May be you can read this paper I publish many years ago..... Below the abstract, and attached the pdf.

Good luck

Jean

Bulletin of Mathematical Biology (2000) 62, 657-674 Statistical analysis and parsimonious modelling of dendrograms of in vitro neurones J.M. Devaud 1*, B. Quenet 2*, J. Gascuel 3 and C. Masson 3 1 Instituto Cajal (CSIC) Avenida Doctor Arce, 37 28002 Madrid – Spain 2 Laboratoire d’Electronique (ESPCI) 10 Rue Vauquelin, 75005 Paris – France 3 Centre d’Etude des Sciences du Goût (CNRS) 15 Rue Hugues Picardet, 21000 Dijon - France * These authors contributed equally to this work. ABSTRACT The processes whereby developing neurones acquire morphological features that are common to entire populations (thereby allowing the definition of neuronal types), are still poorly understood. A mathematical model of neuronal arborisations may be useful to extract basic parameters or organisation rules, hence helping to achieve a better understanding of the underlying growth processes. We present a parsimonious statistical model, intended to describe the topological organisation of neuritic arborisations with a minimal number of parameters. It is based on a probability of splitting which depends only on the centrifugal order of segments. We compare the predictions made by the model of several topological properties of neurones with the corresponding actual values measured on a sample of honeybee (olfactory) antennal lobe neurones grown in primary culture, described in a previous study. The comparison is performed for three populations of segments corresponding to three neuronal morphological types previously identified and described in this sample. We show that simple assumptions together with the knowledge of a very small number of parameters, allow the topological reconstruction of representative (bi- dimensional) biological neurones. We discuss the biological significance (in terms of possible factors involved in the determinism of neuronal types) of both common properties and cell-type specific features, observed on the neurones and predicted by the model.

Also for acuisition you can use fiji (a Image J plugin). However I would be please to know if you use another software may be more convenient (powerfull?)

Hi Marion,

The type of analysis of cell anatomy is usually dependent on your question. What do you want to know? For example, do you simply want to compare the lengths of dendrites? If so, a Sholl's analysis might work best, especially on isolated neurons.

Can you be more specific about your scientific/experimental question?

Andrew

@Jean Gascuel: Thank you very much for the paper. I am not yet sure, if I can use, but the abstract sounds definately very interesting.

I also did not use the fiji plugin yet and will try it out. However, all automized plugins or programs I have used so far caused either more or the same amount of work or simply did not work well.

@Andrew Tan: I want to compare the lengths and ideally the amount of branching of neurons in a primary cell culture. Thus, its (unfortunately) not one specific cell type and in contrast to in vivo (hippocampus slices, which I also use), neurons in vitro are not growing linear, but a bit like the line in the 'at' symbole (@) Therefore the sholl analysis is biased.

Maybe this paper is helpful.

Front Neuroanat. 2011 Feb 17;5:5.

Morphological development of thick-tufted layer v pyramidal cells in the rat somatosensory cortex.

Romand S, Wang Y, Toledo-Rodriguez M, Markram H.

Source

Blue Brain Project, École Polytechnique Fédérale de Lausanne Lausanne, Switzerland.

With Fiji you can using the menue plugin/ segmentation/ simple neurite tracer, mesure highly complex arborisation, then get the following data of tolal length, nb of branches etc... in excel file. You can also get a 3d representation and make movies (this may be without any interest if you work on cell cultures. You can also make direct scholl analysis.

Best

Jean

Neurite Tracer plugin (Image J) might be of help. I have been using it to investigate the effects of growth factors on neurite growth in cortical neurons. For further information you may check: http://fournierlab.mcgill.ca/neuritetracer.html.

Hi Marlon, I have found some goodness in this guys proposition: "One rule to grow them all: a general theory of neuronal branching and its practical application" by Cuntz H, Forstner F, Borst A, Häusser M. (PLoS Comput Biol. 2010 Aug 5;6(8). pii: e1000877.) Problem: Matlab skills are demanded.

Dear Marion, my suggestion considers specific antibodies to probe young cells and growth mechanisms followed by a stereology analysis software such as Zeiss microscopy section developed. Kind regards, cris

Cristofer,

Could you precise which zeiss software you are talking about?

JEan

To provide you with ideas you have to tell us

1. What method do you use to visualize your dendrites - DIC, phase, dark-field, immunohistochemistry...

2. It would also be helpful toknow what the results of your statistics are. The questions you asked suggest that you haven't obtained a significant result. If so, looking for a different statistic probably will not help.

3. I assume (naively) that if the total area occupied by dendrites does not differ between the different conditions, then you may be dealing with a very subtle effect or no effect.

4. You don't have to reveal your independent variables. However, you tell us whether your paradigm is a between or within subjects design

Dear Jean, I´ve used for many years the software of universal imaging under metamorph microscope analysis, but now I've changed to a Zeiss package because it's more use friendly, particularly to pre-selected cells in order to " teach" the software regarding count parameters according to a specific morphology. Very useful!

Stereology has some good methods to quantify almost everything. Some time ago I was helping a friend that needed to measure bone branching during healing after a fracture in rat and I found some good papers in stereology field. See the reference below. Maybe it can give you some insights about quantification.

http://www.sciencedirect.com/science/article/pii/S0968432805001113

Good luck!

Thank you all for the suggestions. I am currently out of office,

but will have a closer look at those as soon as I get back on monday.

Again thanks!

To suggest a method for evaluation, it would be helpfull to know how the neurons are stained (e.g. filled with tracer, GFP-flourescent, Golgi-impregnated or treated with antibodies for visualization). To get very detailled data, I would suggest to use a reconstruction software like NeuroLucida (MBF) or comparable sotware (or ImgaeJ plugins). Indded, the neurons display different morphologies in-vivo or in cell culture.

Thus, using cell culture, only the data from different parameters obtained (either Sholl analysis, branching pattern, dendritic spines) can be compared with neurons from un-treated cell cultures, but not with data from in-vivo neurons.

@Yun Wang: If I am not mistaken you used 3D recontructions in your analysis. I am working with 2D pictures, mainly due to practical reasons. But thank you very much for the interesting paper.

@Jean Gascuel: I read your paper, but I am not sure whether I can use it in my case. I also tried fiji, but despite the fact that I already did a sholl analysis of all cells, the program points out the weaknesses of 2D analysis. Quite often neurites overlap so its hard to impossible to differentiate between the two. Thus I am using rather rough measurements such as the area or robust (i. e. the total length of all neurites).

@Andrew Tan: I just want to know whether the dendritic tree differs depending on a condition (in this case a protein knock out). I did use sholl analysis, but this method has its short coming in vitro, since the dendritic tree grows 'in a circle', meaning that the number of intersections is biased.

@Amalia Dolga: I tried neurite tracer already before. I have the same problems as described above concerning overlapping neurites. Thus, I am sticking to manual analysis. But thank you!

@Rodrigo Herrera-Molina: I scaned through the paper. It is definately an interesting approach. I haven't yet made my mind about whether I can use it.

@Cristofer Caous: You have a point, but I am done with the experiments and so can only optimize the analysis at this point. And my name is Marlon not Marion. Thank you.

@Caroline Fernandes-Santos: I must admit I never heard before of stereology, but as far as I understood its ment for 3D analysis and I work with 2D pictures. However, thank you very much for introducing me to this interesting field.

@Alan Springer: 1. I transfected my cells with EGFP and used an monoclonal antibody to enhcae the signal in the flourescence microscope.

2. and 3. I do have significant results. So I find significant differences, especially in the more sensitive total length analysis. My main reason to ask here was to collect ideas for further analysis criterias and statistical methods. 4. I analysed cells from different animals, grouped together based on the genotype and comapred also different preparations/animals with each other.

@Oliver von Bohlen und Halbach: As I described above due to overlapping neurites and very detailed analysis is not possible.

I want to thank all suggestions once more. It is a great help to reflect my steps so far and get new input.

Hi Marlon! Actually stereology works with 2-D images, and it makes assumptions regarding the spatial distribution in 3-D :)

Thanks Caroline for the clarification. I will have to do more reading about this as it seems. :)

No problem! If you are interested about using it in the future I can recommend you some good readings. Good luck on your experiment!

Actually I am interested, cause I try to figuring out how I can use it for neurons and what I found so far are just systems to purchase. So what are the prerequisits to use it?

There are some softwares for Stereology, but you do not really need of them. In general, you just need to open your image on a monitor screen and count! What you are going to count depends on your goal. For instance, I can estimate the volume of fat inside the whole liver (volume density) just by counting the number of points (the test-system is an image printed in a transparent sheet that you overlap over the image. In this case it might have about 36 points) that hits the fat droplets in an liver section. Maybe you can send me a image that you have and then I see what would fits to your goal.

My email: [email protected]

Do not know the statistic in this area

Pick up Peter Mouton's book, "Principles and Practices of Unbiased Stereology" and you should be set. Microbrightfield makes a very comprehensive stereology package. For your 2D analyses it seems that you would only need to look at one of the "line crossing" algorithms that Mouton describes to get measures of dendritic spread/extent.

@Christopher Wilson: Thanks for the recommendation.

If a filament tracing is not working, then maybe better to look outside of the "box" or rather inside of the pixel or voxel box. If you divide the space into small enough arbitrary boxes by using some 2d or 3d software with a grid tool then convert into binary image and then measure how much space (pixels or voxels) is taken by neuronal processes per grid square. Stereology principles from Gunderson etc are not needed here, it is not a tissue. I would still use radial statistics using the cell body and zero point to show the distribution, and of course, unbiased sampling.