BATF3 programs CD8+ T cell memory

Ataide, Marco A.; Komander, Karl; Knöpper, Konrad; Peters, Annika E.; Wu, Hao; Eickhoff, Sarah; Gogishvili, Tea; Weber, Justus; Grafen, Anika; Kallies, Axel; Garbi, Natalio; Einsele, Hermann; Hudecek, Michael; Gasteiger, Georg; Hölzel, Michael; Vaeth, Martin; Kastenmüller, Wolfgang

doi:10.1038/s41590-020-0786-2

Download PDF

Article
Published: 28 September 2020

BATF3 programs CD8⁺ T cell memory

Nature Immunology volume 21, pages 1397–1407 (2020)Cite this article

13k Accesses
48 Citations
132 Altmetric
Metrics details

Subjects

Abstract

Antiviral CD8⁺ T cell responses are characterized by an initial activation/priming of T lymphocytes followed by a massive proliferation, subset differentiation, population contraction and the development of a stable memory pool. The transcription factor BATF3 has been shown to play a central role in the development of conventional dendritic cells, which in turn are critical for optimal priming of CD8⁺ T cells. Here we show that BATF3 was expressed transiently within the first days after T cell priming and had long-lasting T cell–intrinsic effects. T cells that lacked Batf3 showed normal expansion and differentiation, yet succumbed to an aggravated contraction and had a diminished memory response. Vice versa, BATF3 overexpression in CD8⁺ T cells promoted their survival and transition to memory. Mechanistically, BATF3 regulated T cell apoptosis and longevity via the proapoptotic factor BIM. By programing CD8⁺ T cell survival and memory, BATF3 is a promising molecule to optimize adoptive T cell therapy in patients.

Main

Upon encounter with dendritic cells (DCs) in the context of infections, naive antigen-specific CD8⁺ T cells undergo initial activation (priming) and consecutively, a massive expansion to generate the required effectors-to-target ratio to effectively eliminate infected cells in the body. At the peak of the response, expanded CD8⁺ T cells comprise a spectrum of differentiation states ranging from short-lived effector cells to long-lived memory cells¹. Notably, the quality of the developing memory T cell response is already determined during the first days of activation, among other factors by cognate CD4⁺ T cells², which transmit their signals via a subset of conventional DCs (cDC1)^3,4. Revealing the underlying programs and the transcriptional elements that promote CD8⁺ T cell fitness, functionality and memory, ultimately determines the efficacy of successful immunotherapy in humans.

The family of AP-1 (activator protein-1) transcription factors comprises numerous basic region-leucine zipper (bZIP) proteins that belong to different families (JUN, FOS, ATF, basic leucine zipper transcriptional factor ATF-like (BATF) and MAF) and form heterodimers to execute transcriptional activity. AP-1 transcription factors play a central role in every tissue and cell type and regulate cell death, survival and proliferation⁵. The overall regulation of AP-1 activity and its molecular elements is highly complex and involves transcriptional coactivators, modulation by MAP kinases and transcriptional inhibitors such as the BATF family of proteins⁶. BATF proteins play important roles in the development and function of both myeloid and lymphoid cell populations⁷. They act as negative regulators of the AP-1 complex, but also promote transcriptional activity by interacting with interferon (IFN) regulatory factor (IRF) family members⁸. BATF3 was first identified in human T cells and was later found to play a critical role for the development of the cDC1 subset of conventional DCs^9,10,11,12. Batf3-deficient animals lack CD103-expressing migratory DCs and a varying fraction of splenic CD8α DCs that are dependent on the genetic background (129 versus C57BL/6), the commensal communities and the infection history¹³. Thus, Batf3-deficient mice became a key model for elucidating the in vivo functions of cDC1.

In murine lymphoid cells, Batf rather than Batf3 plays a central role in their function, development and differentiation. In particular, BATF regulates various CD4⁺ helper T cell subsets such as interleukin (IL)-17-expressing helper T cells, follicular helper T cells and T regulatory type 1 cells^14,15,16,17. In cytotoxic CD8⁺ T cells, BATF is required for their expansion and differentiation, but during chronic antigen exposure BATF also drives functional exhaustion^18,19,20. Mechanistically, BATF functions within the AP-1 complex, interacts with cJUN and executes transcriptional activity together with IRF4, leading to reduced anabolic metabolism and a lowered mitochondrial membrane potential. Thereby, BATF regulates a large network of key transcription factors for CD8⁺ T cell differentiation, epigenetic modification, T cell metabolism and survival^21,22.

In this study we identified a specific and CD8⁺ T cell–intrinsic role for BATF3. In contrast to Batf-deficiency, loss of Batf3 did not impact CD8⁺ T cell differentiation or their capacity to produce inflammatory cytokines. Instead, Batf3-deficient CD8⁺ T cells showed a specific defect in T cell memory formation following viral or bacterial infections. The underlying molecular mechanism was due to aggravated apoptosis caused by a dysregulated expression of the proapoptotic factor BIM (Bcl2l11) in Batf3-deficient T cells. Conversely, overexpression of BATF3 augmented CD8⁺ T cell metabolic fitness, inhibited cell death via suppression of BIM and promoted their memory development. By increasing the lifespan of both murine and human CD8⁺ T cells, BATF3 is a promising candidate to enhance adoptive immunotherapy against cancer and infections.

Results

BATF3 plays a CD8⁺ T cell–intrinsic function

Batf3-deficient mice are widely used to investigate the function of cDC1. However, cDC2 and lymphocytes also express BATF3. Therefore, we speculated that the overall effects seen in Batf3-deficient mice in various infection and tumor models might reflect a combined effect involving both DCs and CD8⁺ T cells. To test this hypothesis, we first infected wild-type or Batf3^–/– mice with Vaccinia virus (VV) intraperitoneally (i.p.) and analyzed the CD8⁺ T cell response directed against the immunodominant epitope B8R₂₀ using flow cytometry on day 8 (d8) after primary infection and day 6 (d6) after recall infection²³. There was no significant difference in the B8R-specific CD8⁺ T cell response in Batf3^–/– as compared to wild-type mice (Fig. 1a,b). By contrast, after re-challenging mice with VV on day 30 (d30) after primary infection we found a striking difference when comparing the B8R-specific immune response between wild-type and Batf3^–/– animals (Fig. 1a,b). In line with the critical function of cDC1 to mediate CD4⁺ helper T cell signals, Batf3^–/– mice showed a substantially impaired CD8⁺ T cell recall response^3,4. To test an additional T cell–intrinsic function of BATF3 we first established its expression pattern in CD8⁺ T cells, which peaked shortly after activation (48 h) (Fig. 1c). Next, we generated wild-type:Batf3^–/– mixed bone marrow (BM) chimeric mice and infected them with VV (Extended Data Fig. 1). Probing B8R-specific CD8⁺ T cells in the blood, we detected a similar contribution of wild-type and Batf3^–/– cells at the primary response against VV on d8 post-infection (p.i.) (Fig. 1d,e). By contrast, after recall infection B8R-specific CD8⁺ T cells originating from wild-type cells dominated the overall immune response (Fig. 1d,e). These results pointed to a T cell–intrinsic function of Batf3 that seemed critical for optimal CD8⁺ T cell memory development or recall capacity. To further characterize the T cell–intrinsic role of BATF3, we crossed Batf3^–/– mice with T cell antigen receptor (TCR)-transgenic OT-I mice to perform competitive (50/50) transfer experiments with wild-type and Batf3^–/– OT-I mice, allowing for a side-by-side, quantitative and qualitative comparative analysis over time. Recipient animals were either infected with VV-OVA or L.m.-OVA (Listeria monocytogenes) and analyzed on d8 p.i. (prime) or were rechallenged heterologously on d30 and analyzed on d6 post-secondary infection (recall). In line with our initial results, we detected similar amounts of wild-type and Batf3^–/– OT-I T cells in the spleen on d8 after either VV-OVA or L.m.-OVA primary infection (Fig. 1f,g). By contrast, during recall responses Batf3^–/– OT-I T cells were largely outcompeted by their wild-type counterparts (Fig. 1f,g). Despite these quantitative differences, the remaining Batf3^–/– OT-I T cells had a similar capacity to produce effector cytokines, such as IL-2, tumor necrosis factor (TNF) and IFN-γ, compared to wild-type cells (Fig. 1h,i). Likewise, CD8⁺ T cell differentiation based on KLRG1/CD127 expression pattern on d8 after primary infection did not reveal any differences between wild-type and knockout OT-I T cells (Fig. 1j). Thus, Batf3^–/– CD8⁺ T cells have a normal phenotypical and functional differentiation pattern, despite a defective memory formation.

**Fig. 1: BATF3 has a CD8⁺ T cell–intrinsic function.**

Batf3-deficiency impacts on CD8⁺ T cell survival and recall capacity

Having established that Batf3^–/– CD8⁺ T cells failed to develop a proper memory response, we next addressed whether this phenotype was based on an impaired transition to memory during the contraction phase or due to a specific defect during recall responses or both. To this end, we performed a kinetic analysis of wild-type and Batf3^–/– OT-I T cells in the blood and the spleens of L.m.- and VV-infected animals (Fig. 2a,b and Extended Data Fig. 2). Notably, we detected an aggravated contraction of Batf3^–/– as compared to wild-type OT-I T cells, resulting in significantly reduced numbers by d30 p.i. Next, we analyzed the expression of CX3CR1 and IL-7R on d30 p.i. to discern central memory, effector memory (T_EM) and the recently described peripheral memory CD8⁺ T cell subsets^24,25. We found that all memory subsets were reduced in absolute numbers in Batf3^–/– as compared to wild-type OT-I T cells. Notably, this effect was most pronounced within CX3CR1⁺ T_EM cells as indicated by the additional relative loss of this subset (Fig. 2c,d). To complete the analyses on the memory T cell compartment, we generated tissue-resident memory (T_RM) CD8⁺ T cells by intradermal infection of the ear. On d40 after infection, we observed a pronounced reduction of Batf3^–/– versus wild-type OT-I cells using either two-photon in situ imaging or FACS analyses of tissue homogenates of the skin (Fig. 2e,f). Based on these results we conclude that all memory CD8⁺ T cell subsets are affected by the loss of Batf3.

**Fig. 2: BATF3 regulates memory transition and survival of CD8⁺ T cells.**

To test the recall capacity of Batf3^–/– OT-I memory T cells, we isolated splenic memory populations (wild-type and Batf3^–/–) and transferred equal numbers into naive wild-type recipients and infected them with VV. On d6 after infection we detected a small but important reduction of Batf3^–/– as compared to wild-type OT-I T cells in the spleens (Fig. 2g). Thus, Batf3 is critical for both the transition to and the quality of memory CD8⁺ T cells.

Batf3-deficiency impacts on CD8⁺ T cell fitness

Since CX3CR1-expressing T_EM cells were severely affected by the loss of Batf3, we further analyzed CX3CR1 expression on wild-type and Batf3^–/– OT-I T cells between d8 and d15 after infection, a time-frame during which we had observed an enhanced contraction of Batf3^–/– T cells. Consistently with our initial results, we found the most aggravated loss of Batf3^–/– OT-I T cells within the CX3CR1-expressing fraction (Fig. 3a,b). The overall pronounced loss of Batf3^–/– OT-I T cells may be based on a combination of increased cell death, reduced proliferation rates and subset trans-differentiation. Indicative for reduced proliferation rates, we found a significantly smaller fraction of Ki67-positive cells within Batf3^–/– versus wild-type OT-I T cells at d12 after infection (Fig. 3c). As apoptotic cells are rapidly removed in vivo, we decided to measure the mitochondrial membrane potential as a proxy for cellular fitness. To this end, we applied two different dyes, a strict staining protocol and parallel measurement of wild-type versus Batf3^–/– OT-I T cells within the same sample on d8 and d12 after infection. We then detected no differences on d8 (Fig. 3d and Extended Data Fig. 3), while on d12 p.i. we found a significantly higher fraction of tetramethyl rhodamine ethyl ester (TMRE) (Fig. 3e) and MitoTracker Deep Red (Extended Data Fig. 3) low cells in Batf3^–/– compared to wild-type OT-I T cells. This difference was particularly pronounced within CX3CR1-expressing OT-I T cells in line with the most aggravated loss of this subset by d30 after infection. Together, these results indicate that Batf3-deficient CD8⁺ T cells show reduced proliferation rates and an elevated cell death, which is accompanied by impaired mitochondrial functionality.

**Fig. 3: The lack of BATF3 impacts CD8⁺ T cell fitness.**

BATF3 programs T cell survival via the proapoptotic factor BIM

The critical role of IL-15 to regulate T cell contraction and memory T cell homeostasis is well established^26,27.To address whether IL-15 signaling is involved in the aggravated contraction seem in Batf3^–/– CD8⁺ T cells we co-transferred wild-type versus Batf3^–/– OT-I T cells into Il15^–/– mice and infected them with VV-OVA. In Il15^–/– mice, we detected an aggravated contraction in Batf3^–/– CD8⁺ T cells compared to wild-type cells (Extended Data Fig. 4a), which argues against a critical role for IL-15 signals as an underlying mechanism.

To gain mechanistic insights we first decided to characterize the expression pattern of Batf3 over time using ex vivo isolated TCR-transgenic T cells after VV infection as a basis for consecutive RNA-seq studies. Reflecting our in vitro data, Batf3 expression peaked early after T cell activation, but was absent during the effector or memory phase (Fig. 4a,b). Notably, this pattern was similar to the expression dynamics of transcripts of the related transcription factor Batf (Fig. 4b and Extended Data Fig. 4b). Next, we performed an RNA-seq analysis comparing wild-type versus Batf3^–/– OT-I T cells 48 h and 72 h post-in vitro activation (Extended Data Fig. 4c and Supplementary Table 1). Unsupervised gene set enrichment analysis revealed an involvement of mitochondrial pathways of apoptosis, such as ‘mitochondrial pathway of apoptosis: caspases’ and ‘mitochondrial pathway of apoptosis: Bcl-2 family’ (Fig. 4c). BIM (Bcl2l11) is a mitochondria-associated protein of the Bcl-2 family and a well-known proapoptotic factor that regulates CD8⁺ T cell contraction. We therefore investigated the protein levels of BIM in wild-type and Batf3^–/– OT-I T cells over time. Batf3^–/– and wild-type naive CD8⁺ T cells expressed similar levels of BIM. However, during the course of infection Batf3^–/– CD8⁺ T cells expressed higher levels as compared to wild-type CD8⁺ T cells (Fig. 4d). Based on this correlation, we asked whether BIM is functionally linked to the pronounced contraction seen in Batf3^–/– CD8⁺ T cells. Using short hairpin RNA (shRNA)-mediated knockdown we aimed to rescue the pronounced contraction of Batf3-deficient T cells in vivo by normalizing BIM gene expression to that seen in wild-type T cells (Fig. 4e and Extended Data Fig. 4d). To this end, we transduced wild-type with control shRNA and Batf3^–/– OT-I T cells with control shRNA or shRNA targeting BIM, transferred them into mice at a 50/50 ratio. As expected, on d7 after infection we found tenfold more wild-type than Batf3^–/– OT-I T cells that were transduced with control shRNA (Fig. 4f). By contrast Batf3^–/– OT-I T cells transduced with retrovirus expressing shRNA against BIM outcompeted wild-type OT-I T cells that expressed control shRNA (Fig. 4g). Therefore, we conclude that the pronounced T cell contraction and impaired memory CD8⁺ T cell formation observed in the absence of BATF3 is at least in part based on a dysregulation of BIM.

**Fig. 4: BATF3 regulates CD8⁺ T cell survival via the BIM.**

BATF3 overexpression promotes CD8⁺ T cell survival

Having established how Batf3 deficiency leads to impaired CD8⁺ T cell memory we asked whether gain-of-function of BATF3 could further improve CD8⁺ T cell memory generation. To this end, we transduced naive CD8⁺ T cells retrovirally with a BATF3-overexpressing vector (pMIG expressing BATF3–IRES–GFP) or an empty control plasmid (pMIG expressing only ametrine) (Extended Data Fig. 5). When culturing these cells at a 50/50 ratio in vitro, we found that BATF3- (Fig. 5a), but not BATF-overexpressing CD8⁺ T cells almost completely outcompeted their empty-vector controls over time after re-stimulation with anti-CD3/CD28 at different concentrations of IL-15 (Extended Data Fig. 6a,b). Additionally, BATF3 overexpression rescued re-stimulated Batf3-deficient CD8⁺ T cells in vitro (Extended Data Fig. 6c). In line with our previous results, we detected significantly lower levels of the proapoptotic molecule BIM within BATF3-overexpressing cells, suggesting resistance to apoptosis (Fig. 5b). Similar to Batf3-deficient cells, BATF3-overexpression did not perturb the production of cytokines, such as IFN-γ, TNF and IL-2 (Fig. 5c). Next we wanted to address whether BATF3-overexpression also endows CD8⁺ T cells with a superior capacity to survive in vivo. Therefore, we co-transferred BATF3-overexpressing and empty-vector control OT-I T cells into naive mice and infected them with VV-OVA. By d7 p.i., BATF3-overexpressing GFP⁺ OT-I cells outcompeted their ametrine⁺ OT-I control cells in the blood (Fig. 5d). After 31 d, the memory pool of OT-I cells was almost exclusively composed of BATF3-overexpressing T cells in both blood and spleens of previously infected mice (Fig. 5d,e). After re-stimulating splenocytes ex vivo, we found that the BATF3-overexpressing OT-I cells had a slightly higher capacity to produce IFN-γ and a trend to produce more TNF than their empty-vector controls (Fig. 5f). Together these results show that BATF3 overexpression reduces BIM expression and promotes CD8⁺ T cell function and survival in vitro and in vivo.

**Fig. 5: BATF3 gain-of-function promotes CD8⁺ T cell survival in vitro and in vivo.**

BATF3 overexpression promotes CD8⁺ T cell memory

Having established that both loss- and gain-of-function experiments indicate a critical role of BATF3 in regulating CD8⁺ T cell fitness and memory, we next applied RNA-seq to gain an unbiased view on the molecular changes induced by BATF3. Specifically, we compared the transcriptome of resting or anti-CD3/CD28 re-stimulated OT-I T cells that were transduced with either BATF3-overexpressing or control vectors. We detected more than 180 differentially regulated genes (Fig. 6a and Supplementary Table 1). Some of the observed upregulated genes play a particular role during transition and/or maintenance of memory CD8⁺ T cells. Among them were genes such as Il7r, Sell and Il21 with known functions in lymphocyte survival, recirculation capacity and metabolism. Additionally, we detected upregulation of transcriptions factors (Runx1, Prdm1), co-stimulatory molecules (Cd40lg, Ctla4, Icos) and effector molecules (Gzma). Among the downregulated genes we found anti-inflammatory genes (Il10, Il24) and specific proteases like Gzmd and Gzmg. In summary, BATF3 overexpression upregulates multiple genes that are connected to CD8⁺ T cell function and memory, whereas it downregulates anti-inflammatory mediators.

**Fig. 6: BATF3 gain-of-function promotes CD8⁺ T cell fitness.**

Next we wished to confirm the detected messenger RNA differences on a protein level for some selected genes with a known function in T cell biology. Using FACS analysis, we found a significantly higher mean fluorescence intensity and a larger fraction of cells that expressed IL-7R within the BATF3-overexpressing versus empty vector transduced CD8⁺ T cells at resting and d4 time points in vitro (Fig. 6b and Extended Data Fig. 7a). Similar results were obtained for inducible co-stimulator (ICOS), which was highly upregulated within BATF3-overexpressing cells upon re-stimulation (Fig. 6c and Extended Data Fig. 7b). With regard to co-inhibitory molecules we confirmed higher intracellular levels of CTLA-4 (Fig. 6d and Extended Data Fig. 7c), whereas Tim-3 was significantly downregulated upon cell activation in BATF3-overexpressing as compared to control cells (Fig. 6e and Extended Data Fig. 7d).

CD8⁺ T cell longevity and memory is further characterized by an altered metabolism compared to (short-lived) effector T cells²⁸. Quiescent and memory T cells utilize preferentially oxidative mitochondrial metabolism and fatty acid oxidation and show relatively little glycolytic activity to fuel their bioenergetic demand^29,30. To address whether BATF3 impacts on the metabolism of CD8⁺ T cells, we analyzed aerobic glycolysis and oxidative phosphorylation (OXPHOS) in BATF3 overexpression and control cells before and after anti-CD3/CD28 re-stimulation using a Seahorse flux analyzer (Fig. 6f–i and Extended Data Fig. 7e,f). The glycolytic efflux rate (glycoPER) and the oxygen consumption rate (OCR) increased after re-stimulation with anti-CD3/CD28 in both BATF3 expressing and control cells (Fig. 6f,g and Extended Data Fig. 7e,f). Under basal conditions (before oligomycin treatment) we noticed a small but reproducible increase of OCR in BATF3 transduced CD8⁺ T cells compared to control cells (Fig. 6g), suggesting that BATF3 promoted elevated mitochondrial metabolism. Notably, we observed significantly higher OCR:glycoPER ratios in BATF3-overexpressing T cells compared to control cells both under resting conditions (Fig. 6h) and after re-stimulation with anti-CD3/CD28 (Fig. 6i), demonstrating that BATF3 promotes a metabolic program that is highly reminiscent of memory CD8⁺ T cells³⁰.

BATF3 is a potential target for immunotherapy in humans

The transcriptional changes induced by ectopic BATF3 expression in CD8⁺ T cells and the enhanced in vivo persistence of adoptively transferred CD8⁺ T cells overexpressing BATF3, suggested a possible therapeutic application. To move toward this goal, we first wanted to address whether BATF3-overexpressing CD8⁺ T cells also have a competitive advantage in an environment with persistent antigen exposure (a situation typically seen in adoptive T cell therapy approaches). Therefore, we infected wild-type mice with lymphocytic choriomeningitis virus (LCMV) clone 13 (Cl13) and 30 d later when chronicity was fully established we co-transferred LCMV-specific TCR-transgenic CD8⁺ T cells (P14) transduced with BATF3 or control vector. As anticipated, we detected significantly more BATF3-overexpressing cells compared to control cells 1 week after transfer in the spleens of chronically infected mice (Fig. 7a). Both transferred cell populations were similarly able to produce inflammatory cytokines, yet the absolute number of IFN-γ-producing cells was significantly higher within the BATF3-overexpressing population (Fig. 7b).

So far, we have shown that BATF3 promotes cellular fitness and memory formation in murine models after viral infection, but whether BATF3 has a similar role in human CD8⁺ T cells and is therefore a potential target for immunotherapy is not known. Therefore, we transduced human CD8⁺ T cells from healthy donors with BATF3 and empty-vector expression plasmids similar to our experiments with murine T cells. FACS-sorted BATF3-overexpressing and control CD8⁺ T cells were co-cultured and re-stimulated with anti-CD3/CD28 beads for up to 7 d. Similarly to murine CD8⁺ T cells, human CD8⁺ T cells overexpressing BATF3 also outcompeted their control cells over time (Fig. 7c) demonstrating that BATF3 has a conserved role across species to promote T cell longevity. When analyzing cytokine production, we detected a small but consistent increase in the capacity of BATF3-overexpressing human CD8⁺ T cells to produce inflammatory cytokines such as IFN-γ, TNF and IL-2 (Fig. 7d). From these data we conclude that BATF3 overexpression leads to increased amounts of human CD8⁺ T cells in vitro upon activation, while fully preserving their functional capacity to produce cytokines. Together these results warrant further investigation of the therapeutic potential of BATF3 for adoptive immunotherapy in humans.

Discussion

Here, we identified a physiological, cell-intrinsic function of BATF3 in regulating CD8⁺ T cell survival, proliferation and transition to memory. Batf3^–/– mice are a widely used model to study the role of cDC1 in the context of tumors and infections. Our study does not question the well-established role of cDC1 in cross-presenting antigens, driving antiviral immunity and serving as a critical platform to mediate CD4⁺ helper T cell signals. However, results obtained with Batf3-deficient mice should be carefully interpreted under the light of our work. Future studies aiming to address the role of cDC1 should apply cell-specific models that are based on, for example, Xcr1 (refs. ^31,32,33,34).

In our study we found that in CD8⁺ T cells, BATF3 is transiently expressed early after T cell activation but has long-lasting effects on T cell contraction, cellular/mitochondrial fitness and longevity, and therefore on CD8⁺ T cell memory. Both intrinsic and extrinsic factors regulate T cell contraction and longevity³⁵. Extrinsic factors are homeostatic cytokines such as IL-7 and IL-15 or inflammatory cytokines such as IL-12, type I IFNs and IFN-γ. These factors do not seem to be involved in regulating the aggravated contraction seen in the context of Batf3 deficiency. Well-established intrinsic factors of T cell contraction are the transcription factors T-bet, eomesodermin, Id2 and Bcl-6. While these factors share a function with BATF3 to regulate CD8⁺ T cell survival, they differ with regard to their impact on T cell differentiation. Transcription factors that regulate the balance of effector (KLRG1^hi CD127^lo) versus memory T cell (KLRG1^lo CD127^hi) development naturally impact on the overall size of the long-term T cell response. By contrast, BATF3 deficiency did not alter CD8⁺ T cell differentiation but regulated the survival of the entire CD8⁺ T cell population, notably most pronounced in CX3CR1-expressing T_EM cells. On a mechanistic level, we identified BIM downstream of BAFT3 as central regulator of apoptosis and T cell contraction^36,37. BATF3 has been shown to bind to the BIM promoter region in cDC1 and as a transcriptional repressor, it could negatively regulate BIM expression³⁸. Consistently, we were able to rescue Batf3 deficiency in vivo by knockdown of BIM. As BIM is dynamically regulated over time it is currently not possible to fully normalize BIM in Batf3-deficient OT-I T cells to physiological levels seen in wild-type OT-I T cells. Therefore, the fact that Batf3-deficient CD8⁺ T cells with a knockdown of BIM outcompeted wild-type CD8⁺ T cells in vivo likely reflects a reduction of BIM below wild-type levels. BIM expression is regulated on multiple levels and is influenced by a whole range of transcription factors and epigenetic modifiers^5,39,40. In vivo the dysregulation of BIM in Batf3-deficient CD8⁺ T cells becomes apparent after the physiological expression of BATF3 in wild-type CD8⁺ T cells has ceased. This argues for epigenetic changes that are imprinted by BATF3 within the first days of T cell activation with long-lasting effects on cell survival and proliferation, which are indeed key functions of AP-1 (refs. ^40,41). Alternatively, there may be temporal spikes in BATF3 expression and/or within a fraction of contracting antigen-specific CD8⁺ T cells that are below our detection limit.

Our study has revealed a physiological function of BATF3 in CD8⁺ T cells by regulating memory development and long-term survival. Vice versa, we were able to demonstrate that BATF3 overexpression further enhanced T cell persistence in vivo, making it a suitable target to optimize adoptive T cell therapy approaches. As seen in Batf3-deficient CD8⁺ T cells, BATF3-overexpression also induced complex context-dependent and time-dependent transcriptional changes.

A recent study that screened for factors that regulate T cell exhaustion in the context of cancer identified REGNASE-1 (ref. ⁴²), Regnase-1-deficient CD8⁺ T cells had an enhanced capacity to eliminate tumor cells in vivo. Notably, the authors showed that Regnase-1 deficiency increased the expression of the BATF and found that BATF overexpression induced metabolic changes and increased the number of tumor-infiltrating CD8⁺ T cells in vivo. These data are reminiscent with the impact of BATF3 overexpression on CD8⁺ T cell abundance observed in our study. Yet, BATF in contrast to BATF3 overexpression did not provide a competitive advantage in CD8⁺ T cells in vitro in our hands, underscoring nonoverlapping functions between these two related molecules.

Another study that investigated the molecular basis of exhaustion of chimeric antigen receptor (CAR) T cells identified a dysbalance in the AP-1-IRF network as an underlying mechanism of exhaustion⁴³. The authors found that the overexpression of cJUN could rebalance these factors likely by forming additional cJUN/FOS dimers. As a consequence, cJUN-overexpressing CAR T cells remained functional and acquired a resistance toward exhaustion. Overall the AP-1-IRF family network seems to be a suitable target to reinvigorate exhausted CD8⁺ T cells. The fact that BATF family members cooperate with IRFs (transcriptional activation) and multiple other AP-1 family members such as cJUN (transcriptional repression) and that all these factors are dynamically regulated during T cell activation, poses a significant challenge in delineating how quantitative changes of these molecules contribute to the overall outcome of the network and ultimately CD8⁺ T cell fitness.

A general caveat of modulating the AP-1-IRF network is the potential induction of malignant transformation⁴⁴. Indeed, transgenic mice overexpressing human BATF showed the development of lymphoproliferative disease²¹. Similarly, overexpression of human BATF3 in murine lymphocytes causes B cell lymphomas in mice⁴⁵. While this underscores the need to introduce a safety switch in genetically modified adoptively transferred T cells, BATF3 overexpression in CD8⁺ T cells seems to be an interesting approach to enhance adoptive cell-based therapy, such as CAR T cells. This strategy could be particularly helpful in the context of immunotherapy against solid tumors where high amounts of tumor-reactive cytotoxic T cells are required.

In summary, we identify a physiological role of BATF3 in regulating CD8⁺ T cell memory by modulating the proapoptotic factor BIM. We further demonstrate that this function of BATF3 can be harnessed to improve T cell survival and longevity in vivo. Given its function in promoting CD8⁺ T cell memory in both murine and human CD8⁺ T cells, BATF3 is a highly suitable candidate for the development of optimized adoptive T cell therapies against cancer.

Methods

Animals

WT C57BL/6J and TCR-transgenic OT-I and Batf3^–/– mice¹² were originally purchased from Jackson or Janvier Laboratories. P14 and XCR1-DTR-Venus mice³¹ were provided by H.C. Probst and T. Kaisho, respectively. All mice used in that study were female and maintained in specific-pathogen-free conditions at an Association for Assessment and Accreditation of Laboratory Animal Care-accredited animal facility. All procedures were approved by the North Rhine-Westphalia State Environment Agency and/or the Lower Franconia Government.

Generation of mixed bone marrow chimeric mice

The isolated bone marrow cells were mixed in a 1:1 ratio and a total of 2 × 10⁶ cells of each type (WT and Batf3^–/–) and transferred i.v. into the 9 Gy irradiated WT mice. Chimerism was confirmed 8 weeks after transplantation.

Treatment of mice

For LCMV clone 13 infection, C57BL/6J mice were i.p. injected with 300 μg of anti-CD4 clone GK1.5 (BioXCell) twice before infection (d−2 and d0).

Infections

A total of 10⁶ plaque-forming units (p.f.u.) VV-OVA, and 10⁵ colony-forming units (CFU) LM-OVA for primary and 10⁴ for secondary infection were diluted in PBS and injected intraperitoneally, i.p., and intravenously, i.v., respectively. Then, 2 × 10⁶ LCMV clone 13 were diluted in PBS and injected i.v. Before LCMV infection, mice were i.p. injected twice with anti-CD4 (300 μg) with 1 d of interval.

Adoptive naive T cell transfer

WT C57BL/6J and Batf3^–/– naive OT-I CD8⁺ T cells were sorted using a MACS CD8⁺ T cell negative selection kit (Miltenyi) combined with biotinylated anti-CD44 (IM7, BD Biosciences). A total of 5 × 10³ cells were transferred i.v.

Retroviral vectors and shRNAs

pMIG–Batf3–IRES–GFP was applied for overexpression of BATF3. Batf3 was subcloned from a synthetic gene fragment plasmid (courtesy of Invitrogen GeneArt) into a pRP backbone. From the resulting pRP–Batf3 plasmid, Batf3 was cut using single cutting restriction enzymes BglII and XhoI and ligated into pMIG, generating pMIG-Batf3-IRES-GFP. The pMIG was a gift from W. Hahn (Addgene plasmid 9044; RRID: Addgene_9904). pMIG–Batf–IRES–GFP was applied for overexpression of BATF. Batf was subcloned from pCMV6–Entry–Batf (MR222114, OriGene) by PCR and ligated into pMIG opened with restriction enzymes XhoI and EcoRI. MigR1–mAmetrine was used as an empty-vector control. For the knockdown of BIM, shRNAs were used. The shRNAs for Bcl2l11 as well as a nontargeted control shRNA were expressed by the pLMPd–ametrine backbone (TRMSU2002, TransOmics). The applied shRNAs had the following sequences:

shRNA 5′-TGCTGTTGACAGTGAGCGAAAGCAACCTTCTGATGTAAGATAGTGAAGCCACAGATGTATCTTACATCAGAAGGTTGCTTGTGCCTACTGCCTCGGA-3′.

Retroviral transduction, adoptive transfer of transduced OT-I and P14 T cells and in vitro co-cultures of murine CD8⁺ T cells

Total CD8⁺ T cells were isolated using a mouse CD8⁺ T Cell Negative Enrichment kit (STEMCELL Technologies) and stimulated with 1 μg ml⁻¹ plate-bound anti-CD3e (clone 2C11) plus 1 μg ml⁻¹ anti-CD28 (clone 37.51, both BioXCell) at a density of 2 × 10⁶ cells ml⁻¹ in 12-well plates. T cells were transduced 24 h after stimulation by spin infection (2,500 r.p.m., 32 °C, 90 min) in the presence of frozen retroviral supernatant with 10 μg ml⁻¹ polybrene (SantaCruz Biotechnologies). Retroviral supernatant was produced in the Platinum-E retroviral packaging cell line. Platinum-E cells were transfected by lipofection using GeneJet (Signagen) with various retroviral expression plasmids and the amphotrophic packaging vector pCL-10A1 (Addgene). The supernatant was collected 2 d and 3 d after transfection, filtered and stored in aliquots at −80 °C. After spin infection, viral supernatant was removed from the CD8⁺ T cells and replaced by fresh medium. At 24 h after transduction, T cells were transferred into six-well plates with fresh medium containing 50 ng ml⁻¹ IL-15 and maintained for 2 d at a density of 0.5 × 10⁶cells ml⁻¹. Transduced cells were FACS-sorted using a sterile FACSAria III cell sorter (BD Biosciences) and cultivated for additional 48 h in fresh medium containing 50 ng ml⁻¹ IL-15 at a density of 0.5 × 10⁶ cells ml⁻¹. At d6 of culture, pMIG–ametrine, pMIG–BATF3–GFP, pMIG–BATF–GFP or pLMPd–shRNA–ametrine transduced memory CD8⁺ T cells were mixed at a 1:1 ratio and re-stimulated with 1 μg ml⁻¹ plate-bound anti-CD3ε (clone 2C11) plus 1 μg ml⁻¹ anti-CD28 (clone 37.51, both BioXCell) for the in vitro experiments or transferred (P14 tdTomato cells) i.v. into LCMV-infected C57BL/6J host mice. Transduced OT-I CD45.1 were transferred into naive C57BL/6J mice followed by VV-OVA i.p. infection. A total of 1.5 × 10⁶ transduced CD8⁺ T cells (50/50; empty vector to BATF3) were transferred into the hosts in both, P14 and OT-I adoptive transfer experimental setup.

Isolation and cultivation of primary human samples

Blood samples were obtained from healthy donors after provided written informed consent to participate in research protocols approved by the Institutional Review Board of the University Hospital of Würzburg. Peripheral blood mononuclear cells were isolated by Biocoll centrifugation (Biochrom, Merck Millipore). Bulk CD8⁺ T cells were isolated by untouched magnetic cell separation (Miltenyi Biotec). Primary human T cells were cultured in RPMI 1640 (supplemented with 1% GlutaMAX, 50 mM 2-mercaptoethanol, 100 U ml⁻¹ penicillin/streptomycin (all Gibco, Thermo Fisher Scientific) and 10% human serum (Deutsches Rotes Kreuz) containing 50 U ml⁻¹ human IL-2 (Miltenyi Biotec).

Retroviral transduction of human CD8⁺ T cells

Retroviral supernatant was produced as described previously, using the Platinum-A packaging cell line to obtain amphotropic retrovirus (see above). CD8⁺ T cells were isolated and stimulated for 18 h using human T-Activator CD3/CD28 Dynabeads (Thermo Fisher Scientific). On the next day, CD8⁺ T cells were collected, resuspended in retroviral supernatant containing 10 μg ml⁻¹ polybrene (Merck Millipore) and transduced by spin infection (800g, 32 °C, 45 min). At 24 h after transduction, a half medium change was performed. CD8⁺ T cells were maintained in fresh medium at a density of 2 × 10⁶cells ml⁻¹. CD3/CD28 beads were removed 96 h after transduction and transduction efficiencies were determined by flow cytometry using a FACS Canto II (BD Bioscience). On d7, CD8⁺ T cells were sorted for their respective reporter gene expression using a FACSAria III cell sorter (BD Bioscience) and cultivated for another several days in fresh medium containing 50 U ml⁻¹ human IL-2 plus at a density of 0.5 × 10⁶ cells ml⁻¹. Cells were cultured in a 1:1 ratio (empty vector to mBATF3) and resting or with CD3/CD28 Dynabeads.

Murine T cell cultures

CD8⁺ T cells were isolated from single-cell suspensions of spleens and lymph nodes by negative separation using a mouse CD8⁺ T cell Enrichment kit (STEMCELL Technologies). A total of 2 × 10⁶ T cells were stimulated in 2 ml RPMI 1640 medium (supplemented with 10% FBS, 2 mM l-glutamine, 50 mM 2-mercaptoethanol and 100 U ml⁻¹ penicillin/streptomycin; all Gibco) with 1 μg ml⁻¹ plate-bound anti-CD3 (clone 2C11) plus 1 μg ml⁻¹ anti-CD28 (clone 37.51, both BioXCell) in 12-well plates. At 48 h after activation, T cells were transferred into 96-well plates with fresh medium containing 50 ng ml⁻¹ IL-15 (PeproTech) and maintained for several days at a constant density of 0.5 × 10⁶ cells ml⁻¹ with fresh medium and cytokines.

Extracellular flux analysis (Seahorse assay)

OCR and glycoPER were measured using an oxygen-controlled XFe96 Extracellular Flux Analyzer (Seahorse Bioscience). A total of 1.5 × 10⁵ CD8⁺ T cells per well were seeded on Cell-Tak (Corning) in 6–8 replicates in XF medium (Seahorse Biosciences) supplemented with 10 mM glucose (Sigma Aldrich), 2 mM GlutaMAX (Gibco) and 1 mM sodium pyruvate (Corning). The cells were incubated for 1 h in a non-CO₂ incubator at 37 °C before oxygen consumption and extracellular acidification were analyzed under basal conditions and after the following treatments: ATP synthase inhibitor oligomycin (2 μM), 2-deoxy-d-glucose (50 mM) to inhibit glycolysis, the protonophore FCCP (1 μM) to uncouple mitochondria, the mitochondrial complex I inhibitor (rotenone) (500 nM) and the mitochondrial complex III inhibitor antimycin A (500 nM). Basal OCR was calculated by subtracting the OCR after rotenone and antimycin A treatment from the OCR before oligomycin treatment. Maximal OCR was calculated by subtracting the OCR after rotenone and antimycin A treatment from the OCR measured after addition of FCCP. Basal glycoPER was calculated by subtracting the values after 2-DG from the glycoPER values before oligomycin treatment. Maximal glycoPER was calculated by subtracting values after 2-DG from the glcoPER values following oligomycin treatment. Ratios of OCR to glycoPER were calculated under basal and maximal conditions to determine the metabolic phenotypes of CD8⁺ T cells.

Mitochondrial potential measurement by flow cytometry

Cells were stained with MitoTracker Deep Red FM (250 nM) or TMRE (70 nM) (Thermo Fisher Scientific) in an incubator at 37 °C (5% CO₂) for 30 min before cell surface staining. Cells were then washed three times with FACS buffer (PBS 2% FCS) and followed by cell surface staining for further FACS analysis. FCCP (1 μM) treatment was used to disrupt the mitochondrial membrane potential and then serving as negative control.

Flow cytometry (surface and intracellular staining)

Single-cell suspensions obtained from spleens or blood were used for flow cytometry analysis. Cells were surface stained with anti-CD8⁺ (BioLegend, clone 53-6.7), anti-CD44 (BioLegend, clone IM7), anti-CD127 (eBioscience, clone A7R34), anti-KLRG1 (eBioscience, clone 2F1), anti-CX3CR1 (BioLegend, clone SA011F11), anti-ICOS (BD Biosciences, clone C398.4A) and anti-Tim-3 (BioLegend, clone RMT3-23). For intracellular staining, cells were first fixed with IC Fixation Buffer (eBioscience 00-8222-49) for 30 min on ice. Antibodies were diluted in Permeabilization Buffer 1× (eBioscience 00-8333-56), anti-IFN-γ (BioLegend, clone XMG1.2), anti-TNF (BioLegend, clone MD6-XT22), anti-IL-2 (BioLegend, clone JES6-5H4), anti-CTLA-4 (BioLegend, clone UC10-4B9), anti-BIM (Cell Signaling, clone C34C5) plus anti-rabbit (Life Technologies, clone A-31573). For anti-Ki67 staining (BD Biosciences, clone B56), cells were fixed with eBioscience Foxp3/Transcription Factor Staining Buffer (00-5523-00).

Real-time qRT–PCR

RNA was isolated from CD8⁺ T cells (see above) using the RNeasy Micro Kit (QIAGEN) following the manufacturer’s instructions. RNA was converted to complementary DNA using the iScript cDNA Synthesis kit (Bio-Rad 1708891) according to the manufacturer’s instructions. Real-time qRT-PCR was performed on the Bio-Rad CFX Connect Real-Time PCR instrument (Bio-Rad). Figure 1d; BATF3 (catalog no. 4331182, assay ID Mm01318274_m1) and 18S ribosomal RNA (catalog no. 4331182, assay ID Mm03928990_g1) TaqMan Gene Expression Assay. Figure 4a,b and Extended Data Fig. 4b; SYBR Green primers sequences: 18s forward (5′-CGGCGACGACCCATTCGAAC-3′);

18s reverse (5′-GAATCGAACCCTGATTCCCCGT-3′);

Batf3 forward (5′-CCACGAGGAGCACGAGAGC-3′);

Batf3 reverse (5′-GGGGTGTCATCGTGGTAGAC-3′);

Batf forward (5′-CTGGCAAACAGGACTCATCTG-3′);

Batf reverse (5′-GGGTGTCGGCTTTCTGTGTC-3′);

Prf1 forward (5′-AGCACAAGTTCGTGCCAGG-3′);

Prf1 reverse (5′-GCGTCTCTCATTAGGGAGTTTTT-3′).

RNA sequencing

For the analysis of Batf3^–/– cells, enriched CD8⁺ T cells from WT and Batf3^–/– spleens were stimulated with 1 μg ml⁻¹ plate-bound anti-CD3ε (clone 2C11) plus 1 μg ml⁻¹ anti-CD28 Abs (clone 37.51, both BioXCell) in vitro for 48 h and 72 h. Four technical replicates were generated per condition.

For the analysis of Batf3-overexpressing cells, pMIG–ametrine and pMIG–BATF3–GFP transduced memory CD8⁺ T cells were maintained at a 0.5 × 10⁶ cells ml⁻¹ density (see above) for 6 d of culture. Three technical replicates were kept resting or re-stimulated with 1 μg ml⁻¹ plate-bound anti-CD3ε (clone 2C11) plus 1 μg ml⁻¹ anti-CD28 (clone 37.51, both BioXCell) in vitro for 6 h.

Cells were collected and total mRNA was isolated with the RNeasy Mini Kit (QIAGEN) following the manufacturer’s instructions. RNA quality was checked using a 2100 Bioanalyzer with the RNA 6000 Pico kit (Agilent Technologies). The RNA integrity number for all samples was ~7.6 or higher. DNA libraries suitable for sequencing were prepared from 100 ng of total RNA with oligo-dT capture beads for poly-A-mRNA enrichment using the TruSeq Stranded mRNA Library Preparation kit (Illumina) according to the manufacturer’s instructions. After 15 cycles of PCR amplification, the size distribution of the barcoded DNA libraries was estimated at ~320 bp by electrophoresis on Agilent High Sensitivity Bioanalyzer microfluidic chips. For Batf3-overexpressing and according control samples, sequencing of pooled libraries, spiked with 1% PhiX control library, was performed in single-end mode on the NextSeq 500 platform (Illumina) with the High Output kit v.2.5 (75 cycles). Batf3^–/– samples and WT controls were sequenced on the NovaSeq 6000 (Illumina) in paired-end mode with 100-bp strand length.

Demultiplexed FASTQ files were generated with bcl2fastq2 v.2.20.0.422 (Illumina).

RNA-seq data processing and analysis

The obtained FASTQ files were controlled for quality with FastQC, the Batf3^–/–samples with according WT controls trimmed for adaptor sequences with Skewer⁴⁶ (v.0.2.2) and all files were aligned using STAR (v.2.7) to the GRCm38.98 reference genome using standard settings⁴⁷. The aligned data were counted using the featureCounts function of the Rsubread package in R⁴⁸. Differential expression analysis based on the raw count-matrix was performed using DESeq2 (ref. ⁴⁹). Significant differentially expressed genes were defined as having an adjusted P value <0.01 and log₂FoldChange > 1.5 and log₂FoldChange < −1.5. All significant differentially expressed genes were visualized using the pHeatmap package in R (https://CRAN.R-project.org/package=pheatmap). For unsupervised pathway analysis, the significant differentially expressed genes comparing WT and Batf3^–/– CD8⁺ T cells at 48 h and 72 h after activation were analyzed using the Enrichr tool^50,51. The BioPlanet database⁵² was used as a reference to test for pathway enrichment with standard Enrichr settings. The P value was calculated from a Fisher’s exact test. Combined scores were calculated by taking the log of the P value and multiplying that by the z score of the deviation from the expected rank.

Generation of CD8⁺ T_RM cells

Naive (WT tdTomato and Batf3^–/– GFP) T cell receptor transgenic CD8⁺ T cells (OT-I) were transferred into WT recipient mice followed by 1 × 10⁶ p.f.u. VV-OVA i.p. infection (prime). At the peak of T cell response (d8), both ears were i.d. infected with 1 × 10⁷ IU MVA-OVA (pull). T_RM cells were analyzed by FACS and under two-photon microscopy at d40 after MVA-OVA infection. Lymphocyte isolation from skin was performed by applying an enzymatic digestion-based protocol.

Statistical analysis

Apart from the genomic data, all biological data were analyzed using Prism 8 software (GraphPad) by a two-tailed paired Student’s t-test, two-tailed unpaired Student’s t-test or one- and two-way analysis of variance.

Reporting Summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The sequencing data are available at NCBI Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo) under accession numbers GSE143504 and GSE153341. The data that support the findings of this study are available from the corresponding authors upon request.

References

Kaech, S. M. & Cui, W. Transcriptional control of effector and memory CD8⁺ T cell differentiation. Nat. Rev. Immunol. 12, 749–761 (2012).
Article CAS PubMed PubMed Central Google Scholar
Bedoui, S., Heath, W. R. & Mueller, S. N. CD4⁺ T-cell help amplifies innate signals for primary CD8⁺ T-cell immunity. Immunol. Rev. 272, 52–64 (2016).
Article CAS PubMed Google Scholar
Eickhoff, S. et al. Robust anti-viral immunity requires multiple distinct T cell–dendritic cell interactions. Cell 162, 1322–1337 (2015).
Article CAS PubMed PubMed Central Google Scholar
Hor, J. L. et al. Spatiotemporally distinct interactions with dendritic cell subsets facilitates CD4⁺ and CD8⁺ T cell activation to localized viral infection. Immunity 43, 554–565 (2015).
Article CAS PubMed Google Scholar
Shaulian, E. & Karin, M. AP-1 as a regulator of cell life and death. Nat. Cell Biol. 4, E131–E136 (2002).
Article CAS PubMed Google Scholar
Karin, M., Liu, Z. & Zandi, E. AP-1 function and regulation. Curr. Opin. Cell Biol. 9, 240–246 (1997).
Article CAS PubMed Google Scholar
Murphy, T. L., Tussiwand, R. & Murphy, K. M. Specificity through cooperation: BATF–IRF interactions control immune-regulatory networks. Nat. Rev. Immunol. 13, 499–509 (2013).
Article CAS PubMed Google Scholar
Li, P. et al. BATF–JUN is critical for IRF4-mediated transcription in T cells. Nature 490, 543–546 (2012).
Article CAS PubMed PubMed Central Google Scholar
Iacobelli, M., Wachsman, W. & McGuire, K. L. Repression of IL-2 promoter activity by the novel basic leucine zipper p21SNFT protein. J. Immunol. 165, 860–868 (2000).
Article CAS PubMed Google Scholar
Grajales-Reyes, G. E. et al. Batf3 maintains autoactivation of Irf8 for commitment of a CD8α⁺ conventional DC clonogenic progenitor. Nat. Immunol. 16, 708–717 (2015).
Article CAS PubMed PubMed Central Google Scholar
Schlitzer, A. et al. Identification of cDC1- and cDC2-committed DC progenitors reveals early lineage priming at the common DC progenitor stage in the bone marrow. Nat. Immunol. 16, 718–728 (2015).
Article CAS PubMed Google Scholar
Hildner, K. et al. Batf3 deficiency reveals a critical role for CD8α⁺ dendritic cells in cytotoxic T cell immunity. Science 322, 1097–1100 (2008).
Article CAS PubMed PubMed Central Google Scholar
Edelson, B. T. et al. Peripheral CD103⁺ dendritic cells form a unified subset developmentally related to CD8α⁺ conventional dendritic cells. J. Exp. Med. 207, 823–836 (2010).
Article CAS PubMed PubMed Central Google Scholar
Betz, B. C. et al. Batf coordinates multiple aspects of B and T cell function required for normal antibody responses. J. Exp. Med. 207, 933–942 (2010).
Article CAS PubMed PubMed Central Google Scholar
Karwacz, K. et al. Critical role of IRF1 and BATF in forming chromatin landscape during type 1 regulatory cell differentiation. Nat. Immunol. 18, 412–421 (2017).
Article CAS PubMed PubMed Central Google Scholar
Punkenburg, E. et al. Batf-dependent Th17 cells critically regulate IL-23 driven colitis-associated colon cancer. Gut 65, 1139–1150 (2016).
Article CAS PubMed Google Scholar
Schraml, B. U. et al. The AP-1 transcription factor Batf controls T_H17 differentiation. Nature 460, 405–409 (2009).
Article CAS PubMed PubMed Central Google Scholar
Kuroda, S. et al. Basic leucine zipper transcription factor, ATF-like (BATF) regulates epigenetically and energetically effector CD8 T-cell differentiation via Sirt1 expression. Proc. Natl Acad. Sci. USA 108, 14885–14889 (2011).
Article CAS PubMed PubMed Central Google Scholar
Kurachi, M. et al. The transcription factor BATF operates as an essential differentiation checkpoint in early effector CD8⁺ T cells. Nat. Immunol. 15, 373–383 (2014).
Article CAS PubMed PubMed Central Google Scholar
Quigley, M. et al. Transcriptional analysis of HIV-specific CD8⁺ T cells shows that PD-1 inhibits T cell function by upregulating BATF. Nat. Med. 16, 1147–1151 (2010).
Article CAS PubMed PubMed Central Google Scholar
Logan, M. R., Jordan-Williams, K. L., Poston, S., Liao, J. & Taparowsky, E. J. Overexpression of Batf induces an apoptotic defect and an associated lymphoproliferative disorder in mice. Cell Death Dis. 3, e310 (2012).
Article CAS PubMed PubMed Central Google Scholar
Man, K. et al. Transcription factor IRF4 promotes CD8⁺ T cell exhaustion and limits the development of memory-like T cells during chronic infection. Immunity 47, 1129–1141.e1125 (2017).
Article CAS PubMed Google Scholar
Tscharke, D. C. et al. Identification of poxvirus CD8⁺ T cell determinants to enable rational design and characterization of smallpox vaccines. J. Exp. Med. 201, 95–104 (2005).
Article CAS PubMed PubMed Central Google Scholar
Gerlach, C. et al. The chemokine receptor CX3CR1 defines three antigen-experienced CD8⁺ T cell subsets with distinct roles in immune surveillance and homeostasis. Immunity 45, 1270–1284 (2016).
Article CAS PubMed PubMed Central Google Scholar
Bottcher, J. P. et al. Functional classification of memory CD8⁺ T cells by CX₃CR1 expression. Nat. Commun. 6, 8306 (2015).
Article PubMed CAS Google Scholar
Yajima, T. et al. IL-15 regulates CD8⁺ T cell contraction during primary infection. J. Immunol. 176, 507–515 (2006).
Article CAS PubMed Google Scholar
Rubinstein, M. P. et al. IL-7 and IL-15 differentially regulate CD8⁺ T-cell subsets during contraction of the immune response. Blood 112, 3704–3712 (2008).
Article CAS PubMed PubMed Central Google Scholar
van der Windt, G. J. et al. Mitochondrial respiratory capacity is a critical regulator of CD8⁺ T cell memory development. Immunity 36, 68–78 (2012).
Article CAS PubMed Google Scholar
Pearce, E. L. et al. Enhancing CD8⁺ T-cell memory by modulating fatty acid metabolism. Nature 460, 103–107 (2009).
Article CAS PubMed PubMed Central Google Scholar
O’Neill, L. A., Kishton, R. J. & Rathmell, J. A guide to immunometabolism for immunologists. Nat. Rev. Immunol. 16, 553–565 (2016).
Article PubMed PubMed Central CAS Google Scholar
Yamazaki, C. et al. Critical roles of a dendritic cell subset expressing a chemokine receptor, XCR1. J. Immunol. 190, 6071–6082 (2013).
Article CAS PubMed Google Scholar
Alexandre, Y. O. et al. XCR1⁺ dendritic cells promote memory CD8⁺ T cell recall upon secondary infections with Listeria monocytogenes or certain viruses. J. Exp. Med. 213, 75–92 (2016).
Article CAS PubMed PubMed Central Google Scholar
Mattiuz, R. et al. Novel Cre-expressing mouse strains permitting to selectively track and edit type 1 conventional dendritic cells facilitate disentangling their complexity in vivo. Front Immunol. 9, 2805 (2018).
Article CAS PubMed PubMed Central Google Scholar
Kitano, M. et al. Imaging of the cross-presenting dendritic cell subsets in the skin-draining lymph node. Proc. Natl Acad. Sci. USA 113, 1044–1049 (2016).
Article CAS PubMed PubMed Central Google Scholar
D’Cruz, L. M., Rubinstein, M. P. & Goldrath, A. W. Surviving the crash: transitioning from effector to memory CD8⁺ T cell. Semin. Immunol. 21, 92–98 (2009).
Article PubMed PubMed Central CAS Google Scholar
Pellegrini, M., Belz, G., Bouillet, P. & Strasser, A. Shutdown of an acute T cell immune response to viral infection is mediated by the proapoptotic Bcl-2 homology 3-only protein BIM. PNAS 100, 14175–14180 (2003).
Article CAS PubMed PubMed Central Google Scholar
Hildeman, D. A. et al. Activated T cell death in vivo mediated by proapoptotic Bcl-2 family member BIM. Immunity 16, 759–767 (2002).
Article CAS PubMed Google Scholar
Tussiwand, R. et al. Compensatory dendritic cell development mediated by BATF–IRF interactions. Nature 490, 502–507 (2012).
Article CAS PubMed PubMed Central Google Scholar
Sionov, R. V., Vlahopoulos, S. A. & Granot, Z. Regulation of BIM in health and disease. Oncotarget 6, 23058–23134 (2015).
Article PubMed PubMed Central Google Scholar
Hajji, N. & Joseph, B. Epigenetic regulation of cell life and death decisions and deregulation in cancer. Essays Biochem. 48, 121–146 (2010).
Article CAS PubMed Google Scholar
Shaulian, E. & Karin, M. AP-1 in cell proliferation and survival. Oncogene 20, 2390–2400 (2001).
Article CAS PubMed Google Scholar
Wei, J. et al. Targeting REGNASE-1 programs long-lived effector T cells for cancer therapy. Nature 576, 471–476 (2019).
Article CAS PubMed PubMed Central Google Scholar
Lynn, R. C. et al. c-Jun overexpression in CAR T cells induces exhaustion resistance. Nature 576, 293–300 (2019).
Article CAS PubMed PubMed Central Google Scholar
Lopez-Bergami, P., Lau, E. & Ronai, Z. Emerging roles of ATF2 and the dynamic AP1 network in cancer. Nat. Rev. Cancer 10, 65–76 (2010).
Article CAS PubMed PubMed Central Google Scholar
Weiser, C. et al. Ectopic expression of transcription factor BATF3 induces B-cell lymphomas in a murine B-cell transplantation model. Oncotarget 9, 15942–15951 (2018).
Article PubMed PubMed Central Google Scholar
Jiang, H., Lei, R., Ding, S. W. & Zhu, S. Skewer: a fast and accurate adapter trimmer for next-generation sequencing paired-end reads. BMC Bioinf. 15, 182 (2014).
Article Google Scholar
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Article CAS PubMed Google Scholar
Liao, Y., Smyth, G. K. & Shi, W. The R package Rsubread is easier, faster, cheaper and better for alignment and quantification of RNA sequencing reads. Nucleic Acids Res. 47, e47 (2019).
Article CAS PubMed PubMed Central Google Scholar
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014).
Article PubMed PubMed Central Google Scholar
Kuleshov, M. V. et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 44, W90–W97 (2016).
Article CAS PubMed PubMed Central Google Scholar
Chen, E. Y. et al. Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinf. 14, 128 (2013).
Article Google Scholar
Huang, R. et al. TheNCATS BioPlanet: an integrated platform for exploring the universe of cellular signaling pathways for toxicology, systems biology, and chemical genomics. Front. Pharmacol. 10, 445 (2019).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank the Core Unit for FACS and the Core Unit SysMed of the IZKF Würzburg for supporting this study, H.C. Probst (Institute for Immunology, Mainz University) and T. Kaisho (Immunology Frontier Research Center, Osaka University) for providing the P14 and XCR1-DTR-Venus mice³¹, respectively. The B8R tetramer was obtained through the National Institutes of Health Tetramer Core Facility. This work was supported by grants through the German Research Foundation, SFB TR 221 (project A03 to M. Hudecek and H.E.), Germany’s Excellence Strategy – EXC2151 – (390873048 to M. Hölzel) and German Research Foundation Emmy Noether programme (GA2129/2-1 to G.G.) and by grants of the European Research Council to W.K. (819329- STEP2) and G.G. (759176-TissueLymphoContexts). W.K., G.G. and M.V. are supported by the by the Max Planck Society (Max Planck Research Groups).

Author information

These authors contributed equally: Marco A. Ataide, Karl Komander.

Authors and Affiliations

Würzburg Institute of Systems Immunology, Max-Planck Research Group, University of Würzburg, Würzburg, Germany
Marco A. Ataide, Karl Komander, Konrad Knöpper, Annika E. Peters, Hao Wu, Sarah Eickhoff, Tea Gogishvili, Anika Grafen, Georg Gasteiger, Martin Vaeth & Wolfgang Kastenmüller
Medizinische Klinik und Poliklinik II, Universitätsklinikum Würzburg, Würzburg, Germany
Justus Weber, Hermann Einsele & Michael Hudecek
Department of Microbiology and Immunology, Peter Doherty Institute for Infection and Immunity, University of Melbourne, Melbourne, Victoria, Australia
Axel Kallies
Institute of Experimental Immunology, Medical Faculty, University of Bonn, Bonn, Germany
Natalio Garbi
Institute of Experimental Oncology, Medical Faculty, University of Bonn, Bonn, Germany
Michael Hölzel

Authors

Marco A. Ataide
View author publications
You can also search for this author in PubMed Google Scholar
Karl Komander
View author publications
You can also search for this author in PubMed Google Scholar
Konrad Knöpper
View author publications
You can also search for this author in PubMed Google Scholar
Annika E. Peters
View author publications
You can also search for this author in PubMed Google Scholar
Hao Wu
View author publications
You can also search for this author in PubMed Google Scholar
Sarah Eickhoff
View author publications
You can also search for this author in PubMed Google Scholar
Tea Gogishvili
View author publications
You can also search for this author in PubMed Google Scholar
Justus Weber
View author publications
You can also search for this author in PubMed Google Scholar
Anika Grafen
View author publications
You can also search for this author in PubMed Google Scholar
Axel Kallies
View author publications
You can also search for this author in PubMed Google Scholar
Natalio Garbi
View author publications
You can also search for this author in PubMed Google Scholar
Hermann Einsele
View author publications
You can also search for this author in PubMed Google Scholar
Michael Hudecek
View author publications
You can also search for this author in PubMed Google Scholar
Georg Gasteiger
View author publications
You can also search for this author in PubMed Google Scholar
Michael Hölzel
View author publications
You can also search for this author in PubMed Google Scholar
Martin Vaeth
View author publications
You can also search for this author in PubMed Google Scholar
Wolfgang Kastenmüller
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

M.A.A., K. Komander, M.V. and W.K. conceptualized the study and analyzed data. M.A.A., K. Komander, K. Knöpper, A.E.P., H.W., S.E., T.G, J.W. and A.G. planned and performed experiments and analyzed the data. K. Knöpper and M. Hölzel analyzed transcriptome data. A.E.P. generated critical reagents. M. Hudecek provided critical reagents. A.K., N.G., H.E., M. Hudecek, G.G., M. Hölzel provided intellectual input and gave conceptual advice. M.A.A. and W.K. wrote the manuscript with input from all authors. W.K. and G.G. provided research funds.

Corresponding authors

Correspondence to Marco A. Ataide or Wolfgang Kastenmüller.

Ethics declarations

Competing interests

M.A.A., M. Hudecek and W.K. are currently applying for patents relating to the content of this manuscript.

Additional information

Peer review information Peer reviewer reports are available. L. A. Dempsey was the primary editor on this article and managed its editorial process and peer review in collaboration with the rest of the editorial team.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Batf3^–/– and WT leukocyte chimerism.

Relative abundance of leukocytes after prime (d8; VV) and recall response (d6; VV) of bone marrow chimeric mice. Data are representative of two (n = 9 at prime and n = 7 at recall) independent experiments. Error bars indicate the mean ± SEM. Comparison between groups was calculated using two-tailed unpaired Student’s t-test.

Extended Data Fig. 2 OT-I CD8⁺ T cell immune response during L.m. OVA infection.

Kinetic analysis of co-transferred WT (GFP) and Batf3^−/− (CD45.1) OT-I cells in the blood of WT recipients after L.m.-OVA i.p. Data are representative of three independent experiments (n = 4). Error bars indicate the mean ± SEM. Comparison between groups was calculated using the One sample t-test. ***, p = 0.0003.

Extended Data Fig. 3 Measurement of mitochondrial potential during the contraction phase of CD8⁺ T cell immune response.

Representative dotplots, histograms and statistical analysis of mitochondrial membrane potential of the co-transferred WT (GFP) and Batf3^−/− (CD45.1) OT-I cells. Cells were analyzed at d8 and d12 post- VV-OVA infection. Data are representative of two independent experiments (n = 4). Error bars indicate the mean ± SEM. Comparison between groups was calculated using the paired Student’s t-test. * = p ≤ 0.05.

Extended Data Fig. 4 BATF3 programs T cell survival via the proapoptotic factor BIM.

(a) Kinetic analysis of co-transferred WT (CD45.1 / CD45.2) and Batf3^−/− (CD45.1) OT-I cells in the blood of IL15 deficient (IL15-/-) recipients after or VV-OVA i.p. infection. (b) qRT-PCR of Batf in ex vivo isolated WT and Batf3^−/− CD8 OT-I cells after VV-OVA infection, or 48 h hours post anti-CD3/CD28 in vitro stimulation. (c) Heatmap showing the 166 differentially expressed genes by comparing WT versus Batf3^−/− CD8 T cells in vitro stimulated with anti-CD3/CD28 for 48 h or 72 h. Columns represent technical replicates. (d) Sequence and schematic of the retroviral plasmid carrying the Bcl2l11 shRNA. pLMPd-shRNA-Ametrine were provided by Transomic®. The shRNA position is marked in red and reported by Ametrine (yellow). The cassette carrying the shRNA is flanked by LTRs (dark grey) and MESV Psi (light grey). Ametrine (yellow) is active by a PGK promoter (light grey). LTR: long terminal repeat; MESV: murine embryonic stem cell virus; Psi: RNA target site for packaging; PGK: PGK promoter. Data are representative of one experiment (n = 4). Error bars indicate the mean ± SEM. Comparison between groups (n = 4) was calculated using the One sample t-test. **, p = 0.007.

Extended Data Fig. 5 Schematics of the retroviral vectors.

For both retroviral over-expression plasmids pMIG was used as a backbone. In both cases the inserts, Batf3-IRES-GFP (dark blue and green, respectively) and Batf-IRES-GFP (light blue and green, respectively), were inserted between the LTRs (dark grey) and downstream of MESV Psi (light grey) and gag (grey-blue). As mock control pMIGR1-Ametrine was used. LTR: long terminal repeat; gag: structural precursor protein for packaging; MESV: murine embryonic stem cell virus; Psi: RNA target site for packaging.

Extended Data Fig. 6 BATF3-, but not BATF- overexpression promotes CD8⁺ T cell survival in vitro.

(a, b) Naïve CD8 T cells were isolated, activated and retrovirally transduced with pMIG- Ametrine (empty vector) or carrying the murine Batf3, pMIG-BATF3-GFP (BATF3) or the murine Batf, pMIG-BATF-GFP (BATF). Cells were resting for 48 h in culture after transduction and for additional 48 h after sorting before in vitro activation with anti-CD3/CD28. (a) Representative dotplots and statistical analysis of the in vitro kinetics of WT CD8 T cells coculture (50:50 – empty vector: BATF3) and (b) 50:50 – empty vector: BATF) in two different concentrations of IL-15. (c) Representative dotplots and statistical analysis of the in vitro kinetics of WT versus Batf3-/- CD8 T cells coculture (50:50 – empty vector: empty vector or empty vector: BATF3). Data are representative of one (a; n = 4) or two (b, c; n = 4) independent experiments. Error bars indicate the mean ± SEM. Comparison between groups was calculated using the One sample t-test. **, p = 0.007.

Extended Data Fig. 7 BATF3 overexpression programs CD8⁺ T cell stemness.

(a-d) Representative dotplots and gating strategy for (a) IL-7R, (b) ICOS, (c) intracellular CTLA-4 and (d) Tim-3 expression analysis in resting and after anti-CD3/CD28 stimulation of 50:50 (empty vector: BATF3) CD8⁺ T cell coculture. (e) Statistical analysis of Glycolitic Proton Eflux Rate (GlycoPER) readouts and (f) Oxygen Consuption Rate (OCR) parameters measured in resting state and 24 h after anti-CD3/CD28 stimulation. (a-f) Data are representative of two independent experiments (n = 3). Error bars indicate the mean ± SEM. Comparison between groups was calculated using two-tailed unpaired Student’s t-test.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Ataide, M.A., Komander, K., Knöpper, K. et al. BATF3 programs CD8⁺ T cell memory. Nat Immunol 21, 1397–1407 (2020). https://doi.org/10.1038/s41590-020-0786-2

Download citation

Received: 22 January 2020
Accepted: 17 August 2020
Published: 28 September 2020
Issue Date: November 2020
DOI: https://doi.org/10.1038/s41590-020-0786-2

This article is cited by

Chronic lymphocytic leukemia presence impairs antigen-specific CD8+ T-cell responses through epigenetic reprogramming towards short-lived effectors
- Anne W. J. Martens
- Inga Kavazović
- Arnon P. Kater
Leukemia (2023)
TET2 guards against unchecked BATF3-induced CAR T cell expansion
- Nayan Jain
- Zeguo Zhao
- Michel Sadelain
Nature (2023)
Remodeling and Restraining Lung Tissue Damage Through the Regulation of Respiratory Immune Responses
- Young Jin Pyung
- Da-Jeong Park
- Cheol-Heui Yun
Tissue Engineering and Regenerative Medicine (2023)
RNA polymerase II pausing factor NELF in CD8+ T cells promotes antitumor immunity
- Bogang Wu
- Xiaowen Zhang
- Rong Li
Nature Communications (2022)
A vaccine targeting resistant tumours by dual T cell plus NK cell attack
- Soumya Badrinath
- Maxence O. Dellacherie
- Kai W. Wucherpfennig
Nature (2022)