I agree that route of injection should be largely similar with a few caveats.  We use IV injection because it is clear that the entire dose volume has been delivered with 100% efficiency (i.e. if you missed it, you know it).  With IP injections, some fraction of the cells / volume can be delivered into the fat pad or other unwanted areas even with skilled hands.  For transfer colitis studies, there is a threshold of cells that need to be transferred so small issues with an individual injection may manifest as variability.

There's also a kinetic aspect to it, depending on the in vivo manipulations being performed after introduction of the cells.  IV is a bit faster towards systemic distribution than IP.

Sure! In my practice I could not see any significant difference between IV and IP routes of adoptive transfer of lymphocytes. Also, I could not see difference between results of IV and IP immunization. I think, many scientists tend to keep intact their skills obtained in scientific schools and carefully reproduce methods used.

I agree that route of injection should be largely similar with a few caveats.  We use IV injection because it is clear that the entire dose volume has been delivered with 100% efficiency (i.e. if you missed it, you know it).  With IP injections, some fraction of the cells / volume can be delivered into the fat pad or other unwanted areas even with skilled hands.  For transfer colitis studies, there is a threshold of cells that need to be transferred so small issues with an individual injection may manifest as variability.

There's also a kinetic aspect to it, depending on the in vivo manipulations being performed after introduction of the cells.  IV is a bit faster towards systemic distribution than IP.

One has to consider whether there are other manipulations of the mice in the model of interest. If you also have to deliver other things, e.g. drugs, via the iv route one has to think about to do it twice within a short time frame. Therefore we tested it out and came to the conclusion that in our model (EAE) iv against ip administration of the cells does not matter. For me this was rather surprising given the huge amount of literature using exclusively iv administration, but as discussed before by Dmitry poeple probably rather stick to old protocols.

IV vs. IP can matter when using allogeneic cells (recipient NK cells kill allo cells transferred in IV but not those given IP, i.e. those cells that remain in the peritoneum survive). There is a fairly extensive literature of NK killing of transferred cells even when they are syngeneic.  The question of route of administration in these effects is not yet adequately studied as far as I am aware. 

Dear Colin, can you undouptedly say that these killers are NK cells instead of "natural" CTL?

Hi Dmitry, It was NK cells and not T cells, in that the studies with allo cells were done in Rag KO recipient mice (see PMID: 17437644; Figs 4 and 5) and NK depleting antibody prevented the killing. We have also seen greater recovery of syngeneic T cells transferred into NK depleted Rag KO recipients.

Thanks a lot, Colin! I asked this, because we observed decrease in "spontaneous" killing after injection of anti-CD8 mAbs to WT B10 recipient mice. Your work deals with rejection of cardiac transplants. I think, lymphocytes should be resistant to NK-cells.

At least in C57BL/6 mice, using the EAE model, I've noticed that adoptive transfer of in vitro restimulated (w/MOG35-55 peptide) T cells via the i.p. route leads to disease, while intravenous injection does not.  The route thus does appear to matter in that model.  Whether that is a quantitative difference (ie: more cells somehow survive when delivered i.p.) vs. a difference in lymphatic trafficking etc. is unstudied.  However, I am not the only one who has found that difference...

Most welcome Dmitry. One additional anecdote (not sure how reproducible): when transferring lymphocytes IP, it seemed a large fraction just hung around for quite awhile in the peritoneum (most studies of recovery post IV vs IP don't actually look in the peritoneum post transfer, they look for migration to other sites)

Well! Second additional anecdote (well reproducible): when transferring lymphocytes IV, wast majority of them is trapped by liver and their subsequent fate is not studied too.

I agree with Jason,our adoptive transfer system run on IV vs IP ,rout of cell transfer seems not mater except the challenging after that with pathogen or specific antigen need to be start couple of day late in IP cell transfer. 

I think it's a mistake to think of these two very different protocols as interchangeable (though they may in fact be so for certain models). The technical limitations of each are real - stress to the animal for IV injections (though with an experienced operator this is greatly minimized) and definitely stress to the operator (!) if they lack experience; inconsistency, particularly loss of cells, using IP injection that can mean larger numbers of animals are needed; the number of transfers needed in a single animal, or administration of other cells/drugs in series or simultaneously. But the bigger issue is that cells traffic differently with each protocol. So you need to think about your model carefully - what is physiologically relevant (since you are using an in vivo model, presumably this is important). Are your cells activated or naive? Is uptake into the lymphatic circulation preferable or is egress from the circulatory system preferable? Do they have surface markers that will allow uptake/egress or those that will promote clearance or retention? Where did the cells you are transferring originate and where/how were they activated (in vivo or in vitro)? It is a mistake to take the easier path because the correct path is more difficult or time-consuming (speaking from experience). If you can't do IV injections, it is worth a hands-on tutorial from someone who is good at it. It might make sense to compare both methods in a pilot study as there are so many variables. Understanding that material is limiting and expensive, you might be able to figure out which to use from the literature (e.g. prior studies all used IV injection - if you get a different result with IP is it the method or your intended experimental variable?). If your study is new you can try your preferred method and if it gives the desired result stick with it. 

Hi Adam,

In my experience, primarily transferring naive DO11.10 TCR transgenic CD4 T cells by intravenous injection led to a greater average and smaller variation of the donor cells in the lymph nodes and spleen than transfer into the peritoneal cavity. Since very few naive lymphocytes are present in the peritoneal cavity, I assume they don't express the corresponding chemokine and adhesion receptors to migrate into that location - or out again. On the other hand, they certainly do express the right combinations of receptors needed to enter and exit the blood. Moreover, I think it's reasonable to think of the peritoneal cavity as a specialized compartment into which certain types of leukocytes are recruited and retained. Still, you could always try yourself and choose whichever method you find the better approach.

Good luck with your experiments,

Luke

Another important point to consider is the localisation of cells after injection. We have seen that most of injected cells in IV route are localised into lungs, while those injected by IC route are localised into spine (bones).

Which percentage of labeled T cells should I find in the peripheral lymph nodes after adoptive transfer of say 5-10 million labeled T cells? I find intraperitoneal injections to be very reproducible and reliable, whereas tail-vein injections can be quite variable from mouse to mouse. We just transferred 35 million CFSE-labeled CD3+ cells by intraperitoneal injection to 12 mice and we observed between 0.4-0.7% labeled cells in the popliteal and cervical lymph nodes 48 h later. I was expecting to find about 10 times that number... What would you expect?

Dear Jeremias, we transferred 10 million of GFP+ T cells i.p. from transgenic B10.GFP to intact B10 wild type mice, with or without simultaneous immunization by allogeneic tumor cells . Then, GFP+ cells were detected by FACS in popliteal lymph nodes of recipients. Results - Control (Immunization): 1st day -  0-0.02%(0.25%); 3rd day - 0.01-0.53%(0.28-0.56%); 7th day - 2.46%(2.01%); 9th day - 2.09-2.98%(1.47-3.03%).

Thank you, Dmitry! That is exactly what I needed to know. We didn´t do so bad then. Would you say that the increasing number of GFP+ cells in the lymph nodes of the recipients as time passes is due to accumulation or actual T cell expansion? I'd expected the system to "equilibrate" after 24-48 h, but I guess it takes longer.

Dear Jeremias, I do not see any reasons for T cell expansion in this experimental setting. Expansion takes place under lymphopenia condition if you irradiate recipient mice.

Dear Dmitry, I totally agree with you, it must be T cell recirculation. Your findings (and some of my preliminary data as well) suggest that adoptive transfer by i.p. does take longer to reach stable levels in lymph nodes, at least in some systems. It is something that I need to take into account, or maybe switch to the i.v. route. Thanks again!

This is a great thread. I'm wondering if anyone has updates. Specifically, I'm wondering about the time it takes T cells to get to the site of interest in i.p. vs IV. Also, does one method seem to be more reliable or effective than the other? I don't have experience with tail vein or RO injections, so I was hoping to stick with i.p. Thanks!