Distinct palmitoylation of Foxp3 regulates the function of regulatory T cells via palmitoyltransferases

Regulatory T cells (Tregs) play pivotal roles in maintaining immune homeostasis and preventing excessive immune responses in vivo. As key players in suppressing immune reactions, they help to preserve immune tolerance and are thus crucial for preventing autoimmune diseases. Within the tumor microenvironment (TME), Tregs facilitate tumor immune evasion by impairing the activity of effector cells via multiple mechanisms, thus contributing to the initiation and progression of tumors [1]. To perform this regulatory function, Tregs depend on the master transcription factor forkhead box protein p3 (Foxp3) [2]. The expression of Foxp3 has emerged as indispensable for the development and optimal function of Tregs [3]. Interestingly, different posttranslational modifications, including phosphorylation, ubiquitination, glycosylation, and acetylation, exert significant regulatory effects on the biological function of Foxp3 [4–8].

Palmitoylation is a reversible lipid posttranslational modification characterized by the addition of a 16-carbon palmitoyl group to a cysteine residue via a thioester bond. This modification is crucial for modulating various characteristics of proteins, including their subcellular localization, protein‒protein interactions, stability, and signal transduction activity [9–11]. Recent studies have elucidated key players in the palmitoylation process. Notably, a family of 24 mammalian palmitoyltransferases, known as Asp-His-His-Cys (DHHC) motif-containing proteins, has attracted considerable interest. These enzymes are responsible for catalyzing the addition of the palmitoyl group to a target protein. Additionally, the identification of multiple depalmitoylating enzymes, such as APT1/2, PPT1/2, and ABHD17a/b/c, has revealed the roles of these enzymes in removing palmitoyl groups. The dynamic balance between palmitoylation and depalmitoylation is essential for modulating protein function and coordinating cellular signaling. However, whether the function of Foxp3 is regulated by its palmitoylation under physiological  conditions and in the TME is unknown. Therefore, our study aimed to detect palmitoylated Foxp3 and to determine the effects of Foxp3 palmitoylation on antitumor immune responses.

To determine whether Foxp3 undergoes palmitoylation, we engineered a Foxp3-Myc plasmid and overexpressed it in HEK293T cells. The cells were harvested 24 hours after transfection for protein extraction. Subsequently, both ABE (acyl-biotin exchange) and Acyl-RAC (acyl resin-assisted capture) experiments were performed to determine the palmitoylation status of Foxp3. Our results confirmed that palmitoylation of Foxp3 occurs in vitro (Fig. 1A, left). To detect palmitoylated Foxp3 in primary cells, lymph nodes were collected from C57BL/6 mice for protein extraction and ABE assays. Western blot analysis of the collected samples indicated that Foxp3 was palmitoylated in the lymph nodes (Fig. 1A, right). In further experiments, to determine the specific palmitoylation sites in Foxp3, we designed and expressed a Foxp3-Flag plasmid and then purified the expressed protein for mass spectrometry (MS) analysis. It is recognized that palmitoylation can increase the molecular weight of a protein by 238 Da [12], and the data from our MS analyses suggested that Foxp3 may be palmitoylated at five cysteine residues—Cys204, Cys218, Cys280, Cys281, and Cys424—and that each palmitoylation event results in a mass shift of 238 Da per peptide segment (Fig. 1B and Supplementary Fig. S1). Next, to validate the MS results, various Foxp3 mutants were generated, expressed in HEK293T cells, and subjected to Acyl-RAC to measure the level of palm-Foxp3. The results of these experiments indicated that alanine substitutions at cysteine residues, specifically the C204/218A, C280/281A, and C424A mutations, which prevent the addition of the 16-carbon palmitoyl group, led to a decrease in the level of palm-Foxp3. In contrast, the 5CA mutant, which had a simultaneous mutation of all five cysteine residues, exhibited almost complete abrogation of palmitoylation (Fig. 1C).

Fig. 1.

Palmitoylation of Foxp3 regulates the function of Tregs via palmitoyltransferases under steady-state conditions and in the TME. A ABE and Acyl-RAC assays were performed to detect the palmitoylation of Foxp3 both in vitro and ex vivo. B MS analysis was used to determine the specific palmitoylation sites of Foxp3. C ABE assays and western blot analysis were conducted to measure the palmitoylation levels of Foxp3-WT and Foxp3 mutants. D Immunoblot analysis was performed to examine the interaction between Foxp3 and DHHC2 in HEK293T cells transfected with specific expression vectors. Histograms showing the mean MFI ± SEM of Foxp3 in Tregs isolated from the peripheral lymph nodes (E) and spleens (F) of WT and Dhhc2^−/− mice (n = 5). Histograms showing the mean MFI ± SEM of Foxp3 in Tregs isolated from the peripheral lymph nodes (G) and spleens (H) of Dhhc2^Flox/Flox and Dhhc2-T cell KO mice (n = 5). I Tumor growth was evaluated in Dhhc2^Flox/Flox and Dhhc2-T cell KO mice using the YUMM3.3 melanoma model (n = 9 ~ 10). J Histogram showing the mean MFI ± SEM of Foxp3 in Tregs isolated from the tumors of both Dhhc2^Flox/Flox and Dhhc2-T cell KO mice (n = 6). In this study, intranuclear staining of Foxp3 with a fluorophore-conjugated anti-mouse Foxp3 antibody was performed using the eBioscience Foxp3 Transcription Factor Staining Buffer Set to quantitate nuclear proteins. Statistical significance was determined using an unpaired two-tailed Student’s t-test. Significance levels are indicated as *p < 0.05 and **p < 0.01

The DHHC family palmitoyltransferases are responsible for the palmitoylation of substrate proteins. To identify the palmitoyltransferases contributing to the palmitoylation of Foxp3, a coimmunoprecipitation (co-IP) assay was conducted. First, the pCMV-Foxp3-Myc plasmid was separately cotransfected with each pCMV-DHHC-HA plasmid into HEK293T cells. Then, 48 hours post-transfection, the cells were harvested for protein extraction, and a co-IP assay was performed. The results confirmed the physical interactions between Foxp3 and multiple palmitoyltransferases, including DHHC2, DHHC3, DHHC7, DHHC13, DHHC17, DHHC19, and DHHC23 (Fig. 1D and Supplementary Fig. S2A). To investigate the expression patterns of these seven Dhhc family genes in murine Tregs, we sorted Tregs from both spleen and tumor tissues harvested from Foxp3-eGFP knock-in mice. Quantitative real-time PCR (qRT‒PCR) analysis revealed abundant Dhhc2, Dhhc3, and Dhhc7 expression in the purified Tregs (Supplementary Fig. S2B). To further validate their role as palmitoyltransferases for Foxp3, DHHC2, DHHC3, and DHHC7 were selected for further investigation. For this purpose, we subjected coexpressed protein samples to Acyl-RAC. The results indicated that individual overexpression of DHHC2, DHHC3, or DHHC7 significantly increased Foxp3 palmitoylation (Supplementary Fig. S2C, D). These findings suggest that DHHC2, DHHC3, and DHHC7 are involved in palmitoylating the Foxp3 protein.

To further investigate the function of Foxp3 palmitoylation in vivo, we generated Dhhc2^−/−, Dhhc3^−/−, and Dhhc7^−/− mice using CRISPR/Cas9 genome editing (Supplementary Fig. S3). We first analyzed T cells isolated from peripheral lymphoid organs of both wild-type (WT) and Dhhc2^−/− mice by flow cytometry, and the expression of Foxp3 in T cells was evaluated by intranuclear staining with a fluorophore-conjugated anti-mouse Foxp3 antibody. Interestingly, systemic knockout of Dhhc2 led to a decreasing trend in Foxp3 protein expression in Tregs isolated from the lymph nodes and spleen (Fig. 1E, F). To elucidate the role of two other palmitoyltransferases, Dhhc3 and Dhhc7, in the regulation of Tregs, we analyzed splenic T cells from WT, Dhhc3^−/−, and Dhhc7^−/− mice via intranuclear staining of Foxp3 and flow cytometric analysis. Surprisingly, we found that the loss of Dhhc3 or Dhhc7 in mice resulted in significantly decreased Foxp3 protein expression, as determined by the decrease in the mean fluorescence intensity (MFI) of Foxp3 in Tregs (Supplementary Fig. S4). A published study demonstrated that DHHC2 localizes predominantly to the endoplasmic reticulum, whereas DHHC3 and DHHC7, the closest homologs within the DHHC protein family, are localized in the Golgi apparatus [13]. Interestingly, our immunofluorescence staining results (Supplementary Fig. S5) were consistent with these findings, which may partially explain why Dhhc2, Dhhc3, and Dhhc7 play synergistic and nonredundant roles in regulating Tregs. In addition, qRT‒PCR analysis revealed that among the genes examined, only Dhhc2 exhibited significantly greater expression in Tregs than in conventional CD4⁺ T cells (Supplementary Fig. S6), indicating a potentially important role for Dhhc2 in regulating Treg function. Notably, in Foxp3-Cre mice, a significant decrease in the expression of Foxp3 was observed in Tregs (Supplementary Fig. S7A, B), a phenomenon that was not observed in CD4-Cre mice (Supplementary Fig. S7C, D). Thus, we crossed CD4-Cre mice with Dhhc2 floxed mice to generate mice with T-cell-specific knockout of the Dhhc2 gene (Dhhc2-T cell KO mice). We then obtained spleen and lymph node samples from both Dhhc2^Flox/Flox and Dhhc2-T cell KO mice. Flow cytometric analysis revealed that the loss of Dhhc2 in T cells significantly decreased the intranuclear Foxp3 level in Tregs both in the lymph node and in the spleen (Fig. 1G, H). Given the immunosuppressive function of Tregs in vivo, previous studies established a positive correlation between the Foxp3 expression level and the inhibitory function of Tregs [8, 14]. Hence, we aimed to understand the impact of Dhhc2 deficiency-induced decrease in Foxp3 expression in Tregs on the activation of CD4⁺ and CD8⁺ T cells. The results of flow cytometric analysis indicated that the absence of Dhhc2 significantly increased the proportions of activated CD4⁺ and CD8⁺ T cells (CD62L^low CD44^high) in the mouse spleen (Supplementary Fig. S8). These findings led us to hypothesize that the palmitoylation of Foxp3 may affect the function of Tregs and antitumor T-cell responses.

We assessed the influence of Dhhc2 on the function of Tregs in the TME by subcutaneously injecting mice with YUMM3.3 melanoma cells. Notably, mice with T-cell-specific loss of Dhhc2 exhibited suppressed tumor growth (Fig. 1I). By Day 23, the tumor size in these mice was approximately half that in the control mice (Supplementary Fig. S9A, B). Moreover, analysis of immune cell infiltration within tumors revealed that Dhhc2 deficiency led to a significant decrease in intranuclear Foxp3 expression in tumor-infiltrating Tregs from Dhhc2-T cell KO mice, as shown by the MFI (Fig. 1J). Previous studies demonstrated that tumor-infiltrating Tregs promote tumor growth and development by suppressing the function of effector T cells, including CD8⁺ T cells [15]. Therefore, we examined IFNγ secretion from tumor-infiltrating CD8⁺ T cells. Interestingly, we observed a significant increase in the secretion of IFNγ from tumor-infiltrating CD8⁺ T cells in Dhhc2-T cell KO mice compared to Dhhc2^Flox/Flox control mice (Supplementary Fig. S9C–E).

In summary, our study revealed for the first time that Foxp3 is palmitoylated and that its palmitoylation is catalyzed by multiple members of the DHHC family of palmitoyltransferases. Additionally, the loss of DHHC proteins, which regulate the palmitoylation of Foxp3, significantly diminishes the intranuclear expression of Foxp3 in peripheral immune organs and tumor-infiltrating Tregs, ultimately suppressing the functions of Tregs within the TME.

Author contributions

BZ, EK, and YL designed the experiments. BZ, MZ, HM, YW, JQ, and YL performed the experiments and collected the data. LL, TL, LZ, YG, and RH assisted in the experiments. BZ analyzed the data and wrote the original manuscript. YL revised the manuscript.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41423-024-01166-6.