AFFILIATIONS
1 Broad Institute, Cambridge, United States, Dana-Farber Cancer Institute, Boston, United States
2 Broad Institute, Cambridge, United States, Dana-Farber Cancer Institute, Boston, United States
3 Broad Institute, Cambridge, United States, Dana-Farber Cancer Institute, Boston, United States, Harvard Medical School, Boston, United States
4 Broad Institute, Cambridge, United States, Dana-Farber Cancer Institute, Boston, United States
5 Broad Institute, Cambridge, United States, Dana-Farber Cancer Institute, Boston, United States
6 Dana-Farber Cancer Institute, Boston, United States, Harvard Medical School, Boston, United States
7 Broad Institute, Cambridge, United States
8 Dana-Farber Cancer Institute, Boston, United States, Harvard Medical School, Boston, United States
9 Dana-Farber Cancer Institute, Boston, United States
10 Dana-Farber Cancer Institute, Boston, United States
11 Dana-Farber Cancer Institute, Boston, United States, Harvard Medical School, Boston, United States
12 Broad Institute, Cambridge, United States
13 Broad Institute, Cambridge, United States
14 Broad Institute, Cambridge, United States
15 Dana-Farber Cancer Institute, Boston, United States, Harvard T. H. Chan School of Public Health, Boston, United States
16 Boston University, Boston, United States, Dana-Farber Cancer Institute, Boston, United States
17 Broad Institute, Cambridge, United States, Dana-Farber Cancer Institute, Boston, United States, Harvard Medical School, Boston, United States
18 Broad Institute, Cambridge, United States, Dana-Farber Cancer Institute, Boston, United States, Harvard Medical School, Boston, United States
ABSTRACT
Targeting the HLA-E/NKG2A immune checkpoint holds substantial therapeutic promise, yet its regulatory mechanisms, cancer-specific vulnerabilities, and optimal therapeutic approaches remain incompletely defined. We discovered that HLA-E was highly abundant in diffuse large B cell lymphoma (DLBCL) through spatial proteomic characterization and hence were motivated to identify new targets to overcome this potential immune evasion mechanism. Indeed, through genome-wide CRISPR loss-of-function screens in DLBCL cell lines, we discovered that HLA-E was uniquely regulated by SEC62, an endoplasmic reticulum (ER) membrane protein that facilitates post-translational protein import into the ER. We prioritized SEC62 for further characterization because its depletion consistently and selectively reduced HLA-E expression across all tested cell lines, had no previously established role in classical HLA class I (HLA-I) regulation, and had minimal effects on overall cellular viability.
To dissect how SEC62 regulates HLA-E, we established isogenic SEC62 knockout (KO) lines in HBL1, SU-DHL-4, and TMD8 DLBCL cells. These lines maintained a 42-58% reduction in HLA-E surface expression (p<0.005) without altering HLA-I. Intracellular levels of HLA-E were similarly reduced (24-47%, p<0.001), but HLA-E transcript was unchanged consistent with a post-transcriptional regulatory mechanism. Using subcellular fractionation, we found that SEC62 was required for HLA-E ER transport. SEC62 KO increased levels of a cytoplasmic precursor form of HLA-E, which was subsequently degraded by the proteasome. This level of regulation was dependent upon the 20 amino acid leader sequence of HLA-E (i.e. the signal peptide). We exchanged the HLA-E signal peptide with alternative leader sequences derived from HLA-I and CD8α, which rescued ER transport and restored HLA-E surface expression in SEC62 depleted cells. These data support HLA-E as a direct client of SEC62, which mediates ER transport via the signal peptide sequence.
In addition to this direct regulatory mechanism, we explored secondary effects of SEC62 loss that may impact HLA-E regulation in DLBCL. We compared bulk transcriptomic, proteomic, and ubiquitinomic profiles between sgRNA controls and SEC62 KO isogenic cell lines. This identified global shifts in the proteome with upregulation of pathways involved in ER associated protein degradation, protein processing and N-glycosylation when SEC62 was depleted. Notably, ERAP1, an important aminopeptidase involved in trimming peptides for HLA presentation, was post-transcriptionally reduced through increased ubiquitination (padj=0.029). We subsequently found that SEC62 depletion shifted the HLA-E peptide repertoire and reduced the presentation of the canonical 9-mer peptide (‘VL9’) derived from cleaved leader sequences of other HLA class I molecules.
Functionally, SEC62 depletion reduced engagement with the NKG2A receptor, as measured by a Jurkat NKG2A reporter, and enhanced NK cell mediated cytotoxicity in co-culture with primary NK cells (≥3 donors, p<0.05). SEC62 KO also augmented antibody-dependent cellular cytotoxicity, resulting in greater sensitivity to rituximab. Finally, we assessed whether Sec62 depletion enhanced antitumor immunity in the immunocompetent A20 lymphoma mouse model. Analysis of tumor infiltrating lymphocytes demonstrated that 66% of NK cells and 29.2% of CD8+ T cells expressed NKG2A. Despite similar growth kinetics in vitro, Sec62 KO impaired tumor growth (p=1.02x10-9) and prolonged overall survival in vivo (p=0.0013), an effect which was dependent upon NK cells. Furthermore, Sec62 KO tumors also contained fewer terminally exhausted PD1+ TIM3+ CD8+ T cells relative to control (26.3% vs. 12%, p=0.01), consistent with a more favorable antitumor immune microenvironment.
Together, our findings illuminate new HLA-E regulatory mechanisms through SEC62 and define the HLA-E/NKG2A axis as a therapeutically actionable immune checkpoint in DLBCL.