Defective kinesin heavy chain behavior in mouse kinesin light chain mutants

Abstract

Conventional kinesin, kinesin-I, is a heterotetramer of two kinesin heavy chain (KHC) subunits (KIF5A, KIF5B, or KIF5C) and two kinesin light chain (KLC) subunits. While KHC contains the motor activity, the role of KLC remains unknown. It has been suggested that KLC is involved in either modulation of KHC activity or in cargo binding. Previously, we characterized KLC genes in mouse (Rahman, A., D.S. Friedman, and L.S. Goldstein. 1998. J. Biol. Chem. 273:15395-15403). Of the two characterized gene products, KLC1 was predominant in neuronal tissues, whereas KLC2 showed a more ubiquitous pattern of expression. To define the in vivo role of KLC, we generated KLC1 gene-targeted mice. Removal of functional KLC1 resulted in significantly smaller mutant mice that also exhibited pronounced motor disabilities. Biochemical analyses demonstrated that KLC1 mutant mice have a pool of KIF5A not associated with any known KLC subunit. Immunofluorescence studies of sensory and motor neuron cell bodies in KLC1 mutants revealed that KIF5A colocalized aberrantly with the peripheral cis-Golgi marker giantin in mutant cells. Striking changes and aberrant colocalization were also observed in the intracellular distribution of KIF5B and beta'-COP, a component of COP1 coatomer. Taken together, these data best support models that suggest that KLC1 is essential for proper KHC activation or targeting.

Figure 1

Generation of KLC1 gene-targeted mice. A, Restriction enzyme map of the genomic DNA of KLC1 and the resulting targeting vector. Arrowheads indicate location of PCR primers for detecting homologous recombination events. Restriction enzyme sites are as follows: H, HindIII; S, SacI; X, XbaI; B, BglII; EV, EcoRV; E, EcoR0109; P, PstI; and N, NotI. The thicker black boxes represent exons. B, Southern analysis of clones that tested positive by PCR for recombination events. The recombination event produces a larger SacI fragment when probed with a 240-bp SacI/BglII fragment residing outside the targeting vector. The clones A1, A2, E2, F1, F11, and H4 all tested positive for a recombination event. A4 is another G418 resistant colony that was not positive for homologous recombination at the correct locus. Germline transmission was obtained from lines E2 and F11. C, Offspring from matings of KLC1 heterozygous mice were genotyped by PCR. PCR was performed on genomic DNA isolated from mouse tail clippings. Primers were designed to amplify 800- and 210-bp fragments for the recombinant and wild-type KLC1 genes, respectively.

Figure 2

Analysis of basic phenotypes of KLC1 mutant mice. A, All 98 offspring from the first 12 litters of F1 heterozygous × F1 heterozygous matings were weighed at days 10, 14, 18, and 21 after birth. 25 adult mice of each genotype (caged together) were also weighed at ∼1 year of age as a final data point. The average weights in each category are represented in this graph. Homozygous mice were consistently smaller than their heterozygous or wild-type littermates. B, 15 wild-type, heterozygous, and homozygous mice were tested for sensorimotor defects. Mice were allowed to hang upside down from chicken wire (1-cm gauge) attached to a bell jar ∼1.5–2 ft above a surface. They were subsequently timed until they could no longer maintain grip and fell. Timing was cut off at 2 min for wild-type and mutant heterozygotes. C, Picture of wild-type and homozygous mice from the same litter at ∼3-wk old. The wild-type mouse (agouti coat color) is noticeably larger than the KLC1 homozygous mouse (black coat color). D, Equal amounts of total protein (40 μg) from brain extracts were loaded for Western analysis. KIF5A and KIF5B antibodies were used to detect the two KHC components, and 63-90 was used to detect both KLC2 (upper band) and KLC1 (lower band). Actin was used as an internal loading control for the blots. The intensity of the bands was quantitated using NIH Image and the ratios calculated as shown in Table .

Figure 3

Biochemical analysis of KLC1 mutant mice. A, KIF5A or KIF5B antibodies were used in immunoprecipitation experiments and then probed with KIF5A, KIF5B, or 63-90 antibodies to assess the association of KHC and KLC forms in wild-type, heterozygous, and homozygous mutant genotypes. B, Sucrose gradient analysis of high-speed supernatant of brain extracts from wild-type (open circles) and mutant KLC1 (closed circles) was done using 5–20% linear sucrose gradients. 16 fractions were collected (fraction 1 is top of gradient), and equal volumes were loaded for Western analysis. The relative intensity of the bands in each fraction was calculated as described in Fig. 2 D and plotted. Control protein markers were run in parallel gradients; the enzyme activity of alcohol dehydrogenase (7 S) was at fraction 6, catalase (11.3 S) was at fraction 9-10, and β-galactosidase (16 S) was at fraction 13-14 (data not shown).

Figure 3

Biochemical analysis of KLC1 mutant mice. A, KIF5A or KIF5B antibodies were used in immunoprecipitation experiments and then probed with KIF5A, KIF5B, or 63-90 antibodies to assess the association of KHC and KLC forms in wild-type, heterozygous, and homozygous mutant genotypes. B, Sucrose gradient analysis of high-speed supernatant of brain extracts from wild-type (open circles) and mutant KLC1 (closed circles) was done using 5–20% linear sucrose gradients. 16 fractions were collected (fraction 1 is top of gradient), and equal volumes were loaded for Western analysis. The relative intensity of the bands in each fraction was calculated as described in Fig. 2 D and plotted. Control protein markers were run in parallel gradients; the enzyme activity of alcohol dehydrogenase (7 S) was at fraction 6, catalase (11.3 S) was at fraction 9-10, and β-galactosidase (16 S) was at fraction 13-14 (data not shown).

Figure 4

Abnormal accumulation of KIF5A on Golgi structures in KLC1 mutants. A, Abnormal localization of KIF5A in the motor neuron cell bodies on structures that costain with the peripheral cis-Golgi marker, giantin. KIF5A staining is diffuse in wild-type cells, but accumulates on structures staining for a Golgi marker in KLC1 mutants. B, KIF5A localization in or near structures staining for giantin in KLC1 mutants is also observed in the cell bodies of sensory neuron in the DRG. C, There are no detectable differences in the staining pattern of KIF5A between wild-type and heterozygous DRG sensory and motor neurons. Bars, 10 μm.

Figure 5

Immunofluores-cence studies of sensory neuron cell bodies of the DRG. A, Giantin recognizes similar structures as MannII in DRG sensory neuron cell bodies. B, There is a marked depletion of KLC2 in cell bodies of sensory neurons of the DRG in KLC1 mutant mice. The punctate perinuclear staining observed in KLC1 −/− cells is also detected in wild-type cells, although not as clearly because of the enhanced cytoplasmic staining. C, KLC1 staining in wild-type, heterozygous, and homozygous KLC1 mutants. D, KIF5B staining changes from an overall diffuse pattern in wild-type cells to a punctate form in KLC1 −/− cells. E, β′-COP (COP1) staining as visualized by CM1A10 is disrupted in KLC1 mutant mice. The staining pattern with CM1A10 resembles Golgi apparatus-like structures in wild-type cells, however, this pattern becomes punctate in KLC1 mutant cells. F, Colocalization of KIF5B and β′-COP (CM1A10) in aberrant accumulations in KLC1 mutants. Bars, 10 μm.

Figure 6

Sciatic nerve ligations detect plus end directed movement of both KIF5A and KIF5B in KLC1 mutant mice. Accumulations on the proximal side of the ligations were comparable between wild-type, heterozygous, and mutant mice. Bar, 100 μm.