga, Chlamydomonas, experiments have demonstrated that kinesin II is essential both for formation and maintenance of flagella. ⁴ In this species, kinesin II mediates intraflagellar transport (IFT) of material required for flagellar assembly in large protein “raft” complexes to the tip of the flagella. Similarly, kinesin II is crucial for proper cilia formation in nodal cells during mammalian development. Kinesin II subunit knockout mice contain nodal cells with abnormal cilia that can no longer function properly to generate normal right–left asymmetry early in embryogenesis. ^{2 3}  

Retinal photoreceptors possess a modified nonmotile cilium that connects the specialized phototransduction apparatus contained within the outer segment to the remaining portion of the cell. Outer segment proteins translated in the inner segment portion of the photoreceptor must be transported to the outer segment through the connecting cilium. The connecting cilium is comprised of microtubules arranged with their plus ends oriented toward the distal end of the connecting cilium, and kinesin transport of proteins would therefore be directed from the inner segment to the outer segment. ⁹ Evidence supporting the hypothesis that kinesin II mediates inner to outer segment transport of IFT particles across the connecting cilium has been suggested by results in mice with a mutation in an IFT particle that results in abnormal outer segment formation and eventual photoreceptor degeneration. ¹⁹ In addition, several IFT particle subunits as well as kinesin II subunits have been immunolocalized to the connecting cilium. ^{6 7 8}  

In this study, tagged and untagged fragments of the two Xenopus kinesin II subunits were expressed alone or in combination in Xenopus rods using the Xenopus rod opsin promoter. The effects of the kinesin II XKlp3B-ST fusion protein as a dominant negative mutation on endogenous kinesin II has already been demonstrated by other investigators researching the role of kinesin II in pigment granule dispersion in Xenopus melanophores ¹² and in transport between the endoplasmic reticulum-Golgi apparatus in Xenopus cultured cells. ¹¹ They showed that the EGFP-XKlp3B-ST subunit associates with endogenous XKlp3A subunits. Because two functional motor subunits are required for kinesin II motility, endogenous kinesin II dimerization partners are inactivated through their association with the overexpressed coiled coil kinesin II domains in the stalk domain. 

Transgenic kinesin II proteins tagged with EGFP were expressed only in rod photoreceptors of the retina at the approximate time of opsin gene expression onset, as has been documented in other studies in which the rod opsin promoter was used for transgenesis. In addition, the distribution of tagged kinesin II protein expression was identical with that reported for endogenous kinesin II subunits; green fluorescence was displayed in the cytosolic compartments of rods (inner segment, connecting cilium, and synapse). ⁸ However, the efficiency of transgene expression was variable between animals, as would be expected in founder animals, because of the timing of transgene integration. In addition, the levels of transgene expression between rods within a single retina also varied, as has been reported by Moritz et al. ²⁰ Expression of the transgene remained mosaic, even in F1 progeny of an EGFP/XKlp3B-ST parent, despite transgene transmission through the germline. This cell-to-cell variability in expression has been attributed to heterochromatin-associated position-effect variegation. ²⁰  

After the onset of the fluorescently tagged kinesin II transgene expression, accumulations of the fusion protein begin to appear in the inner segment and synapse of rods that represent the first signs of photoreceptor degeneration. Another unexpected observation was that kinesin II transgene expression in tadpoles was transient. Approximately 5 to 7 days after the onset of EGFP fluorescence, only low to background levels of green fluorescence were detected in the tadpole retina. Our demonstration that the number of rods was reduced compared with the number of cones indicates the disappearance of fluorescence from the kinesin II transgenics was due to the loss of subsets of rods expressing the dominant negative-acting transgene. Therefore, our observations indicate that disruption of kinesin II function in rods leads to cellular apoptosis, thereby reducing the number of rods, whereas the number of cones that were not affected by the transgene remained the same. This effect of a reduced rod–cone ratio was found in retinas examined 7 to 9 days after DNA injection in tadpoles containing either the EGFP/XKlp3B-ST transgene alone or in concert with an untagged XKlp3A-ST subunit. 

Transgenic tadpoles expressing EGFP alone exhibited very high levels of EGFP fluorescence in their cytosol, yet never contained bright accumulations of the fluorescent protein and had an average rod–cone ratio that was equivalent to that found in wild-type tadpoles. It should be noted that the average rod–cone ratio from 7- to 9-day old wild-type (1.6 ± 0.052) and control EGFP (1.7 ± 0.095) retinas in this study were slightly higher than the ratio reported in adult retinas (1.2). ²¹ In transgenic animals expressing dominant negative-acting DNA constructs of both kinesin II subunits, there was a greater reduction of the rod–cone ratio, caused by increasing the amount of dominant negative-acting subunits in the rods and/or increasing the number of rods expressing sufficient amounts of the transgenic protein to disrupt kinesin II function. It should be noted that rod degeneration in transgenics containing either one or both of the kinesin II transgenes was insufficient to cause complete rod degeneration in the retina in most cases, probably because of the mosaic nature of the transgenics, which is attributable to the time of integration for founder animals and chromosomal position effects in transgenic kinesin II animals. In addition, rods expressing high enough quantities of the dominant negative-acting transgene were lost through apoptosis when endogenous kinesin II was disrupted, but continual replacement of rods from stem cells existing at the lateral margins of the Xenopus retina was also occurring. Therefore, the effects of the transgene are difficult to detect in later stages of tadpoles, especially those containing a small percentage of rods expressing high enough levels of the kinesin II transgene to cause apoptosis, because the degenerating rod population was renewed by the lateral margin cells. The observation that only rods near the lateral margins of the retina expressed the fluorescent tag in later stages of tadpoles or in tadpoles with lower expression levels supports this premise. 

Our results show that disruption of kinesin II function in rods caused a loss of rods from the retinas of transgenic kinesin II tadpoles. These observations are consistent with results from previous experiments examining the effects of a kinesin II subunit knockout in mouse photoreceptors using the CRE-lox system, in which recombination to eliminate the mouse KIF3B subunit was controlled by the mouse IRBP promoter. ¹⁰ Massive photoreceptor degeneration occurred in mouse retinas containing the photoreceptor-specific knockout. Rod degeneration was accompanied by aberrant opsin distribution. Marszalek et al., ¹⁰ attribute the effects in the kinesin II knockout to the disruption of opsin transport across the connecting cilium, which is normally mediated by kinesin II. It should be noted that transgenic kinesin II tadpoles were examined for opsin mislocalization, but no detectable differences in opsin immunostaining were detected between wild-type rods and fluorescent and nonfluorescent rods within retinas containing the kinesin II transgene. 

We used a different experimental strategy to compromise kinesin II function in rods, and rod degeneration also occurred in our system when kinesin II was disrupted. It should be noted that the promoters used to generate the kinesin II transgenic Xenopus and the kinesin II knockouts in mouse photoreceptors are not activated until after the connecting cilium has already formed. Whether kinesin II plays a role in the formation of the connecting cilium during photoreceptor morphogenesis, similar to its role in ciliogenesis in other systems, remains to be determined. 

 Supported in part by National Eye Institute Grants EY03575 (BB), EY06859 (JL-J), EY11256 (BEK), and EY12975 (BEK); a Research to Prevent Blindness grant to State University of New York Upstate Medical University; and Foundation for Fighting Blindness.

 Submitted for publication February 14, 2003; accepted March 19, 2003.

 Disclosure: J. Lin-Jones, None; E. Parker, None; M. Wu, None; B.E. Knox, None; B. Burnside, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Localization of transgenic EGFP/XKlp3B-ST and control EGFP proteins in Xenopus laevis retina. (A) A cryosection of a tadpole retina containing the EGFP/XKlp3B-ST transgene exhibits fluorescence in the rod inner segment (IS), connecting cilium (arrowhead), and synapse. (B) Nomarski image of the same EGFP/XKlp3B-ST retinal section as shown in (A). (C) Fluorescence from expression of the EGFP transgene was found within the inner segment (IS), connecting cilium (arrowhead), nucleus (N), and synapse of rods. (D) Nomarski image of the EGFP retina shown in (C). (A, arrows) Accumulations of the fluorescently tagged kinesin II fusion protein that were sometimes seen in the inner segment and synapse of kinesin II transgenics but never in transgenic EGFP control retinas. Notice the mosaic nature of transgene expression and the variability in levels of transgenic protein expression in both the kinesin II and EGFP transgenic tadpoles. OS, outer segment. Scale bar, 10 μm.

Figure 2.

 

Retinal sections from kinesin II transgenic and control Xenopus tadpoles are examples of retinal sections used to count the number of rod and cone photoreceptors in transgenic Xenopus tadpoles containing the EGFP/XKlp3B-ST transgene (A), the EGFP/XKlp3B-ST and the XKlp3A-ST transgenes (B), or the EGFP transgene alone (C) or in wild-type tadpoles (D) fixed 7 to 8 days after fertilization. The photoreceptor counts were used to determine the average rod-cone ratio for each of the different types of Xenopus tadpoles and are compared in Figures 3 and 5 . (A, areas located between the filled arrows) A section of the retina from an EGFP/XKlp3B-ST tadpole that displayed a reduced number of rod photoreceptors compared with the remainder of the retina and with the control retinas (C and D; areas located between the open arrows). Notice that the retina in (B) expressing the dominant negative-acting transgenes of both kinesin II subunits consists predominantly of cones (containing lipid droplets) and contains substantially fewer rods than the retina expressing only one of the kinesin II transgenes (A). Scale bar, 100 μm.

Figure 3.

 

The average rod-cone ratio in kinesin II transgenic tadpole retina was reduced compared with control Xenopus retinas fixed 7 to 9 days after fertilization. Three to five 0.5-μm sections from Xenopus tadpole retinas were used to count the number of rod and cone photoreceptors and to determine the average rod-cone ratio for each tadpole analyzed. The overall average rod-cone ratio was determined for multiple retinas fixed 7 to 9 days after fertilization. There was a reduction in the number of rods compared with the number of cones in retinas of tadpoles containing either the EGFP/XKlp3B-ST transgene alone (▪) or in combination with the XKlp3A-ST transgene ( ). A statistically significant reduction was found between rod-cone ratios of tadpoles containing the EGFP/XKlp3B-ST transgene alone or the EGFP/XKlp3B-ST +XKlp3A-ST transgenes and control tadpoles containing either the EGFP transgene alone ( ) or no transgene (wild-type; □; P < 0.001). In addition, the mean rod-cone ratio of the tadpoles containing both kinesin II transgenes is significantly lower than the ratio from tadpoles expressing only the EGFP/XKlp3B-ST transgene (P < 0.005). There was no statistical difference between rod-cone ratios of tadpoles containing the EGFP transgene alone and wild-type tadpoles. Error bars, SEM.

Figure 4.

 

A tadpole containing EGFP/XKlp3B-ST and XKlp3A-ST transgenes possessed very few rods due to rod degeneration. (A) A retinal section from a tadpole containing both EGFP/XKlp3B-ST and XKlp3A-ST transgenes was composed predominantly of cone photoreceptors. A degenerating rod with dense vesicular accumulations (arrow) in the inner segment was seen in the kinesin II transgenic retina and not in the rods of a transgenic tadpole expressing EGFP alone (B). In addition, the rod outer segments in the kinesin II transgenic retina were highly disorganized, whereas the rods of the EGFP tadpole contained an extensive number of ordered disks in their outer segments. The nucleus of the rod (N) in the double kinesin II transgenic retina was condensed compared with the cone nuclei in the outer nuclear layer and with the underlying nuclei in the inner nuclear layer. The cones (⋆ in the lipid droplets) had a relatively normal arrangement in the retina of the kinesin II transgenic compared with the EGFP transgenic retina. Scale bar, 2 μm.

Figure 5.

 

The EGFP/XKlp3B-ST transgene was transmissible to F1 progeny and caused a reduction in the number of rods in the transgenic retina. An adult male frog containing the EGFP/XKlp3B-ST transgene was mated to a wild-type female to generate F1 progeny. Tadpoles expressing the transgene were selected for the presence of EGFP eye fluorescence and fixed 7 days after fertilization. Retinal sections from the F1 progeny were analyzed for photoreceptor number and an average rod-cone ratio was determined for each. A statistically significant reduction in the rod-cone ratio of EGFP/XKlp3B-ST F1 progeny (▪), compared with control EGFP (▧) and wild-type (□) tadpoles of the same age, was observed (P < 0.001). Error bars, SEM.

The authors thank Nathalie Le Bot and Isabelle Vernos for providing the EGFP/XKlp3B-ST and XKlp3A plasmids, Ming Ji for help with the initial transgenic frogs, Andréa Dosé for critical reading of the manuscript, and members of the Burnside Lab for useful comments.