Automated screening of neurite outgrowth

Outgrowth of neurites in culture is used for assessing neurotrophic activity. Neurite measurements have been performed very slowly using manual methods or more efficiently with interactive image analysis systems. In contrast, medium-throughput and noninteractive image analysis of neurite screens has not been well described. The authors report the performance of an automated image acquisition and analysis system (IN Cell Analyzer 1000) in the neurite assay. Neuro-2a (N2a) cells were plated in 96-well plates and were exposed to 6 conditions of retinoic acid. Immunofluorescence labeling of the cytoskeleton was used to detect neurites and cell bodies. Acquisition of the images was automatic. The image set was then analyzed by both manual tracing and automated algorithms. On 5 relevant parameters (number of neurites, neurite length, total cell area, number of cells, neurite length per cell), the authors did not observe a difference between the automated analysis and the manual analysis done by tracing. These data suggest that the automated system addresses the same biology as human scorers and with the same measurement precision for treatment effects. However, throughput of the automated system is orders of magnitude higher than with manual methods.     

A simple, flexible, nonfluorescent system for the automated screening of neurite outgrowth

Measurement of neurite outgrowth is a common assay of neurotrophic activity. However, currently available techniques for measuring neurite outgrowth are either time or resource intensive. The authors established a system in which chronic treatment of a subcloned SH-SY5Y cell line with aphidicolin and various concentrations of nerve growth factor (NGF) induced discernable alterations in proliferation and differentiation. Cells were fixed, labeled with a nonfluorescent dye, and evaluated both manually and with an automated analysis system. NGF increased multiple parameters of differentiation, including neurite length, the proportion of cells extending neurites, and branching, as well as promoting cellular survival/proliferation. Interestingly, although NGF treatment increased the total number of branches, it actually decreased the proportion of branches per neurite length. The authors observed no differences in results obtained using the manual and automated systems, but the automated system was orders of magnitude faster. To demonstrate the flexibility of the system, the authors also show that they could measure changes in differentiation induced by a small-molecule Rho kinase inhibitor, as well as by retinoic acid cotreatment with brain-derived neurotrophic factor. In addition to this flexibility, this system does not require specialized equipment or fluorescent antibodies for analysis and therefore provides a less resource-intensive alternative to fluorescence-based systems.     

WIS‐neuromath enables versatile high throughput analyses of neuronal processes

Automated analyses of neuronal morphology are important for quantifying connectivity and circuitry in vivo, as well as in high content imaging of primary neuron cultures. The currently available tools for quantification of neuronal morphology either are highly expensive commercial packages or cannot provide automated image quantifications at single cell resolution. Here, we describe a new software package called WIS‐NeuroMath, which fills this gap and provides solutions for automated measurement of neuronal processes in both in vivo and in vitro preparations. Diverse image types can be analyzed without any preprocessing, enabling automated and accurate detection of neurites followed by their quantification in a number of application modules. A cell morphology module detects cell bodies and attached neurites, providing information on neurite length, number of branches, cell body area, and other parameters for each cell. A neurite length module provides a solution for images lacking cell bodies, such as tissue sections. Finally, a ganglion explant module quantifies outgrowth by identifying neurites at different distances from the ganglion. Quantification of a diverse series of preparations with WIS‐NeuroMath provided data that were well matched with parallel analyses of the same preparations in established software packages such as MetaXpress or NeuronJ. The capabilities of WIS‐NeuroMath are demonstrated in a range of applications, including in dissociated and explant cultures and histological analyses on thin and whole‐mount sections. WIS‐NeuroMath is freely available to academic users, providing a versatile and cost‐effective range of solutions for quantifying neurite growth, branching, regeneration, or degeneration under different experimental paradigms. © 2012 Wiley Periodicals, Inc. Develop Neurobiol, 2013     

A method for the quantitative analysis of nerve growth in vitro

Summary A method is described for the quantitative analysis of the nerve-growth-promoting activity of biological molecules in tissue culture. The criteria used for the evaluation of this activity is based on the neurite length as well as the total number of neurites produced by the explant of whole dorsal root ganglia from 12-d-old chick embryos. A nerve growth index (NGI) is given to each ganglion during each of a 5-d culture period. The NGI is defined as the product of average neurite length in millimeters and the total number of neurites. We report that with increasing concentrations of fetal bovine serum, there was a proportional increase in NGI due to increased neurite density while the neurite length was not greatly affected. The NGI of several proteins with known nerve growth promoting activity, namely nerve growth factor, insulin, transferrin, and fibronectin were investigated for their activity and compared with that of fetal bovine serum. This work was supported in part by grant GM-10374 from the National Institutes of Health, Bethesda, MD.     

The effects of nerve growth factor and dibutyryl cyclic AMP on cytoskeletal densities in cultured sensory ganglia.

The effects of nerve growth factor (NGF) and dibutyryl cyclic AMP (DBC) on the density of cytoskeletal structures in cultured dorsal root ganglia were examined using morphometric techniques. After 24 hr in culture, NGF-treated neurites were longer than either DBC-treated or control neurites. At 48 hr, neurites produced in response to NGF and DBC were of equivalent length, while controls were considerably shorter. Comparison of electron micrographs of neuritic profiles revealed some differences of area and cytoskeletal density between treatment groups. Morphometric analysis was used to determine these differences under several growth conditions, at various rates of elongation and at different neurite lengths. As shown by analysis of variance, both NGF-treated and control neurites tapered in diameter at 48 hr in vitro, while DBC-induced neurites increased in area. An increase in cytoskeletal density for all treatment groups indicated that density was not always correlated with changes in area. An increased density of microtubules as compared to neurofilaments was seen at 24 hr, with equal densities of both cytoskeletal elements present after 48 hr in vitro. Comparisons between individual groups of data indicated that NGF-treated neurites relied primarily on microtubular density at 24 hr in vitro, when NGF induced longer, faster growing neurites. At 48 hr, there was an increase in neurofilaments proximal to the explant in the presence of DBC, implying that DBC may cause increased synthesis and/or transport of these structures. A comparison of microtubule to neurofilament ratios indicated that at 24 hr, there was always a greater density of microtubules. However, after 48 hr, neurofilament density increased such that there were equivalent densities of both cytoskeletal elements, possibly due to the overall increase in length observed in each treatment group. These data imply that 1) neurites with different rates of elongation may exhibit differences in cytoskeletal density; 2) neurites of equivalent lengths may be of differing stabilities; 3) NGF and DBC produce neurites with different cytoskeletal densities, implying divergent mechanisms of neurite induction; 4) the presence or absence of NGF may be partially responsible for variations in cytoskeletal densities observed between peripheral and central processes of DRG during development.     

Automatic tracking of medial gastrocnemius fascicle length during human locomotion

During human locomotion lower extremity muscle-tendon units undergo cyclic length changes that were previously assumed to be representative of muscle fascicle length changes. Measurements in cats and humans have since revealed that muscle fascicle length changes can be uncoupled from those of the muscle-tendon unit. Ultrasonography is frequently used to estimate fascicle length changes during human locomotion. Fascicle length analysis requires time consuming manual methods that are prone to human error and experimenter bias. To bypass these limitations, we have developed an automatic fascicle tracking method based on the Lucas-Kanade optical flow algorithm with an affine optic flow extension. The aims of this study were to compare gastrocnemius fascicle length changes during locomotion using the automated and manual approaches and to determine the repeatability of the automated approach. Ultrasound was used to examine gastrocnemius fascicle lengths in eight participants walking at 4, 5, 6, and 7 km/h and jogging at 7 km/h on a treadmill. Ground reaction forces and three dimensional kinematics were recorded simultaneously. The level of agreement between methods and the repeatability of the automated method were quantified using the coefficient of multiple correlation (CMC). Regardless of speed, the level of agreement between methods was high, with overall CMC values of 0.90 ± 0.09 (95% CI: 0.86-0.95). Repeatability of the algorithm was also high, with an overall CMC of 0.88 ± 0.08 (95% CI: 0.79-0.96). The automated fascicle tracking method presented here is a robust, reliable, and time-efficient alternative to the manual analysis of muscle fascicle length during gait.     

Area-based analyzing technique at cell array experiment using neuronal cell line

We propose a simple procedure for the identification and quantitative analysis of neurite outgrowth in neuronal cell lines that were uniformly differentiated. Upon stimulation most neuronal cell lines extend neurites in the differentiation process, resulting, according to our observation, in the increase of cell surface area. This increase is dependent on the length and the number of extended neurites. Furthermore, we use this method for the phenotype analysis of cell array experiments to perform large-scale functional evaluation of genes involved in the neurite outgrowth during neuronal differentiation.     

Automatic dendritic spine analysis in two-photon laser scanning microscopy images.

Dendritic spine expression plays an important role in the central nervous system. Modern fluorescence microscopy and green fluorescent protein technology have facilitated the research on spines. To quantitatively analyze the spines in fluorescence microscopy images, an automatic dendritic spine analysis method is proposed. Because of the limit of axial resolution, our method is designed to process the projection image along the z-axis and analyze the lateral spines. The method can automatically extract the dendrite centerlines and segment the spines along the dendrites according to width-based criteria. The criteria utilize a common morphological feature of the spines. It can detect some shapes of spines missed by previous methods. In addition, the proposed method is automatic once a few parameters are set. Spine numbers, lengths, and densities, which biologists are interested in, are analyzed both manually and automatically. The results of the two methods match well. The proposed method provides automatic and accurate dendritic spine analysis. It can serve as a useful tool for spine image analysis to avoid tedious manual labor.     

Automated fluorescent in situ hybridization (FISH) analysis of t(9;22)(q34;q11) in interphase nuclei.

BACKGROUND: For chronic myeloid leukemia, the FISH detection of t(9;22)(q34;q11) in interphase nuclei of peripheral leukocytes is an alternative method to bone marrow karyotyping for monitoring treatment. With automation, several drawbacks of manual analysis may be circumvented. In this article, the capabilities of a commercially available automated image acquisition and analysis system were determined by detecting t(9;22)(q34;q11) in interphase nuclei of peripheral leukocytes. METHODS: Three peripheral blood samples of normal adults, 21 samples of CML patients, and one sample of a t(9;22)(q34;q11) positive cell-line were used. RESULTS: Single nuclei with correctly detected signals amounted to 99.6% of nuclei analyzed after exclusion of overlapping nuclei and nuclei with incorrect signal detection. A cut-off value of 0.84 mum was defined to discriminate between translocation positive and negative nuclei based on the shortest distance between signals. Using this value, the false positive rate of the automated analysis for negative samples was 7.0%, whereas that of the manual analysis was 5.8%. Automated and manual results showed strong correlation (R(2) = 0.985), the mean difference of results was only 3.7%. CONCLUSIONS: A reliable and objective automated analysis of large numbers of cells is possible, avoiding interobserver variability and producing statistically more accurate results than manual evaluation.     

Changing interactions between astrocytes and neurons during CNS maturation

The environments of the developing brain and injured adult brain differ in their abilities to support axonal growth. To determine if astrocytes contribute to this difference, neurons were plated onto astrocytes cultured from the neonatal rat cortex and from the injured adult brain. Two patterns of neurite growth were observed in these two astrocyte culture systems. Neurons contacting the neonatal astrocytes had neurites that were twice as long as those contacting the injured adult astrocytes. Furthermore, in cultures with neonatal astrocytes, neurites faithfully followed the astrocytic processes, maximizing their contact, while in cultures of injured adult astrocytes, the neurites had a tendency to cross the processes orthogonally, minimizing their interaction with the astrocytes. When neurons were grown suspended over either neonatal or injured adult astrocytes, no difference in neurite length or the pattern of neurite growth was observed, indicating that neurite growth was not differentially affected by soluble factors released from the two populations of astrocytes. The addition of fetal calf serum, which is known to contain protease inhibitors, did not alter neurite growth when compared to serum-free medium, suggesting that a substantial difference in protease activity does not account for the variations in neurite length observed. Based on these results, it appears that the molecular components of the external surface of injured adult astrocytes do not support neurite growth to the same extent as those found on neonatal astrocytes. The differing abilities of these two populations of cultured astrocytes to support neurite growth in culture may reflect a change in the functional role of these cells that occurs during the development of the central nervous system.     

The Effect of Including the C2 Insert of Nonmuscle Myosin II-C on Neuritogenesis

The functional role of the C2 insert of nonmuscle myosin II-C (NM II-C) is poorly understood. Here, we report for the first time that the expression of the C2 insert-containing isoform, NM II-C1C2, is inducible in Neuro-2a cells during differentiation both at mRNA and protein levels. Immunoblot and RT-PCR analysis reveal that expression of NM II-C1C2 peaks between days 3 and 6 of differentiation. Localization of NM II-C1C2 in Neuro-2a cells suggests that the C2 insert-containing isoform is localized in the cytosol and along the neurites, specifically at the adherence point to substratum. Inhibition of endogenous NM II-C1C2 using siRNA decreases the neurite length by 43% compared with control cells treated with nonspecific siRNA. Time lapse image analysis reveals that neurites of C2-siRNA-treated cells have a net negative change in neurite length per minute, leading to a reduction of overall neurite length. During neuritogenesis, NM II-C1C2 can interact and colocalize with β1-integrin in neurites. Altogether, these studies indicate that NM II-C1C2 may be involved in stabilizing neurites by maintaining their structure at adhesion sites.     

Tension and compression in the cytoskeleton of PC-12 neurites. II: Quantitative measurements

We assessed the mechanical properties of PC-12 neurites by applying a force with calibrated glass needles and measured resulting changes in neurite length and deflection of the needle. We observed a linear relationship between force and length change that was not affected by multiple distensions and were thus able to determine neurite spring constants and initial, nondistended, rest tensions. 81 out of 82 neurites showed positive rest tensions ranging over three orders of magnitude with most values clustering around 30-40 mu dynes. Treatment with cytochalasin D significantly reduced neurite rest tensions to an average compression equal to 14% of the former tension and spring constants to an average of 17% of resting values. Treatment with nocodazole increased neurite rest tensions to an average of 282% of resting values but produced no change in spring constant. These observations suggest a particular type of complementary force interaction underlying axonal shape; the neurite actin network under tension and neurite microtubules under compression. Thermodynamics suggests that microtubule (MT) assembly may be regulated by changes in compressive load. We tested this effect by releasing neurite attachment to a polylysine-coated surface with polyaspartate, thus shifting external compressive support onto internal elements, and measuring the relative change in MT polymerization using quantitative Western blotting. Neurons grown on polylysine or collagen without further treatment had a 1:2 ratio of soluble to polymerized tubulin. When neurites grown on polylysine were treated with 1% polyaspartate for 15-30 min, 80% of neurites retracted, shifting the soluble: polymerized tubulin ratio to 1:1. Polyaspartate treatment of cells grown on collagen, or grown on polylysine but treated with cytochalasin to reduce tension, caused neither retraction nor a change in the soluble:polymerized tubulin ratio. We suggest that the release of adhesion to the dish shifted the compressive load formerly borne by the dish onto Mts causing their partial depolymerization. Our observations are consistent with the possibility that alterations in MT compression during growth cone advance integrates MT assembly with the advance.     

Morphological Identification and Development of Neurite in Drosophila Ventral Nerve Cord Neuropil

In Drosophila, ventral nerve cord (VNC) occupies most of the larval central nervous system (CNS). However, there is little literature elaborating upon the specific types and growth of neurites as defined by their structural appearance in Drosophila larval VNC neuropil. Here we report the ultrastructural development of different types VNC neurites in ten selected time points in embryonic and larval stages utilizing transmission electron microscopy. There are four types of axonal neurites as classified by the type of vesicular content: clear vesicle (CV) neurites have clear vesicles and some T-bar structures; Dense-core vesicle (DV) neurites have dense-core vesicles and without T-bar structures; Mixed vesicle (MV) neurites have mixed vesicles and some T-bar structures; Large vesicle (LV) neurites are dominated by large, translucent spherical vesicles but rarely display T-bar structures. We found dramatic remodeling in CV neurites which can be divided into five developmental phases. The neurite is vacuolated in primary (P) phase, they have mitochondria, microtubules or big dark vesicles in the second (S) phase, and they contain immature synaptic features in the third (T) phase. The subsequent bifurcate (B) phase appears to undergo major remodeling with the appearance of the bifurcation or dendritic growth. In the final mature (M) phase, high density of commensurate synaptic vesicles are distributed around T-bar structures. There are four kinds of morphological elaboration of the CV_I neurite sub-types. First, new neurite produces at the end of axon. Second, new neurite bubbles along the axon. Third, the preexisting neurite buds and develops into several neurites. The last, the bundled axons form irregularly shape neurites. Most CV_I neurites in M phase have about 1.5–3 µm diameter, they could be suitable to analyze their morphology and subcellular localization of specific proteins by light microscopy, and they could serve as a potential model in CNS in vivo development.     

Neurite development in vitro. I. The effects of adenosine 3′5′-cyclic monophosphate (cyclic AMP)

Elevated levels of 3′5′ adenosine monophosphate (cyclic AMP) stimulate a wide variety of cellular events including aggregation, differentiation, morphological expression, pigment migration, and secretion. The role of cyclic AMP in these events prompted our present study of embryonic chick dorsal root ganglia. Test substances were applied to cultures during the routine feeding procedure. Their development was quantitatively evaluated on the basis of explant size, length of glial-like outgrowth, distribution of growth, neurite number, length, diameter, and degree of arborization. These parameters were all shown to be independent of each other. The high variability of in vitro neurite development necessitated the use of over 100 cultures per treatment group. Cultures treated with 5′ AMP exhibited no significant differences from controls. Those treated with cyclic AMP, dibutyryl cyclic AMP, or Nerve Growth Factor (NGF) exhibited statistically significant increases in area of outgrowth, the number of neurites per culture, and in diameters, lengths, and degree of neurite arborization. The growth promoting activity of dibutyryl cyclic AMP and NGF were greater than those of cyclic AMP. Electron microscopic study shows neurites formed under the influence of cyclic AMP or its dibutyryl derivative to resemble those grown in NGF. These studies suggest the possibility that cyclic AMP stimulates neurite growth by mediating the process of microtubule (MT) assembly. They further prompt us to speculate that one way NGF enhances neurite development is by stimulating MT assembly via a “Second Messenger System”.     

Neuronal differentiation of NTE-deficient embryonic stem cells

Organophosphates induce neurological disorders. One of the enzymes inhibited by these compounds is neuropathy target esterase (NTE). In vitro, inhibition of NTE activity by organophosphates is correlated with inhibition of neurite initiation and reduction of neurite length, supporting the hypothesis that organophosphate-induced neurological disorders are caused by inhibition of NTE activity. However, there is no direct evidence for the involvement of NTE in organophosphate-induced impairment of neurites in vitro. To examine the role of NTE, we have generated NTE-deficient mouse embryonic stem cells. These cells can differentiate into neuron-like cells. Although NTE-deficient cells exhibited a delay in neurite initiation in vitro, both the proportion of neuron-like cells which initiated neurites and the elongation of these neurites occurred at the normal rate. These results demonstrate that NTE activity is not required for neurite initiation or elongation per se, but is essential for the optimal rate of neurite initiation.     

Steps in the formation of a bipolar outgrowth pattern by cultured neurons,and their substrate dependence

We studied the steps in the formation of the bipolar outgrowth pattern of cultured adult Anterior Pagoda (AP) neurons of the leech growing on a central nervous system (CNS) homogenate as substrate. This pattern, which consists of two primary neurites directed in opposite directions plus some bifurcations, resembles their embryonic pattern but is different from the patterns they develop in culture on leech laminin or Concanavalin A as substrates. In eight neurons that were studied, one primary neurite formed and branched several hours before the second one. Time‐lapse video analysis showed that between 12 and 36 h of growth, the more proximal branch of the early neurite migrated retrogradely, rotated, and formed the second primary branch. Both neurites elongated until the total neurite length reached 130–160 μm, when the elongation of primary neurites became synchronous with the retraction of secondary processes, suggesting competition. The substrate dependence of these events was tested by plating AP neurons on leech laminin. On this substrate AP neurons produced multiple independent primary neurites with branches. Retraction of some large branches was followed by their regrowth, and did not correlate with the changes in other neurites. We propose that the dynamics in the formation of the bipolar outgrowth pattern of AP neurons arise from inhibitory extracellular matrix molecules, which reduce the synthesis of precursors for neurite formation. © 2002 Wiley Periodicals, Inc. J Neurobiol 50: 106–117, 2002; DOI 10.1002/neu.10017     

Evidence that Protein Kinase C Activities Involved in Regulating Neurite Growth Are Localized to Distal Neurites

Abstract: Previously, we observed that long-term treatment of distal nerve fibers of rat sympathetic neurons in compartmented cultures with phorbol 12-myristate 13-acetate (PMA) caused a reduction in the rate of neurite elongation by >50%. In the present report we show that protein kinase C (PKC) activity could be measured in extracts of distal neurites by an assay of the Ca²⁺-dependent phosphorylation of a PKC-specific octapeptide substrate. We found that local application of 1 µ M PMA for 24 h to distal neurites caused nearly complete down-regulation of Ca²⁺-dependent PKC activity measured in this manner. We determined that the inhibition of neurite elongation by PMA was mediated by local mechanisms in the neurites because local application of PMA to center compartments containing cell bodies and proximal neurites did not inhibit the rate of elongation of distal neurites. We then investigated the effects of the recently available PKC inhibitors, calphostin C and chelerythrine, finding that, like PMA, these inhibited the growth of distal neurites when applied locally to them, and had no effect when applied to cell bodies and proximal neurites. However, the inhibition of neurite growth by calphostin C occurred at a concentration far below its IC₅₀ value for protein kinase inhibition, and both calphostin C and chelerythrine inhibited distal neurite growth even in neurons pretreated with PMA. Thus, it appears that these agents do not all inhibit neurite growth through the same mechanisms. Although the PKC activities involved in neurite elongation in sympathetic neurons have not been precisely defined, these data presented in this study indicate that protein kinases localized to growth cones play a complex and important role in regulating axonal growth.     

Steps in the formation of a bipolar outgrowth pattern by cultured neurons, and their substrate dependence.

We studied the steps in the formation of the bipolar outgrowth pattern of cultured adult Anterior Pagoda (AP) neurons of the leech growing on a central nervous system (CNS) homogenate as substrate. This pattern, which consists of two primary neurites directed in opposite directions plus some bifurcations, resembles their embryonic pattern but is different from the patterns they develop in culture on leech laminin or Concanavalin A as substrates. In eight neurons that were studied, one primary neurite formed and branched several hours before the second one. Time-lapse video analysis showed that between 12 and 36 h of growth, the more proximal branch of the early neurite migrated retrogradely, rotated, and formed the second primary branch. Both neurites elongated until the total neurite length reached 130-160 microm, when the elongation of primary neurites became synchronous with the retraction of secondary processes, suggesting competition. The substrate dependence of these events was tested by plating AP neurons on leech laminin. On this substrate AP neurons produced multiple independent primary neurites with branches. Retraction of some large branches was followed by their regrowth, and did not correlate with the changes in other neurites. We propose that the dynamics in the formation of the bipolar outgrowth pattern of AP neurons arise from inhibitory extracellular matrix molecules, which reduce the synthesis of precursors for neurite formation.     

Molecular weight characterization of high molecular weight dextran with multiangle light scattering in on‐line and off‐line mode

This work reports the molecular weight (MW) analysis of high MW dextran using multiangle light scattering (MALS) in both chromatography and automated batch measurement mode. The results show that the chromatographic columns alter the high MW native dextran and cause underestimation of the MW as a consequence. Alternatively, a batch MALS measurement (without columns) provides more accurate MW values. The batch MALS measurement was automated with the incorporation of an automatic sample dilution and injection device. This automation reduces the sample preparation time and minimizes concentration errors introduced by manual sample dilution. To the best of our knowledge, this is the first study using an automated batch MALS in the analysis of high MW dextran. © 2015 Wiley Periodicals, Inc. Biopolymers 103: 387–392, 2015.     

Influence of pulsed electromagnetic field with different pulse duty cycles on neurite outgrowth in PC12 rat pheochromocytoma cells

The influence of low frequency (50 Hz repetition rate) pulsed electromagnetic field (EMF) on PC12 cell neurite outgrowth in vitro was investigated in this study. We studied the percentage of neurite bearing cells, average length of neurites, and directivity of neurite outgrowth in PC12 cells cultured for 96 h in the presence of nerve growth factor (NGF). PC12 cells were exposed in one incubator to pulsed EMF at 1.36 mT (peak value) generated by a pair of Helmholtz coils, and the control samples were placed in another identical incubator. We found that the pulse duty cycle had significant effect on neurite outgrowth. Low (10%) pulse on-time significantly inhibited the percentage of neurite bearing cells, but at the same time increased the average length of neurites, while 100% on-time (DC) had exactly the opposite effects. Furthermore, we found that neurites were prone to extend along the direction of pulsed EMF with 10% pulse on-time. Our studies show that neurite outgrowth in PC12 cells is sensitive to the pulse duty and this sensitivity was associated with NGF concentration.