New therapeutics targeting immune checkpoint proteins have significantly advanced treatment of non-small cell lung cancer (NSCLC), but protein level quantitation of drug targets presents a critical problem. We used multiplexed, targeted mass spectrometry (MS) to quantify the immunotherapy target proteins PD-1, PD-L1, PD-L2, IDO1, LAG3, TIM-3, VISTA, GITR, and CD40 in formalin-fixed, paraffin-embedded (FFPE) NSCLC specimens. Immunohistochemistry (IHC) and MS measurements for PD-L1 were weakly correlated, but IHC did not distinguish protein abundance differences detected by MS. PD-L2 abundance exceeded PD-L1 in over half the specimens and the drug target proteins all displayed different abundance patterns. mRNA correlated with protein abundance only for PD-1, PD-L1, and IDO1 and tumor mutation burden did not predict abundance of any protein targets. Global proteome analyses identified distinct proteotypes associated with high PD-L1-expressing and high IDO1-expressing NSCLC. MS quantification of multiple drug targets and tissue proteotypes can improve clinical evaluation of immunotherapies for NSCLC.

Targeted MS analyses were performed on an Orbitrap Fusion Lumos Tribrid instrument (Thermo Scientific, Bremen, Germany) equipped with an Easy nLC 1200 liquid chromatograph and a Nanospray Flex ion source (Thermo Scientific, San Jose, CA). Reverse phase liquid chromatography was done with a PepMap RSLC C18-3 micron column, 75 micron x 15 cm, eluted at 250 nL/min with a mobile phase gradient consisting of solvent A (0.1% aqueous formic acid) and solvent B [(0.1% formic acid in water/acetonitrile (1:4, v/v)]. The mobile phase was initially 5% B and then programmed to 20% B over 18 min, to 35% B over 14 min and finally to 95% B over 5 min before recycling to starting composition. Targeted MS analysis was done by parallel reaction monitoring on the Lumos. The acquisition method consisted of a full scan selected ion monitoring event followed by targeted MS2 scans as triggered by a scheduled inclusion list, with a 5 minute retention time window containing the precursor m/z values. Retention times were determined from prior analyses of synthetic peptide standards. The MS1 scan was collected at a resolution of 30,000, an automatic gain control (AGC) target value of 5e4, and a scan range from m/z 350–1000. MS1 data were recorded in profile mode. The MS1 scan was followed by targeted MS2 collision induced dissociation scans at a resolution of 30,000, an AGC target value of 5e4, 1.6 m/z isolation window, activation Q of 0.25 and an optimized collision energy for each target of 30%. MS2 data were recorded in profile mode. Parallel reaction monitoring transitions were extracted from raw datafiles and analyzed with Skyline44. Peptide peak areas were calculated as the sum of three most abundant transitions. Peptide abundance was calculated from the ratio of peak area for the unlabeled endogenous peptide to peak area and spike amount for the labeled internal standard.

FFPE tissue blocks of NSCLC tissue were commercially acquired from iSpecimen (Lexington, MA, USA) and Individumed (Frederick, MD, USA). Vendor source and tissue diagnosis are provided in Table S5. Microtomy was performed immediately prior to experiment initiation. Tissue sections were cut at 5µm thickness and allowed to dry at room temperature overnight.