DAB quantification

[Panayiota Ploutarchou]

Hi everyone,

Would love your feedback on an issue I'm having.

So, I've done Immunohistochemistry on tissue sections, and finished with DAB staining (which basically gives brown staining at the locations where a reaction with the antibody took place). Now, I would like to quantify that, but the pixel intensity function of Image J seems to give weird results. So basically firstly I would like to know what this function actually does (i.e. does it measure dark pixels only? what about taking into account the shades of colour?) and also how would you suggest I go about quantifying the DAB staining as I would need to compare staining in control and treated sections.

Many thanks for your help,
Yiota

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[hesham977]

very interesting questions
please we in-need the reply in steps
we need to follow steps from the start of counting the positive cells in
the immunohistochemistry slides until the full results
Thanks in advance


On Fri, Jan 17, 2014 at 6:42 PM, Panayiota Ploutarchou <
[hidden email]> wrote:

> Hi everyone,
>
> Would love your feedback on an issue I'm having.
>
> So, I've done Immunohistochemistry on tissue sections, and finished with
> DAB staining (which basically gives brown staining at the locations where a
> reaction with the antibody took place). Now, I would like to quantify that,
> but the pixel intensity function of Image J seems to give weird results. So
> basically firstly I would like to know what this function actually does
> (i.e. does it measure dark pixels only? what about taking into account the
> shades of colour?) and also how would you suggest I go about quantifying
> the DAB staining as I would need to compare staining in control and treated
> sections.
>
> Many thanks for your help,
> Yiota
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--













*Dr. Hesham Noaman Abdelraheem MustafaAssistant Professor,Faculty of
Medicine, King Abdulaziz University,e-Library Consultant,Student Support
Unit Member,Nutrition & Metabolism Module Coordinator.Address: P.O.Box
80205               Jeddah, 21589               Saudi
Arabia.Phone:           (Office):  [phone] Ext. 22016
(Mobile): [phone]           (Egypt):  [phone]*

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Gabriel Landini]

On Friday 17 Jan 2014 15:42:22 Panayiota Ploutarchou wrote:
> So, I've done Immunohistochemistry on tissue sections, and finished with DAB
> staining (which basically gives brown staining at the locations where a
> reaction with the antibody took place). Now, I would like to quantify that,
> but the pixel intensity function of Image J seems to give weird results

Unfortunately IHC is not stoichiometric and so you won't get a reliable
measure of how much product you have. Most stains are not "quantitative" (with
some exceptions like Feulgen and phalloidin for example).

>  and also how would you suggest I go about quantifying
> the DAB staining as I would need to compare staining in control and treated
> sections.

You can't, immunostains are not a quantitative technique. What you see as dark
has no accurate relation to the amount of antigen present, which is what the
purpose of the quantification is.

Cheers

Gabriel

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[G. Esteban Fernandez]

Are your slides also stained with a counterstain like hematoxylin or only
with DAB?

-Esteban
On Jan 17, 2014 7:54 AM, "Panayiota Ploutarchou" <
[hidden email]> wrote:

> Hi everyone,
>
> Would love your feedback on an issue I'm having.
>
> So, I've done Immunohistochemistry on tissue sections, and finished with
> DAB staining (which basically gives brown staining at the locations where a
> reaction with the antibody took place). Now, I would like to quantify that,
> but the pixel intensity function of Image J seems to give weird results. So
> basically firstly I would like to know what this function actually does
> (i.e. does it measure dark pixels only? what about taking into account the
> shades of colour?) and also how would you suggest I go about quantifying
> the DAB staining as I would need to compare staining in control and treated
> sections.
>
> Many thanks for your help,
> Yiota
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Panayiota Ploutarchou]

Hi Esteban,

Just DAB!

Yiota
________________________________
From: G. Esteban Fernandez [[hidden email]]
Sent: Friday, January 17, 2014 4:19 PM
To: [hidden email]; [hidden email]; Panayiota Ploutarchou
Subject: Re: DAB quantification


Are your slides also stained with a counterstain like hematoxylin or only with DAB?

-Esteban

On Jan 17, 2014 7:54 AM, "Panayiota Ploutarchou" <[hidden email]<mailto:[hidden email]>> wrote:
Hi everyone,

Would love your feedback on an issue I'm having.

So, I've done Immunohistochemistry on tissue sections, and finished with DAB staining (which basically gives brown staining at the locations where a reaction with the antibody took place). Now, I would like to quantify that, but the pixel intensity function of Image J seems to give weird results. So basically firstly I would like to know what this function actually does (i.e. does it measure dark pixels only? what about taking into account the shades of colour?) and also how would you suggest I go about quantifying the DAB staining as I would need to compare staining in control and treated sections.

Many thanks for your help,
Yiota

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Panayiota Ploutarchou]

Hi Gabriel,

Thanks for your answer. I agree that I can't quantify, but doesn't darker (and more) DAB staining correlate to more antigen present? I was just wondering whether there is some sort of semi-quantitative way to go about doing this...

Many thanks for your help,
Yiota
________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Gabriel Landini [[hidden email]]
Sent: Friday, January 17, 2014 4:19 PM
To: [hidden email]
Subject: Re: DAB quantification

On Friday 17 Jan 2014 15:42:22 Panayiota Ploutarchou wrote:
> So, I've done Immunohistochemistry on tissue sections, and finished with DAB
> staining (which basically gives brown staining at the locations where a
> reaction with the antibody took place). Now, I would like to quantify that,
> but the pixel intensity function of Image J seems to give weird results

Unfortunately IHC is not stoichiometric and so you won't get a reliable
measure of how much product you have. Most stains are not "quantitative" (with
some exceptions like Feulgen and phalloidin for example).

>  and also how would you suggest I go about quantifying
> the DAB staining as I would need to compare staining in control and treated
> sections.

You can't, immunostains are not a quantitative technique. What you see as dark
has no accurate relation to the amount of antigen present, which is what the
purpose of the quantification is.

Cheers

Gabriel

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Gilbert Bigras]

Hello Yiota


There is indeed proportionality between the intensity ant the amount of antigen but the proportionality is unpredictable for the following reasons:

1) as Gabriel wrote, most stains are not stoichiometric and the main reason I believe is the more or less random (or not controlled) amplification methods used by detection kit used in IHC lab (polymer or multimer attached with multiple enzymes like Horseradish Peroxidase for DAB). So the proportion of one antigenic site and the number of one (chomogens) signals is unknown


2) If you deal with a stoichiometric stain, the relationship between the amount of antigen and the intensity is not linear... but logarithmic. (See Beer-Lambert law and related integrated optical density IOD)

3) finally DAB is the worst stain for quantification: it violates the Beer-Lambert law: in addition to absorb light like others stains, when abundantly present there are clots of the chomogen which diffract light which therefore provides a stronger intensity than explained by light absorption for a given amount of antigen or substrate.

 
Best


Gilbert Bigras, M.D., Ph.D., FRCPC (Path)


Cross Cancer Institute
11560 University Avenue
Edmonton, Alberta
Canada T6G 1Z2
 
Associate Clinical Professor
Department of Laboratory Medicine & Pathology
University of Alberta, Canada



________________________________
 From: Panayiota Ploutarchou <[hidden email]>
To: [hidden email]
Sent: Friday, January 17, 2014 9:24 AM
Subject: Re: DAB quantification
 

Hi Gabriel,

Thanks for your answer. I agree that I can't quantify, but doesn't darker (and more) DAB staining correlate to more antigen present? I was just wondering whether there is some sort of semi-quantitative way to go about doing this...

Many thanks for your help,
Yiota
________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Gabriel Landini [[hidden email]]
Sent: Friday, January 17, 2014 4:19 PM
To: [hidden email]
Subject: Re: DAB quantification

On Friday 17 Jan 2014 15:42:22 Panayiota Ploutarchou wrote:
> So, I've done Immunohistochemistry on tissue sections, and finished with DAB
> staining (which basically gives brown staining at the locations where a
> reaction with the antibody took place). Now, I would like to quantify that,
> but the pixel intensity function of Image J seems to give weird results

Unfortunately IHC is not stoichiometric and so you won't get a reliable
measure of how much product you have. Most stains are not "quantitative" (with
some exceptions like Feulgen and phalloidin for example).

>  and also how would you suggest I go about quantifying
> the DAB staining as I would need to compare staining in control and treated
> sections.

You can't, immunostains are not a quantitative technique. What you see as dark
has no accurate relation to the amount of antigen present, which is what the
purpose of the quantification is.

Cheers

Gabriel

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html


--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Gabriel Landini]

On Friday 17 Jan 2014 16:24:44 Panayiota Ploutarchou wrote:
> Thanks for your answer. I agree that I can't quantify, but doesn't darker
> (and more) DAB staining correlate to more antigen present? I was just
> wondering whether there is some sort of semi-quantitative way to go about
> doing this...

No, that is what "non-stoichiometric" means. It is worse than that because DAB
which is used to visualise the result, scatters light rather than transmitting
it and so Beer-Lambert law (to compute the amount of dye based on the optical
density) does not follow well either (so you have 2 problems).

Search the list archives. This has been discussed lots of times. An
interesting post is this one:
https://list.nih.gov/cgi-bin/wa.exe?A2=ind0902&L=IMAGEJ&P=R18412

Possible solution: describe what you find with IHC (which cells appear stained
or how many appear stained at a particular antibody dilution) but do not be
tempted to put a numeric quantity to the *intensity* of the stain. Explain why
quantification of DAB intensity is not a good idea and try to find a different
way to quantify what you are after.
 
That being said, IHC is very useful and has helped immensely in pathology with
tumour typing and clarifying difficult diagnoses, etc. but putting numbers to
the intensity of the stain is stretching the technique too far.

Hope this helps and avoids future headaches.

Gabriel

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Michelle Naugle-2]

Hi There,
If you want to quantify with DAB you can only count the # of cells that are
stained or not (use a nissl counterstain to coun unlabeled cells). You can
use stereology to do this quantification in an unbiased way. For more info
on this method look up:
Uylings HB: Morphometry of size/volume variables and comparison of their
bivariate relations in the nervous system under different conditions. J
Neuroscience Methods 1986
Gundersen HJGJ, E.B. : The efficiency of systematic sampling in stereology
and its prediction. J. Microsc 1987
West MJ, : Unbiased stereological estimation of the total number of neurons
in the subdivisions of the rat hippocampus using the optical fractionator.
Anat Rec 1991
Glaser EM: The coefficient of error of optical fractionator population size
estimates: a computer simulation comparing three estimators. J Microscopy
1998


On Fri, Jan 17, 2014 at 10:51 AM, Gilbert Bigras
<[hidden email]>wrote:

> Hello Yiota
>
>
> There is indeed proportionality between the intensity ant the amount of
> antigen but the proportionality is unpredictable for the following reasons:
>
> 1) as Gabriel wrote, most stains are not stoichiometric and the main
> reason I believe is the more or less random (or not controlled)
> amplification methods used by detection kit used in IHC lab (polymer or
> multimer attached with multiple enzymes like Horseradish Peroxidase for
> DAB). So the proportion of one antigenic site and the number of one
> (chomogens) signals is unknown
>
>
> 2) If you deal with a stoichiometric stain, the relationship between the
> amount of antigen and the intensity is not linear... but logarithmic. (See
> Beer-Lambert law and related integrated optical density IOD)
>
> 3) finally DAB is the worst stain for quantification: it violates the
> Beer-Lambert law: in addition to absorb light like others stains, when
> abundantly present there are clots of the chomogen which diffract light
> which therefore provides a stronger intensity than explained by light
> absorption for a given amount of antigen or substrate.
>
>
> Best
>
>
> Gilbert Bigras, M.D., Ph.D., FRCPC (Path)
>
>
> Cross Cancer Institute
> 11560 University Avenue
> Edmonton, Alberta
> Canada T6G 1Z2
>
> Associate Clinical Professor
> Department of Laboratory Medicine & Pathology
> University of Alberta, Canada
>
>
>
> ________________________________
>  From: Panayiota Ploutarchou <[hidden email]>
> To: [hidden email]
> Sent: Friday, January 17, 2014 9:24 AM
> Subject: Re: DAB quantification
>
>
> Hi Gabriel,
>
> Thanks for your answer. I agree that I can't quantify, but doesn't darker
> (and more) DAB staining correlate to more antigen present? I was just
> wondering whether there is some sort of semi-quantitative way to go about
> doing this...
>
> Many thanks for your help,
> Yiota
> ________________________________________
> From: ImageJ Interest Group [[hidden email]] on behalf of Gabriel
> Landini [[hidden email]]
> Sent: Friday, January 17, 2014 4:19 PM
> To: [hidden email]
> Subject: Re: DAB quantification
>
> On Friday 17 Jan 2014 15:42:22 Panayiota Ploutarchou wrote:
> > So, I've done Immunohistochemistry on tissue sections, and finished with
> DAB
> > staining (which basically gives brown staining at the locations where a
> > reaction with the antibody took place). Now, I would like to quantify
> that,
> > but the pixel intensity function of Image J seems to give weird results
>
> Unfortunately IHC is not stoichiometric and so you won't get a reliable
> measure of how much product you have. Most stains are not "quantitative"
> (with
> some exceptions like Feulgen and phalloidin for example).
>
> >  and also how would you suggest I go about quantifying
> > the DAB staining as I would need to compare staining in control and
> treated
> > sections.
>
> You can't, immunostains are not a quantitative technique. What you see as
> dark
> has no accurate relation to the amount of antigen present, which is what
> the
> purpose of the quantification is.
>
> Cheers
>
> Gabriel
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
Michelle M. Naugle

Neuroscience PhD Candidate
College of Pharmacy
The University of Texas at Austin
Telephone: [phone]
Fax: [phone]
Email: [hidden email]

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[sarahmmph]

Hi everyone,

i have the same problem. i have samples stained with H\DAB and need to compare the results of treated and control samples. so i have 2 questions:

1- can i use threshold to determine brown area stained with DAB, then compare them?

2- which stain can be used (to avoid the DAB issue with quantification) to quantify a marker in treated and untreated samples?

Sarah M. Mosaad, M.Sc