Long-term in vitro maintenance of clonal abundance and leukaemia-initiating potential in acute lymphoblastic leukaemia

Lack of suitable in vitro culture conditions for primary acute lymphoblastic leukaemia (ALL) cells severely impairs their experimental accessibility and the testing of new drugs on cell material reflecting clonal heterogeneity in patients. We show that Nestin-positive human mesenchymal stem cells (MSCs) support expansion of a range of biologically and clinically distinct patient-derived ALL samples. Adherent ALL cells showed an increased accumulation in the S phase of the cell cycle and diminished apoptosis when compared with cells in the suspension fraction. Moreover, surface expression of adhesion molecules CD34, CDH2 and CD10 increased several fold. Approximately 20% of the ALL cells were in G0 phase of the cell cycle, suggesting that MSCs may support quiescent ALL cells. Cellular barcoding demonstrated long-term preservation of clonal abundance. Expansion of ALL cells for >3 months compromised neither feeder dependence nor cancer initiating ability as judged by their engraftment potential in immunocompromised mice. Finally, we demonstrate the suitability of this co-culture approach for the investigation of drug combinations with luciferase-expressing primograft ALL cells. Taken together, we have developed a preclinical platform with patient-derived material that will facilitate the development of clinically effective combination therapies for ALL.

Transmitochondrial embryonic stem cells containing pathogenic mtDNA mutations are compromised in neuronal differentiation

Objectives:  Defects of the mitochondrial genome (mtDNA) cause a series of rare, mainly neurological disorders. In addition, they have been implicated in more common forms of movement disorders, dementia and the ageing process. In order to try to model neuronal dysfunction associated with mitochondrial disease, we have attempted to establish a series of transmitochondrial mouse embryonic stem cells harbouring pathogenic mtDNA mutations.

Materials and methods: Transmitochondrial embryonic stem cell cybrids were generated by fusion of cytoplasts carrying a variety of mtDNA mutations, into embryonic stem cells that had been pretreated with rhodamine 6G, to prevent transmission of endogenous mtDNA. Cybrids were differentiated into neurons and assessed for efficiency of differentiation and electrophysiological function.

Results:  Neuronal differentiation could occur, as indicated by expression of neuronal markers. Differentiation was impaired in embryonic stem cells carrying mtDNA mutations that caused severe biochemical deficiency. Electrophysiological tests showed evidence of synaptic activity in differentiated neurons carrying non‐pathogenic mtDNA mutations or in those that caused a mild defect of respiratory activity. Again, however, neurons carrying mtDNA mutations that resulted in severe biochemical deficiency had marked reduction in post‐synaptic events.

Conclusions:  Differentiated neurons carrying severely pathogenic mtDNA defects can provide a useful model for understanding how such mutations can cause neuronal dysfunction.

Modeling channel properties in vestibular calyx terminals

Two types of hair cell have been described in the vestibular end organs of amniotes, but the coding of sensory signals by different hair cell types is not well understood. Type I hair cells are contacted by cup-shaped afferent calyx terminals, whereas type II hair cells are contacted by bouton terminals. The whole cell patch-clamp technique has been used to record voltage-dependent ionic currents from calyx terminals. Type I hair cells along with their calyces were non-enzymatically dissociated from the semicircular canals and utricles of Mongolian gerbils. Voltage-dependent currents identified in whole cell voltage-clamp included transient inward sodium currents and outward potassium currents. Potassium currents were pharmacologically blocked by cesium in the patch electrode solution. The NEURON simulation environment was used to model the properties of the calyx terminal. A series of interconnected cylindrical compartments was designed to represent the inner and outer calyx membrane, the base of the calyx and a short axon segment. Kinetic parameters for the Na+ current were optimized with a genetic algorithm to match kinetic data from the whole cell recordings.

Mechano-electrical transduction in the turtle utricle

Two classes of mechano-sensory hair cell are present in the vestibular system of mammals, birds and reptiles. Type I hair cells are bottle-shaped and make synaptic contact with afferent calyx terminals. Type II hair cells are cylindrical and contact small bouton afferent terminals. Voltage-dependent basolateral currents have been found to differ between the two cell types and these properties are believed to contribute to the shaping of primary afferent responses. Type I hair cells have a low input resistance, which may be related to the low gain of calyx afferents. In the turtle utricle, type I vestibular hair cells are found only in a narrow band of the sensory epithelium called the striola, whereas type II hair cells are found in both striolar and extrastriolar regions. We have made whole cell patch clamp recordings from type I hair cells, type II hair cells and calyx fibers in order to better understand the processing of vestibular signals. Here we describe responses of hair cells to hair bundle displacement with a stiff glass probe. Mechano-electrical transduction (MET) currents were largest in type I hair cells, where the mean peak amplitude was approximately 500 pA. MET currents in all hair cells showed rapid and slow adaptation.

Return of potassium ion channels in regenerated hair cells: possible pathways and the role of intracellular calcium signaling

Recent electrophysiological studies in pigeon have demonstrated that potassium channels are completely functional in regenerated type II hair cells at 21 days post-treatment (PT) with ototoxic doses of streptomycin. The currents return in the order they appear during development. The mixture of ionic currents in a regenerated type II hair cell in a particular region of the neuroepithelium is the same as in its ancestor in that region. The return of currents in regenerated type I hair cells is more complicated. The dominant conductance gKI is not present until after 70 days PT. Before 70 days, the ionic currents in type I hair cells resemble those of regenerated type II hair cells, suggesting that the ionic currents in type II hair cells might be precursors of the ionic currents in regenerated type I hair cells. New data show that at one year PT, the kinetics and drug sensitivity of the dominant K+ conductance in type I hair cells are identical to gKI. Supporting cells, believed to be the precursors of regenerated type II hair cells, have effectively no voltage-gated outward potassium channels, suggesting that regenerated type II hair cells must develop these channels de novo. The next step is to understand the mechanisms by which the potassium channel protein is synthesized, migrates through the cytosol, and is inserted into the plasmalemma of regenerating hair cells. These mechanisms are unknown. We propose that intracellular calcium is involved in this process, as well as in the differentiation, proliferation, and gene regulation of precursor cells fated to become hair cells.

Effects of KCNQ channel blockers on K(+) currents in vestibular hair cells

Linopirdine and XE991, selective blockers of K(+) channels belonging to the KCNQ family, were applied to hair cells isolated from gerbil vestibular system and to hair cells in slices of pigeon crista. In type II hair cells, both compounds inhibited a slowly activating, slowly inactivating component of the macroscopic current recruited at potentials above -60 mV. The dissociation constants for linopirdine and XE991 block were <5 microM. A similar component of the current was also blocked by 50 microM capsaicin in gerbil type II hair cells. All three drugs blocked a current component that showed steady-state inactivation and a biexponential inactivation with time constants of approximately 300 ms and 4 s. Linopirdine (10 microM) reduced inward currents through the low-voltage-activated K(+) current in type I hair cells, but concentrations up to 200 microM had little effect on steady-state outward K(+) current in these cells. These results suggest that KCNQ channels may be present in amniote vestibular hair cells.

Effects of cationic substitutions on delayed rectifier current in type I vestibular hair cells

The resting potassium current (I(KI)) in gerbil dissociated type I vestibular hair cells has been characterized under various ionic conditions in whole cell voltage-clamp. When all K(+) in the patch electrode solution was replaced with Na(+), (Na(+))(in) or Cs(+), (Cs(+))(in), large inward currents were evoked in response to voltage steps between -90 and -50 mV. Activation of these currents could be described by a Hodgkin-Huxley-type kinetic scheme, the order of best fit increasing with depolarization. Above approximately -40 mV currents became outward and inactivated with a monoexponential time course. Membrane resistance was inversely correlated with external K(+) concentration. With (Na(+))(in), currents were eliminated when K(+) was removed from the external solution or following extracellular perfusion of 4-aminopyridine, indicating that currents flowed through I(KI) channels. Also, reduction of K(+) entry through manipulation of membrane potential reduced the magnitude of the outward current. Under symmetrical Cs(+), 0 K(+) conditions I(KI) is highly permeable to Cs(+). However, inward currents were reduced when small amounts of external K(+) were added. Higher concentrations of K(+) resulted in larger currents indicating an anomalous mole fraction effect in mixtures of external Cs(+) and K(+).

Evidence for an Na(+)-K(+)-Cl- cotransporter in mammalian type I vestibular hair cells

In amniotes, there are two types of hair cells, designated I and II, that differ in their morphology, innervation pattern, and ionic membrane properties. Type I cells are unique among hair cells in that their basolateral surfaces are almost completely enclosed by an afferent calyceal nerve terminal. Recently, several lines of evidence have ascribed a motile function to type I hair cells. To investigate this, elevated external K+, which had been used previously to induce hair cell shortening, was used to induce shape changes in dissociated mammalian type I vestibular hair cells. Morphologically identified type I cells shortened and widened when the external K+ concentration was raised isotonically from 2 to 125 mM. The shortening did not require external Ca2+ but was abolished when external Cl- was replaced with gluconate or sulfate and when external Na+ was replaced with N-methyl-D-glucamine. Bumetanide (10-100 microM), a specific blocker of the Na(+)-K(+)-Cl- cotransporter, significantly reduced K(+)-induced shortening. Hyposmotic solution resulted in type I cell shape changes similar to those seen with high K+, i.e., shortening and widening. Type I cells became more spherical in hyposmotic solution, presumably as a result of a volume increase due to water influx. In hypertonic solution, cells became narrower and increased in length. These results suggest that shape changes in type I hair cells induced by high K+ are due, at least in part, to ion and solute entry via an Na(+)-K(+)-Cl- cotransporter, which results in cell swelling. A scheme is proposed whereby the type I hair cell depolarizes and K+ leaves the cell via voltage-dependent K+ channels and accumulates in the synaptic space between the type I hair cell and calyx. Excess K+ could then be removed from the intercellular space by uptake via the cotransporter.

Vestibular type I and type II hair cells. 1: Morphometric identification in the pigeon and gerbil

Classically, type I and type II vestibular hair cells have been defined by their afferent innervation patterns. Little quantitative information exists on the intrinsic morphometric differences between hair cell types. Data presented here define a quantitative method for distinguishing hair cell types based on the morphometric properties of the hair cell's neck region. The method is based initially on fixed histological sections, where hair cell types were identified by innervation pattern, type I cells having an afferent calyx. Cells were viewed using light microscopy, images were digitized, and measurements were made of the cell body width, the cuticular plate width, and the neck width. A plot of the ratio of the neck width to cuticular plate width (NPR) versus the ratio of the neck width to the body width (NBR) established four quadrants based on the best separation of type I and type II hair cells. The combination of the two variables made the accuracy of predicting either type I or type II hair cells greater than 90%. Statistical cluster analysis confirmed the quadrant separation. Similar analysis was performed on dissociated hair cells from semicircular canal, utricle, and lagena, giving results statistically similar to those of the fixed tissue. Additional comparisons were made between fixed tissue and isolated hair cells as well as across species (pigeon and gerbil) and between end organs (semicircular canal, utricle, and lagena). In each case, the same morphometric boundaries could be used to establish four quadrants, where quadrant 1 was predominantly type I cells and quadrant 3 was almost exclusively type II hair cells. The quadrant separations were confirmed statistically by cluster analysis. These data demonstrate that there are intrinsic morphometric differences between type I and type II hair cells and that these differences can be maintained when the hair cells are dissociated from their respective epithelia.

Vestibular type I and type II hair cells. 2: Morphometric comparisons of dissociated pigeon hair cells

Morphometric properties of solitary hair cells dissociated from the semicircular canals (SCC), utricles (UTR), and lagenas (LAG) of adult white king pigeons, Columbia livia, were compared. Measurements were made of the cell body, cuticular plate and hair bundle. Cells were divided into two groups: type 1 (group 1) was predominantly type I hair cells, and type 2 (group 3) was primarily type II hair cells. Comparisons are made initially between end organs for each group. A subpopulation of short otolith hair cells was identified. Quantitative comparisons between isolated type 1 and type 2 hair cells demonstrated that type 1 hair cells were more homogeneous both within and between vestibular end organs; while they had shorter, thinner neck regions, narrower apical surfaces, with longer and thinner bodies than did type 2 hair cells. Generally, for both type 1 and type 2 hair cells, two different hair bundle shapes were present, those (unimodal) with a single sharp taper from longest to shortest stereocilia, and those (bimodal) with an initial steep taper followed by a less steep taper. An additional subtype of type 1 hair cells with short hair bundles was identified. SCC hair cells have fewer hair bundles with bimodal tapers across all cell groups when compared to UTR or LAG. All cell subtypes identified for dissociated hair cells were corroborated using histologic sections.

Electrical filtering in gerbil isolated type I semicircular canal hair cells

1. Membrane potential responses of dissociated gerbil type I semicircular canal hair cells to current injections in whole cell current-clamp have been measured. The input resistance of type I cells was 21.4 +/- 14.3 (SD) M omega, (n = 25). Around the zero-current potential (Vz = -66.6 +/- 9.3 mV, n = 25), pulsed current injections (from approximately -200 to 750 pA) produced only small-amplitude, pulse-like changes in membrane potential. 2. Injecting constant current to hyperpolarize the membrane to around -100 mV resulted in a approximately 10-fold increase in membrane resistance. Current pulses superimposed on this constant hyperpolarization produced larger and more complex membrane potential changes. Depolarizing currents > or = 200 pA caused a rapid transient peak voltage before a plateau. 3. Membrane voltage was able to faithfully follow sine-wave current injections around Vz over the range 1-1,000 Hz with < 25% attenuation at 1 kHz. A previously described K conductance, IKI, which is active at Vz, produces the low input resistance and frequency response. This was confirmed by pharmacologically blocking IKI. This conductance, present in type I cells but not type II hair cells, would appear to confer on type I cells a lower gain, but a much broader bandwidth at Vz, than seen in type II cells.

The delayed rectifier, IKI, is the major conductance in type I vestibular hair cells across vestibular end organs

Hair cells were dissociated from the semicircular canal, utricle, lagena and saccule of white king pigeons. Type I hair cells were identified morphologically based on the ratios of neck width to cuticular plate width (NPR < 0.72) as well as neck width to cell body width (NBR < 0.64). The perforated patch variant of the whole-cell recording technique was used to measure electrical properties from type I hair cells. In voltage-clamp, the membrane properties of all identified type I cells were dominated by a predominantly outward potassium current, previously characterized in semicircular canal as IKI. Zero-current potential, activation, deactivation, slope conductance, pharmacologic and steady-state properties of the complex currents were not statistically different between type I hair cells of different vestibular end organs. The voltage dependence causes a significant proportion of this conductance to be active about the cell's zero-current potential. The first report of the whole-cell activation kinetics of the conductance is presented, showing a voltage dependence that could be best fit by an equation for a single exponential. Results presented here are the first data from pigeon dissociated type I hair cells from utricle, saccule and lagena suggesting that the basolateral conductances of a morphologically identified population of type I hair cells are conserved between functionally different vestibular end organs; the major conductance being a delayed rectifier characterized previously in semicircular canal hair cells as IKI.

Potassium currents in mammalian and avian isolated type I semicircular canal hair cells

1. Type I vestibular hair cells were isolated from the cristae ampullares of the semicircular canals of the Mongolian gerbil (Meriones unguiculatus) and the white king pigeon (Columba livia). Dissociated type I cells were distinguished from type II hair cells by their neck to plate ratio (NPR) and their characteristic amphora shape. 2. The membrane properties of gerbil and pigeon type I hair cells were studied in whole-cell voltage- and current-clamp using the perforated patch technique with amphotericin B as the perforating agent. 3. In whole-cell current-clamp, the average zero-current potential, Vz, measured for pigeon type I hair cells, was -70 +/- 7 (SD) mV (n = 18) and -71 +/- 11 mV (n = 83) for gerbil type I hair cells. 4. At Vz, for both gerbil and pigeon type I hair cells, a potassium current (IKI) was > or = 50% activated. This current deactivated rapidly when the membrane potential was hyperpolarized below -90 mV. 5. IKI was blocked by externally applied 4-aminopyridine (4-AP) (5 mM) and by internally applied 20 mM tetraethylammonium (TEA). It was also reduced when 4 mM barium was present in the external solution. The degree of block by barium increased as the membrane potential became more positive. External cesium (5 mM) blocked the inward component of IKI. When IKI was pharmacologically blocked, Vz depolarized by approximately 40 mV. Therefore IKI appears to be a delayed rectifier and to set the more negative Vz noted for isolated type I hair cells when compared to isolated type II hair cells, which do not have IKI. 6. A second, smaller potassium current was present at membrane potential depolarizations above -40 mV. This current was blocked by 30-50 mM, externally applied TEA, 100 microM quinidine, 100 nM apamin, but not 100 nM charybdotoxin, indicating that this is a calcium-activated potassium current, IK(Ca), different from the maxi-K calcium-activated potassium current found in most other hair cells.

Effects of extracellular ATP on hair cells isolated from the guinea-pig semicircular canals

Externally applied ATP (100 microM) induced membrane currents in type I and type II vestibular hair cells enzymatically isolated from guinea-pig semicircular canals. In whole-cell voltage-clamp and with 140 mM K+ in the pipette solution, ATP evoked an inwardly directed current in 58% of the cells when held at potentials below -40 mV. In the remainder, external ATP produced an outward current. After block of the K currents, an inward current activated by ATP was revealed at -50 mV. Intracellular Ca2+ levels were monitored using the Ca2+ indicator Fura-2 and were found to rise in both hair cell types in response to ATP. These results strongly suggest that ATP directly controls the entry of Ca2+ into crista hair cells which can then further modulate K+ currents.

Ionic currents in isolated vestibular hair cells from the guinea-pig crista ampullaris

Ionic currents have been recorded under whole cell patch clamp in cells isolated from the guinea-pig vestibular system. Type I and type II cells were separately identified. Type II cells were further classified as short (less than 15 microns in length) or tall (greater than 15 microns). Under whole cell voltage clamp, cells showed an outward current which activated at potentials above about -50 mV, and tail currents which reversed near the potassium equilibrium potential. The outward current was reduced in the presence of external 10 mM tetraethylammonium or cadmium ions and when calcium was removed from the external medium. A small cadmium-sensitive transient inward current, a putative calcium current, was observed in cells loaded with caesium from the patch pipette. In 27 out of 64 cells a component of the recorded outward current inactivated. Such current components were most common in tall type II cells. This inactivating component was blocked by 4-aminopyridine and removed by depolarizing prepulses consistent with it being an A-type potassium current. Type I cells, on the other hand, showed mainly a non-inactivating outward current which slowly relaxed on repolarization to resting potentials. When membrane potentials were measured under current clamp, injections of less than 100 pA produced a single, highly damped transient followed by a plateau in Type II cells. No such transient was present in Type I cells. There is thus little evidence for an electrical resonance in these cells.