My guess is that you see more RNA copies than infectious particles (TCID50)? With TZM-bl cells, you measure the infectivity of the viral preparation or sample. With RT-PCR, you measure the number of RNA copies. A lot (a majority) of viral particles produced are defective in one way or another so that RNA is there but they cannot necessarily infect cells productively. So you don't get the same number because you don't measure the same thing. :)

RT-PCR is a much more sensitive measure of viral particles and it does not take into account that some or most of the HIV particles in a patient are non-infectious. It gives you the total amount of viral RNA that is present per ml of blood in the patient. It includes all particles that contain the HIV RNA- exosomal vesicles, virus like particles, immature viral particles, viral particles without Env, mutated and non-functional viral particles etc... The infectivity assay actually tells you how much of the virus is actually infectious and is a more relevant measure in my opinion.

Knowing how wimpy HIV-1 really is and also knowing that HIV-1 Env is not that great a protein to provide optimal functions, I am not surprised that you have a 5 fold difference in viral load vs TCID50. I am surprised that the difference is 5 fold though. A 3-fold difference is what I had envisioned in my mind. I could be wrong. You could refer to papers that show longitudinal studies of patients and how the viremia changes in them. I am not sure if they have done a concomitant infectivity assay with them though.

Viral load determined by RT-PCR and TCID50 cannot be compared: by RT-PCR you measure all the viral particles in your supernatant, whereas TCID50 is a measure of infectious viral particles, which represent a very small percentage of the total viruses.

It's pretty reliable. It also does really help to use Dextran, it increases the titers about 20 Fold. Also remember that the measurement of viral RNA by PCR does not necessarily correlate with the amount of infectious viruses, as many viral particles are non-infectious, so you would expect a lower TCID50.

Unlike the infectivity assays, the RT-PCR is supposed to quantify both infectious and non-infectious virus and viral partciles. So no surprise if you get a higher RT-PCR quantity than TCID50. Hover the 5Log difference appears to be high. I am just curious what volume of samples you used in both assays? What volumes you used to extarct the RNA? Volume used in infectivity assays?

To determine the titre of HIV isolate, I would prefer to use PBMC instead of TZM-bl cells unless your future study of the isolate will be conducted in TZM-bL cells.

Maintain awareness that whatever result you get, the TCID50 values refer to TZM-Bl cells, and may or may not correlate with infectivity of other target cell types/populations (e.g. PBMC).

One question, what's the HIV isolate you are working on? Is it R5 or X4 virus? TZM-bL seems work better for R5 virus.

Sometimes, unfortunately, expression level of CD4/CXCR4/CCR5 may be reduced. TZM-bl naturally exppresses CXCR4, therefore, CD4 and CCR5, which are artidfically expressed by transduction, are major determinats. In my experience, most of TZM-bL, which are less expressing CD4, almost all cells contained fully functional LTR system. Thus, please check expression level of CD4/CCR5 and compare with that of CXCR4 as a control.

The difference is very common. Tzm-bl cells express luciferase after HIV entry this cell. Therefore, Tzm-bl cells don't determine non-infectious HIV. But real-time RT-PCR does not differ infectious HIV from total HIV, so RT-PCR determines all HIV.

Thanks to all of you.

I am aware of the fact that RT-PCR reflects total number of viral particles and that some isolates have more or less infectious particles, which is precisely the reason why I measure TCID50.

However, I have been told that the difference ranges around 3 log though I get 5 log difference. But thanks a lot for your advice of looking at CD4/CCR5 expression!

I am also wondering if the excel sheet I use is "working well" and I'm struggling a bit with the math of it to be honest. Would someone have one to recommend?

I have definitely seen >4-log differences between particle count and infectivity.

For the record, even fully competent virions can fail to establish infection (else why would the titre of the same viral stock differ when evaluated in different target cells?). Think of infection as a multi-step process, where each step has a less than 1 chance of happening, and a lot of randomness comes into play.

TZM-BL cells are a very convenient and pretty reproducible way of evaluating the infectivity of your virus stocks (unlike donor PBLs), but you need (as Ed Berger says) to establish the relationship between infectivity in these cells and infectivity in the cell type you care about, if you want to control your multiplicity of infection.

You can do limiting dilution end-point assays in the latter using a sensitive read-out (Matt Collin's in the mid-1990s is a good example).

For software to calculate true infectivity (and the related ID50) from limiting dilution assays, we've used John Spouge's vacman (https://www.ncbi.nlm.nih.gov/CBBresearch/Spouge/html_ncbi/html/software/program.html?uid=0) but there may be more recent and better programs. I see a convenient Excel file at http://www.klinikum.uni-heidelberg.de/fileadmin/inst_hygiene/molekulare_virologie/Downloads/TCID50.xls

For TCID calculation, try "Brett D. Lindenbach Infectivity" excel sheet. That's a good one.

I think you can use the attached excel file for TCID50 calculation. Its good.

The 5 log difference is too high. We do expect difference between TZM-bl estimated viral titer and RT- , p24 ELISA or Electron microscopy-based data. The reasons are obvious and have been explained above. The simple answer is - yes TZM-bl method is reliable. Also, it is important to define for what purpose you need TCID50. If you are strictly following the recommendations for the TZM-bl-based TCID50 protocol, you will have reproducible and, thus, reliable data. However, I have to stress again that  5 log difference between rt- and TZM-bl is more than we expect (usually). Either your TZM-bl is underestimating the titer or your RT is giving you too high counts. I would suggest to try the digital droplet PCR (ddPCR) and compare with your RT.

Hi Raphaelle,

This is a good question and I agree with Klintsy Torres, Yonghou Jiang and Edward Berger. TCID50 represents a virus amount correlated with that kind of cells you tested for, thus TCID50 calculated from TZM-bl cannot be used for PBMCs (TCID50 should be calculated with cells you're intending to use in your final experiments). I don't like to use TZM-bl for determining HIV-1 TCID50 for general purpose because these cells have high expression of CD4 and CCR5 (for CCR5 is about 10^5 molecules per cell), thus becoming very susceptible to infection with very low concentration of HIV (e.g. in my hands, 50 to 70 pg p24/ml of BaL is able to infect TZM-bl but it doesn't do the same on primary CD4 T cells at this same concentration). Thus, TCID50 calculated on TZM-bl is not a good tool if you intend to use the virus on primary cells.