Determining the RT concentration in HIV-1 samples through Sandwich ELISA - Not seeing any infection/RT activity/indication of virus (Oct/01/2007 )
Hi everyone!
I'm doing (or trying to do) quantification of HIV-1 in infected cell samples. The problem is that I'm not seeing any indication that the cells are actually infected by the virus...
I've tried everything and anything I could think of, including using different cells, different virusses at different concentrations, changing the protocol parameters (media, incubation temp, time, etc). I've even tried different ways of enriching the virus - ultra-centrifugation @ 39 000 x g and PEG precipitation.... I'm probably doing something obscenely fundamental wrong, BUT WHAT?!
I've been doing a standard infection as per the NIH protocol and using the "Reverse Transcriptase Assay, colorimetric" ELISA kit from Roche. So it should work..!
If anyone has any suggestions or a shotgun, please respond... 
What cells are you infecting? How do you quantify your "concentration of virus" (RT-activity, p24, TCID50,...)
Tell us your exact protocol please, because it's hard to tell what's going wrong.
One thing is to make sure your cells are in exponential phase when infecting them, or that they are seeded not too long before (if using stably transfected adherent cells), and if you are using suspension cells: to split them regularly after infection so that the cells are nicely growing (this allows for more viral production/allows better infection per cell).
I'm doing acute infection of CEM.SS suspension cells with HIV-1 subtype C, and using the RT activity as a measure of the virus concentration (RT kit from Roche).
My method (and parameters I've checked):
- Plate CEM.SS cells (PBMCs) in logphase growth (P3-P6) at 10 000 cells per well, using heat-inactivated (non-inactivated) 10% RPMI (0.25% Strep/Pen)
- Prepare a 1 x 10^3 TCID50 virus working dilution from a 1 x 10^6 TCID50 HIV-1 B (HIV-1 C) virus stock (1 ul in 1000 ul media)
- Add 100 ul virus working dilution to each well to obtain an MOI of 0.01 (MOI = 0.001, 0.002, 0.005)
- Incubate the plate at 37°C (4°C, 25°C) for 2 hours (1 hr, 1.5 hrs, 4 hrs, 24 hrs)
- Wash the cells with fresh media to remove unbound virus (no washing step)
- Incubate supposedly infected cells @ 37°C for 10 days, taking 100 ul aliquots of the supernatant at days 3, 6, 8, and 10 (24 hrs, 2 days), and replacing it with 100 ul fresh media.
- Test viability of cells via MTT (MTS) - cells are always >90% viable
- Centrifuge the supernatant for 30 minutes @ 2000 x g and 4°C to remove cellular debris
- Enrich viral particles by ultracentrifugation @ 40 000 x g and 4°C for 1.5 hrs (overnight PEG precipitation)
Use the RT kit for the following steps: - Add lysis buffer (provided in RT kit) to the pellet to lyse the viral particles and incubate @ rt for 30 minutes (1 hr)
- Make up RT standards (provided in kit)
- Add the template/primer hybrid and labelled nucleotides (provided) to the samples and standards and incubate @ 37°C for 1 hr (4 hrs, 15 hrs)
- Transfer the nucleotide reaction mixture to the MP modules (provided) and incubate for 1 hr (specified)
- Remove the solution and wash x 5 with 250 ul washing buffer (30 seconds per wash)
- Add 200 ul freshly made up anti-DIG-POD solution (provided) per well and incubate @ 37°C for 1 hr (specified)
- Remove the solution and wash x 5 with 250 ul washing buffer (30 seconds per wash)
- Add 200 ul freshly made up substrate solution (reagents provided) per well and incubate @ rt for 15 minutes or until colour change
- Measure absorbance @ A405/A490 nm and determine the RT concentration in the sample wells from the standard curve equation.
No matter how I combine the different test parameters, the result is always the same - the standards turn a beautiful green colour and the standard curve is basically perfect, but the samples show absolutely no RT activity (no colour change at all). This makes me think that the problem is not with the kit, but with the virus or the virus preparation prior to the RT reaction (where we start using the kit).
As you'll see, one of the things I tried was to leave the initial washing step out completely. So, even if there's no infection, the virus should still be present in the supernatant and show up during the RT determination. Still nothing...
Protocol itself seems OK
Haven't worked with CEM.SS cells, do they have CCR5-expression? Subtype C virusses more often use CCR5 exclusively (compared to subtype B virusses). What cells was your titer determined on? (since you know your TCID50)? Certain cells are less efficiently infected than others, so MOI 0,01 can be MOI 0,0001 for the same stock on a different cell.
Do you have a subtype B control?
Is your virus a laboratory adapted strain or a clinical isolate (clinical isolates are harder to grow).
Do you have any other HIV-suscptible cell line to check your virus stock (MT-4, C8166, CEMx174, U87's,..)?
If you don't wash your cells, after 3 days you shouldn't have any RT-activity left as your virus doesn't survive eternally and the RT itself probably (haven't tested myself) isn't too stable in your medium. Have you checked RT of your virus stock itstelf?
Thanks for your quick response vairus!
The CEM.SS cells definitely express CCR5 and CXCR4...
We got the cells from another lab... The only thing we know for certain is that the TCID50 is 1 x 10^6, and we're pretty sure that this was determined in MT-2 cells... We don't have any other appropriate cell lines at the moment to test the TCID50 ourselves - sounds like that should've been our first step 
Could we use syncitia-forming cells to determine the TCID50..?
Is there some way to correlate the infectivity in MT-2 cells to that in CEM.SS cells..?
All of the virusses I've been using are clinical isolates...
I'll try your suggestion of determining the RT activity of the stock and get back to you
THANKS!!!!
Is there some way to correlate the infectivity in MT-2 cells to that in CEM.SS cells..?
Yes, you can determine TCID50 by checking for syncitia formation (or RT-activity, of p24-elisa), but you're first problem seems to be that you're not noticing any infection at all... I'd do an overinfection, of say 100µl of your pure virus stock (if you can afford to miss that much of course) and make a serial dilution. See what happens.
I'd only determine TCID50 on one cell line and use amounts of virus to infect the same cell line based on this titration, and I only do this for a short period after determination of TCID50 as cells change by passaging them.
Thanks for your suggestions, I'll try them all! Any kind of a result would be better than what I'm getting now..! 
So, if you're doing acute infection you have to determine the TCID50 on all of the cell lines that you want to infect, and recheck it regularly..? It sounds much less labour-intensive to just get a chronically infected cell line and work with that... (Or am I being naive... )
That depends on your purpose of course.
Working with different cell lines tends to get labour intensive indeed, even though there are publications where they do titration on one cell line and do another experiment on a different cell line. This might be a safe way if you are only studying the effect of a certain mutation in a "standard virus" (mostly NL4.3 or so), but when working with clinical isolates is a different thing.