Salih,

Your questions are valid regarding the measurement of p24, which is an archaic method to look at vector titer. The p24 levels are completely arbitrary from prep to prep, so you are better off ignoring the calculation of transduction efficiency (i.e., TU/ng p24) and just use real-time PCR after collecting the genomic DNA from limiting diluted vector-infected cells or if you have a transgene like beta-gal or a fluorescent protein for analysis on the cells to calculate titer.

In my experience, lentiviral vector production can be readily achieved in the 10e6-10e7 TU/mL range (from conditioned media) using the standard DMEM/10% FBS/penn-strep-glutamine. You can try other variations, but the use of antifungals seems unnecessary especially if you are using sterile techniques in the cell culture area. I am a minimalist in terms of adding extra solutions to the media. You can try to use less FBS (like 1-2% or even zero %), but what I found is to use very low passage 293T cells (>5 passages) and you should be able to achieve relatively high titers in your lab. Heat inactivation of FBS is not necessary to make good vector preps.

Frank Park

Non-heat inactivated serum (active serum) is ok to be used for 293T cells. As for the concentration of FBS to be used, it is better that standard concentration is used before and during transfection. After transfection, 5% FBS would be fine. As for the use of 1% pen+strep, it is not required for vector production but it can help to reduce the chance of contamination by microbes. If you can make sure the cell culture will not be contaminated, you do not have to use.

I've not used roller bottles to grow up the host/producer cells (I typically use 40 x 15cm plates), but I don't see why this wouldn't be an excellent way to efficiently boost cell numbers for vector amplification. I've only used the 293T cells in 10% FBS, and not heat inactivated. As a grad student we used heat inactivated horse serum to avoid cytopathic effects (PC 12 cells), but never needed to heat treat the FBS. As to the optimization of the infection, etc, the stress applied by the infeciton may help or interfere, and in all liklihood will need to be determined empirically. When we make Adeno vectors with 293s the infection is done in serum free DMEM low glucose for 4 hours, as they seem to do better with the stress, but other vectors (MMTV derivitives) seem to do better with full serum. Altering the CO2 will certainly change the pH, a different stressor, and again hard to predict if it would be better or worse. I have not seen a difference in aBx manipulation. In the end, the nitty gritty details of the CaPO4 transfection will likely not play as big a role as the secondary infection you obtain when the primarily CaPO4 transfected 293T cells (perhaps 5-10% of the culture) release the first wave of recombinant virions into the cultre, that then infect the rest of the 293Ts for the "full on" vector production. It is perhaps more improtant to closely observe the culture for evidence of when to harvest (early in the process when they get sick and start to detach). If you harvest too early or too late, your yield will drop precipitously. Another suggestion would be to do the primary infection in a more manipulatable culture container (T-flask or 10cm plate) which may be easier than the roller bottle; once the cells start to look sick in the smaller container, transfer the supernatant to the roller flask and start up the "secondary infection" in a more controlled manner.

Agree with most answers. No need for heat-inactivated FBS, antibiotics of any kind, or playing around with CO2. HEPES-buffered GlutaMAX media also work well, and is more stable than no buffer and added glutamine. The cells matter a lot though. We have tested various 293T (incl. FT) cells from different sources, and they all differed in terms of viral yield (all other conditions equal). FT was not among of the best.

In our hands, CaPO4 transfection robustly yielded functional (=infectious) titers of factor 5 to 10 lower than we regularly get with any 293-specialized transfection reagent. Also, CaPO4 titers were fluctuating a lot between preps. Perhaps we didn't do it right, but it was frustrating and we went back to reagents for consistent results, despite the costs. What about comparing your functional viral yield between CaPO4 and , all other conditions equal. You might be able to substantially cut back on culture area, for the same viral yield.

For functional titering of VSVG-pseudotyped virus, conveniently use the same 293T cells as the indicator cell line. We plate clonal dilutions of viral stock (1:10e4 to1:10e8) together with cells in suspension, into 48 well plates at ca 30-35% confluence. Using unconcentrated supernatant, start with 1:10 down to 1:10e4. Let cells grow for 3 doubling times (approx. 3 days = over the weekend), then fix and immunostain for vector-expressed protein. Count immunopositive clones (2-8 cells) per well with a regular 10x inverted objective. Simply view dry samples, no mounting necessary. For most accurate results, count samples containing 20 - 100 positive clones. Calculate clone numbers back to transducing units/ml in the viral stock. Good preps (concentrated by ultracentrifugation) will yield n x 10e10 to 10e11 tu/ml. Sups should be ca. n x 10e7 to 10e8 tu/ml.

Not every cell (clone) carrying a single-copy insertion will express above immunodetection level, so this method underestimates the real functional titer. Still it is a pedestrian but good way of evaluating any given viral prep for infectivity, also to compare different preps with each other.

Limitations: 1. Indicator cell type must be receptive for the pseudotype you are using. 2. Only VSVG pseudotype is able to infect freshly trypsinized cells. 3. The internal promoter used in your construct must be active in the indicator cell type.

I appreciate all of you who contributed to this discussion...It was great help...

Just add one important point: Calcium phosphate transfection is toxic to cells, though it might be adequate for some purpose, but to achieve high titter, better use liposome transfection, like lipofectamine. I did side-by-side comparison to find this, though you can also figure it out by yourself.

It is a good point but the reason why people use calcium phosphate traditionally is because of the cost of lipofectamine considering large culture volumes...

Salih,

it might still pay off. You need large culture volumes for large viral yields. But, if you can harvest 10 x more virus/area through higher transfection efficiency, you may scale down your cultures, and you will save a lot of running costs just by scaling down (plus, less work). We have calculated both scenarios down to the individual pipette tip. For us, higher transfection efficiency actually pays off economically - provided we get a good offer on the transfection reagent.

Lipofectamine is one of the most expensive transfection reagents, but definitely not the best for this purpose. Look for reagents specializing in 293-type cells. We are currently using https://www.mirusbio.com/products/transfection/transit-293-transfection-reagent, with very good results. You may use half the amount given in the protocol (co-downscale DNA), and it's still a lot better than any CaPO4 result we have ever managed.

Good luck!

I use DMEM + 10% FBS (heat inactivated) + pen strep in 5% CO2 for high-titer lentiviral vector production. I make 70μl lentiviral vectors with 10e9 to 10e11 TU/ml from 2 of 100 mm dishes every week. Antibiotics (pen strep) were not influenced the lenti- or AAV-viral titers, and calcium-phosphate transfection were not so toxic to HEK293T cells (Thermo).

I believe that there are other 3 important points to obtain high titer lentiviral vectors.

One point is 293T cell's condition. The status which increasing 5-fold in 48 hours or 12-fold in 72 hours are the best condition of 293T cells. These cells have high metabolic functions, therefore they can take calcium-phosphate particles very well and make lentiviral vectors in cytoplasm well. Ｔｈｅ best thing is that membrane of 293T cells is used to compose particle of lentiviral vectors. The 293T cells become best condition over 1 to 3 month culture from liquid nitrogen, you can use countinuously them over 6 months. If your cells were dead or not proliferated by calcium-phosphate transfection, the cells were very weakly.

Second point is the serum lot check. Previous time of it, I checked proliferation speed of 293T cells in vitro and strength of gene expression in vivo. Because selected 293T cells having good proliferation speed not sometimes have the productive ability of lentiviral vectors. And the serum also influence the components of cell membrane of 293T cells, so serum influence the viral tropism. If you will use high titer lentiviral vectors in experimental animals, you should check them in vivo.

Last point is plasmids. I obtained good titer from pCL20c plasmids and Dr. Miyoshi's plasmids (RIKEN) but not good results from pLenti plasmids. One reason is probably that packaging plasmids and expression plasmids interact to produce the vectors. So you should check in other's papers whether your plasmid set could product high titer lentiviral vectors.