You could start here: Determination of intracellular cytokines IFN-gamma and IL-4 in canine T lymphocytes by flow cytometry following whole-blood culture. Can J Vet Res. 2009 Apr;73(2):137-43. PMID: 19436583
The article describes some optimization steps using non-specific stimulation with PMA/monensin, but the titration and optimization steps would be applicable. Many folks substitute ionomycin for monensisn, or use Concanavalin A to induce proliferation, so consider that as well.
Since you're presumably not interested in using PMA/ionophore stimulation other than as a positive control, I'll suggest that ex vivo stimulation of canine T cells can be relatively straight-forward or complicated depending on the the nature of the antigen.
The easiest method is to use the specific peptide or pooled peptides (or drug metabolite) to which to expect a response. This would be most likely to allow for detection of cytokine production during a stimulation interval similar to PMA/ionophores reported in the paper.
It becomes more complicated with complex antigens such as whole bacteria/lysates, or drugs with high molecular weights. The more complex the antigen, the more processing required before it can be recognized by a TCR. Whole cell lysates may require 12, 24, or 48 hours for optimum stimulation, and require the addition of golgi-inhibitors to be deferred to the last few hours of stimulation. Also, kinetics of memory CD8 T cells can differ from memory CD4 T cells, so separate assays with different periods of stimulation and antigen concentrations may be needed. Other authors and labs with which we have collaborated have found IL-4 detection in canine CD4 T cells to be challenging. Interferon gamma and TNFa work more reliably.
The nice folks in the Petersen lab at the U of Iowa describe their protocol for canine T cell stimulation with leishmania antigens in papers such as this one:
Programmed death 1-mediated T cell exhaustion during visceral leishmaniasis impairs phagocyte function. Esch KJ1, Juelsgaard R, Martinez PA, Jones DE, Petersen CA.J Immunol. 2013 Dec 1;191(11):5542-50. doi: 10.4049/jimmunol.1301810. Epub 2013 Oct 23. PMID: 24154626
I will read it with a high interest. In 2009 I also published a very similar paper with human T-cell from PBMC short time (6h) culture with rDer p 10 (the mite recombinant tropomyosin) under co-stimulation with CD28 and inhibition of intracelular cytokine transportation with brefeldin A. In fact IFN, IL-2, IL-5 and IL-13 presented good and predictable response, while IL-4 did not respond so good, even with a control stimulation with PHA or with PMA+I. Although, I preferred PHA, since PMA+I downregulates CD4 expression and we will need to referentiate CD4 by CD3+ CD8-.
I will try now with dog cells, with the perspective, especially from the first paper you told me.