Research Article: Optimal Isolation of Functional Foxp3+ Induced Regulatory T Cells Using DEREG Mice

Date Published: September 5, 2012

Publisher: Public Library of Science

Author(s): Abdul Mannan Baru, Christopher Untucht, Venkateswaran Ganesh, Christina Hesse, Christian T. Mayer, Tim Sparwasser, Roland Lang.


Foxp3 reporter mice including DEREG (DEpletion of REGulatory T cells) mice have greatly helped in exploring the biology of Foxp3+ Tregs. DEREG mice express a DTR-eGFP fusion protein under the control of a bacterial artificial chromosome (BAC)-encoded Foxp3 promoter, allowing the viable isolation and inducible depletion of Foxp3+ Tregs. Adaptive Tregs differentiated in vitro to express Foxp3 (iTregs) are gaining high interest as potential therapeutics for inflammatory conditions such as autoimmunity, allergy and transplant rejection. However, selective isolation of Foxp3+ iTregs with a stable phenotype still remains to be a problem, especially in the human setting. While screening for culture conditions to generate stable CD4+Foxp3+ iTregs from DEREG mice, with maximum suppressive activity, we observed an unexpected dichotomy of eGFP and Foxp3 expression which is not seen in ex vivo isolated cells from DEREG mice. Further characterization of eGFP+Foxp3− cells revealed relatively lower CD25 expression and a lack of suppressive activity in vitro. Similarly, eGFP− cells isolated from the same cultures were not suppressive despite of a broad CD25 expression reflecting mere T cell activation. In contrast, eGFP+Foxp3+ iTregs exhibited potent suppressive activity comparable to that of natural eGFP+Foxp3+ Tregs, emphasizing the importance of isolating Foxp3 expressing iTregs. Interestingly, the use of plate-bound anti-CD3 and anti-CD28 or Flt3L-driven BMDC resulted in considerable resolution of the observed dichotomy. In summary, we defined culture conditions for efficient generation of eGFP+Foxp3+ iTregs by use of DEREG mice. Isolation of functional Foxp3+ iTregs using DEREG mice can also be achieved under sub-optimal conditions based on the magnitude of surface CD25 expression, in synergy with transgene encoded eGFP. Besides, the reported phenomenon may be of general interest for exploring Foxp3 gene regulation, given that Foxp3 and eGFP expression are driven from distinct Foxp3 loci and because this dichotomy preferentially occurs only under defined in vitro conditions.

Partial Text

Foxp3 is an established marker for the identification of both natural and induced CD4+ regulatory T cells (Tregs) [1]–[4], yet it is inaccessible to reagents for their viable isolation or depletion. To overcome this limitation, DEREG mouse was generated which report Foxp3 promoter activity by the expression of a DTR-eGFP fusion protein from an ectopic bacterial artificial chromosome (BAC)-encoded Foxp3 locus [5]. The use of Foxp3 reporter mice, including DEREG mice have firmly established the crucial and non-redundant role of CD4+Foxp3+ Tregs in preserving the immune homeostasis and maintaining immunological self/tumor-specific tolerance [6]–[11]. Consequently, CD4+Foxp3+ Tregs are gaining impetus as prophylactics or therapeutics in order to regulate various immune disorders such as transplant rejection, autoimmunity and allergy. Nevertheless, in many instances the number of Tregs required for an effective intervention proves to be a limitation for their application. Recent advances pertaining to ex vivo induction and expansion of Foxp3+ Tregs (iTregs) from naïve CD4+Foxp3− T cells in the presence of TGF-β and retinoic acid (RA) [12],[13] could potentially surmount this bottleneck. Isolation and transfer of Tregs on the basis of classical Treg surface markers (e.g. CD25), which simultaneously get strongly up-regulated on conventional T cells during in vitro activation, poses a potential risk for their clinical application. Additionally, employment of polyclonal or antigen-specific Foxp3+ iTregs as potential therapeutics is a matter of debate [14],[15]. Besides, Foxp3+ iTregs generated in vitro tend to rapidly lose Foxp3 expression and concomitantly their suppressive activity following adoptive transfer [16],[17]. Consequently, culture conditions favoring the induction of stable Foxp3 expression as well as strategies for the selective isolation of Foxp3+ iTregs from these cultures remain to be established. In this study, we report protocols for the optimal generation and isolation of functional eGFP+Foxp3+ iTregs using DEREG mice.

Specialized dendritic cells (DCs) can endogenously generate Foxp3+ iTregs and DC-derived signals have been implicated to contribute to a stable Foxp3 expression [18]. By using both DC-supplemented and APC-free in vitro cultures we aimed to define conditions resulting in differentiation of Foxp3+ iTregs with maximum suppressive capacity and comparative stability. CD4+eGFP− T cells sorted from DEREG mice to a high purity (Figure S1) were used to generate eGFP+Foxp3+ iTregs that could be easily isolated by FACS sorting on the basis of eGFP expression for their functional analysis. We have recently employed a similar approach to generate and characterize CD8+Foxp3+ T cells [19]. While the vast majority of ex vivo isolated Foxp3+ Tregs co-express eGFP in DEREG mice (Figure 1A, left panel) [5], we surprisingly detected a sizeable fraction of eGFP+Foxp3− and eGFP−Foxp3+ populations in iTreg cultures supplemented with transforming growth factor-β (TGF-β, retinoic acid (RA), soluble anti-CD3 antibody and GM-CSF derived BMDC (Figure 1A). Albeit the frequency of eGFP+Foxp3+ cells peaked by day 3 of the differentiation, we could obtain maximum absolute numbers of eGFP+Foxp3+ cells by day 4, and further differentiation led to a drastic decline in eGFP+Foxp3+ cell frequencies (Figure S2).

During the screening for conditions that result in the generation of stable and highly suppressive CD4+Foxp3+ iTregs using DEREG mice, we observed unexpected eGFP+Foxp3− and eGFP−Foxp3+ populations. eGFP+Foxp3− T cells expressed lower levels of CD25 when compared with their eGFP+Foxp3+ counterparts, consistent with CD25 being a direct target gene of Foxp3 [20]. We could thus utilize the intensity of CD25 expression to isolate and characterize eGFP+Foxp3− T cells. In contrast to eGFP+CD25hi iTregs, eGFP+CD25lo T cells lacked significant suppressive activity. This is consistent with previous studies demonstrating that, Foxp3 is essential to confer suppressive activity. Given that IL-2 is an important T cell growth factor, increased consumption of IL-2 by eGFP+CD25hi cells could be a simple explanation for their higher suppressive activity. Similarly, the eGFP− fraction, predominantly comprising of CD25lo T cells, lacked suppressive activity. However, as CD25 is only one of about 700 Foxp3 target genes that may be involved in conferring suppressive properties [21], additional mechanisms might be involved. The complete lack of suppressive activity by the eGFP− population, albeit containing a small fraction of eGFP−Foxp3+ T cells could be simply explained by the minor proportion of Foxp3+ cells present in this fraction. Alternatively, a non-suppressive behavior of the eGFP−Foxp3+ population could be hypothesized, as despite the successful Foxp3 induction, these cells may not have fully established the Foxp3-dependent suppressive program. In line with this notion, small populations of non-suppressive Foxp3+ T cells have been reported in mice [22].