A Suction Blister Protocol to Study Human T-cell Recall Responses In Vivo

This manuscript describes a practical procedure for the study of human immune memory responses in vivo, using cutaneous antigen recall and cell harvest by suction blister induction. TST was used as an example of intradermal antigen deposition and BCG-vaccine recall. Finally, an example of SB specimen characterization by flow cytometric and multiplex cytokine analysis was provided, demonstrating that roughly a third of the SB cells were antigen-specific polyfunctional T-cells of the effector memory phenotype.

The critical steps of this protocol include the intradermal injection technique, the suction blister induction, and the blister puncture. Firstly, a correct intradermal deposition requires trained personnel. An incorrect deposition can lead to suboptimal results. PPD is generally a well-known and safe cutaneous antigen, but its composition may vary between manufacturers, limiting comparability^13,20. This report provides instructions for a single intradermal deposition of 2 TU, and optionally two parallel depositions (2 x 2 TU), which may require additional ethical approval. However, other studies have used 10 TU, which increase the likelihood of a strong skin reaction^10,21. During the SB induction step, small stepwise increases in the negative pressure will reduce the risk of hemorrhage and blister rupture. The chances of contamination with red blood cells or leucocytes from the bloodstream are generally low². The aseptic puncture technique prevents microbial contamination and avoiding contact between the puncture needle and the dermal blister floor reduces impurities of debris or resident skin cells. However, some researchers prefer to harvest SB fluid by applying a rolling pressure to the punctured blister¹⁰. It may be necessary to puncture the septa within the blister. The SB technique itself is minimally invasive; collapsed SBs heal without scarring and infections are very rare.² However, some degree of hypo-pigmentation may occur and SB induction should probably be avoided in people with a history of colloid scarring.

Technical limitations include low total cell yields and consequently limited options for long-term storage. Relationships between the leucocyte yield, clinical TST responses, the blister size, and erythrocyte contamination has been thoroughly described elsewhere^2,10. In the BCG-recall experiments presented here (Figure 1), the median total yield from each blister was 50,000, implying a small-scale experimental set-up using these cells, especially if SBCs are studied alone¹⁵. However, the specific T-cell population in an SB sample mostly exceeds what is found in PBMC samples in both relative and absolute cell counts. In the example presented in Figure 2, the number of triple-positive PPD-specific T-cells in a sample of 100,000 SB cells were more than 2x greater than the number found in 1 x 10⁶ PBMCs from the same donor (data not shown). A visual scoring of the TST reaction is the most common clinical method for the evaluation of M. tuberculosis memory. Of note, the underlying adaptive immunological reaction does not peak at the same time as the clinical skin reaction^2,21. Not all BCG-vaccinated or Mtb-exposed individuals will develop strong TST reactions and strategies for classification, and a handling of samples from TST non-responders, as well as a preexisting mycobacterial sensitization in the study population, need to be considered before initiating a recall trial using this method. Also, in both SB sampling and visual scoring, there is a potential for a theoretical bias in individuals with reduced skin responses due to global or skin-related impaired immunity as seen, for instance, in HIV-infection and certain age groups^2,13,18,20. In addition, an immunological boosting of the TST reaction with repeated testing is well known^20,22. However, the SB cell phenotypes remain rather constant when TSTs are repeated¹⁰. This observation supports the role of SB recall in longitudinal studies of cellular immune responses.

For T-cell immunologists, SB recall allows for the harvest of high fractions of antigen-specific cells. However, the timing of SB for a sampling from a PPD-deposition is critical as both the cellular composition and the cytokine microenvironment change over time. In this protocol, blisters were induced 7-day post-challenge and the isolated cells were primarily CD4+ effector memory T-cells including high fractions of cells with a co-expression of IFN-γ, TNF-α, and IL-2. These observations are in line with previous studies describing how T-cell activation, proliferation, and differentiation occurs in PPD-primed skin^2,15,21. Kinetic SB studies in PPD-sensitized individuals suggest that the very early skin response is unspecific; however, already within the first days following the PPD-challenge, the response becomes dominated by CD4+ T-cells (both central memory and effector memory phenotypes have been described) and, after 3 days, the response is dominated by high fractions of PPD-specific cytokine producing CD4+ T-cells^2,8,10,15,21. This adaptive cytokine response remains detectable for more than 2 weeks^2,8,10,23. Day 7 post-TST appears to be the most optimal time point for the collection of cytokine producing memory CD4+ T-cells². However, other time points may, of course, be relevant depending on the antigen and response of interest. Of note, in one study, the adaptive PPD response has been reported to peak a little earlier with a decrease in the pro-inflammatory cytokine secretion already at 5 days post-TST¹⁰.

SB fluid contains high levels of both pro-inflammatory cytokines and other proteins shown to be representative of the skin microenvironment^15,24. Kinetics studies have shown that TNF-α and IFN-γ levels in SB fluid peak after 3 days while IL-2 levels peak after 7 days^2,10, probably reflecting the dominance of adaptive responses at this later stage. Because SB cells have been activated in vivo, they exhibit a high spontaneous cytokine release as well as a potential for a specific release upon restimulation (Figure 3 and 4).

This report focuses on CD4+ T-cell immunity. As demonstrated in Figure 2, the majority of the isolated T-cells were indeed CD4+, while there were almost no CD8+ T-cells in the suction blister fluid. This is in line with other SB PPD-recall studies reporting low proportions and an inferior migratory capacity of CD8+ T-cells compared to CD4+ T cells^2,8,10,23. A further characterization of the CD8+ contribution would be preferable; however, this is beyond the scope of this report.

T-cells are considered essential for the immune control of Mtb; however, it has been difficult to identify a reliable correlate of the protection reflected in the adaptive immune response from the blood^25,26. This roadblock severely hinders the development of new TB vaccines, as there currently is no alternative to large and very costly efficacy trials²⁷. TB vaccine developers determine vaccine immunogenicity by assessing small and transient changes in the vaccine-specific T-cell populations in the blood^28,29. However, it is questionable whether the small fraction of circulating antigen-specific T-cells found in the blood is relevant (i.e., capable of migrating to the site of an infection and representative of the T-cell-rich microenvironment controlling Mtb)^27,30,31.

Based on previous studies and the data presented herein, the SB cutaneous recall model represents an untapped potential for the study of vaccine-specific T-cells. Not only does the model enable a recall of a vaccine response generated decades ago, it also allows the evaluation of the true memory potential in a tissue-specific context^15,18,19,21. Novel, specific skin tests, which include antigens also comprised of candidate subunit TB vaccines, suggest new opportunities for vaccine evaluation using this model^28,32. Furthermore, transcriptomic analyses suggest that a cell-mediated immune response generated in PPD-challenged skin is similar to the response found in the Mtb-infected lung¹⁸.

While skin punch biopsies also allow for a cell harvest from the skin and provide spatial information, compared with SB sampling, the method is more invasive and requires enzymatic or mechanic processing to prepare single-cell suspensions³³. Measurements of cytokines and cell markers are comparable between the two methods^2,10.

The suction blister method has already been applied in many areas of medical research besides dermatology, either alone or in combination with a systemic or local skin challenge. Examples include studies of sepsis, Epstein-Barr virus-associated lymphoproliferative disease, diabetic neuropathy, glucocorticoid intake effects, and human trials testing therapeutic antibodies or models for T-cell-targeted therapies^{10,34,35,36,37,38}.

From a therapeutic point of view, the SB cutaneous recall method offers unique advantages to study the central T-cell memory potential and-from both a biological and technical point of view-the skin seems to provide a relevant sampling compartment^{2,15,18,19,21}. In particular, compared to traditional, passive sampling of circulating PBMCs, the SB cutaneous recall method allows for the study of T-cells that have proven the ability to migrate from the lymph node in response to their relevant antigen, and complete the local expansion and differentiation in a tissue-specific in vivo context^15,18,19,21.

In conclusion, the model demonstrated here could be relevant for researchers within the field of human adaptive immunity and T-cell targeting agents (i.e., in infectious disease vaccinology or cancer research). The TST model applied in this protocol is, of course, of special relevance in the field of TB vaccinology. However, the basic concept of this model is highly applicable in other fields of research.