|
Figure 1. KCNMB3 N-terminal variation. (A) Predicted β3a (exon 1a) and the initial segment of exon 2 were determined for a range of vertebrate species from searches of publicly available databases, as described in the Materials and methods. Sequences are from the top: human (Hs: Homo sapiens), chimpanzee (Pt: Pan troglodytes), rhesus monkey (Mm: Mulatta macaca), marmoset (Ct: Callithrix jacchus), bushbaby (Og: Otolemur garnettii), treeshrew (Tb: Tupaia belangeri), thirteen-lined squirrel (St: Spermophilus tridecemlineatus), cow (Bt: Bos taurus), horse (Ec: Equus caballus), mouse (Mm': Mus musculus), rat (Rn: Rattus norvegicus), cat (Fc: Felix catus), dog (Cf: Canus lupis familiaris), opossum (Md: Monodelphis domestica), and duckbilled platypus (Oa: Ornithorhynchus anatinus). Exon 1a when attached to exon 2 encodes the β3a isoform. The β3b isoform is generated when exon 1b (not depicted), corresponding to a single methionine, is attached to Exon 2. Vertical red line indicates junction between N-terminal exons and exon 2. Gray highlights residues that are identical among most species. Unshaded residues are unique, while yellow, blue, or green indicate residues shared among a smaller subset of species. (B) Potential N termini with homology to human β3c are aligned. Horizontal blue line separates primate (above) and nonprimate (below) species. The marmoset (Ct) open reading frame (ORF) requires a change in reading frame (indicated by **), while the cat (St) ORF is in a different reading from exon 2. (C) Potential N termini with homology to human β3d are aligned with a horizontal blue line separating primates from other mammals. The cat putative β3d N terminus contained frame shifts and stop codons in the reading frame (**). (D) KCNMB3 gene maps for human and mouse are compared. In contrast to human, the putative mouse 1b coding exon (1b?) that was identified by cloning is positioned upstream of the conserved 1a exon.
|
|
Figure 2. The KCNMB3 1a exon is highly conserved in mammals. Segments of nucleotides spanning the putative 1a exon of various placental mammals were aligned using the UCSC genome browser. Displayed nucleotides are complementary to those for mRNA and are displayed in reverse order to that typically used for amino acid alignments. The top two tracks show, respectively, the nucleotide positions on human chr. 3 and the presumptive open reading frame for the identified human 1a exon. The PhastCons plot shows scores for 17 placental mammals plus human over this genome segment. For the alignments, the segment boxed in red indicates the highly conserved consensus GT intron–exon splice site preceding the ORF (reverse complement: GTAAGT). Numbers on the Gaps track indicate number of inserted residues for particular species indicated by vertical orange lines. Double lines indicate segments with one or more unalignable bases. Single lines indicate positions with no alignable bases.
|
|
Figure 3. The mouse β3a subunit produces a use-dependent tail current enhancement similar to hβ3a, but mβ3a inactivation is more rapid and more complete. (A) Currents arising from α+mβ3a subunits were activated by the indicated protocol. (B) Currents resulting from α+hβ3a subunits show slower and more incomplete inactivation. Currents in both A and B were activated by the indicated voltage protocols with 10 μM cytosolic Ca2+. (C) Inactivation time constants measured over a range of potentials are displayed for both hβ3a and mβ3a. (D) Currents arising from channels containing either mouse β3a subunits (top) or human β3a subunits (bottom) are displayed on a slower time base to emphasize the slow tail currents. Currents were activated by the indicated voltage protocol with 10 μM Ca2+. Control currents are in red, while currents following brief application of 0.1 mg/ml trypsin are in black. Trypsin removes inactivation and the slow tail currents. (E) Activation steps of differing duration to +180 mV were used to elicit α+mβ3a currents with 10 μM Ca2+. As the command step duration is increased, the tail current switches from exclusively fast deactivating to slow deactivating. Intermediate duration command steps elicit tail currents containing distinct slow and fast time constants. Below, a brief (black, τf) and more prolonged (red, τs) step are compared emphasizing the differences in tail currents in each case. (F) The fraction of slow time constant (τs: red circles) as a function of command step duration is plotted along with the time course of inactivation onset, showing the close correlation of the two. In G, the temporal development of the slow component of tail current is compared for mouse (red circles) and human (black circles) β3a subunits.
|
|
Figure 4. Primate and candidate rodent KCNMB3 1b exons probably arose independently and neither appear to be conserved among mammals. In A, the indicated genomes were compared using the UCSC genome browser for homologies to a segment from the cloned human KCNMB3 1b exon (Uebele et al., 2000) but also including 10 bases preceding the beginning of the presumed 1b ORF. Segments of homology are only observed in chimp and rhesus monkey, both of which also share a consensus sequence (red box) appropriate for an intron–exon junction just preceding the ORF. In B, the indicated genomes were compared for segments homologous to the 122 base pairs in the cloned putative mouse 1b exon, with no segments of homology found in any other mammalian genome. The candidate mouse 1b exon is preceded by an appropriate intron–exon splice site.
|
|
Figure 5. The rat genome contains a potential homologue of the mouse KCNMB3 1b' exon. (A) A lower stringency conservation screen, using the candidate mouse KCNMB3 1b' sequence (build mm8), identified a segment of potential homology in the rat. The rat segment contains a 120-bp insert of no homology that connects two segments sharing homology with the mouse 1b' exon. This rat 1b' homologue occurs in a position similar to that in the mouse within the overall KCNMB3 gene map (Fig. S5). (B) The cloned mouse 1b' exon is aligned directly with the candidate rat 1b' exon highlighting the insert (yellow) in the rat sequence. Both candidate exons share a similar consensus sequence for an appropriate intron–exon splice site. The 120-bp insert appears to be a repeatable genetic element occurring on multiple chromosomes in the rat. A homologue of ∼90% identity is also found in the mouse. The position of the mouse 1b exon is, in build mm8, chr3:32,400,082-32,400,204, while in build mm8, the position is chr3:32,692,709-32,692,831. Alignments based on mm8 use a 17-way vertebrate species set while alignments based on mm9 use a 30-way Multiz alignment. The UCSC alignment procedures use an additional filtering step in the generation of the 30-way conservation track that reduces the number of paralogues and pseudogenes from the high-quality assemblies.
|
|
Figure 6. The mouse candidate β3b' subunit is functionally distinct from the human β3b subunit, failing to produce trypsin-sensitive inactivation of outward current. (A) Currents from α+mβ3b' subunits were activated with the indicated voltage protocol either with 0 (A1) or 10 (A2) μM cytosolic Ca2+. (B) Currents were acquired from the same patch after application of 0.5 mg/ml trypsin for a period sufficient to remove inactivation mediated by hβ3b subunits. (C) Currents resulting from α+hβ3b subunits are displayed, with the inset in C2 showing currents at 10 μM at higher magnification to show the rapid inactivation at positive command potentials. (D) After application of trypsin, α+hβ3b currents are markedly increased both at 0 (D1) and 10 (D2) μM, due to removal of fast block at positive activation potentials.
|
|
Figure 7. Both residue and length variation contribute to differences in inactivation behavior between mouse β3b' and human β3b subunits. (A) BK channels containing hβ3b subunits were activated by 10 μM Ca2+ with voltage steps applied in 20-mV steps up to +180 mV with tail currents at −120 mV. Currents in the left column were obtained in control saline and, on the right, after brief application of 0.1 mg/ml trypsin. (B) Currents arising from a construct containing a substitution of phenylalanine with leucine (hβ3b-F4L) are shown before and after trypsin. Inactivation is less ineffective but still persists. (C) In construct hβ3b(Δ10-15), residues 10–15 were deleted from hβ3b, resulting in some attenuation of fast inactivation, but also conferring resistant of the residual inactivation to digestion by trypsin. (D) Leucine 4 in mβ3b' was mutated to phenylalanine (mβ3b-L4F), resulting in slight restoration of block at more positive voltages. Consistent with the absence of trypsin sensitivity of the shortened hβ3b(Δ10-15) in C, the block of outward currents in mβ3b'-L4F shows only weak sensitivity to digestion by trypsin.
|
|
Figure 8. The mouse β3b' subunit produces leftward gating shifts at all Ca2+ concentrations. (A) GV curves were generated from measurement of tail currents following activation protocols as in Fig. 6, either for patches expressing α subunits alone (n = 5; black symbols) or for patches with α+mβ3b' subunits (n = 5; red symbols) for cytosolic Ca2+ concentrations from 0 to 300 μM. (B) Vh estimated from a fit of a Boltzmann equation (Eq. 1) to GV curves at each Ca2+ is plotted as a function of applied [Ca2+]i. (C) Time constants of activation and deactivation for α + mβ3b' channels are plotted as a function of voltage for four Ca2+ concentrations, highlighting a large change in activation rates for an increase in Ca2+ from 1 to 10 μM, and the more gradual slowing in tail current deactivation as Ca2+ is increased. (D) Time constants for activation and deactivation are compared for α alone (black symbols) and α+mβ3b' (red symbols), showing the similarity in activation rates at positive potentials, but marked changes in deactivation. (E) Red symbols plot single channel Po as a function of membrane potential for six single channel patches with Po approaching 0.8 at +200 mV. For comparison, the GV curve at 0 Ca2+ normalized to a maximum saturating fractional conductance of 0.95 is plotted. For comparison, the single channel Po was determined for five patches expressing single Slo1 α channels at +200 mV and 0 μM Ca2+ (solid black circle).
|
|
Figure 9. Mouse β3b' produces weaker instantaneous outward current rectification than human β3b. In A, tail currents at voltages from +180 through −180 mV following an activating voltage step to +180 mV are shown for α+mβ3b' currents. (B) Tail currents measured immediately following the repolarizing step are plotted for Slo1 α alone (black circles), α+mβ3b' (red circles), and α+hβ3b (black diamonds, following trypsin-mediated removal of hβ3b-mediated inactivation) with currents normalized to the amplitude at +100 mV. (C) Openings of single α+mβ3b' channels are shown at potentials from +160 to −160 mV. O and C indicate open and closed current levels, respectively. Ca2+ was varied from 0 μM (+160 and +100 mV), 10 μM (−100 mV), and 300 μM (−160 mV) to obtain a sufficient number of openings at each voltage. The dotted lines correspond to a 200-pS single channel conductance level. (D) Single channel currents were plotted as a function of voltage (red circles), while the instantaneous tail current amplitudes, normalized to the single channel current level at +100 mV is also shown (open circles).
|
|
Figure 10. KCNMB3 exons 1c and 1d exhibit weak conservation among mammals. (A) A segment from human KCNMB3 containing the end of shared exon 2 and exon 1c was used to search for homology among placental mammals. The translated open reading frame is shown at the top. Both exon 1c and the end of exon 2 exhibit low PhastCons scores indicative that there has been little selective pressure in this region among placental mammals. Many species also lack a triplet (CAT complementary to AUG) encoding an initiation methionine or contain gaps or inserts that disrupt the ORF (see Fig. S6 for translations of different reading frames). (B) A segment spanning the presumed human 1d exon also shows minimal conservation with other placental mammals, despite strong sequence identity in the higher primates. The intronic side of the presumed GT 5′ intron–exon splice junction is noted in the red block. The 5′ intron–exon splice site, which is conserved in primates, exhibits a number of substitutions across species, although the initial GT motif is shared among most of the displayed genomes. The presence of inserts, gaps, and missing CAT methionine-encoding triplets suggests that, other than in primates, this region may also not contribute valid ORFs in most mammals.
|
